Home

Funcom Hardware Statistics System Vision Document

1. Activity Person Sondre Sverre Kent Kim Dag 18 25 10 10 10 10 11 8 6 6 10 4 2 4 Total 33 25 18 20 18 20 51 5 1 2 December Activity Person Sondre Sverre Kent Kim Dag 8 4 2 2 2 37 5 31 28 18 5 32 1 5 4 3 4 5 1 5 2 2 5 Total 48 40 5 36 26 39 5 51 5 1 3 January Activity Person Sondre Sverre Kent Kim Dag 12 75 7 5 7 5 5 14 75 8 5 12 38 8 49 37 39 25 42 10 25 3 8 5 12 5 1 3 5 20 20 27 24 19 24 2 75 29 6 Total 109 5 92 5 115 107 97 51 5 1 4 February Activity Person Sondre Sverre Kent Kim Dag 15 8 75 5 3 5 4 0 25 9 5 4 7 5 25 25 1 8 2 5 9 5 54 75 49 5 61 20 48 10 25 14 5 1 8 30 2 75 4 3 2 Page 80 of 194 Project FHaSS Project Plan 2012 Total 108 25 83 25 79 63 79 51 5 1 5 March Activity Person Sondre Sverre Kent Kim Dag tings amp plann 5 7 5 3 5 3 5 3 5 2 5 8 7 10 7 1 5 1 1 5 88 75 53 5 42 33 5 49 5 2 5 5 0 5 5 12 21 5 20 22 21 22 th 1 5 1 2 0 5 Total 123 25 95 77 5 82 92 51 5 1 6 April In the month of April the project team had one 1 whole week off from the project to load the batteries with some Easter vacation Activity Person Sondre Sverre Kent Kim Dag Meetings amp planning 4 75 2 5 1 2 0 5 2 1 5 12 75 5 8 5 44 3
2. 194 Page 186 of 194 Project FHaSS Reflection Document 2012 60 Introduction The purpose of this document is to view our work and implementation period in retrospect giving a short explanation of the implemented functionalities and features evaluate what was excluded from the extensions and go into detail about what could be done differently and explain the workarounds we had to undergo with to get the FHaSS system to the usable state it is in now 61 Challenges 61 1 Loading time When we first had a meeting with the product owner we were told that 2 seconds load time was the requirement and that it was a quite feasible time to retrieve the entries from the database Looking at this in retrospect we now understand that this wouldn t be doable given that there are far beyond four million entries in the hardware statistics database and that this number will increase each day At the start of our project we had huge problems with the loading time which required us to constantly try to optimize the database fetching queries and algorithms In the first version of the system we handled everything on demand per request Everything in this context means handling the request from the users web browser setting up the connection with the WCF web service retrieving the hardware statistics from the database process and prepare the data and return them to the MVC layer for display This solution gave us very lon
3. 41 ES Oe ele Ile Un 41 Ska BE ln ET 42 352 Testatnalyst ic penetret re ast d deno ledende ute s dense diet 42 35 3 EE EE 42 Page 37 of 194 Project FHaSS Test Strategy 2012 36 SOUCO S mane 42 30 Requirements for testing The main purpose for testing is to ensure all specifications in the requirement specification are implemented and approved The tester should distinguish between constructive and destructive approaches to the test Think about how the unit to be tested may fail but do not be so critical that the unit never measure up and is seen as unacceptable It is important when the testing is executed that the person doing the test is not the same person who has written the code to be tested This is because the person who has written the code often gets blind on their own work and does not want to find any errors in the code If the person who executes the test has not written the code the person goes deeper into the code and may think of other ways to test the code that the developer may not have thought of 31 Test strategy This will be a system were the importance of stability and reliability is high at any time The producer of the game needs to get the information they want from the bug system The tests need to be accurate and test the system in all ways possible Our main strategy of testing will be a gradual step by step approach critical to less impor
4. Ungrouped Unrecognized registered hardware shall be grouped in an Unknown group in the statistics A NF 1 Programming language A A NF 1 Meetings Product Backlog amp Technology Document The whole system is to be implemented using the programming language C on top of the ASP NET framework 27 10 11 A NF 2 Model View Controller A A NF 2 11 12 11 Product Backlog The code of the web interface shall be designed and implemented following the ASP NET Model View Controller pattern A NF 3 Database type A A NF 3 11 12 11 Product Backlog The system must be compatible with a database of the type Oracle A NF 4 Database connection A A NF 4 11 12 11 Product Backlog The system that is both the data collection from database and the parser must make use of nHibernate for connecting to and accessing the database A NF 5 Database access A A NF 5 11 12 11 Lorenzo Aurea The web interface shall be given an interface to the database from a Windows Page 51 of 194 Project FHaSS Requirements Specification 2012 Communication Foundation web service In other words it shall not interface with the database directly A NF 6 Web page loading time A A NF 5 11 12 11 Product Backlog The web page displaying
5. Tester Sverre C Thune WSS ial paliareres Function test black box Test approach 3 Load the web page containing the software statistics 4 Compare the calculated percentages to the manually calculated percentages Test criteria e Database and website is up and running e Test dataset has been made and inserted into database Expected results e The percentages the system produces and displays on the web page based on the test data set are similar to those calculated manually Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors Outcome 56 5 T R 6 Future hardware prediction After discussion with the project manager we decided to not perform this test due to the time it would take to create the data for the test data set and this being a C requirement The functionality is implemented and debug tested locally by the creator to ensure that it works but it is not documented Future hardware prediction est ID T R 6 Requirement ID C F 1 ussee rsessMe Test to see if an extension of requirement A F 2 works The extension is to predict the future trend of hardware usage for the players of Funcom games When the chart for requirement A F 2 is shown it will be possible to toggle the prediction on and off The prediction will extend the current lines for hardware evolution in the chart and predict the future evolution Testdate
6. Kim Richard Johansen WSS ie galsereres Function test black box Page 180 of 194 Project FHaSS Test log 2012 Test approach 6 Putin the amount of time into the future the system is to predict the hardware evolution for 7 Load the web page containing one of the possible charts containing future hardware prediction 8 Toggle the chart to display the future hardware prediction chart lines if toggled off 9 Check that the hardware evolution prediction lines are proportional 10 Compare the updated percentages in proximity to the chart to the manually calculated percentages This stage corresponds to testing the last time point on the chart to the manually calculated percentages for the whole test data set for the whole time period put in 1 Test criteria e Database and website is up and running e Test dataset has been made and inserted into database Expected results e That the hardware evolution prediction lines are proportional e The percentages the system produces and displays on the web page based on the test data set are similar to those calculated manually Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors Outcome 57 Verification tests This is meant as a short summary for each of the verification tests performed during our project period 57 1 Code review Most of the code revie
7. Test criteria what type of criteria Expected results e Expected errors and results Test creator name date updated by name date Errors e Encountered errors Outcome What happened 47 2 Validation tests 47 2 1 List of validation tests e T R 1 Display hardware software statistics e T R 1 1 Display hardware software statistics CPU information e T R 1 2 Display hardware software statistics GPU information e T R 1 3 Display hardware software statistics BAM information e T R 1 4 Display hardware software statistics OS information e T R 1 5 Display hardware software statistics Direct X information e T R 2 Display hardware software evolution charts e T R 3 1 Filter statistics e T R 3 2 Filter statistics user interface test e T R 3 3 Filter statistics game filter e T R 3 4 Filter statistics server universe filter e T R 3 5 Filter statistics time period filter e T R 3 6 Filter statistics geographical location filter e T R 4 1 Hardware type management and grouping e T R 4 2 Hardware type management and grouping new hw type e T R 5 Display software statistics e T R 6 Future hardware prediction Page 60 of 194 Project FHaSS Test Specification 2012 47 2 2 Tests Display hardware software statistics est ID T R 1 REOS RS A F 1 and B F 1 ES ESddaielk Test to see if tables are created and contains accurate information about the d
8. HOGSKOLEN i Buskerud Funcom Hardware Statistics System Vision Document 00000 Projectname Acronym Y Project name Acronym FHaSS Sensors amp Supervisors Internal sensor Olaf Hallan Graven Internal supervisor Aurilla Aurelie Arntzen External supervisor Rui M Monteiro Casais Group members Sondre Bjerkerud Sverre Christoffer Thune Kim Richard Johansen Funcom Hardware Statistics System Client Funcom N V 10 01 2012 Dag Hem Kent Brian Dreyer Project FHaSS Vision Document 2012 Vision Document 14 General Document Information Deliverable nr D1 3 3 1 Deliverable type Research Release Public Work package 1 Responsible Sondre Bjerkerud 14 1 Declaration of intention This document is intended to give the reader a fundamental insight into our project assignment by giving an introduction to our client and their area of business It will give a more in depth explanation of the context and background of the system to give the reader an understanding of the necessity of the system Last but not least this document will go through what to improve about the current system how to improve it and the benefits our client will obtain by using the new system 14 2 Definitions and Acronyms MMORPG Massive Multiplayer Online Role Playing Game Crash A crash is a term for when a software system fails A crash usually occur because of incompatibility be
9. 64 2 View raw data view full screen graphs Above the graphs there are two links which opens in a new browser tabs The linked views contains raw graph data showing all the entries in a HTML table for each date The other link extends the desired graph in to a full screen window for in depth analysis To store the filter information for use in a new View we store the settings in Session variables These variables timeouts and resets after 20 min of inactivity but can be increased in binding options We experimented putting this information as cookies but were deemed useless as the Session variables work better keeping the time out in mind These functionalities were a request from the product owner but prior to this already noted as an extension task 64 3 X axis zoom for graphs The graphs support zooming by clicking and dragging the mouse over the desired period This works in the Home index as well as in the full screen extension We experimented with making this zoom compatible for both X and Y axis but since the graphs are of the type stacked Percent graph zoom for the Y axis would be irrelevant as the data must be compared to the other hardware entries thus we did not include the y axis zoom functionality 64 4 Exporting the graphs In the top right corner of the graphs we implemented a functionality to export the graphs to an image file PNG JPEG as well as support for exporting it to a PDF file This was a fe
10. Added new database and parser design Sverre Added CSS amp UI Design Dag 5 Added design for GPU manufacturer and vram 23 02 2012 6 Added design for memory and os 29 02 2012 7 Added Redesign of responsibilities of WCF and Parser and GPU user 01 03 2012 story Sondre 8 Added HWParser Interface design Dag 01 03 2012 9 Update design with CPU model 14 03 2012 10 Updates after code review before first deployment 19 03 2012 11 HighCharts 18 04 2012 Page 132 of 194 Project FHaSS Design Document 2012 Geographical Location Dag 12 0 Windows service 24 04 2012 13 0 General model relation search manufacturer 27 04 2012 14 0 Hard disk information 04 05 2012 15 0 Filter storage and URL exception handling and logging caching 09 05 2012 16 0 Final updates after document review 28 05 2012 Page 133 of 194 Project FHaSS Design Document 2012 31 Table of contents 30 General Document Information 132 30 1 Declaration of intentio rete it rere reet rent tei v ab eee ve eR ee tee vie yea 132 30 2 Definitions and Acronyms issesesesseeseeeeeennnnn enne nennen iinne naiean seran sans asses nean 132 30 3 RR le elle Tele 132 31 Tableof contents tree ete tire irent eee ee ine EEN kimen 134 32 Introduction unernuses D 135 33 CT cami Ll TT 135 34 CPU Speed grouping sicccccscsueceases
11. 16 Deliverable guidelines This section covers the structure layout and routines of the group s deliverables it also defines the quality process and timelines that any group member should follow 16 1 Deliverable distribution Deliverable distribution between members of the project group will usually happen using an ad hoc communications channel and as a benefit of using the Dropbox file sharing system we will always be able to reach our documents on our workstations Any completed deliverable should be moved from its temporary folder to the Deliverables folder and then to the correct Workpackage Distributing a deliverable to either supervisor requires that the document has been completed and reviewed by a fellow project member See section16 2 3Pre completion process before delivering the document either by e mail cd dvd or a paper copy If a deliverable is to be delivered in paper format it should be printed out in color and stapled together at the top left corner also review the document and check if any page layout errors are visible Page 117 of 194 Project FHaSS Quality Assurance Document 2011 16 1 1 Deliverable list An updated list off all deliverables and their respective delivery date is found in our Gantt diagram which all project members and supervisors have access to 16 1 2 Naming convention Any completed deliverable should be named after the following format D Workpackage nr Deliv
12. H GSKOLEN i Buskerud Funcom Hardware Statistics System Final Project Documentation Funcom Hardware Statistics System Client Acronym Funcom N V Sensors amp Supervisors Internal sensor Olaf Hallan Graven Internal supervisor Aurilla Aurelie Arntzen External supervisor Rui M Monteiro Casais Author Sondre Bjerkerud Sverre Christoffer Thune Kim Richard Johansen Dag Hem Kent Brian Dreyer Table of documents User Requrement Spechftcation neneeese ee sett testet teser tssttttssttttssttttssttntssttttssttntsstennesten ns stene 3 Programming Language Document 7 DEI P 16 Current System Description iiie nice EA ha tiene na pine raa sn de ea pin EERRE ee aa Funde eaa dva de na pince ea ande 26 INQUIETA ELITR UL II Le 35 Requirements Specification oteen cre te esa cerea eed EE Ee 43 Test Specification oe rere eoe rt eae erue eskere 55 dri darn EAEEEdE 74 Risk Analysis Doc mlent 5 2 1 1 01 retorica ce Fe toc koe anu Fe tace od uan Fe eur DRe Foe naue Fe our aad d n 87 Project Model Document seseseesseeesesnressetsstrrsssttssttttssttttsst tnte stt ts stnn s Stt tss tt tnst neesst en nnst nense en nne 101 IDE Document d deeg Ed aie epa erre to pete had dd EAEEREN EEE Ea 107 Quality Assurance Document 112 Version Control Document 124 Design Docutmnent oer Ea VE ee np
13. Test description Test date Testing method Test approach Test criteria Display software statistics est ID T R 5 B F 2 Test to see if the tables are created and contain the same percentage value as manually calculated from the test data set about the different software stored in the database 14 04 2012 Tester Sverre C Thune Function test black box 1 Loadthe web page containing the software statistics 2 Compare the calculated percentages to the manually calculated percentages e Database and website is up and running Page 70 of 194 Project FHaSS Test Specification 2012 e Test data set has been made and inserted into database Expected results e The percentages the system produces and displays on the web page based on the test data set are similar to those calculated manually Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors e Encountered errors Outcome What happened Future hardware prediction est ID T R 6 Requirement C F 1 ussee sessMe Test to see if an extension of requirement A F 2 works The extension is to predict the future trend of hardware usage for the players of Funcom games When the chart for requirement A F 2 is shown it will be possible to toggle the prediction on and off The prediction will extend the current lines for hardware evolution in the chart a
14. distribution ended up to be per month and for the whole project period 49 2 Definitions and Acronyms Product Owner JS HTML CSS Scrum project model The stakeholder s representative when following the JavaScript Hyper Text Markup Language Cascading Style Sheets 49 3 Document History Version 1 Description First version created Time estimations done and Gantt diagram made Date 16 12 2011 Project document layout adopted Resources responsibility areas and requirements breakdown sections added 04 01 2012 Modified section Personnel under the main section Resources Added section Remake of the Requirement breakdown under the main section Requirement breakdown Added section Remake of the Gantt diagram under the section Gantt diagram 19 03 2012 Added Activity description and Actual resource usage sections under the 28 05 2012 Page 75 of 194 Project FHaSS Project Plan 2012 Resources section Also added Milestones accomplishment section under Milestones section 50 Table of contents 49 General Document Information 75 49 1 Declaration of intention eere err teer ana aaa aaa e aea ada 75 49 2 Definitions and Acronyms rrrrnnvvrnnnvnnsnvrrsnvvrrnannrnnnvrrnnvrrnan enne innen sn nene en nnns nennen nnns 75 49 3 Document HIStOFV uradel yet be wens v Ee a k
15. e Meetings amp Planning o Meetings and preparation for meetings with supervisors and or sensors as well as internally in the project team o Completion of minutes Page 78 of 194 Project FHaSS Project Plan 2012 o Mailing and Skype chatting in conjunction with meetings and or planning o Scrum task breakdown estimation and picking e Documentation o Project documentation o Wikipages documentation o Project documentation research e Research amp Set up Learning Reading Lessons Installation of software Server set up O O O 0 0 Fixing of software problems e Implementation o Coding C C JS HTML CSS debugging building and test executing new code o Uploading files to file repository and resolving file merging issues o Code review o Bug fixing o Stylecop ReSharper code standard maintenance e Design o UML design o Ul design o Design discussion verbally and text chat o Conduction o Preparation o Logging e Presentation o Conduction o Preparation e Other Deployment to Funcom Test deployment Contracts Room ordering Prepare documentation for delivery Etc O 0 OO 0 Page 79 of 194 Project FHaSS Project Plan 2012 51 5 Actual resource usage 51 5 1 Per month 51 5 1 1 September October and November
16. Internal sensor Olaf Hallan Graven Internal supervisor Aurilla Aurelie Arntzen 24 05 2012 Author External supervisor Rui M Monteiro Casais Sondre Bjerkerud Sverre Christoffer Thune Kent Brian Dreyer Kim Richard Johansen Project FHaSS Test Specification Test Specification 44 General Document I nformation 2012 Deliverable nr D1 7 4 Deliverable type Report Release Public Wo Res rkpackage 1 ponsible Kim Richard Johansen 44 1 Declaration of intention This is a document containing information about all the tests that are to be done All the tests are made with basis upon the requirements specification that contains a list of all the requirements 44 2 Definitions and Acronyms Bug An er ror in the programming code Parser Searches a data file for key words and copies the values it finds CPU Central Processing Unit Graphical Processing Unit Random Access Memory DirectX Hard Disk Drive Solid State Disk A type of HDD Operating System Cond ition of Satisfaction 44 3 Document History Version Description Date 1 First version created 15 12 2011 2 Document template adopted 05 01 2012 Information reviewed and updated 3 Updated with new small scale 16 03 2012 tests to be used to test sprint 24 04 2012 results 4 Updated doc
17. Universes GroupCpuSpeed speed int CpuSpeedGroup GroupCpuCores cores int CpuCoresGroup GroupVram vram int Vram Group GroupRam ram int RamGroup GroupGpuModellmodelName string GeneralGpuModel GroupCpuModel modelName string GeneralCpuModel Figure 26 HWParser classes Page 155 of 194 Project FHaSS Design Document 2012 Some unnecessary methods have been removed from the parser classes after rewriting parts of the code CountCpuManufacturers hardwaselnfolist IListcHardwarelnfo gt FetchResultPart i CountGpuManuf infolist IList Hardwarelnfo FetchResultPart Count VRam hardwarelnfolist IListeHartwarelnfo FetchResultPart Figure 27 Service classes A third service has been created to contain the code for communication with the parser object The diagram below shows the main components for the current system componenb gj Gamer PC Figure 28 Overall system components Page 156 of 194 Project FHaSS Design Document 2012 43 HighCharts For creating the charts we were recommended to use a JavaScript API called HighCharts This API contains a lot of different functionalities and has a lot of customization options available to generate whichever charts necessary for displaying the relevant information The way they are generated is by dynamically generating JavaScript with Razor Ctt syntax using loops and conditions HighCharts are struct
18. database hw info v games Y W id INTEGER id INTEGER Q username VARCHAR Q game VARCHAR Q universe VARCHAR Q os VARCHAR Universe ToGame Q cpu VARCHAR ram VARCHAR Q vcard VARCHAR Q vram VARCHAR universes v Q driver VARCHAR id INTEGER Q bits VARCHAR V games id INTEGER FK Q count INT Q universe VARCHAR Q resolution VARCHAR LA universes FKIndex1 Q timestamp TIMESTAMP V games id Figure 1 Database tables The overall system will contain two applications The WCF Service Application where the database queries and calculations will take place and the ASP NET MVC3 Application which is the web side application that displays the information from the database The WCF Service Application contains the IReadDataService cs which is the interface of the service and the ReadDataService svc which implements the interface There are also the mapped classes which nHibernate uses to communicate with the database tables Below figure 2 shows the classes that will be in the ReadDataService application Page 135 of 194 Project FHaSS Design Document 2012 Figure 2 ReadDataService class diagram By finding the universes that exist in the HW INFO table the system won t need to be updated in the future as it will automatically detect new universes when they are added to the database You only need to set up the new relations between the games and the universes so that application will take the new server into account T
19. it should be printed that it is an SSD disk While the design is quite simple the task is quite a challenge because it needs to be written in C without the use of NET libraries or CLR so that it eventually can be included in the The Secret World game client This will require finding and learning to use low level functions in FE 2 Find disks windows 1 GetHddinfo For each disk Pu 3 GetModelNumber FE 4 GetFree Space r 5 GetTotalSpace P 6 PrintHddinfo Figure 31 Sequence for finding hard disk information Page 159 of 194 Project FHaSS Design Document 2012 48 Filter storage and URL There are some filters the users of the system would like to be able to view easily and fast because it is information they regularly want to review or show to different people Two new functionalities will give the users the ability to do this The first one is the filter storage The user should be able to select what he wants in the filter menu and the give the filter a name and save it The filter should then appear in a secondary menu that all users of the system can see When the name of the filter is clicked the filter will be loaded To store the filter some new database tables are needed One table will store the name of the filter and the timer period Another will store countries connected to saved filters and the last will store the universes connected to the saved filters Figur
20. self explanatory and intuitive yet expansive Changes in this section however might require reparsing and regrouping of the database structure which can be time consuming for the back end of the system For this reason we display warnings before any changes are applied The users should be aware of this and the administration section should not be used to frequent as it is designed for larger changes in the system Page 191 of 194 Project FHaSS Reflection Document 2012 We feel that we have covered every necessary option in this section the Funcom employees has not raised any additional requests to this part There was problems getting all of this completely debugged but as of now none are reported 66 Reflection Looking back at the last months of the development period it is easy to see what mistakes the project team has done and what could have been done differently 66 1 Requirements capturing In the startup of the project the team should have put more time into understanding exactly what way and to what purpose the FHaSS was going to be used A better understanding of this would have saved us very much time On the other side we would have had to put in more time in the pre face of the project and probably much more time in a pre face design period which is not according to how the Scrum project model suggests the project workflow 66 2 Differences in development and production environments Ever since the first att
21. 10 1 Isit possible to use this programming language to solve our bachelor task 12 10 2 Does it work well with other planned programming languages cssccccessseceeeesteeeeseaaes 12 11 PHP Hypertext Preprocessor iet eee eege 13 11 1 Isit possible to use this programming language to solve our bachelor task 13 11 2 Does it work well with other planned programming languages cssccccessseeesesteeeeeenaes 13 12 n COR UE Eegeregie dcc pU E cute uie deed 14 12 1 Languages used for database retrieving and web development 14 12 2 Languages used for database parsing essssssesseeeeeeee eene nnne nennen 14 METIRI III LLLI LL 14 13 1 FR 14 dE DE Ium E 15 13 3 de ln EE 15 Page 9 of 194 Project FHaSS Programming Language Document 2011 7 Introduction Through this document we will review different computer languages for the development of our project task We need to find the technology that meets the necessary requirements needed Our client has already requested that we use C with ASP NET so that the system is easy maintainable and accessible for future expansions 8 Python Python is a general purpose high level programming language whose design philosophy emphasizes code readability This object orientated language is quite similar to PHP in its ability to create dynamic web pages and various types of software web applications One of Pythons
22. 11 1 Is it possible to use this programming language to solve our bachelor task Yes but it would not be optional over C with ASP NET because of its inability to easily debug and its bad error handling Security is not much of an issue since our system will run on a closed intranet neither is the server side preprocessing There will never be enough users using this system at the same time 11 2 Does it work well with other planned programming languages As C with ASP NET would also be used as our web programming language they would fulfill the same role Our PHP supported web page would also have no trouble communicating with a database built up from either a C or a Python developed parser Benefits Drawbacks Less server memory usage than ASP NET Lacks API Consistency Less learning curve Scripted language not compiled Hard to debug Bad Cross platform capable error handling Easy access to the database Not thread safe Does not have native support for Unicode or multibyte strings No debugger Page 13 of 194 Project FHaSS Programming Language Document 2011 12 Conclusion 12 1 Languages used for database retrieving and web development The choice stands between ASP NET PHP and Python They are all pretty similar in functionality and all of them are capable of doing what is needed in our bachelor assignment though ASP NET is dependent of C to perform database calls What separates these languages is
23. 2 VAR vs specified type ssssssssssseseseeeeee ener nenne nennt nasse s nnne ti asas asser na nan 121 18 23 LINQUSABC 18 ea 121 Page 114 of 194 Project FHaSS Quality Assurance Document 2011 18 2 Versioning Control Pertorce 122 EAE II 122 1331 Sle Opies guten tyeecceanyendece yeeiGenneces 122 18 32 E ET 122 ake ReELENCES ROTER DD CDU QI LL CILLIT III 123 Page 115 of 194 Project FHaSS Quality Assurance Document 2011 14 Introduction As our group consists of persons from very different backgrounds we therefore have many different views on document layout document and code quality timeliness communication and information exchange This gives us the need for a collected set of rules and guidelines so that we are all on the same level regarding document code and communication standards To ensure the highest possible quality in all our written forms Deliverables documents code syntax and other publications every member of the project group shall read and apply the rules discussed in this document to their work 15 Communication and Information Exchange This section is divided into two parts and contains information regarding the communication flow and information exchange between the members of the project group the supervisors and our client The communications section handles how the group will communicate with others and what communication channels that are used while the information e
24. 27 04 2012 Tester Kim Richard Johansen WSS ie glsierere Function test black box Test approach 1 Putin the time period for which the statistical data shall be produced with basis upon 2 Choose to get to see statistical information for GPU 3 Wait for the web page containing the information to load 4 Compare the resulting percentages to the percentages that have been manually calculated from the test data set Test criteria e Database and website is up and running Page 64 of 194 Project FHaSS Test Specification 2012 e Test data set has been made and inserted into database Expected e The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Kim Richard Johansen 16 03 2011 Errors e Encountered errors Outcome What happened Display hardware software evolution charts est ID T R 2 REOG DA A F 2 EEE Test to see if evolution charts are created and to see if the contained information is accurate against the test data set Test date 02 05 2012 Kim Richard Johansen N Ne Function test black box Test approach 1 Putinthe time period for which the evolution charts are to be produced with basis upon 2 Choose to get to see one of the possible evolution charts CPU GPU CPU cores or Windows OS version and DirectX ve
25. CPU usage e No compilation no debugger Open source Can be used as both our database parser and our web development language Page 10 of 194 Project FHaSS Programming Language Document 2011 9 Ct C is a simple modern general purpose object oriented programming language developed by Microsoft running on the NET Framework CH is part of the Common Language Runtime and the most widely used one The CLR is the virtual machine component of Microsoft s NET framework and is responsible for managing the execution of NET programs Visual C Project Creates IL metadata amp references loaded by CLR NET Framework Converted to native machine code Figur 1 A graphical representation from the initial execution of CH code and down the NET framework hierarchy CH uses visual studio as its native IDE also developed by Microsoft for easy development for Windows operating systems Very similar to other object oriented programming languages i e java and C CH is however not a cross platform language so in order to make this run on Linux Mac etc a 3 party framework called Mono is used The stated purpose of Mono is not only to be able to run Microsoft NET applications cross platform but also to bring better development tools to Linux developers Mono can be run on Android BSD iOS Linux Mac OS X Windows Solaris and Unix operating systems as well as some game console operating systems
26. Document history uausaneradesamepudenuesuaekeentdenteenanesgmeridelntiet 88 59 Table of contents erri paces vaeayedeanedheestvagntsabssanevedessites socuyaevsasseViebastvesesesseersvoanbextes 89 60 Risk atialysls cac Dages ed Eerad Tage eo ce Rue 90 60 1 RISK MaAMGlinG iere concede iter erre eerte Let ee eese dd ES eek decedens TAO 90 60 2 Risk table template 5 eire repere et roe ERR pep dle cured ie peer aa 90 OMNE Lc 91 61 1 Requirements related risks rrrannrrrnrannvrvnrannvrvnrensrrvnrensrrensensrvnnssnsrnnnssnsrnnnssnvnnnssnrsnnssnnnneesene 92 61 2 Technicalrisks riter nece the SEET eege eaa 93 61 3 LIEN GER 94 61 4 Other risks eterne ni e E en 97 62 ee lg Ro LEE 98 62 1 Requirement related siente lerne eere dee resi ri sa eerie Tra pu eet rr Po euo dg 98 62 2 Technical ier HE rte Ed EHE E idee ange 98 623 MEM rue 99 62 4 Other ie ettet Enero ge et ath wa poe ees cer A eed uen Edd 100 Page 89 of 194 Project FHaSS Risk Analysis Document 2012 60 Risk analysis 60 1 Risk handling Risks are a large part of project management and as the project grows larger so does the amount of risks we are going face As this is a small scale project with only a handful of developers we do not expect to be overwhelmed by many risky decisions but it is nevertheless important to have a basic knowledge of the risks we are up against and how to handle them in the best possible manner The understanding of risk re
27. GSKOLEN i Buskerud Funcom Hardware Statistics System Project Model Document Funcom Hardware Statistics System Client Funcom N V Acronym Sensors amp Supervisors Internal sensor Olaf Hallan Graven 10 01 2012 Author Sondre Bjerkerud Internal supervisor Aurilla Aurelie Arntzen External supervisor Rui M Monteiro Casais Sondre Bjerkerud Dag Hem Sverre Christoffer Thune Kent Brian Dreyer Kim Richard Johansen Project FHaSS Project Model Document 2012 Project Model Documents 1 General Document Information Deliverable nr D2 3 1 PENE ETER Research Release Public Work package WP2 Responsible Sondre Bjerkerud 1 1 Declaration of intention This document will explain several project models that our team has considered working according to as well as a conclusion discussing which model we believe will fit our project the best and why 1 2 Definitions and Acronyms UP Unified process XP Extreme programming 1 3 Document History Version Description Date 1 0 First version created 19 12 2011 1 1 Peer reviewed Document 10 01 2011 template adopted Page 102 of 194 Project FHaSS Project Model Document 2012 2 Table of contents 1 General Document Informatio sristi adegit aede nne Fee ERR ede ERR onte ANE aen aba Fede 102 1 1 Declaration of ihtentiOD cie ere
28. P PUER ere Cha d Ge Ene eeu ud 160 49 Exception handling and logging 160 S ME S el aA I T LE 161 cueste 162 Page 134 of 194 Project FHaSS Design Document 2012 32 Introduction The purpose of this document is to present our design for user stories in the sprints Each section of the document contains the design for a specific user story or larger overall updates to the design This document will be expanded when we create new designs throughout the project period so one can see how the system has evolved UML diagrams are used to describe the system so general knowledge of UML is recommended to fully understand this document 33 Game filter Our first objective is to complete the Game filter user story The goal of this user story is to display how many entries there are per Funcom game in the HW INFO table There is no column in the HW INFO table that tells us which game the row of information is connected to There is however a column called Universe which stores the name of the universe server We can use the universe name to find out the game To do this we will create a table that contains different games and a table that contains the different universesand relations between these two tables so that when one searches for one specific game it will search through all the universes connected to that game Below figure 1 you can see the tables that will be in the
29. Page 11 of 194 Project FHaSS Programming Language Document 2011 10 C with ASP net ASP NET is part of the NET Framework and when coding ASP NET applications you have access to classes in the NET Framework You can code your applications in any language compatible with the CLR including Microsoft Visual Basic and C These languages enable you to develop ASP NET applications that benefit from the common language runtime type safety inheritance and so on By compiling C you get compatible ASP NET code C may also be compiled as a file that can be executed from a website This gives great flexibility and makes code modification easy 10 1 Is it possible to use this programming language to solve our bachelor task Yes to build a parser create threads and to make SQL queries it will prove very useful for this task including the NET debugger it will be easier to build solid code than with Python Including the fact that we can generate the full web representation by calling ASP NET from C simplifies the development process greatly C with ASP NET was also highly recommended by our client Funcom for possible future expansion 10 2 Does it work well with other planned programming languages C is interoperable with the rest of the NET framework ASP NET and by using this we can develop everything we would need for the given task with C only C is something that that the team hasn t worked with yet but it merg
30. ReadDataService had to evolve to supply more and more data using less and less time and is solely used to fetch sort count and group statistical information It contains the following functionality when fetching information it always accounts for the filtered options such as universes country and time span e Fetching Sorting Counting Grouping information for every hardware part from the database e Fetching Sorting Counting Grouping information for a single hardware part from the database e Fetching information for a month and storing this in a cache e Sorting Counting Grouping information for every hardware part from the cache e Sorting Counting Grouping information for a single hardware part from the cache e Estimate future values for a set of data 64 Graphs and visual presentation of data 64 1 HighCharts For creating the visual graphs we were recommended to use a JavaScript API called HighCharts This API contains a lot of different functionalities and has a lot of customization options available to generate whichever graph necessary for displaying the relevant information Our graphs are of the type Stacked Percent Area which was a request from the product owner This type of graph displays the percentage of entries for a hardware part based on the other hardware parts entries We started out with having a fixed resolution of 12 x axis points where we printed out the date from and date to summarizing the entries betwee
31. Scrum team responsibilities Following the Scrum project model a project team will be doing a standard meeting before each sprint iteration In this sprint startup meeting the team will together with the Product Owner the stakeholders representative decide which user stories requirements to implement in the following sprint All the user stories agreed upon goes into the Sprint Backlog where they are broken down into tasks It is then up to the team members to pick the tasks they would like to do in the following sprint In our project team we have somewhat clearly defined fields of responsibility as given in the section Fields of responsibility These responsibilities will be the basis for task picking in the sprint startup meeting For instance Sverre will mainly pick designing tasks while Kent will mainly pick implementation tasks In the case when there will be an overweight of tasks relating to the fields of responsibility of only a part of the team the other members will help the responsible persons by picking tasks from their Page 82 of 194 Project FHaSS Project Plan 2012 field of responsibility However the person with the field of responsibility relating to that task will have the head responsibility of the task independently of who is actually doing it For instance can Dag be implementing some functionality of the system but it is still Kent who has implementation as one of his fields of responsibility who wil
32. Systems ETOILE OLUUULMA 129 27 e elle TN ET 129 pueden 130 29 SOUT IDIOT 130 Page 126 of 194 Project FHaSS Version Control Document 2012 22 Introduction In this document we will take into account that IDE document our deliverable about which IDE we are going to use concludes that we are going to use Visual Studio 2010 Ultimate to write code So a prerequisite for this document is that we are using the Visual Studio 2010 Ultimate IDE and therefore need a version control system that works well with Visual Studio To work simultaneously on the same Visual Studio project we need a version control system The purpose of this document is to decide which version control system is best to use for our project Certain operations that we need to do will be tested in the different version control systems This includes installing the server and client adding files checking out files checking in files merging files viewing changes between revisions and reverting to an older revision 23 Version Control Version control is also known as revision control and source control A version control system is software that keeps track of the history of your project including code files documents and any other type of file that you want to track It is also a tool for sharing a project so that multiple people a team can work on the same project even the same files simultaneously This is done by having a server which hosts the main
33. Test to see if tables are created and contains accurate information about the different OS information stored in the database It should be possible to toggle between number of entries and COS Show the OS from all systems in db and group them under different groups Data to be shown e OS grouped under Windows 7 64 bit Windows 7 32 bit Windows Vista 64 bit Windows Vista 32 bit Windows XP 64 bit Windows XP 32 bit Unknown Test date 14 03 2012 Tester Kim Richard Johansen WSS ial paljeareles Function test black box Test approach 5 Put in the time period for which the statistical data shall be produced with basis upon 6 Choose to get to see statistical information for GPU 7 Wait for the web page containing the information to load 8 Compare the resulting percentages to the percentages that have been manually calculated from the test data set Test criteria e Database and website is up and running Page 172 of 194 Project FHaSS Test log 2012 e Test data set has been made and inserted into database Expected e The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Kim Richard Johansen 13 03 2011 Errors e None Outcome After the test data set had been parsed and grouped the information displayed on the website was correct according to the
34. The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Kim Richard Johansen 13 03 2011 Errors e Zero errors Outcome All the different universes gave the same results and they corresponded to the test data set Page 174 of 194 Project FHaSS Test log 2012 EN est Failure 55 8 Story 9 Geographical location filter Filter statistics geographical location filter est ID T R 3 6 RGT Pias A F 3 usse sese Test to see if the geographical location filter combinations are working correctly It should be possible to choose between continent and country as a filtering option COS That the data shown after choosing a continent or country corresponds to the manually calculated results Continents to be able to choose from North America South America Antarctica Africa Europe Asia Australia Test date 27 04 2012 Tester Kim Richard Johansen WSS ie plsereles Function test black box Test approach 4 Setthe filter combination to be tested 5 Wait for the web page containing the filtered statistics to load 6 Compare the produced filtered percentage statistics to the manually calculated percentages Test criteria e Database and website is up and running e Test dataset has been made and inserted int
35. YER EF ee dba Ee Us WERE ee 131 ient me 163 Project Reflection Document 5 ntt oth inkaenes sendes dene se nane Gea dn ase guia Gee aaia 183 H GSKOLEN i Buskerud Funcom Hardware Statistics System User Requirement Specification Funcom Hardware Statistics System Client Acronym Funcom N V Sensors amp Supervisors Internal sensor Olaf Hallan Graven Internal supervisor Aurilla Aurelie Arntzen External supervisor Rui M Monteiro Casais 10 01 12 Author Sondre Bjerkerud Sondre Bjerkerud Sverre Christoffer Thune Kim Richard Johansen Dag Hem Kent Brian Dreyer Project FHaSS User Requirement Specification 2011 User Requirement Specification 1 General Document Information Deliverable nr D1 1 1 ADMIN Report Release Public Workpackage 1 Responsible Sondre Bjerkerud 1 1 Declaration of intention This document contains the functional and non functional requirements captured during two meetings with Funcom e g September at Funcom Oslo 27 10 2011 Skype meeting This document gives a quick overview of what we are going to do achieve in our project 1 2 Definitions and Acronyms DxDiag Diagnostics file containing hardware and software information 1 3 Document History Version Description Date 1 0 First version created 28 10 2011 1 5 Template added 10 01 2011 1 6 Updates after document review 2
36. a saved filter would produce the user would get very bad loading times for a simple operation having to wait for all the hardware statistics to load once more The solution that the project team has come up with to remove the need for the single Saved Filters administration page is to implement the administration functionality into the Filter Menu like explained above by using AJAX By using AJAX the web browser can send HTTP requests without having to reload the page and thereby good response times could be kept up while at the same time offering a good user interface to the Funcom employees 67 4 Graph resolution The graphs on the Home page of the web page interface of FHaSS is currently displaying one data point for each day for each of the top six 6 groups of the hardware category If the user choose to get statistics from the last year for instance by using the Time Period Filter the charts would display 365 6 more than 2000 data points which is quite a lot Given that the charts are displayed side by side the amount of space they ve got for displaying their data is limited and with many data points Page 193 of 194 Project FHaSS Reflection Document 2012 the charts are looking very spiky The charts could potentially be more informative if the chart resolution number of data points was either static variable and calculated based on the charts available space or given by the user With some more fixing in the bac
37. actually do The amount of work the team accepts in this meeting shall and will be expected to be finished at the end of the sprint 56 1 2 2 Sprint review meeting The purpose of this meeting is for the project team to show off to the Product Owner what has been developed in the past sprint The newly produced functionality shall either be accepted or denied by the Product Owner or eventually form the basis for an eventual changes in the Product Backlog 56 1 2 3 Sprint retrospective meeting The purpose of this meeting is for the project team to discuss what went well and what went not so well in the past sprint The conduction of such a discussion is supposed to improve the effectiveness and efficiency of the work of the project team 56 2 Updates to the project plan The Gantt diagram will typically be updated if needed after each sprint every second week The work velocity of the project team is very uncertain at this point but estimates will probably improve drastically after just a few sprints as well as the speed itself The same goes for the requirements implementation time estimates The result is that the project plan will probably have to be updated often reflecting both a new work velocity estimate and the new requirements implementation time estimates Page 85 of 194 Project FHaSS Project Plan 2012 56 3 File repository The project team will be using two types of file repositories in this project 1 Dropbox will b
38. add ons to make sure our code follows our client s standard Each student using ReSharper is advised to use theconfiguration filelocated in Other ReSharper CodeStyleSettings v6 Page 122 of 194 Project FHaSS Quality Assurance Document 2011 19 References e http blogs msdn com b sourceanalysis Last visited 24 05 2012 e http www jetbrains com resharper Last visited 24 05 2012 e http www perforce com Last visited 24 05 2012 e https www dropbox com home Last visited 24 05 2012 Page 123 of 194 AP HOGSKOLEN i Buskerud Funcom Hardware Statistics System Version Control Document Project name Funcom Hardware Statistics System Client Acronym Funcom N V FHaSS Date Sensors amp Supervisors 29 05 2012 Internal sensor Olaf Hallan Graven O Internal supervisor Aurilla Aurelie Arntzen Sverre C Thune External supervisor Rui M Monteiro Casais Sondre Bjerkerud Dag Hem Sverre Christoffer Thune Kent Brian Dreyer Group members Kim Richard Johansen Project FHaSS Version Control Document 2012 Version Control Document 20 General Document Information Deliverable nr D2 6 1 1 PENE Er ER Research Release Public Workpackage 2 Responsible Sverre C Thune 20 1 Declaration of intention The purpose of this document is to decide which version control system is best to use for our project 20 2 Definitio
39. and multi monitor resolution Test date 27 04 2012 Kim Richard Johansen USS ial galjieteres Function test black box Test approach 1 Putinthe time period for which the statistical data shall be produced with basis upon 2 Choose to get to see statistical information for GPU 3 Wait for the web page containing the information to load 4 Compare the resulting percentages to the percentages that have been manually calculated from the test data set Test criteria e Database and website is up and running e Test data set has been made and inserted into database Expected e The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Kim Richard Johansen 23 12 2011 Errors e Encountered errors Outcome What happened Display hardware software statistics RAM information est ID T R 1 3 Reese RES A F 1 and B F 1 EEE Test to see if tables are created and contains accurate information about the different RAM stored in the database It should be possible to toggle between number of entries and COS Show the RAM from all systems in the DB and group them under different range groups Page 62 of 194 Project FHaSS Test Specification 2012 Data to be shown e RAM grouped under Less than 513 MB 513 MB to 999 MB 1GB 2 GB 3 GB 4 GB 5GB and higher Unknown Tes
40. black box 1 Start the filter combination to be tested 2 Wait for the web page containing the filtered statistics to load 3 Compare the produced filtered percentage statistics to the manually calculated percentages e Database and website is up and running e Test dataset has been made and inserted into database e That the produced filtered percentages based on the test data set are similar to the manually calculated percentages for the tested filter combination Test creator Kim Richard Johansen 31 01 2011 Errors e Encountered errors Outcome What happened Filter statistics server filter est ID T R 3 4 Requirement ID Test description Test date Testing method Test approach Test criteria Expected results A F 3 Test to see if the server filter combinations are working correctly e Test if results after filtering show Funcom servers and entries for each of them 26 02 2012 Kim Richard Johansen Function test black box 1 Start the filter combination to be tested 2 Wait for the web page containing the filtered statistics to load 3 Compare the produced filtered percentage statistics to the manually calculated percentages e Database and website is up and running e Test dataset has been made and inserted into database e That the produced filtered percentages based on the test data set are similar to the manually calculated
41. bug report functionality in their games the players themselves can report back errors that has to be fixed By taking this advantage even further and opening up the game for one or several pre release beta testing periods Funcom can get bug and crash information sent to them even before the official release of the game This results in Funcom having to spend less time on testing and quality assurance for the games themselves that is the costs of the testing phase is automatically reduced Such companies can then use more time actually fixing the bugs rather than finding them The fact that Funcom employees will have to spend a lot of time fixing bugs and hardware incompatibilities has resulted in a demand of a system that effectively handles the information gathered about the faults in their games Funcom is currently using a system for this purpose that is both not precise enough for their use nor hold the required functionality that Funcom needs 17 4 Hardware trends The evolution of the computer technology is as discussed earlier very fast paced This is resulting in a problem for game development companies that it can be hard to determine what type of or more precisely the capabilities of the hardware that will be mainstream when their new game is planned to be released Without this knowledge the companies would have to base this solely on guessing which could result in a released game that is technically outdated or on the contrary a g
42. by only tiny factors it s very hard to get out the exact wanted information For instance it is usually more informational for the employees to know the percentage difference in amounts of players that are using nVidia or ATI GPU s rather than the amount that are using specific graphics cards from specific manufacturers It all comes down to the ability for the employees to filter what type of information they would like to see Some information can be more informal to the user when they have been structured grouped organized in a certain way The current system is also lacking the future hardware evolution trend prediction functionality described in the Context of the study part of this document This is the functionality that the employees can use to get a prediction of what hardware will be mainstream at a time in the future for instance on the release date of a new game By getting such information employees can plan for optimizing the game towards the future mainstream hardware Further on there is currently no functionality for filtering the statistical data That is the statistics are based on absolutely all data stored in the database This is an enormous amount of data and is gathered from players all over the world for many years and from several games Funcom is therefore in need of functionality for filtering the statistical data based on certain criteria such as geographical location of the players time period game title an
43. combinations are working correctly e Test if results after filtering universe show entries Test date 26 02 2012 Kim Richard Johansen WSS ia paljereres Function test black box Test approach 1 Start the filter combination to be tested 2 Wait for the web page containing the filtered statistics to load Page 167 of 194 Project FHaSS Test log 2012 3 Compare the produced filtered percentage statistics to the manually calculated percentages Test criteria e Database and website is up and running e Test data set has been made and inserted into database Expected e That the produced filtered percentages based on the test data set are results similar to the manually calculated percentages for the tested filter combination Test creator Kim Richard Johansen 31 01 2012 Errors e None Tried each filter combination and the right amount of entries did show for each of them Outcome Tried each filter combination Each of the 23 universes showed 30 entries and that is correct with the manual calculations in the test data set Combining universes also worked showed and combined entries correctly At this stage in the project CPU results did show for the universe filtering but due to the parser not being 100 at the moment results did not match the test data set Test Failure 55 3 Story 5 CPU information Display hardware software statistics CPU in
44. decision that the Parser Page 151 of 194 Project FHaSS Design Document 2012 will now take over the grouping functionality it would also be a breach of the separation of concerns to still have these functions still running in the WCF layer In meetings with Funcom employees we have found out that they want to be able to control almost everything related to the systems grouping functionality manually Since the only interface our system has with the user is the MVC and the only interface between the MVC and the database is the WCF the WCF is of course then the part that will have to handle this database table updating The WCF has earlier only been used for retrieving data from the database and sending it to the MVC for display Since the WCF has now been extended with this new administration functionality it has become obvious that the most appropriate action is to separate the earlier service methods and the new admin methods into their own services The old functionality will still be stored in the ReadDataService service while the new admin functionality will be stored in the AdminService service Figure 22 Admin Service class 40 GPU user story Because of a large variety in the different GPU models that is registered in the hw info database a bit more sophisticated filtering and grouping functionality had to be used for this If we had gone with the same functionality as used in earlier user stories the hardware statist
45. development periods Let s take a look at each part and discuss which drawbacks can appear when reducing focus on them specifically The inception of an idea is a more or less instant process and can be neglected regarding time consumption The game design is a more or less time consuming activity but requires just a single or a few employees to be done In addition the idea inception and the game design are tasks that can be done an unlimited time in advance to actual development of the game So to some extent those two activities can be disregarded Further on regarding the game design it is now more and more common that game development companies work according to an agile project model which usually emphasizes the pre development design phase a lot less than the models mainly used in the past This reduces the focus on the game design part automatically without the administration having to actively reduce it Reducing the focus on the implementation phase will result in all or some of the functionality being of lesser quality than originally intended or some earlier designed functionality not being developed at all Consequently the game title will not be as good as it was supposed to be potentially resulting in less games sold and thereby less income for the company As a result most game development companies which are aiming for a large amount of sold games and having their existence depending upon their games selling well like Funcom
46. diode 117 16 1 Deliverable distribution esee eene enne nnn nnne nennen nnn nnn nennen 117 16 11 JDeliverable list ttr tee e eee eet e eee EE eet visne 118 16 12 Naming CONVENTON iiss ecce tee enne ne enu geg Eua eaae eene E nean esee teda caue Nee Dua aea RYE 118 16 13 E el E itor rede eren terere E oce Poo ene de eas dichas e scuto e ete 118 16 2 Deliverable quality process sssessessee eene enne enne nnn enne nnne ena 118 16 2 1 Document creation and Structure 119 16 2 2 TIMENE S c i eer tetendit eh tester tete ME Errat uberes eae ees cas 119 16 2 2 1 Timeliness amp exceptions sese eee eene enne enne nnne 119 16 2 3 Pre completion process cce tiec e haee Cea ee d e ae Ee Ce Eee uen Madre EAR 119 16 2 4 Reviewing adeliverable seen enne enne nennen nnns 120 17 Other publications and releases ei ceeee decine eter eei cen oe cree sa d idevassaldeususdceeeavebeccesuaces 120 17 1 Notice of meeting Agenda and Minutes guidelines eese 121 17 1 1 Notice of meeting amp Agenda 121 17 12 MINUut s n sk c a eceene sed ee 121 18 Coding standards and plugins ssrrrerannrvvnrannrvrnrennvvrnrsnsrvenssnsrvenssnsrnnnssnsnnrnsssennrnsssennnnsssennrnsssennnnee 121 18 1 Coding guidelini8s ree rented date tan oe ue ese rta ete Eae le T ede ete eo eee d 121 18 1 1 Naming conmventions nennen nennen ennt nnns nist estne tnn nr E nnne 121 18 1
47. done editing a file or multiple files you need to check them in just like when adding new files you then have to submit the change in the pending list As mentioned earlier to submit you are required to add a comment describing what the change is This way it is easy to track the history since you can view the comments of every change submitted If two different users have checked out the same file and both wants to check them in a conflict occurs and it has to be resolved This can be done be merging the files P4 has a tool for this which works very well You can view the changes from both of the users and compare them to both the current revision on the server and the automatically merged file If you are happy with the result accept the merge and re submit the merged result You can also edit the merge to your liking You can compare different revisions of files by using the diff function in the both visual studio and the P4V GUI client Changes between the revisions are displayed very well and it is easy to understand It is also no problem reverting to an older revision of a file Deleting files locally can create problems You have to delete files with the P4 GUI client so the server knows If you delete a file locally it will not be downloaded again on successive updates because the server thinks you already have it If this happens you may need the clean your local copy and download it again which can be time consuming By just ma
48. each of them based on the total This test will also contain e Test if registered but unrecognized hardware is grouped in the unrecognized category and shows the right percentage Test date 22 05 2012 Tester Sverre C Thune WSS ial paljeerele Function test black box Test approach 1 Getthe statistics page Print out write down etc the groups and their respective percentages 2 Compare the group percentages to percentages manually calculated from the test data set Test criteria e Database and website is up and running e There has to be one or more currently registered but not recognized hardware types e Test data set has been made and inserted into database Expected results e That the system produced percentages is the same as the manually calculated ones Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors e Encountered errors Outcome What happened Page 69 of 194 Project FHaSS Test Specification 2012 Hardware type management and grouping new hw type est ID T R 4 2 Requirement ID Test description Test date Testing method Test approach Test criteria Expected results Test creator Errors Outcome A F 4 and A F 5 Test is run after T R 4 1 Create a new general hardware type and choose which type of hardware it should contain from the unreco
49. filter is available when a specific Funcom game is selected and unavailable when a game isn t selected WS eres Kent Brian Dreyer usus User interface test Black box Test approach 5 Set all filter options blank null 6 Check if game server filter is unavailable 7 Choose one of the available Funcom games as a filter 8 Check if game server filter is available Test criteria e Database and website is up and running e Test data set has been made and inserted into database Expected e Game server filter works as explained in the test description results Page 179 of 194 Project FHaSS Test log 2012 Test creator Kim Richard Johansen 9 01 2012 Errors Outcome 56 4 T R 5 Display software statistics This test was to check if the software programs e g Skype Steam MSN stored on the computer from a player experiencing a game crash would show up correctly in the right groups under the software statistics table This functionality was not implemented due to other priorities and time and therefore there were no need for this test to be executed Display software statistics est ID T R 5 AEEA ek B F 2 x sswersresieusme Test to see if the tables are created and contains the same percentage value as manually calculated from the test data set about the different software stored in the database Testdate
50. five days a week from the end of Easter until project end The project team will be having one full week of Easter vacation Take a look at the Requirements breakdown document attached to this document to get an overview of how the amount of work for the project has been estimated 51 2 Hardware Each of the members of the project team owns their own portable computer These will be used for just about anything regarding the project work development project management web page updates etc The portable computers are of course available at all times 51 3 Software The project team will be using several software applications for the project work e GanttProject Project management e Microsoft Excel Google Docs Spreadsheet Project management e Visual Paradigm UML Design e Visual Studio Coding e Perforce Version control e Microsoft Word Documentation e MySQL Database We will also be using several pieces of software that cannot be considered to be applications The software are either necessary for the system to work or are required by our client to be used e ASP NET with C Programming language amp Compiler e ASP NET MVC Web page framework e nHibernate Database abstraction layer 51 4 Activity description The following is a list of the different activities that the work hours has been categorized in as well as a more detailed list of what types of specific activities that belong to each activity
51. have the last years more and more converted their workflow towards agile practices This goes for Funcom as well working by their own modified version of Scrum Of this simple reason Funcom urged us to use Scrum as our project model When the UP type of requirements are already set from the school there isn t much that the project model choice is actually influencing Even though we will try to take the practices of Scrum to the largest extent as possible both because of the urge from Funcom but also because it seems to be the most appropriate model to work by for this project given the team composition the fluid requirements and the good easy to understand and clearly defined structure of the Scrum model this will probably at most influence the day to day workflow of the project as well as maybe some per iteration planning We will also be adopting the short iterations of only one week that are usually the case when working by XP The reason for this is that is gives the team more frequent chances of steering the development process more frequent and thereby more improvements of time estimation for planning purposes and more frequent evaluation of the development process for the purpose of improving the team s practices All as a result of more frequent sprint planning and sprint evaluation meetings It is however possible that the lengths of the sprints will be increased to two weeks after the initial couple of sprints if we feel tha
52. high amount of crashes experienced An even larger pressure will also be tested to ensure that the system will manage larger work than intended All data inserted into the system is checked up against manually calculated results to ensure that no data has been corrupted due to high system load 33 2 8 Regression test The regression test is meant to see if previously tested functions in our system still work after new code has been implemented or after an update 33 2 9 Performance test This type of testing is to check if the systems performance is good enough e g the time it takes for the website to load show results as charts or tables with information from the database Page 40 of 194 Project FHaSS Test Strategy 2012 34 Test documentation Test documentation will help us to quality assure our project The documents will also give us some indication of the time used to quality assure our project in accordance to the requirements specification 34 1 Test specification The test specification contains information about every test to be done with cross reference to other documents that may be relevant 34 2 Test log The point behind a test log is to document every test that has been executed It will contain e Name of tester e When did the test occur e What was the result e Error list e Evaluation 34 3 Test reports Tests that constitute a module will be reported in a test report The meani
53. in our database including the distinctcpumodel and generalcpumodel tables In addition to the previous updates all the data models that nHibernate uses and the nHibernate helper are moved to its own project 42 Updates after code review before first deployment After reviewing our code before the first deployment better ways to structure our system were discovered These updates are represented in the updated UML diagrams below Page 154 of 194 Project FHaSS Design Document 2012 Id int Universe Universe Name sting GeneralGpuModel GeneralGpuModel Figure 25 nHibernateMappings classes All the MappedClasses created for nHibernate and the helper are moved to the new NHibernateMappings project so that the mappings can be used in both the services and the parser easily Parser nHibernateHelper NHibemateHelper string Parser stringParser Run void GetHardwarelnfoList session ISession limit int IList GetCpuManufacturerList session ISession IList AutoRelateDistinctToGeneral GpuModels void GetGpuManufacturerList session ISession IList AutoRelateDistinctToGeneralCpuModels void GetOsList session ISession IList GetResolutionList session ISession IList ParseString stringToParse string startWo ParseString stringToParse string startWo GetNextUngroupedHardwareList session ISession IListsHardware gt GroupUniverse universeName string
54. list of the requirements as well as a more in depth description of each one After having read this document the reader shall know what features the product is to hold and how the features shall work 37 2 Definitions and Acronyms BugzZilla Bug handling software used by Funcom CPU Central Processing Unit GPU Graphical Processing Unit RAM Random Access Memory DirectX A collection of application programming DX interfaces APIs for handling tasks related to multimedia like for instance game programming Uniform Resource Locator 37 3 Document History Version Description Date 1 Created the document Created Requirements list 15 12 2011 and Requirements details sections 2 Added priority explanation and requirement 19 12 2011 capturing sources Updated origin and date fields of the requirement details tables 3 Rewrote the text about the requirement capturing 3 01 2012 meeting Project document template incorporated 4 Added Courses with Lorenzo Aurea as a 19 03 2012 requirement capturing source Added requirement A NF 5 Rewrote redefined requirements A F 1 A F 4 A NF 2 and A NF 3 5 Added paragraph under the Meetings with 21 05 2012 Page 44 of 194 Project FHaSS Requirements Specification 2012 external supervisor requirement capturing source Added requirements B NF 1 C F 2 C F 3 C F 4 and C NF 1 Rewrote redefined requi
55. lt Story 7 OS ANTONM ATION Ss sui ise de intended drite a EU ese se dees bee e e Deoa ER ERE 172 55 7 Story 8 Memory Information 174 55 8 Story 9 Geographical location Dlter anaa 175 55 9 Story 11 Time p riod filter tte eite ertet ripe Pee ye Ete de agere 176 55 10 Kugeler ee le Reie 176 OM ve 177 56 Validation tests that were not performed and ww 178 56 1 List of validation tests that were not performed rrnrvrrrnnnnornnnnnnnrnrrnannvnrrnnnnvnrrnnnnnnrrnnnnnnrr 178 56 2 T R 1 Display hardware software statistics sese nennen 178 56 3 T R 3 2 Filter statistics user Interface 179 56 4 T R 5 Display software statlstlcs 180 56 5 T R 6 Future hardware predicton 180 57 MORITICATIOM TESTS e 181 SLI ee DEE 181 57 2 Code COMPILATION acc tee testet rare besvarte 181 57 3 Ee EIDEL 181 LE gei 181 57 5 Regression test Ad hoc test 182 RE Oe tel EE 182 Page 165 of 194 Project FHaSS Test log 2012 54 List of validation tests completed Userstory Name of test Test id Date 3 Game filter Filter statistics game filter T R 3 3 02 02 2012 4 Server filter Filter statistics server filter T R 3 4 26 02 2012 5 CPU information Display hardware software T R 1 1 27 04 2012 statistics CPU information 11 03 2012 6 Graphics card information Display hardware software T R 1 2 27 04 2012 statistics GPU information 6 7 Hardware type Hardware type management
56. main strength s is its simple syntax and easy to read code That means even though this might be a new language for the team we will be able to learn the language quite fast Python is a cross platform scripting language and can easily connect to different databases on multiple platforms As mentioned earlier Python is a general purpose programming language this means that we could also use Python as our database parser as we would with C Python is an old language this means that it had time to grow and developers has created a plethora of tools and frameworks for it to make Python perform tasks that were otherwise lacking or hard to achieve 8 1 Is it possible to use this programming language to solve our bachelor task Yes Python is versatile both when it comes to web development and database information retrieval as well as parsing text It would also be simple enough for us to learn without setting us back in productivity The current system was based on Python but the client request was primarily to use C with ASP NET framework 8 2 Does it work well with other planned programming languages Yes if we were to create our database parser in Python it would have a minimal effect on either PHP or ASP NET which would be used as our web development tools We could choose to use Python as the main web development tool and it would have no issues with a parser created from CH Benefits Drawback Very fast little memory
57. minor errors if possible but it is the author s job to be aware of and possibly improve any area commented by the reviewer 17 Other publications and releases In addition to formal deliverables the project group may from time to time publish material that does not count as a document and may even be open for public viewing therefore certain guidelines have been created to keep any of our publications at a professional level with the same high quality standard one would find in our deliverables Any official project group document shall follow the given document guidelines to the extent necessary and work closely with the quality assurance manager in order to create new guidelines and templates if needed If any member of the project group is planning to create a publication or release information under the project group s name they must notify the other members of the group regarding the planned publication and provide the opportunity to both perform a quality check of a reasonable length and gain the approval of the project member s that the publication is bearing the project group s name Page 120 of 194 Project FHaSS Quality Assurance Document 2011 17 1 Notice of meeting Agenda and Minutes guidelines 17 1 1 Notice of meeting amp Agendas Any notice of meeting and agenda should be created using the Notice of Meeting Template found in Meetings Notices and Agendas 17 1 2 Minutes Any minute from a me
58. need for retrieving universes in the game filter user story so most of the logic is already in place Some methods will be created or modified in the ReadDataService to reach the desired information with new queries 34 CPU Speed grouping This user story entails showing CPU speeds by brackets starting at 1 GHz and incrementing with 400Mhz per bracket We want to display both the number of entries within a speed group and the percentage that speed group is of the total Currently the CPU information is stored in the CPU column of the hw info table The information stored is the full line copied from a dxdiag file This means that there is a lot of extra information stored so we need to parse the CPU speed To do this we will create a few new methods in the Page 137 of 194 Project FHaSS Design Document 2012 ReadDataService figure 5 This includes a method to get a list with all the CPU information from the database a method to filter the strings and a method to count for the different speed groups Figure 5 CPU grouping methods for the ReadDataService A new model controller and view figure 6 also needs to be created so that we have a specific page that displays the information in this user story Figure 6 MVC Classes After the second iteration it became clear to us that using the original hw info database table by parsing the columns for the information we want for example CPU speed is very inefficient The p
59. of Funcom After discussion with the Funcom employees we decided to again build more of the work which at that point was done per request into its own part of the system We called the new part the Grouper and its job was to work on the data that was the product of the Parser and make them even more optimized for counting By using a set of supporting database tables giving the restrictions and limits for the groups in each hardware category the Grouper could group every hardware statistics Page 187 of 194 Project FHaSS Reflection Document 2012 entry in to their respective group for all the hardware categories For instance the grouper would group a hardware statistics entry with a CPU speed of 2300MHz into the CPU speed group that has a range from 2000 to 2400MHz and in a highly optimized database table only store the CPU speed group id After implementing the Parser and Grouper the work left to be done per request was reduced a lot 61 1 3 Further loading time improvements The load times decreased even more after the implementation of the new Grouper but still we were experiencing way too long loading times The project team was out of ideas on how to improve the loading times and we had extensive discussions with the Funcom employees on possible solutions to this The Funcom employees came up with several new ideas two of which has been implemented in the current version of FHaSS a single SQL query per hardware category and mem
60. of universes countries and time period should be displayed in the table Site navigation buttons also fit very nicely centered at the top of the page After getting good feedback about the initial design by our client we were confident that our vision of the system matched up with our clients A few more mockups were created to illustrate different solutions we could have chosen for our page which in turn gave us some ideas for where to place the future functionality but the layout itself stayed true to the original designs Page 144 of 194 Project FHaSS Design Document 2012 Funcom Hardware Statistics System Displaying data from All games a All locations From fromtim Paa CPU Cores Graphics Card Vendor Graphs will be implemented later cal Le Hardware Information Sort by Descending Nr of entries 3000 Period testlive 300k tsw 10k server1 5k server2 1K Age of Conan aocliveeu 0 Anarchy Online i B ies LI Most Popular CPU Speeds 2 25GhZ 50 16B 10 2GB 20 HDD Space ipsum dolor sit amet consectetur adipisicing elit sed do eiusmod tempor incididunt ut labore et dolore magna aliqua Ut enim ad minim veniam quis Figure 14 Final iteration The final iteration design became a bit more complex as we factored in functionality that have not yet been planned let alone implemented The Ul s behavior became more defined in this it
61. ona budget The fact that it is open source doesn t concern us as we will not spend time modifying the functionality of the IDE Benefits Drawback e Fast and responsive e Shortcut keys aren t customizable e Builds projects faster than other e No edit and Continue functionality IDE s 9 3 MonoDevelop MonoDevelop is another free alternative to VS and it is cross platform meaning we can use the same environment to develop on Windows Mac OS X and Linux In this project we will exclusively develop on Windows OS for Windows OS so advantages and drawbacks will not be listed as this IDE is severely lacking in the feature department and is not considered a viable option for us 10 Conclusion For this project we will use Visual Studio 2010 Ultimate because of the best integration with the intended language It has most features including debugger extensive auto complete and features we might need in later phases of the implementation The sheer size of the IDE is a minor problem and should not cause any problem for the team As of the reported issues these are considered to unimportant for us to discard VS as the IDE choice 11 References e http www icsharpcode net opensource sd Last visited 13 12 2011 e http community sharpdevelop net blogs mattward pages VisualStudioExpressComparison aspx Last visited 13 12 2011 e http blog codebeside org post 2011 04 19 replacing visual studio 2010 with sharpdevelop
62. project so that the team works with a standardized tool 7 2 Definitions and Acronyms Integrated Developer Environment Visual Studio Operating System 7 3 Document History Version Description Date 1 First version created 15 12 2011 2 Document template added 10 01 2012 Information reviewed and updated Page 108 of 194 Project FHaSS IDE document 2011 8 Table of contents 7 General Document Information eege ad rod er rene ene vanes eoe aeo ineo Fred ento n Tena Roe ERR ERR en 108 7 1 Declaration of intentiOD mereseire uaea ie E Ee eege ee bl dee venue 108 7 2 Definitions and Acronyms sica nennen nnns nnns en nns enses 108 7 3 Document TE ee 108 8 T bleof contents etre ere Mesh tire irent eee ee intr kaken eee eee ERE xe ar MEE He EN 109 ZEND 110 9 1 Visual Studio 2010 Ultimate A 110 9 2 tipp 111 9 3 leie RTE EE 111 10 Conclusioni rerit aet rase tee ola etti ces deer ete tenuis koken 111 MEE ccce 111 Page 109 of 194 Project FHaSS IDE document 2011 9 Introduction C is an official Microsoft computer language and its native IDE is Visual Studio also developed by Microsoft There are also a few other good IDE s for C development but very few presents all the features and content that we require for our project In a sense th
63. report objects of them A bug report contains a lot of information parsed from the XML file see attachment 4 which includes the DxDiag file the file we are initially interested in fcBugDXDiag py This file handles the parsing and storing of hardware information Some properties are taken from the bugReport e Username player login name e Universe game server e Bits System architecture 32bit 64bit Then it collects text lines by parsing the DxDiag document for retrieving the following information e Operating system Processor Computer memory RAM Card name Actual model name of graphic hardware Video memory Dedicated if exists else display memory Screen resolution Driver version for the video card After the information is successfully retrieved from the document it is inserted into the hw info table in the database e insert into hw info username universe os cpu ram vcard vram driver resolution bits count Where count count 1 The insert sentence over displays the type of information that is currently being stored in the hw info table of the database See attachment 5 for an example of the text lines extracted from the DxDiag file Because of the large file size we are not able to open the hw info file in either notepad or PhpMyAdmin for a complete analysis of the table 24 3 2 Miscellaneous Python files There are several other important files in the system that perform different
64. table 39 Redesigned responsibilities of WCF and Parser Because of the bad performance that our system encountered when querying the database for large sets of data we had to look into enhancing the data processing The result we came to was to move the grouping functionality that was currently implemented in the WCF web service layer to the new Parser layer This will make it possible for the system to do almost all the pre display work even before the SQL query has been sent from the WCF Since the filtering and grouping operations of the current hw info database are two separate operations that most often will be run independently and with different intervals in time our solution was to create another hardware table called groupedhardware see database illustration in previous section With this separation of storage the grouping operations will be working on already parsed data stored in the hardware table while the parsing functionality will be working on the raw data from the hw_info table The Parser will be running kind of more in the background than the WCF will do The WCF will be doing its work on demand from the MVC while the Parser will more or less almost always be running Because of these different modes of runtime it is more appropriate that the Parser will be the process running the UpdateDistinctUniverses function that was previously implemented in the WCF and also the new UpdateDistinctGpuModels function Because of the
65. the client resulting in a sub optimal system prevention Keeping to the schedule and deal with abnormalities early will give us the needed time to develop and finish this feature R1 3 Any C Requirement not working Page 92 of 194 Project FHaSS Risk Analysis Document 2012 Probability Crequirements Future hardware evolution prediction By being unable to meet any C requirement in time we are failing to meet a Consequence tertiary requirement set by the client resulting in a still functional system but it will be lacking a highly desired functionality SEE Keeping to the schedule and deal with abnormalities early will give us the d needed time to develop and finish this feature 61 2 Technical risks R2 1 ASP Net nHibernate or C difficulties Probability Since we are using a new programming language as well as a new programming Description standard some trouble is to be expected Can be caused by lacking understanding API or documentation By using too much time learning the language our development timetable may Consequence be affected SE A proper study done in the Programming Language document Spending personal time getting used to the language if personal level is below your peers R2 2 Data loss Probability Impact May cost us varying amounts of time loss to restore the system ETEVERTIORE Keeping backup versions of documents manually using version verifying i software securing persona
66. the statistics defined in requirement A F 1 must load within 10 seconds when the time filter defined in requirement A F 3 is set to span over the last month A NF 7 Wiki pages documentation A A NF 7 11 12 11 Product Backlog The system must be documented using Wiki pages Documentation shall consist of e A user manual e Asystem overview specification description A NF 8 BugZilla interfacing A A NF 8 27 10 11 Meetings The new system has to interface with Bugzilla in the same way the current system does B F 1 Display hardware network statistics B B F 1 11 12 11 Product Backlog This is an extension of requirement A F 1 Two new types of information to be displayed are added to the list The two types of information to be displayed statistics for and their specific grouping structure are as follows a HDD Hard Disk Drive a Capacity In the following groups 10 99GB 100 249GB 250 499GB 500 749GB 750 999GB and above 1TB b Type In groups of SSD or Normal Magnetic Disk b Network download speed In groups of all the different speeds registered in the database All the filter types of requirement A F 3 shall be compatible with the two new types of information as well B F 2 Display software statistics B B F 2 11 12 11 Product Backlog Statistical information about specific software inst
67. the system that may not be related to a specific functionality of the system This can for example be performance or security 32 3 Static and dynamic testing Reviewing reading the code is considered static testing while debugging or executing the code is considered dynamic testing 33 Testing methods In the following are short descriptions of testing strategy and methods we will be using for testing throughout this project 33 1 Strategy 33 1 1 Black box testing The tests performed under this strategy are made directly from the requirements specification Black box testing is a method of software testing for testing the functionality of the system without knowing the structure of the system Expected data on the output is a result of the known input data tests under this strategy are also called functional testing 33 1 2 White box testing This strategy deals with the internal logic and structure of the code It is possible to analyze the test element and decide exactly how the test is built up and what to test When making a test based on white box testing knowledge about the structure for the element to be tested is needed Page 39 of 194 Project FHaSS Test Strategy 2012 33 2 Methods 33 2 1 Function test A function test is written in order to check if a functional completed requirement behaves and works as planned 33 2 2 Code Compilation A successful compile of the code is a basic form
68. the time period for which the statistical data shall be produced with basis upon 2 Choose to get to see statistical information for one of the hardware software categories CPU Graphics Card RAM Windows OS HDD or Network download speed 3 Wait for the web page containing the information to load 4 Compare the resulting percentages to the percentages that has been manually calculated from the test data set Test criteria e Database and website is up and running e Test dataset has been made and inserted into database Expected e The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors Outcome 56 3 T R 3 2 Filter statistics user interface This test ended up being unnecessary due to change in the user interface originally planned Now all the games and underling servers are always shown on the left side and it is now possible to filter with more than one game and not only one The user interface has also been thoroughly tested in test T R 3 1 Filter statistics user interface test est ID T R 3 2 creases DHE A F 3 UsSeelSegieaean Test is run after T R 3 1 Test to see if the user interface for the filter statistics option is working correctly e Check if game server
69. the time to perform a thorough review Completed peer review After the peer has reviewed the document according to the QA specifications See section 16 2 4Reviewing a deliverable he should update the deliverables versionand notify the original author After the original author reviews and accept any changes made by his peer he should place a backup version of the document under the Deliverables Correct WP Old Versions folder before sending it for a final review to our internal supervisor Completed supervisor review When the internal supervisor has reviewed the document and either applied changes or added comments the original author should go through these and make the appropriate adjustments to the deliverable and updating the document version After the corrections have been made the author should update the documents version 16 2 4 Reviewing a deliverable To make sure the reviewing process is performed optimally one should always follow these guidelines e Fix faulty typography spell check punctuation etc and grammar e Be critic about what you read mark statements that lack proper citation e Optimize the document structure layout if possible e Make updates to the document if the document template has been changed e Add comments and notes where needed e Use the Track Changes function of Microsoft Office Word e Update all document references To files headings or web pages The reviewer should fix any
70. to implement our own 62 2 Technical R2 4 Bad internal server performance Sprint recognized Sprintnr3 Severity mer As we fetched more and more data from our database we recognized a steadily Description NE increased load time for the site as data grew due to un optimized code Consequence We have done several updates both to the client side fetching and to database Solution optimization through parsing amp grouping Page 98 of 194 Project FHaSS Risk Analysis Document 2012 R2 4 Bad internal server performance Sprint recognized Sprint nr 9 Severity NA Our system required more and more information from the database as the Description system evolved thus slowing down loading times further We came up with a caching solution together with our external supervisors in Consequence order to greatly reduce loading times for a limited time span The consequence Solution for this was a large rewrite of our fetching mechanisms basically some time loss R2 2 Data loss Setback Sprint recognized Sprintnr7 Severity Due to inexperience with our version verifying software Perforce one of the Description project members inadvertently set an earlier version of the project as the current one overwriting several updated files Consequence After a heroic effort from several team members we were able to restore the Solution system Moderate amount of time loss R2 5 Pe
71. want to be able to see memory information by groups of memory amount The OS user story includes windows version bit version and DirectX version For now we do not have any information stored about DirectX so we will skip it Page 148 of 194 Project FHaSS Design Document 2012 Figure 19 Memory OS and bit methods New filter methods needs to be created to get the data from hw info and into the Hardware table Also the InsertintoHardwareDbTable method needs to be updated Below we can see the new sequence of the methods in the HWParser Page 149 of 194 Project FHaSS Design Document 2012 Main 1 RunParser 1 1 GetHardwarelnfoList 12 GetColumnEntries For each Hardware info entry 12 1 FilterC puManufadurer 12 2 FilterC puSpeed 1 2 3 FilterC puCores 12 4 FilterGpuManufacturer 12 5 FilterVram S 12 6 FillerRam PE 12 7 FilterOs P 1 28 FilterBits 1 2 InsertintoHardwareDbTable D 3 SetParsed 4 Group Figure 20 New HWParser sequence Then we have the updates to the database Page 150 of 194 Project FHaSS Design Document 2012 EL IR Jer EL is B B B B El B Figure 21 Database updates The OS table will contain the different existing OS the hardware table have new columns for ram os and bits Then we have the ramgroup table In the end new MVCs must be created to administrate the OS table and ramgroup
72. 0hgz Inc at 400mhz e Show split between AMD and Intel and other e Show number of cores Test date 11 03 2012 Tester Kim Richard Johansen 27 04 2012 WSS ial galjuereles Function test black box Test approach 1 Putinthe time period for which the statistical data shall be produced with basis upon 2 Choose to get to see statistical information for CPU 3 Wait for the web page containing the information to load 4 Compare the resulting percentages to the percentages that have been manually calculated from the test data set Page 61 of 194 Project FHaSS Test Specification 2012 Test criteria e Database and website is up and running e Test data set has been made and inserted into database Expected e The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Kim Richard Johansen 23 12 2011 Errors e Encountered errors Outcome What happened Display hardware software statistics GPU information est ID T R 1 2 ATENDA A F 1 and B F 1 Ene Test to see if tables are created and contains accurate information about the different GPUs stored in the database It should be possible to toggle between number of entries and COS Show split between nVidia Intel AMD and others Data to be shown e Show video card model like Steam e Show VRAM e Show Primary
73. 1 64 3 X axis zoom for graphs eiie teet eee andere LIH ERE Ee Ru EOS 191 64 4 Exporting the graphs oo rtr tete eate eet ce Rena ER keen ERR baa NER genene 191 64 5 Future prediction of hardware in graphs ssssssesssssrsesrnssrnssrnssrnssrnssrnssrnssrnssrnssrnssressrnssrnne 191 65 Administration side full control to the user 191 S Sii deemed 192 66 1 Requirements capturing nennen nnne ennt en entre nete sten eterne ns sister nna 192 66 2 Differences in development and production environments 192 66 3 Very large amounts of data c ihe d essi the cde qr Eee Pe aaia 192 67 FHaSS extension suggestions 00 0 2 eeeeeeeseeeceeeceeeeeeeaeeeceeeeeeeesaaaecececeseeeeaaaaeeeeeeeeeeeesaaaeceeeeeeeeteaaaeess 192 67 1 CPU cores groups with range cccccccccesseceeseneececeseeececseaaeceeeeaaeeeseeaaeeeeeeaueeesseaaeeeseeaeesenenaes 193 67 2 Physical and virtual CPU cores statistics rrnnnannrnvnnannrnenrennrnenrsnsnnnnrsrsnnvnrsrennvnssnennrnsssennneee 193 67 3 Saved filter administration with AJAX sssssesseesseeseeenee nee nnne entren enne 193 67 4 Graph resolUtion uie deeg seess eessen Ee geed eaae n deo 193 SEBSEEIPISEMEIIMAUILUEMOR Tnt 194 Page 185 of 194 Project FHaSS Reflection Document 2012 67 6 Toggle displaying individual charts sess ennt 194 67 7 Compare newest event log entry from database to local event log 194 67 8 Recommended hardware software to look into these need attention
74. 4 1 aspx Last visited 15 12 2011 e http desipenguin com techblog 2009 07 10 why sharpdevelop is better ide Last visited 11 15 2011 Page 111 of 194 N H GSKOLEN i Buskerud Funcom Hardware Statistics System Quality Assurance Document Funcom Hardware Statistics System Client Funcom N V Acronym Sensors amp Supervisors Internal sensor Olaf Hallan Graven Internal supervisor Aurilla Aurelie Arntzen 29 05 2012 Author External supervisor Rui M MonteiroCasais Sondre Bjerkerud Sverre Christoffer Thune Kent Brian Dreyer Kim Richard Johansen Project FHaSS Quality Assurance Document 2011 Quality Assurance Document 12 General Document Information Deliverable nr D2 5 3 PENE FE BAST Guidelines Release Public Workpackage WP2 Responsible Dag Hem 12 1 Declaration of intention This quality assurance document should teach its readers the quality and standard that any of the project group s deliverables should uphold to reach the high quality standard needed to satisfy our client This document will review the following e Group communication and information exchange e Quality assurance regarding deliverables and documents e Quality assurance regarding other publications such as web blog or posters e Code development standard 12 2 Definitions and Acronyms Quality Assurance Non disclosure agreement Micr
75. 4 5 51 21 42 5 2 7 1 5 24 5 th 10 25 1 2 1 Total 74 25 47 58 48 5 59 5 51 5 1 7 May Activity Person Sondre Sverre Kent Kim Dag 8 25 6 5 3 3 3 26 75 30 39 31 5 32 4 16 5 8 5 40 5 54 72 5 52 61 5 7 4 6 1 8 6 4 25 18 1 11 5 5 Total 96 75 130 115 5 106 116 51 5 1 8 June Due to the fact that the final documentation delivery date is set to the 29 of May 2012 there has of course not been logged any work hours in June at this point However the project team has estimated that each team member will work approximately 40 hours on the Presentation activity in this month 51 5 2 For the whole project period The following table is a summary of the monthly tables given in the section Hour list per month and is thereby an hour list for each team member and the team as a whole for the whole project period Page 81 of 194 Project FHaSS Project Plan 2012 Sondre Sverre Kent Kim Dag 72 46 75 30 5 28 30 93 25 97 92 106 98 93 5 59 5 61 33 5 78 238 25 191 5 229 5 126 5 210 21 75 43 2 1 5 28 0 1 0 94 5 0 87 5 87 86 80 86 27 20 5 40 18 5 14 633 25 546 25 541 490 5 543 52 Responsibility 52 1 Fields of responsibility SondreBjerkerud e Project Management until 7 05 2012 e ScrumMaster Dag Hem e Quality Assurance e Risk Kent B Dreyer e Implementation e Web page Sverre C Thune e Project Management from 7 05 2012 e Design Kim R Johansen e Test 52 2
76. 47 3 3 Debugging Debugging will be used in two ways 1 To identify the problem area s of the code when compilation of the code fails This will help the developer to fix the error warning to make the code compile without faults 2 Inthe case that any of the verification tests fails there will be one or more logical errors in the code that the corresponding test s are testing Debugging can be used to identify these errors by going through the code step by step and examining the values of the variables in the code 47 3 4 User Interface Testing When the user interface for our system has been designed user interface tests can be created to ensure that the user interface behaves the way it is supposed to These tests will be made and carried through when the user interface corresponding to the specific test has been designed and implemented 47 3 5 Ad Hoc test The Ad Hoc test is a test we will be done late in our project when most of our grade A requirements are completed and tested This test is done without any formal test plan and the purpose of this test is to try and break the system by trying different system functionality Since the purpose is to try and break the system negative testing is allowed Negative testing means that the tester can try to find not intended functionality that may break the system Page 72 of 194 Project FHaSS Test Specification 2012 47 3 6 Strain test The Strain test is a
77. 9 05 2012 Page 4 of 194 Project FHaSS User Requirement Specification 2011 2 Table of contents 1 General Document Information a eiue teinte dete teet at Teen Foe ERRARE HRS Rene AER Sage 4 1 1 Declaration of intention oerte retener eal tees ele esee et envie eege 4 1 2 Definitions and Acronvms sess enne nennen entren a enses nnn rentis nns n sens 4 1 3 DOCUMENT HISTON os IEEE 4 2 T bleof contents dicere eter reka eie 5 3 Functional requirements isccs cesccccssesccccvesccedesaaccecteeaceceevencceussadccdstanccedssaacesdevaaccedersaceedeanccedenaacceseeas 6 4 Non functional reguirements 6 Page 5 of 194 Project FHaSS User Requirement Specification 2011 3 Functional requirements 1 Asa user I want to be able to see information about the hardware of the client computers that crashed because I can then see which type of hardware there are most problems with 2 Asa user want to be able to filter crash information based on time period region geographical location continent countries etc and hardware type and all combinations of these 3 Asa user want to get more accurate information from the DxDiag file than the current system because there is currently a lot of unknown hardware 4 Asa user I want to add and remove hardware types to the database because new types of hardware is entering the market as technology evolves 5 Asa user I want the system to group almost identical hardwar
78. Arntzen 28 05 2012 Author Sverre C Thune External supervisor Rui M MonteiroCasais Sondre Bjerkerud Sverre Christoffer Thune Kent Brian Dreyer Kim Richard Johansen Project FHaSS Design Document 2012 Design Document 30 General Document Information Deliverable nr D4 1 16 DAMIER GHE Report Release Public Workpackage Independent Responsible Sverre C Thune 30 1 Declaration of intention The purpose of this document is to present our design for user stories in the sprints Each section of the document contains the design for a specific user story or larger overall updates to the design 30 2 Definitions and Acronyms FHaSS Funcom Hardware Statistics ASP NET ASP NET MVC3 is a framework for how code should be MVC3 structured using the Model View Controller pattern Ne Eee Windows Communication Foundation Service ill lg A library for accessing a database UML Unified Modeling Language OS Operating System 30 3 Document History Version Description Date 1 First version created 26 01 2012 1 5 Updated class diagram after review and feedback of the document 26 01 2012 from Funcom employees Lorenzo Aurea and Hans Birkeland 2 Game filter design added database diagram and server filter design 07 02 2012 3 Added design for CPU speed grouping 12 02 2012 4 Added redesign of WCF web service Sondre 15 02 2012
79. Because the home page uses the last month by default the users will almost always be using the cached information This made our home index pop up almost instantly given the fact that this is also the most relevant information to analyze this was a great solution to our load time issues The drawback is that the cache needs to be built the first time the system starts up and the first time a user uses the page for the first time in a day to let the cache refresh itself this should take as much time as a normal database fetch as if a user wanted information about a time period that goes back more than one month We also considered using a thread that would do this at the turn of a day but that could cause issues with the web client which is a singleton meaning there s only one instance of the client at any given time If several processes would try to access this simultaneously there could be problems Workarounds for this problem could obviously been developed but this was implemented in one of Page 188 of 194 Project FHaSS Reflection Document 2012 the later periods of the project so we didn t have the resources or time to implement this potential improvement 61 2 NHibernate NHibernate is an object relational mapper for the NET framework and it was a requirement from Funcom to use this software for the systems database interaction In essence NHibernate has been a big blunder to the project team NHibernate was very stubbor
80. C s ability to do threading something PHP cannot C with ASP NET is also a compiled language which makes it easier to debug and gives it a big advantage over working with Python and PHP One of Python s main benefits is the fact that it could be used as both a programming language for our parser and as a web development tool this would eliminate the need to learn two languages The system our client is currently using was also created with Python So we could have used all the languages for web development but since our client would prefer the use of CH in combination with ASP NET for further possible expansion ASP NET and CH is the logical choice for us to use as our web development tool 12 2 Languages used for database parsing We have the choice between C and Python as they can both perform the job for parsing our received data into the databases and thread creation which PHP cannot Python s main benefit is the fact that we could use it for both database parsing and web development something our clients have done for their current system Since we have chosen to use ASP NET as our web development tool which is dependent on the C language and our clients have already suggested that we use C with ASP NET it is clear that C would be the best choice for us to write our database parser in 13 References 13 1 PHP e http www pretechno com i oracle comparison of php asp net html Last visited 15 12 2011 e htt
81. C framework communicates in the normal way Figure 10 Class diagram of HWParsers interface system The solution to graceful threading is to use a volatile Boolean to exit the while loop sustaining the thread alive Page 141 of 194 Project FHaSS Design Document 2012 System User 2 index 1 Refresh i 1 1 1 1 GetParseData 1 1 GetParseData 3 StartParsing 3 1 HttpPost 4 StartParsing 4 1 StartParser 4 1 1 ParseData parsing 4 111 1 ParseData 5 1 HttpPost 5 1 1 StopParsing 5 1 1 1 StopParser 5 1 1 1 1 Destroy Figure 11 Sequence diagram of parser interface As we can see from the diagram above the MVC framework communicates through normal means while the parserModel calls method in the parser 35 Redesign of the WCF web service After discussion both internally and externally with the external supervisor and employees at Funcom we came to the solution to add a parsing application layer between the current hw info table and a new hw info table The addition of this parser made it clear that some of the functionality that the WCF service currently possess would be transfered to this new parser layer After our discussion about the parser we also understood that the way the WCF service is currently handling queries to the database could be improved We decided therefore to implement a more gener
82. Each of the filter methods in the HWParser will have specific keywords and this way find the desired result Depending on if the result returned is text or a number we want to return a string or an integer therefore we will need different methods in the stringParser that can return different types Page 139 of 194 Project FHaSS Design Document 2012 Main 1 RunParser Fil 1 1 GetCPUList 1 2 filterCPUList Until CPUList filtered 1 2 1 FilterCPUManufacturer 1 2 2 ParseString ra 1 2 3 FilterCPUSpeed 1 2 4 ParseString m 1 2 5 FilterCPUCores 1 2 6 ParseString pr 2 InsertintoDB Figure 9 HWParser sequence diagram The figure above figure 9 shows how the parser will work The RunParser method starts the sequence First the GetCPUList method is called this is going to return all the rows contained in the CPU column in the original database table hw info This method should use nHibernate to communicate with the database Then the FilterCPUList receives the list of strings containing the CPU information and runs a loop the calls the different filter methods that filters each string in the string list for the desired information After a string is filtered the returned values will be inserted into the new database with the InsertIntoDB method which will also use nHibernate The loop continues filtering each string and inserting the information into the new database u
83. Functionality to automatically restart is included in the windows service Page 160 of 194 Project FHaSS Design Document 2012 P 4 Write log to file and exit Figure 33 Sequence for exception handling The last few entries in the log should be viewable on the administration page under the interface section 50 Caching As our queries to the database took longer and longer as the database grew we had no chance of meeting the loading time requirements with the current system so after a discussion with the product owner and other Funcom employees this requirement has been dropped Good performance is of course one of our top priorities so we with the help of Funcom employees developed a solution which could work as a compromise between extremely large amounts of data as well as fast loading times The compromise would be that this solution only works on a preset time span of data in our case this became one month of information By caching every entry for the last month into the service memory any request regarding information within that time span would be processed without needed to query the database This way we avoid the heavy loading time caused by accessing the database with several million rows about four and a half at the moment The caching is implemented by having the WCF service retrieve the database rows for the previous month and store this in memory Every day it should update the memory so that this stays tr
84. Manufacturer and then show it in tables and charts So to test the grouping functionality we need to put in some smart values that are close to or exactly in the boundary range of the different grouping ranges Page 58 of 194 Project FHaSS Test Specification 2012 46 2 Test types Tests types listed are explained in more detail in the Test strategy document and the procedure for the tests is listed under Tests e Function test e Code Compilation e Debugging e Code Review e User Interface e Ad Hoc test e Strain test e Regression test e Performance test 46 3 Identifying different tests and requirements Short explanation of the ID used for requirements and tests Requirement ID A F 1 Letter explaining the importance of the requirement where A is highest and C lowest Letter explaining the requirement as Functional F or non functional FN Requirement number Test ID T R 1 2 Letter T means test Letter R means requirement The first number equals test number If a second number is present the test is divided into more than one test and the number represents the test number of that test Page 59 of 194 Project FHaSS Test Specification 2012 47 Tests 47 1 Test template Test Name est ID Test ID Requirement ID ID EEE Test description Test date dd mm yyyy Name of tester Sn ae Test type black box white box Test approach 1 approach
85. P address of the server You also have to create an account to connect but it is really fast to do it Just input a username if you want a password and an email address and you are ready to connect You also need to install and set Visual Studio to use the Perforce SCC plug in Setting up a project to use P4 is straightforward First you create a project and then add the solution to source control in Visual Studio After that adding a file to the version control is easy Just create Page 127 of 194 Project FHaSS Version Control Document 2012 the file in Visual Studio and it is automatically added to the pending change list Check in and the new files are added To submit changes you are required to add a comment describing what the change is This way it is easy to track the history since you can view the comments of every change submitted You can also use the P4 GUI client to add files which can be useful for adding other types of files than code to the version control If files have been changed by other users after you refresh your P4 GUI client by pressing F5 you will get a notice that there is a new revision By choosing get latest revision you download the newest changes from the server Checking out files can be done either manually through the P4 GUI client or automatically in Visual Studio just by starting to write in the code file You can even see if files are checked out by other users and by whom After you are
86. R 3 2 Requirement ID Test description Test date Testing method Test approach Test criteria Expected results A F 3 Test is run after T R 3 1 Test to see if the user interface for the filter statistics option is working correctly e Check if game server filter is available when a specific Funcom game is selected and unavailable when a game isn t selected 24 05 2012 Kent Brian Dreyer User interface test Black box 1 Set all filter options blank null 2 Check if game server filter is unavailable 3 Choose one of the available Funcom games as a filter 4 Check if game server filter is available e Database and website is up and running e Test dataset has been made and inserted into database e Game server filter works as explained in the test description Page 66 of 194 Project FHaSS 2012 Test Specification Test creator Errors Outcome Kim Richard Johansen 09 01 2012 e Encountered errors What happened Requirement ID Test description Test date Testing method Test approach Test criteria Expected results Filter statistics game filter est ID T R 3 3 A F 3 Test to see if the game filter combinations are working correctly e Test if results after filtering show Funcom games and entries for each of them 02 02 2012 Tester Kim Richard Johansen Function test
87. T R 4 1 22 05 2012 management and grouping and grouping 6 7 Hardware type Hardware type management T R 4 2 22 05 2012 management and grouping and grouping new hw type 7 OS information Display hardware software T R 1 4 14 03 2012 statistics OS information 7 OS information Display hardware software T R 1 5 27 04 2012 statistics Direct X information 8 Memory information Display hardware software T R 1 3 13 03 2012 statistics RAM information 9 Geographical location filter Filter statistics geographical T R 3 6 27 04 2012 location filter 11 Time period filter Filter statistics time period T R 3 5 27 04 2012 filter 20 Trend charts Display hardware software T R 2 02 05 2012 evolution charts Other Filter statistics T R 3 1 23 05 2012 55 Logs for validation tests performed This is meant as a short overview for each of the validation tests performed Explanation Tests with outcome Test Failure means that the test failed and the system worked as planned Tests with outcome Test Successful means that the test successfully pointed out that something in the system did not work as planned Modifications where needed on the system and a new test was scheduled Page 166 of 194 Project FHaSS Test log 2012 55 1 Story 3 Game filter Filter statistics game filter est ID T R 3 3 REMEG DHE A F 3 EEE Test to see if the game filter combinations are working correctly
88. Test Specification 2012 Test date Testing method Test approach Test criteria Expected results standard value e Test with all filters set The filter values that will be beneficial to put in will depend on the data set For instance should both values that are represented in the data set and values that are not represented in it say for instance a specific Funcom game be filtered upon For the time period filter both the whole period which the test data set is describing and only parts of it and also with time periods where the to and from limits are stretched further into time or farther back in time than the test data set describes 23 05 2012 Tester Kim Richard Johansen Function test black box 1 Putinthe filter combination to be tested 2 Wait for the web page containing the filtered statistics to load 3 Compare the produced filtered percentage statistics to the manually calculated percentages e Database and website is up and running e Test dataset has been made and inserted into database e That the produced filtered percentages based on the test data set are similar to the manually calculated percentages for each of the tested filter combinations Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors e Encountered errors Outcome What happened Filter statistics user interface test est ID T
89. Test with all filters set The filter values that will be beneficial to put in will depend on the data set For instance should both values that are represented in the data set and values that are not represented in it say for instance a specific Funcom game be filtered upon For the time period filter both the whole period which the test data set is describing and only parts of it and also with time periods where the to and from limits are stretched further into time or farther back in time than the test data set describes Test date 23 05 2012 Kim Richard Johansen Page 177 of 194 Project FHaSS Test log 2012 WSS ial plereles Function test black box Test approach 7 Put in the filter combination to be tested 8 Wait for the web page containing the filtered statistics to load 9 Compare the produced filtered percentage statistics to the manually calculated percentages Test criteria e Database and website is up and running e Test data set has been made and inserted into database Expected e That the produced filtered percentages based on the test data set are results similar to the manually calculated percentages for each of the tested filter combinations Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors e None Outcome Mapped all countries to their respective region and servers to their game Tried to use just the time fil
90. The following is only sort of a list of the titles of the requirements organized in their respective priority category Supplementary details for each requirement are given in the section Requirements Details 41 1 A Requirements 41 1 1 Functional A F 1 Display hardware software statistics table A F 2 Display hardware software evolution charts A F 3 Filter statistics A F 4 Hardware type management A F 5 Hardware grouping 41 1 2 Non Functional A NF 1 Programming language A NF 2 Model View Controller A NF 3 Database type A NF 4 Database connection A NF 5 Database access A NF 6 Web page loading time A NF 7 Wiki pages documentation A NF 8 BugZilla interfacing 41 2 B Requirements 41 2 1 Functional B F 1 Display hardware network statistics B F 2 Display software statistics 41 2 2 Non Functional B NF 1 Windows service Page 48 of 194 Project FHaSS Requirements Specification 2012 41 3 C Requirements 41 3 1 Functional C F 1 Future hardware evolution prediction C F 2 Chart raw data C F 3 Chart zoom C F 4 Textual help 41 3 2 Non Functional C NF 1 Filter URLs 42 Requirements details The requirements from the requirement list will in this section be explained in detail Each requirement will be described using the following table An explanation of the fields in the table e Category A Bor C requirement e ID The requi
91. WCF Funcom Hardware Statistics System Standardized image file formats A library for accessing a database Framework library we use for the system Windows Communication Foundation 58 3 Document History Version Description Date 1 First version created 22 05 2012 2 Updated content finalized document 29 05 2012 Page 184 of 194 Project FHaSS Reflection Document 2012 59 Table of contents 58 General Document Information 184 58 1 Declaration of intention orte tr tenete revente iv bn eee ve eR e tee vae avus 184 58 2 Definitions and Acronyms ssseseseseeeeeeeen nennen enne ee nennen nennt neni atas asses esseri iiaiai 184 583 D c ment e Tee 184 59 Table of contents teer eee tire irent ees EET Ce Re RV eue Gear frie 185 60 Introduction mp 187 EH WE en TE 187 611 L ading Lu deus 187 L11 PSE EE 187 61 1 2 Groupe RI 187 61 1 3 Further loading time improvements cccssccccccesessessseeececeseeseseaeeeeeesseeseaeaeeeeeens 188 61 1 3 1 Sigue L 188 61 1 3 2 Metromara s etersged siie eiaeiiio et dee Sede cub ire ade Eder eebe 188 612 NHibernate Huse lades 189 62 E E EE EE 189 ERE CEDERE EDU 189 64 Graphs and visual presentation of data 190 64 1 HighChiarts 1 teo detta eye eat eunti EE coeant esto ated 190 64 2 View raw data view full screen graphs eene 19
92. a released title with lots of bugs and or hardware incompatibilities would not result in a lower release sales number than without except if the game is so full of faults that the customers demand their money returned but it will worsen the game title s and the game company s reputation 6 This in turn could lead to less sold games for the next releases This game fault ignorance is not possible for game development companies that rely heavily on periodic subscription fees like Funcom does Gameplay ruining bugs or large hardware incompatibilities with a game title in this category will result in the players abandoning the game rather quickly potentially resulting in a lot less income for the company than expected Such companies therefore have to firstly release a game without major faults and secondly create frequent updates patches for the game after the release to remove averagely and minor rated faults from the game This will ensure that people don t stop to play the game that is stop paying to play the game because of insufficient testing and bug fixing So what is said above is actually that game development companies relying on periodic subscription fees have to do more complete testing than companies selling games with a one time purchase business model This is not solely true Companies that are producing online games like for instance Funcom use the fact that their games are online as an advantage By implementing a
93. al database fetching method that is responsible for building up a query dynamically The query shall be built based on filtering criteria given by the client MVC as arguments to the fetching method This new method will also have the responsibility of sending the fetched result to private grouping methods in the WCF where the data will handled and grouped according to the requirement specifications Lastly the method is responsible for returning the result back to the client Since the rules of WCF web services sets restrictions on what types of data can be returned to the client we had to think of a good way for returning the fetched and processed data The conclusion we came to was to create two new classes FetchResult and FetchResultPart which has the responsibility of storing the results produced from the grouping methods An instance of FetchResult built up by a set of FetchResultPart s is then returned to the client Page 142 of 194 Project FHaSS Design Document 2012 ReadData Service GetGamesList string GetGameForUniverse string GetUniversesForGame g ameTitle string string SetGameForUniverse universe string gameTitle string void UpdateUniverses void FetchFromDb universes string regions string dateFrom int dateTo int FetchResult GroupCpuS peeds hardwarelnfoList IListsHardwarelnfo gt FetchResultPart GroupCpuCores hardwarelnfoList IList lt Hardwarelnfo gt Fe
94. alled on the players computers shall be displayed in a table The table shall contain entries for the different types of software and the corresponding percentage of players who have that software installed on his her computer The types of software are a Java b Unity c Flash Page 52 of 194 Project FHaSS Requirements Specification 2012 d Web browsers Microsoft Internet Explorer Mozilla Firefox Google Chrome Opera B NF 1 Windows service B B NF 1 24 04 12 Meetings Eventual modules needed to be implemented to support the web interface and the windows communication foundations WCF application layers shall be implemented as a Windows Service C F 1 Future hardware prediction C C F 1 27 10 11 Meetings Calculation of a prediction of the hardware evolution of the players of Funcom games a set amount of time into the future based on the hardware evolution from a set amount of time backwards in time The length of the future time period to be predicted and the backward time period to be the basis of the prediction is to be put in by the user of the system The prediction shall be displayed as extended chart lines in the charts of requirement A F 2 It must be possible to toggle whether or not the future prediction shall be displayed or not C F 2 Chart raw data C C F 2 24 04 12 Meetings The system must be able
95. ame that is Page 21 of 194 Project FHaSS Vision Document 2012 putting too high demands on the hardware components of the players computers Both scenarios resulting in less sold games Funcom is of this reason in demand of a system that can capture the hardware evolution trend for a time period and extend it a certain amount of time into the future The future trends will of course only be predictions but they will undoubtedly limit or even remove the need for guesswork 17 5 Online The fact that Funcom is producing online games gives them a great opportunity of getting information regarding bugs hardware incompatibilities and hardware information in general directly sent to them from the players over the internet The only questions are how and when to send such information Funcom is currently using a system in their games that is identifying when a crash has occurred This system has the responsibility of informing the user that a crash has happened and to ask the user whether or not he wants to report information back to Funcom 17 6 Usage of the system There are different groups of Funcom employees that experience benefits from all of the collected crash data Quality Assurance and several managers watch the system for bugs and hardware incompatibilities and creates tasks for the programmers to fix these issues The programming team use the system to test their patches that is to find out whether the issue was act
96. arsing of strings takes a lot of time so it is impossible to fulfill the performance requirement of max two seconds load time when searching through large amounts of data However if the hardware information we want is already parsed and stored in a way so that we can directly extract it from the database it will go fast Therefore we are going to create a new database with the exact information we want and a parser that will be used to get the information from the original database to the new one This is a temporary solution in the future we will use integrate this into the current system that Funcom have replacing the parser they have now instead of having an extra Page 138 of 194 Project FHaSS Design Document 2012 34 1 The new database Figure 7 Database diagram The new database figure 7 above will contain multiple tables with the main table being the hardware table The columns will as mentioned only contain the exact information we want So the CPUCores will contain e g the integer 4 meaning four CPU cores 34 2 The parser Figure 8 HWParser class diagram The parser will contain the classes in the figure above figure 8 The stringParser will be a class with methods to parse a general string based on a start and end keywords E g as start keyword and GHz as end keyword In some cases the strings stored will need extra checks to be able to find what we want as not every string use the same format
97. ate errors to a separate list in the log 36 Sources e http en wikipedia org wiki Test management Last visited 15 12 2011 e http en wikipedia org wiki Test plan Last visited 15 12 2011 e http www aptest com glossary html Last visited 15 12 2011 e RIVET Test strategy document HIBU Kongsberg library e Dragonfly Test strategy document HIBU Kongsberg library Page 42 of 194 9 HOGSKOLEN i Buskerud Funcom Hardware Statistics System Requirements Specification Funcom Hardware Statistics System Client Funcom N V Acronym Sensors amp Supervisors Internal sensor Olaf Hallan Graven Internal supervisor Aurilla Aurelie Arntzen External supervisor Rui M Monteiro Casais 10 01 2012 Author Sondre Bjerkerud Sondre Bjerkerud Sverre Christoffer Thune Kent Brian Dreyer Kim Richard Johansen Project FHaSS Requirements Specification 2012 Requirements Specification 37 General Document Information Deliverable nr 1 6 5 Deliverable type Research Release Public Work package 1 Responsible Sondre Bjerkerud 37 1 Declaration of intention The intention of this document is to give the reader an understanding of the requirements of the system we are to create The requirements capturing methods we have been using will be explained as well as the priority levels used to organize the requirements The later sections will give both a
98. ature implemented by us if the graphs are to be used in a presentation or similar 64 5 Future prediction of hardware in graphs We wanted to give the users the possibility to apply future dates to the filter menu so instead of giving them some sort of an error or warning we calculate the predicted trend pattern of the hardware parts with the use of linear regression based on previous entries giving the developers an idea of what the coming days or months might bring in terms of hardware crashes It has to be noted that the hardware info table will not get populated with this information since this cannot be monitored in a day to day sense and we do not want the table to yield any false information regarding the crash statistics This functionality was requested by the product owner so we implemented it but it should be used with care by the users since the entries are in a way just fictive predictions 65 Administration side full control to the user In this section the users of the system can modify the structure of the database without the need to manually inject database queries as well as start and stop the grouper and parser This section was created to make the system modifiable and dynamic for future use If there are hardware changes in the market or Funcom releases new titles entries can be added deleted mapped and modified whilst making the parser and grouper aware of these changes We managed to make the interface
99. be able to display a logo for it A simple way to do this is to let the user choose a manufacturer when he creates the general model The general model can contain a database relation to the manufacturer and the manufacturer should also be added to the name of the general model Page 158 of 194 Project FHaSS Design Document 2012 So to summarize to create a new general model the user should specify a name a relation search optional and a manufacturer The relation search is optional because the user should be able to choose to set the relations manually if wanted If this is the case the relation search string should just be left empty and this needs to be handled by the system Figure 30 Sequence of automatically relating distinct to general models 47 Hard disk information This task in our project requires code to be written that will be used in the game client to extract hardware information from the player pc and add this to the bug report that is sent to Funcom The task was to make a prototype program that would find out if the hard disk is a SSD disk or not based on the model number of the disk and also the free and total space available on the disk The prototype should just get the information and then print it out So the design is quite simple All the disks on the pc should be found and then the model number and free total disk space should be extracted and printed If the model number contains the keyword SSD
100. be discovered during testing Prevention oe gt Optimized parser coding and web page towards fewer and smaller data connections Optimized parser coding and web page towards fewer and smaller data Possible connections solution Either upgrade current server or switch to a new one if the problem persists after several iterations with testing R2 5 Personal Computer problems Probability One of the projects member s laptop breaks down leaving him without a Description e workstation at the school other locations Affected person will be unable to work outside of his home for the duration of Consequence the problem Might hinder communication and workflow with other project members Backup personal data keep your system up to date and check your Prevention PP PENN H y warranty guarantee Possible i POUR We may be able to borrow workstations from both the school and our client R2 6 System not viewed correctly by different internet browsers Probability DOSEHBUOR As different internet browsers handle html code differently there may be P i discrepancies between how the page is viewed on the different browsers Minor visual glitches or lack of data on unsupported browsers Writing structured and robust code will always help program stability By testing Prevention our system up against different types of browsers and using code checking software we should be able to detect any malfunctioning browser Possible f
101. c the groups and their respective percentages 4 Compare the group percentages to percentages manually calculated from the test data set Test criteria e Database and website is up and running e There has to be one or more currently registered but not recognized hardware types e Test data set has been made and inserted into database Expected results e That the system produced percentages is the same as the manually calculated ones Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors e None Outcome The statistics in the hardware info table show the same counts and percentages as the test data set All the percentages add up to a hundred percent Test Failure Hardware type management and grouping new hw type est ID T R 4 2 RESSON nA BR A F 4 and A F 5 usse Test is run after T R 4 1 Create a new general hardware type and choose which type of hardware it should contain from the unrecognized hardware group Test date 22 05 2012 Tester Sverre C Thune WSS ie palsuereles Function test black box Test approach 7 Get the statistics page Print out write down etc the groups and their respective percentages 8 Create a new general hardware type of one of the possible categories CPU Graphics Card 9 Choose group one some all of the registered but not recognized hardware types under the newly creat
102. cified variable type but it more often than not complicates code for a reader A non specified variable type should mostly be used where the variable has a simple and specific usage e g the int i in a for loop 18 1 3 LINQ usage LINQ may sometimes be used to refactor simplify certain for foreach loops in our code but it should be noted that LINQ is somewhat hard to read and understand for people with little previous knowledge of the system Only use LINQ where it would be an improvement to the code and comment it properly Page 121 of 194 Project FHaSS Quality Assurance Document 2011 18 2 Versioning Control Perforce Perforce is a piece of version control software to help the project members share and merge their additions to the project in fast and simple manner Perforce keeps the latest deployment version of the system on its depot while each team member must check out any files they wish to edit and make additions to After changes have been made the team member uploads their addition to the project back into the depot Any member should only upload files that contain changes to the original file They should also get the latest revision if any from the depot and thoroughly test this with their own version of the project If everything works after merging with the latest deployment on the depot the user should write a description explaining the changes they have done and the necessary precautions other use
103. combination Test creator Kim Richard Johansen 24 04 2012 Errors e None Outcome Tested information that gives an error in the date fields such as non numeric characters nothing 0 out of bounds such as day nr 32 and month nr 13 It worked well Default values are used when an error occurs Worked Tested predefined dates in from field and to field and checked if the total amount of entries displayed corresponded to the manually calculated results They did Test failure 55 10 Story 20 Trend charts Display hardware software evolution charts est ID T R 2 OEG DH A F 2 EEE Test to see if evolution charts are created and to see if the contained information is accurate against the test data set Test date 02 05 2012 Kim Richard Johansen NES ie ott eteieben Function test black box Test approach 6 Put in the time period for which the evolution charts are to be produced with basis upon 7 Choose to get to see one of the possible evolution charts CPU GPU CPU cores or Windows OS version and DirectX version Page 176 of 194 Project FHaSS Test log 2012 8 Wait for the chart to load 9 Doas exact a comparison between the manually calculated data and the chart The chart shall be compared to the manually calculated data at five 5 times points with equal distance between them The first and the last times points are the same values that was put in at s
104. cted from the hardware database The information is then put through functions which results in the GUI website displaying some general statistics of the hardware information e g how many types of a certain processor In its current state this GUI is very basic and limited Our first task is to create a better and more useful GUI 25 Summary The batch server creates batches of bug files and moves them to the appropriate folder to be downloaded The internal server downloads the files and extracts them The files are parsed bug reports and hardware information is extracted The hardware information is stored in hw_info table 26 Sources amp References http www bugzilla org Last visited 14 12 2011 http www perforce com Last visited 14 12 2011 Funcom s current bug System 27 Attachments Attachment 1 batchdownloader py 1 2 3 4 An argument sets the path of the folder the instance of this process looks through An infinite loop that handles downloading files from where this folder starts Connects to the batch server with user name and password over SFTP Opens the correct folder in batches on the batchserver and creates a list of all the files in the folder Moves the files in the list to the outgoing folder on the batchserver and then downloads them into the incoming folder on the internal server Page 32 of 194 Project FHaSS Current System Description 2011 Deletes the ori
105. d game server 19 How to improve it To address the problem of a lot of unknown hardware we will be adding a functionality for adding hardware information manually by the users When unknown hardware has been recognized by the Page 23 of 194 Project FHaSS Vision Document 2012 system the users are given the possibility of entering the required information about the new hardware type This will then be reflected in the statistics with each entry of that specific hardware no longer being unknown but instead represented by the information the user entered It must also be possible to alter this information after the first time in case of wrong input input errors etc The GUI can be enhanced a lot by generally giving the web pages a good looking and consistent layout Further on the statistical information can be given more life and in fact be easier readable and more containing by displaying it as charts as well as in tables For almost identical hardware types to be represented as one table row we will have to implement some kind of grouping functionality In some way a rule set of which hardware types are corresponding to which table entry must be stored This rule set must also either be automatically or manually maintained or a combination of the two approaches The filtering functionality can somewhat easily be implemented with the help of a bit more critical and dynamic database queries than the current system is usi
106. de sinte say Ree eaa Rees aar pee a REED a arae 22 18 What to Bel ge eod eonim nte eteren td 22 18 1 Crash report parsing oscri etes Trees euet e Fa do pec aou sega de nEe e se syuueeta depre gun Eno EENEG 22 18 2 Graphical User Interface ice reete bee eoe iy te ute tiet 23 18 3 Lack of functionality eeseesssssssseseseeee eene eene nennen nnne nennen nnn nennen innen nnns 23 19 How to Improve EE 23 ETSI EE 24 AME Umm 24 Page 18 of 194 Project FHaSS Vision Document 2012 16 Our client Our client Funcom is Norway s largest game development company and was founded in 1993 in Oslo When internet became public property around 1995 Funcom began to head their game production more and more towards online games especially so called MMORPG s In the year 2001 Funcom released their first game of this type namely Anarchy Online and the second game Age of Conan was released in 2008 The new MMORPG project that Funcom is currently working on is called The Secret World and is planned to be released in April 2012 that is in the middle of our project period 1 17 Context of the study 17 1 Computer technology evolution The evolution in the field of computer technology has in the last decades been enormous with lots of new and improved hardware components entering the market every year New manufacturers enteri
107. does will not reduce the focus on the design and implementation phase This phase is on the contrary the most important part of game production for such companies The more cool features graphics and gameplay the game contains the more games they sell The last part of the game production is the testing phase Since none of the earlier phases have been cut down on it is given that this is the phase that most game development companies are doing the Page 20 of 194 Project FHaSS Vision Document 2012 largest cuts in Less emphasis on the testing phase will result in less code specific testing and less hardware compatibility testing which in turn can result in software code errors so called bugs and hardware incompatibility The word can is important in this relation because it leaves room for the possibility of an untested game being released without problems at all or at least with few enough problems for the game to sell well As mentioned above if the companies were to cut down in the implementation phase it would almost without exception result in fewer sold games If they on the contrary were to cut in the testing phase there is a possibility that it would not have the same effect So in essence the decision is easy to make for game development companies that rely on a lot of sold games they do the cuts in the testing phase 17 3 Fault tolerance For game development companies that develop ordinary standalone offline games
108. does not contain the right percentages CPU core does not contain the right percentages Outcome Information displayed correctly but does not correspond with the data in the test data set The reason for not showing the right percentages could be that the parser is not accurate enough when it comes to extracting the speed manufacturer and cores from CPU information A few entries differed for speed and manufacturer but cores did not even show all the different types Test Successful Page 169 of 194 Project FHaSS Test log 2012 55 4 Story 6 Graphics card information Display hardware software statistics GPU information est ID T R 1 2 REG Mee DA A F 1 and B F 1 EEE Test to see if tables are created and contains accurate information about the different GPUs stored in the database It should be possible to toggle between number of entries and COS Show split between nVidia Intel AMD and others Data to be shown e Show video card model like Steam e Show VRAM e Show Primary and multi monitor resolution Test date 27 04 2012 Kim Richard Johansen eSak Function test black box Test approach 5 Put in the time period for which the statistical data shall be produced with basis upon 6 Choose to get to see statistical information for GPU 7 Wait for the web page containing the information to load 8 Compare the resulting percentages to the percentages that have been manually calculated fro
109. e because in a separate state they would not give any extra relevant information 6 Asa user want to see trends in what hardware the users are using as time passes 7 Asa user want to be able to see a prediction of the hardware in the future based on the trend from a specific time period because this will give me information about what type of hardware we will be developing games for 8 Asa user want to be able to filter the information based on the game that was the source of the crash 4 Non functional requirements 1 It is important that the new system will interface with external systems for instance Bugzilla in the same way as the current systems does 2 The foundation of the system must be built general enough so that the same foundation can be used for the eventual improvement implementation of the Crash Statistics Systems Page 6 of 194 H GSKOLEN i Buskerud Funcom Hardware Statistics System Programming Language Document Funcom Hardware Statistics System Client Funcom N V Acronym Sensors amp Supervisors Internal sensor Olaf Hallan Graven 10 01 2012 Author Kent Brian Dreyer Internal supervisor Aurilla Aurelie Arntzen External supervisor Rui M Monteiro Casais Sondre Bjerkerud Sverre Christoffer Thune Kent Brian Dreyer Kim Richard Johansen Project FHaSS Programming Language Document 2011 Programming Language Doc
110. e 32 Database tables for storing filters The second is that the filter should generate a unique URL which the user can then bookmark and or share with other users This will easily be done by changing theHTML request method of the Filter Menu form from POST to GET This results in a resulting URL including all the forms inputs names and values The URL can for instance be accessed by other users of the system and he she will get the exact same data presented 49 Exception handling and logging Late in the project period we have realized that a lot of the crashes we experienced with the parser grouper are not really possible to always avoid These are for instanceissues like timeouts to the database The only realistic fix is to handle these exceptions that occur because they can t always be avoided and will happen randomly Our solution is to try to run a method again when an exception occurs The event should be logged to the database if possible if this fails then to a file If the exception should occur again on the next attempt then it needs to try again This should repeat until it is successful or it have tried a maximum amount of times four Between each attempt it should wait a short but increasing amount of time before trying again so that the other systems it is relying on can become available e g the database connection If it fails the last attempt the program should log this and then shut down and needs to be restarted
111. e Test if results after filtering show entries for each universe for that game Test date 02 02 2012 Kim Richard Johansen WSS ie galjiereres Function test black box Test approach 4 Start the filter combination to be tested 5 Wait for the web page containing the filtered statistics to load 6 Compare the produced filtered percentage statistics to the manually calculated percentages Test criteria e Database and website is up and running e Test dataset has been made and inserted into database Expected e That the produced filtered percentages based on the test data set are results similar to the manually calculated percentages for the tested filter combination Test creator Kim Richard Johansen 01 31 2011 Errors e None Tried each filter combination and the right amount of entries did show for each of them Outcome Tried each game filter The results after filtering did give the right amount of data from the universes for that game The test data set confirms the result Age of Conan 390 13 universes The Secret World 210 7 universes Global 90 3 universes At this stage in the project CPU results did show for the universe filtering but due to the parser not being 100 at the moment results did not match the test data set Test Failure 55 2 Story 4 Server filter Filter statistics server filter est ID T R 3 4 creases A F 3 usse sese Test to see if the server filter
112. e other IDE s will be matched against VS since it is generally considered the natural choice for C and the NET framework 9 1 Visual Studio 2010 Ultimate Visual Studio is Microsoft s own IDE for development in C with NET framework The licensing for this IDE is not free but via the e Academy agreement with our school we have the possibility to get this license free of charge VS is a large software and thus will require more processing power to run than the other IDE s reviewed in this document This also means that we will be in a safer position later in the implementation phase that we have access to required IDE functionality for our project Benefits Drawback Code Snippet Manager auto Large software generation of code Reports that VS handles certain Database Designer Tools functionality suboptimal or buggy Web Forms Designer Reported file cache problems Superior auto complete with No edit and Continue functionality Visual Studio Intellisense Custom debugger visualizers Native IDE Page 110 of 194 Project FHaSS IDE document 2011 9 2 Develop Develop SharpDevelop is an IDE for Microsoft s C with NET platform It is open source and free Develop is a lighter software to run than VS which can be an important factor in the long run working with it The later stable releases of develop is considered good enough by the community to compete with VS but is really regarded as a free option to VS if you are
113. e used for storing and sharing documentation and planning files 2 Perforce is the versioning software that will be used to handle versioning and storage of files containing code That is system specific files 57 Attachments e Requirements breakdown e Gantt diagram Page 86 of 194 H GSKOLEN i Buskerud Funcom Hardware Statistics System Risk Analysis Document Project name Client Funcom N V 29 05 2012 Author Dag Hem Funcom Hardware Statistics System Acronym Sensors amp Supervisors Internal sensor Olaf Hallan Graven Internal supervisor Aurilla AurelieArntzen External supervisor Rui M Monteiro Casais Sondre Bjerkerud Sverre Christoffer Thune Kent Brian Dreyer Kim Richard Johansen Project FHaSS Risk Analysis Document 2012 Risk Analysis Document 58 General Document Information Deliverable nr D2 2 3 PE MEIERI Guidelines Release Public Workpackage WP2 Responsible Dag Hem 58 1 Declaration of intention The intention behind this document is to give its reader a quick understanding of what risks are and how to handle them properly so that we are better equipped when a risk strikes This document will also list most known and large risk along with a small description of the risks consequence and a possible solution By reading understanding and using this document we will be more aware of the overhanging ris
114. ecific combinations of hardware may result in the game not functioning as it is supposed to For instance outdated hardware will not be able to run or run fast enough software written with the capabilities of the state of the art hardware technology in mind This specific problem is especially important when it comes to the gaming industry because newer games are so dependent on the ability to display complex graphics on a computer screen It is the job of the graphics card to create the graphics to be displayed and with the game developers constantly pushing the limits of graphical complexity in their games graphics cards have a very limited lifetime Several game developer companies have had to shut down their whole production or at least having to do significant cuts in their staff after releasing a game that was not compatible with large parts of the players hardware combinations Funcom has themselves been the victim of such disasters when releasing their first two MMORPG games Anarchy Online 3 and Age of Conan 4 Page 19 of 194 Project FHaSS Vision Document 2012 17 2 Shortening of production time The fact that the technology is evolving with the high velocity that it is is resulting in shorter game development periods per game 5 If game development companies had not shortened their development period their games would have been outdated when it entered the market and thereby no one would have bought the game Short
115. ect FHaSS Risk Analysis Document 2012 e Bad internal server performance e Personal computer problems e Web page not viewed correctly by different internet browsers Human risks e Broken NDA e Miscommunication e Lack of knowledge e Teammate is unavailable of an extended amount of time e Solution disagreement e Internal conflicts e Disencouragement and workload e Major project changes additions e Change of supervisor e Change of project manager Other risks e Acquisition of computer hardware 61 1 Requirements related risks R1 1 Any A Requirement functional amp non functional not working Functional A requirements Display statistics table amp evolution charts filter statistics hardware type management and hardware grouping Description Non functional Programming language MVC Database type connection access loading times wiki documentation BugZilla interface windows service Failing one or more A requirements means our system will not be satisfactory Consequence i for our clients use and will lack crucial functions Planning ahead and keeping to the project plan and delivery dates Detect any Prevention er abnormalities early on and deal with them R1 2 Any B Requirement not working Probability B requirements Display additional hardware network and software statistics ET By being unable to meet any B requirement in time we are failing to meet a q d secondary requirement set by
116. ect FHaSS Test Specification 2012 Unknown Test date 14 03 2012 Kim Richard Johansen eSak Function test black box Test approach 1 Putinthe time period for which the statistical data shall be produced with basis upon 2 Choose to get to see statistical information for GPU 3 Wait for the web page containing the information to load 4 Compare the resulting percentages to the percentages that have been manually calculated from the test data set Test criteria e Database and website is up and running e Test data set has been made and inserted into database Expected e The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Kim Richard Johansen 13 03 2011 Errors e Encountered errors Outcome What happened Display hardware software statistics Direct X information est ID T R 1 5 REOM DHE A F 1 and B F 1 Usa e Segleiaielan Test to see if tables are created and contain accurate information about the different Direct X information stored in the database It should be possible to toggle between number of entries and COS Show the Direct X from all systems in the DB and group them under different groups Data to be shown e Direct X grouped under Direct X 11 1 Direct X 11 0 Direct X 10 1 Direct X 10 0 Direct X 9 0c Unknown Test date
117. ect FHaSS Vision Document 2012 The parser that extracts information does not extract enough nor exact enough information from the report There are even somewhat many cases where the parser is not able to obtain any information at all about specific hardware leaving the database field s as unknown This results in insufficient and inaccurate information for the Funcom employees to view which in turn can result in poor estimates This part of the system will have to be improved 18 2 Graphical User Interface The Graphical User Interface GUI is not user friendly and is not representing the stored data in an easy to understand way For instance is information regarding percentage usage of the different Windows versions displayed in a unaltered boring standard HTML table The overall layout of the web pages are not very good looking Further on all the information for the employees to see is squeezed into one page making the web browser s scroll functionality heavily used This is usually seen on as negative and leaves the users with a feeling of lack of overview and control 18 3 Lack of functionality Some functionality that Funcom employees would like to have in the system is not present in the current system There is no handling of almost identical hardware types This results in large tables of statistics with rows for each and every type of GPU units ever discovered by the system Since many hardware types are varying
118. ed general hardware type 10 Get the updated statistics page Print out write down etc the groups and their respective percentages 11 Compare the new group percentages to percentages produced before adding the hardware type s to the new general hardware type step 1 12 Compare results the group percentages whit percentages manually calculated from the test data set Test criteria e Database and website is up and running e There has to be one or more currently registered but not recognized Page 171 of 194 Project FHaSS Test log 2012 hardware types e Test data set has been made and inserted into database Expected results e That the system produced percentages is the same as the manually calculated ones That is both the basis statistics page step 1 and the updated statistics page step 4 contain the same percentages as the manually calculated percentages for the two instances of the general hardware type grouping Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors e None Outcome After mapping all the distinct GPU models and CPU models to general GPU models and CPU models the values printed in the statistics table corresponded to the test data set Test Failure 55 6 Story 7 OS information Display hardware software statistics COS information est ID T R 1 4 sre Mee ud RS A F 1 and B F 1 En
119. ed in the list of risks A risks ID is written R Category number ID in order to minimize clutter as we add and edit risks Page 90 of 194 Project FHaSS Risk Analysis Document 2012 during the course of this project Every risk will be given a unique risk ID which can be referenced to from other documents in the project if this is needed the risks are also given an evaluation of their probability and impact See fig 2 below R Cat ID Risk name Probability Description If needed A short description clarification of the risk Only if needed Consequence A short description of the consequence of this risk occurring Prevention A short guideline on how to prevent this risk from occurring Possible If there is a simple solution to a risk that has occurred it should be explained solution If any here Major Catastrophic Impact Moderate Minor Insignificant Very unlikely Unlikely Moderate Likely Almost certain Probability Fig 2 Color coded chart describing the total severity of probability and impact 61 Risks Here is a list over all known and identified risks placed into several categories Project requirements related risks e Any A requirement not met e Any B requirement not met e Any C requirement not met Technical amp System risks e ASP net or CH programming language difficulties e Loss of data e Slow web page loading times Page 91 of 194 Proj
120. em that we are producing will be deployed at the office in Oslo of our client Funcom and be taken into use by their employees 54 1 Milestones accomplishment Both the first project presentation milestones 1 and 3 were accomplished as expected Before the 1 project presentation the project team worked hard to be able to reach the first release deployment milestone 2 but unfortunately more and more unforeseen problems appeared and the team realized that the milestone was not completeable at that point However after the 1 presentation the team continued the work to get the 1 version of FhaSS deployed to the Funcom severs Again unfortunately more and more problems occured and the team had to rebuild large portions of the system Due to this development test deployment period the second release deployment milestone was just more or less accomplished The team managed to get a version of the system deployed but it was not completely functional and did not have all of the middle priority requirements implemented In essence the second release deployment milestone was also not accomplished The final release milestone 5 on the other side is considered as accomplished The project team did their last deployment to the Funcom servers on the 10 and 11 of May However the Funcom employees that also served as supervisors for the project team throughout the project period has at later dates done deployments to their own server with some or
121. empt to deploy the FHaSS in the Funcom offices in Oslo the differences between the project team s software development environment and Funcoms production environment has been prominent A lot of thought and work hours have been put into solving problems that only occurred in the production environment throughout the latest half of the project development period We could probably have been able to become better known with the environment differences had we been able to spend more time in the Funcom offices in Oslo However this was not easy to accomplish both due to time and economy 66 3 Very large amounts of data The raw hardware statistics database table that Funcom was already using for their previous hardware statistics system had over four 4 million entries With the help of the Funcom employees we quickly got a dump of this table downloaded to our server to test our system against However testing for instance the parser and grouper discussed in section against such a large number of database table entries takes a lot of time We are talking up to several days to complete such a process As a result the testing and debugging of the system in the development period has mainly been done on a subset of the original database table While this gave us the opportunity to test more frequently and efficiently it introduced another potential area of issues the stability and endurance of the backbone of the system We now see that
122. enesccactuccasadacescdachuceitesanessdectucciiadaseduvechuceataduwessvbedueeatadagese acdeteisades 137 34 1 The new database snsc naa eee aaeei aaee aae ae ae aS Ea begdecsonsabiersevaiees 139 34 2 The Par Sen csere eise e ea eE eE a E E A Ee ee ed 139 34 2 1 The Parser interface ENEE 140 35 Redesign of the WCF web service 142 36 Designing the page layQUt mime agetur eade cett e rade ede oed urea d ee 143 37 GPU Manufacturer and VRAM eanoranononnvnnnvnnnvnnnrnenrnnernsersnenreenresnresnvennnesnneesnrnsnrnenrnsnrnsersserssenseene 146 38 Memory nd OS user Stol Sek geed EENS ENEE NEEN deles VEER dE 148 39 Redesigned responsibilities of WCF and Parser ccccccesscssssscssssecseeeecseeesseeeseseecsesessaeesenseesees 151 40 GPU USE STO eee 152 S DT e EET 153 LHuNEEeUUB DINE 153 42 Updates after code review before first deployment essere 154 43 High Charts iones eoe dee EN era hu E aeu at areae vv RE x pua 157 44 Geographical LoCatlot gece iere tere er eee cere endete rn Eee yx ee e a eee pea eee Ce a rupe Yee Ee Eege 157 45 Windows SErViCe SEA 158 46 General model relation search manufacturer rarsrrvnnnnnvrvnnnnnvrvnnnnnrrnnnsnsvvnnnensvvenssnsrrnnnsnsvveesenn 158 47 Hard diskinformatlorni iie einer ee treinta eroe eene Pane ENNEN HERR ue ERR ER NR save 159 48 Filter sto r ge nd URL iter te en tre eet ege e Ere ER
123. er development periods can be solved in two ways by increasing the number of employees or by using less time on less important parts of the development The first solution will result in larger wage expenses and an increase in staff usually is not proportional to the increase in development velocity because of cooperation overhead Because of the usually unstable and unpredictable economy of game development companies the mentioned increase in staff size isn t always a good or even a possible solution Thus are the companies forced the option of focusing less on some part of the development When choosing this solution the companies are being met with the important decisions regarding what or which parts to prioritize less This brings us to a short explanation of the process of developing a game The source of a game is an idea Based on the idea a game designer is creating documentation describing what the game is supposed to contain This documentation is then used as the foundation for the actual designing and implementation of the game In the end the product has to be tested to ensure that it contains what the design documentation says that all functionality works the way it is supposed to and that it can be run on all the different hardware setups that it is supposed to This short introduction to game development gives an idea of which parts the game development process the companies potentially can reduce focus on when shortening their
124. erable nr Version Document name Ex D2 5 1 Quality Assurance Document Where Field Content Workpackage nr The workpackage the deliverable belongs in Deliverable nr The deliverables number Starts with1 Version Minor changes add 0 5 to the version number If it has gone through major changes or review it is increased by 1 Document name The entire name of the document All the information needed is to be found in the project Gantt diagram where all the documents are listed 16 1 3 File formats Any new deliverable should be created from the pre formatted template See section 16 2 1Document creation and Structure and will therefore be of the Microsoft Office 2007 or later Word format DOCX After a document have been reviewed See section 16 2 3Pre completion process and completed it should be converted to the Adobe PDF format 16 2 Deliverable quality process A deliverable is a document which will be passed to the sensors and supervisors both in electronic and paper form for extensive reviewing it is therefore of the upmost importance that any deliverable holds the highest standard possible The quality process is a guideline to make sure the deliverable meets the minimum of requirements the quality process consists of e Having the correct document structure Introduction Content Summary e Making sure all technical references are correct and valid e Any comparison to o
125. eration when it comes to markings expandable collapsible fields dynamic fields and how information should be displayed All in all the design is quite simple but offers a clean natural look making it intuitive and easy to use for new users Tooltips will also be implemented in a later stage to let users get acquainted with both the advanced features of the system as well as some tips and tricks As the mockup was done the next step was to create a basic CSS layout for the main page which would prepare the page for implementation of the filtermenu amongst other functionality Page 145 of 194 Project FHaSS Design Document 2012 37 GPU Manufacturer and VRAM This user story will add information about the GPU manufacturer and vram for the user to see information about One should be able to see how many players use ATI NVIDEA or Intel GPUs in both numbers and percentage and also how much vram they have The HWParser needs to be updated so it will filter the vcard column in the hw info table for the manufacturer and the vram column for the amount of vram Also the database needs to be extended to include this new information Below are updated diagrams illustrating this Figure 15 GPU methods added to HWParser class The class diagram above figure 11 shows the new GPU related methods and also some changes that were done to the design during implementation of the previous user story Page 146 of 194 Project FHaSS Design Docume
126. es with all of NET Paired with mono and the NET debugger it is extremely versatile and a solid cross compatible language for us to develop in Benefits Drawback language interoperability with NET e IDE choice might be restricted framework Cross platform compatible with Mono Easy accessible SQL querying Great debugger through NET Page 12 of 194 Project FHaSS Programming Language Document 2011 11 PHP Hypertext Preprocessor PHP is a quick and easy language to learn and a sophisticated language to master Most if not all of our groups members have worked with PHP before in different degrees this gives PHP a small advantage over the new languages since its going to take a lot less time for us to start using it for our purpose PHP is mostly run on a Linux environment connected to a MySQL database and its mostly for this reason that it usually uses less memory during runtime than ASP NET running on IIS Internet Information Services This may not however be an issue due to the fact that our system will only be used by a small amount of developers at a time The fact that PHP does not have a debugger present makes it the main reason we will probably choose C with ASP NET as our web development tool PHP can be developed in a handful of IDE s but NetBeans IDE would be the IDE of choice if we decided to use PHP because of its wide range of functions and we already have extensive knowledge of NetBeans IDE
127. eting should be created using the Minutes template Meetings Minutes and one should also refer to the Minutes writing guidelines Meetings Minutes for further assistance 18 Coding standards and plugins Because every member of the project group has taken the same classes during their bachelor it is natural that their way of coding coding standard resembles each other s to some degree but we are still in the need of some external tools and predetermined set of rules to make sure we deliver code with a high standard and a clear conformitythroughout the project The project group will not be weighted down by a large list of specific rules but rather take use of plugins that are tweaked for our specific needs and a small set of guidelines set by this document 18 1 Coding guidelines 18 1 1 Naming conventions It is the project group s intention that our code is easily readable and understandable by external viewers CamelCase The project group will take use of CamelCase on variables methods and classes The initial letter is lower case on variables while methods and classes use upper case on initial letters E g variable ypuModelDictionary and method RefreshCache Descriptive naming Every variable method and class name should be descriptive of their respective purpose without creating unnecessary confusion 18 1 2 VAR vs specified type In many situations it is faster to use a non spe
128. f 194 9 HOGSKOLEN i Buskerud Funcom Hardware Statistics System Project Plan Projectname 0 Project name Funcom Hardware Statistics System Client Funcom N V Acronym Sensors amp Supervisors Internal sensor Olaf Hallan Graven Internal supervisor AurillaAurelieArntzen 20 03 2012 Author SondreBjerkerud External supervisor Rui M Monteiro Casais SondreBjerkerud Dag Hem SverreChristoffer Thune Kent Brian Dreyer Kim Richard Johansen Project FHaSS Project Plan 2012 Project Plan 49 General Document Information Deliverable nr D2 1 3 Deliverable type Research Release Public Work package WP2 Responsible SondreBjerkerud 49 1 Declaration of intention The intention of this document is to give the reader insight into what the project group is supposed to do at which times throughout the project period at a reasonable abstraction level The plan is supposed to work as a guideline for the team in the project period and the project progress shall be measured against it The plan can eventually be updated if it becomes obvious that it won t be possible for the team to follow it This document will also describe how responsibility areas are distributed between the team members as well as give some insight into how the time estimations were done to produce the project plan Lastly this document will summarize how the resource
129. faulty The consequence was iterations on the system as our experience grew In the middle stages we might run into a wall when we failed to properly grasp a specific concept and loose some morale The consequence was to switch tasks and support each other when needed work might have been a bit slower than optimal In the later stages of the project we recognized that some people became a bit more specialized in a certain area the more they worked on it The consequence became that we could get tunnel vision by focusing purely on our own areas loosing perspective Consequence Solution R3 10 Change of Project Manager At the beginning of our g sprint our supervisors and sensors decided to change Description the Project Manager due to some minor misunderstanding between the PM and our client The transition between PM s went very good but we still lost a lot of sorely needed management expertise the original PM had The consequence for this was a severely decreased morale across the board Consequence Solution 62 4 Other R4 2 Relocation of internal server Sprint recognized Sprint nr 1 Severity During our set up sprint the IT Department at HiBu was unable to get our internal server through their firewall in time Consequence We had to relocate the internal server to avoid additional downtime We Solution relocated it to our project manager s apartment Description Page 100 of 194 H
130. formation est ID T R 1 1 REG Mee ud DH A F 1 and B F 1 ES Edda Test to see if tables are created and contains accurate information about the different CPUs stored in the database It should be possible to toggle between number of entries and Data to be shown e CPU speed by brackets like Steam Inc at 400mhz staring at 1 Ohgz e Show split between AMD and Intel and other e Show number of cores Test date 27 04 2012 Kim Richard Johansen ESRA Function test black box Test approach 5 Putin the time period for which the statistical data shall be produced with basis upon 6 Choose to get to see statistical information for CPU 7 Wait for the web page containing the information to load 8 Compare the resulting percentages to the percentages that have been manually calculated from the test data set Test criteria e Database and website is up and running e Test dataset has been made and inserted into database Expected e The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Kim Richard Johansen 23 12 2011 Page 168 of 194 Project FHaSS Test log 2012 Errors e None Outcome Test data set loaded into the data base and went through the parser and grouper functionality successfully The information displayed on the website corresponded to the manually calculated results for d
131. g cache for faster loading times inside the last month Page 162 of 194 9 HOGSKOLEN i Buskerud Funcom Hardware Statistics System Client Funcom N V 29 05 12 Author Project name Funcom Hardware Statistics System nm E I c Acronym Sensors amp Supervisors Internal sensor Olaf Hallan Graven Internal supervisor Aurilla Aurelie Arntzen External supervisor Rui M MonteiroCasais Sondre Bjerkerud Sverre Christoffer Thune Kent Brian Dreyer Kim Richard Johansen Project FHaSS Test log 2012 Test summary 52 General Document Information Deliverable nr D8 2 1 DY ViMele lelismaecee Report Release Public Workpackage Independent Responsible Kim Richard Johansen 52 1 Declaration of intention This is a document containing information about all tests that has been completed successful and unsuccessful Tests that have not been executed are also listed with an explanation for why it has not been completed Verification tests are listed with a short summary 52 2 Definitions and Acronyms Bug An error in the programming code Parser Searches a data file for key words and copies the values it finds CPU Central Processing Unit Graphical Processing Unit Random Access Memory DirectX A collection of application programming interfaces APIs for handling tasks related to
132. g loading times for the final web page displaying the hardware statistics This was especially true for large time spans large amounts of statistics data and we quickly understood that some of the work that back then was done per request had to be done in advance of the actual request instead to decrease the overall loading time 61 1 1 Parser The second version of the systems backbone layer therefore had a new feature implemented the Parser Some of the work that in the previous version was done in the data processing and preparation phase had now been moved to another part of the system Here the work could be done completely independent and in advance of the users requests More specifically the Parsers job is to improve the accuracy of the hardware information stored For instance it is the job of the Parser to retrieve the CPU speed from the original CPU string that is sent by the crashed game clients Having the more accurate data retrieved in advance and stored in another database table one major operation earlier done per request had been removed 61 1 2 Grouper With the wide range of different hardware specifications in terms of memory CPU frequency storage and so forth we needed to group similar hardware in the database to improve the load time and how we retrieve and present the data The new version of the system worked well and the loading time improvements were significant but still they did not match the requirements
133. ges add or delete user stories from the PB and prioritize them This means that the requirements described in this document are by no means final neither are the number organization nor the definition of them The requirements are all fluid and modifications can happen at all times throughout the project period The current requirements will act as a basis for the planning of and the starting work with the project and in the case that no Page 47 of 194 Project FHaSS Requirements Specification 2012 changes are done to the PB they are the requirements we will work towards developing implementing for the whole project period 41 Requirements list Following is a list of the requirements for the Hardware Statistics System The requirements are organized by priority as they are given in the Product Backlog by the external supervisor The Product Backlog is prioritized as a list without categories but there are a couple of somewhat clear but not displayed lines partitioning the user stories with regard to priority Of this reason it felt natural to separate the requirements into three priority categories A requirements are major features necessary for the new system to be able to outdistance the current system B requirements are feature extensions that will further extend the range of use of the system and C requirements are features that can even further extend the range of use of the system but which are of lesser importance
134. ginal authors at least 2 workdays before the planned completion date before it is given to another project member for a peer review See section 16 2 3Pre completion process The peer review should be done at least 2 workdays before the delivery deadline and mailed to the internal supervisor by the original author for a final review 16 2 2 1 Timeliness amp exceptions It is each author s own responsibility to ensure the on time completion of their deliverable and also their responsibility to make sure that each review step is completed on time The project manager should however still keep track of each author s progress and give them a notification if it is believed that their deliverable should take longer than planned If a project member feels that their given time will not be enough to complete the deliverable they should discuss this with the project manager 16 2 3 Pre completion process Completed 1 version When a document or deliverable is nearing completion its author should take his time to do a complete self assessment review before deeming it completeand place it in the Deliverables Correct WP folder After a deliverable is considered completed by its author it should be reviewed by a suitable peer within the project group this individual should ideally be e Knowledgeable about the subject Page 119 of 194 Project FHaSS Quality Assurance Document 2011 e Nota co author of the document e Have
135. ginal file from the outgoing folder on the batchserver Opens TAR files in the incoming folder on the internal server and extracts them to the same folder Closes the TAR file writing to file and the folder paths Closes all contact with the SFTP and sockets Attachment 2 Reportdownloader py 1 2 3 4 oo Sets the path to the complete folder An infinite loop that handles downloading files from this folder starts Connects to the batch server with user name and password over SFTP Opens the correct folder in batches on the batch server and creates a list of all the files in the folder Moves files from the complete folder to the outgoing folder on the batch server Downloads the files from the batch server and moves them to the complete folder on the internal server Deletes the original files in the outgoing folder on the batch server The downloaded files will later on be handled by XMLBugReport described later Attachment 3 XMLBugReport py o amp ommc Ne Handles arguments if they exsist up to three set incomingDir logfile and forceid Infinite loop starts Creates a list of the files in the complete folder on the internal server Moves the file from the complete folder to the processing folder Start to process the file calls fcBugMain ProcessFile which parses the XML file including dxdiag If the processing is successful the file is moved to its own subfolder in the archive old folder If t
136. gives the employees new possibilities of searching for trends and hardware information based on their own criteria For instance can geographically filtered information be very valuable when it comes to porting a game to another country or region like Funcom did with Age of Conan for the Asian market 21 Sources 1 http en wikipedia org wiki Funcom Last visited 13 12 2011 2 http en wikipedia org wiki Graphics processing unit Last visited 21 12 2011 3 http www vg no spill artikkel php artid 182545 Last visited 13 12 2011 4 http www idg no computerworld article145027 ece Last visited 13 12 2011 Page 24 of 194 Project FHaSS Vision Document 2012 5 http en wikipedia org wiki Video game development Last visited 21 12 2011 6 http www shacknews com article 70451 rage pc players report bugs Last visited 21 12 2011 Page 25 of 194 H GSKOLEN i Buskerud Funcom Hardware Statistics System Current System Description FHaSS Sensors amp Supervisors Funcom Hardware Statistics System Client Funcom N V Internal sensor Olaf Hallan Graven Internal supervisor Aurilla Aurelie Arntzen 10 01 2012 External supervisor Rui M Monteiro Casais Group members Sondre Bjerkerud Sverre Christoffer Thune Kim Richard Johansen Dag Hem Kent Brian Dreyer Project FHaSS Current System Description 2011 Current System Description 22 General Document Informati
137. gnized hardware group 22 05 2012 Tester Sverre C Thune Function test black box 1 Getthe statistics page Print out write down etc the groups and their respective percentages 2 Create a new general hardware type of one of the possible categories CPU Graphics Card or Windows OS 3 Choose group one some all of the registered but not recognized hardware types under the newly created general hardware type 4 Getthe updated statistics page Print out write down etc the groups and their respective percentages 5 Compare the new group percentages to percentages produced before adding the hardware type s to the new general hardware type step 1 6 Compare results the group percentages whit percentages manually calculated from the test data set e Database and website is up and running e There has to be one or more currently registered but not recognized hardware types e Test data set has been made and inserted into database e That the system produced percentages is the same as the manually calculated ones That is both the basis statistics page step 1 and the updated statistics page step 4 contain the same percentages as the manually calculated percentages for the two instances of the general hardware type grouping Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 e Encountered errors What happened Requirement ID
138. h the 1 version of the Project Plan The portion of user stories following the Time Period Filter is basically the same functionality that the requirements Display hardware software statistics Display Software Statistics and Future Hardware Prediction are describing Therefore the time estimate for this portion of functionality is the same but with the new version of course distributing the time between the user stories The distribution has been done following our experience and intuition 54 Milestones Milestones functions as intermediate goals for the project work and will help the project manager with the task of knowing whether or not the project work is going according to the plan When milestones are reached can and should the project team gather and celebrate that the goal has been achieved to strengthen the mentality and focus of the team members for the upcoming work period against the next milestone The following will list the milestones and their respective dates that the project team will work towards 1 The 1st presentation 13 01 2012 2 Release Deployment High Priority Requirements 19 03 2012 Page 83 of 194 Project FHaSS Project Plan 2012 The 2nd presentation 22 03 2012 Release Deployment Middle Priority Requirements 5 05 2012 Release Deployment Final release 19 05 2012 The 3 and final presentation 06 06 2012 Qr ur mm owe After each of the release deployment milestones the syst
139. he necessary alterations to the python parser in order to make this information available in our database Researching geotargeting quickly told us that we need some external pieces of software and data to retrieve a country from an IP address Because a country does not have a predetermined range of IPs but rather a set of ranges that may vary you end up with a whole lot of different ranges to different countries luckily this information is collected by interested parties and stored within a geolocation database which is available for free In addition to requiring the data itself we also need to implement an external C class that handles the conversion of IP address to a number IP number which is used as the key for values within the database We settled with MaxMind s free GeolP database as well as the accompanying open source CH CountryLookup class The GeolP data file is to be implemented as an embedded resource while the open source code is to be used as a standalone class in our parser Page 157 of 194 Project FHaSS Design Document 2012 45 Windows service Originally the parser was a console application which had to be started by a exe file A new requirement added late in the project period was that the parser had to be re implemented as a windows service application The reason for this is that it should be installed to run in the background So the parser had to be rewritten to support windows service functionalit
140. he ReadDataService will need a method that updates the Games table with all the distinct games and a method that updates the Universe table with all the distinct universes Also a method that counts how many entries there are per game in the games table and hold these numbers in a list ASP NET MVC3 Application will have models to get the required information from the service and we also need an appropriate controller and a view to finally display the amount of entries We are also going to have a model controller and views to let the user set up new game to universe relations easily through the web interface Below figure 3 you can see a figure of these classes The views are not really classes however it is an important part of the MVC pattern so decided to include it here Page 136 of 194 Project FHaSS Design Document 2012 Figure 3 Game filter class diagram The second user story is about the server universe filter It should function like this the user select the game and then a specific server for that game The view should then display the number of entries for the specific server for the game A server is also called a universe which is the name used in the database tables See figure 2 below for a activity diagram describing the flow of events in this user story Figure 4 Server filter activity diagram This user story will mostly use the same classes as the last one The reason for this is that we already had the
141. he impact depends on the amount of time the missing member is unable to work It is not always possible to foresee or prevent this from occurring but keeping Prevention ME healthy and abiding the law is a good start Possible Some possible solutions could be to either change project scope getting an solution outside workforce or reuse some of the old system R3 5 Project related disagreement Many of the choices in a large project is often planned months ahead or extensively researched if we are unsecure about how to proceed We may however still come across small issues that need to be resolved quickly between us The consequence will be minor because of all the larger issues are already Consequence resolved A conflict like this may take time if the parties does not seek out to find new information or fail to reach an agreement Many of these disagreements can be avoided by assigning a responsible person Prevention i i for a certain task and letting them decide If the parties fail to resolve the issue quickly they could either take their argument to an objective team member or if it seems fitting ask our supervisors Description Possible solution R3 6 Internal Conflicts to internal conflicts Internal conflicts are personal issues between two parties Depending on the issue the impact may range between a team member having Consequence to work at home for a short time to relax away from the group
142. he processing fails the file is moved to a failed folder Attachment 4 List of contents of a bug report IPAddress Date Username Title Type Email Category Universe Body Attachments BodyElements BuzillaCCs ExtraCCs TeleportHistory Attachment 5 mapfileimporter py Runs the function importMapFiles continously makes a directory list of the files in the mapfiles folder Traverses the directory looking for files to import Page 33 of 194 Project FHaSS Current System Description 2011 e Takes all files and calls parseFile function from fcBugSymbols e Moves all files from importMapFiles folder to archive folder Attachment 6 Example of extracted text lines from DxDiag Operating System Windows XP Professional 5 1 Build 2600 Service Pack 3 2600 xpsp sp3 gdr 080814 1236 Processor Intel R Celeron R M processor 1 60GHz Memory 1016MB RAM Card name Mobile Intel R 915GM GMS 910GML Express Chipset Family Current Mode 1280 x 800 32 bit 60Hz Display Memory 96 0 MB Driver Version 6 14 0010 4609 English Page 34 of 194 9 HOGSKOLEN i Buskerud Funcom Hardware Statistics System Test Strategy Funcom Hardware Statistics System Client Funcom N V Acronym Sensors amp Supervisors Internal sensor Olaf Hallan Graven Internal supervisor Aurilla Aurelie Arntzen 29 05 2012 Author External supervisor Rui M Monteiro Casais So
143. ics display would contain lots and lots of GPU models with both big and small differences in their names Our solution to this problem was to create two database tables distinctgpumodel and generalgpumodel see database illustration in section 8 The distinctgpumodel table will be a table similar to the universe table where all distinct registered GPU models in the hardware table are Page 152 of 194 Project FHaSS Design Document 2012 stored Then we let the user of our system create and maintain a set of more general GPU models stored in the generalgpumodel and choosing which distinct GPU models shall be mapped to which general GPU model Our system will then generate all statistics and displays for GPU models as though the distinct GPU models actually was of the general GPU model it is mapped to making the statistics a lot more readable and informing 40 1 Resolution For the resolution parsing filter we will create a new resolution table that will contain the resolutions that are going to be parsed 41 CPU Model Like the GPU user story Funcom wants to see statistics of general CPU models For example they want to see how many percentage of the players have a Core 2 Duo or an i5 processor A distinctcpumodel and a generalcpumodel table will be used and the user of our system will be able to create and maintain these general CPU models just like with the GPU models The user will be able to set the distinct CPU models relat
144. ifferent hardware software stored in the database Test date 09 02 2012 Tester Kim Richard Johansen 25 03 2012 WSS ie paljierele Function test black box Test approach 1 Putin the time period for which the statistical data shall be produced with basis upon 2 Choose to get to see statistical information for one of the hardware software categories CPU Graphics Card RAM Windows OS HDD or Network download speed 3 Wait for the web page containing the information to load 4 Compare the resulting percentages to the percentages that have been manually calculated from the test data set Test criteria e Database and website is up and running e Test dataset has been made and inserted into database Expected e The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors e Encountered errors Outcome What happened Display hardware software statistics CPU information est ID T R 1 1 REG Mee ud ES A F 1 and B F 1 ES Edda Test to see if tables are created and contains accurate information about the different CPUs stored in the database It should be possible to toggle between number of entries and Data to be shown e CPU speed by brackets like Steam Starting at 0 1
145. ifferent models speed groups manufacturer and number of cores Test Failure Display hardware software statistics CPU information est ID T R 1 1 REG MeuEu PES A F 1 and B F 1 EEE Test to see if tables are created and contains accurate information about the different CPUs stored in the database It should be possible to toggle between number of entries and Data to be shown e CPU speed by brackets like Steam Inc at 400mhz staring at 1 Ohgz e Show split between AMD and Intel and other e Show number of cores Test date 11 03 2012 Kim Richard Johansen USS ial palsuereles Function test black box Test approach 1 Putin the time period for which the statistical data shall be produced with basis upon 2 Choose to get to see statistical information for CPU 3 Wait for the web page containing the information to load 4 Compare the resulting percentages to the percentages that have been manually calculated from the test data set Test criteria e Database and website is up and running e Test dataset has been made and inserted into database Expected e The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Kim Richard Johansen 23 12 2011 Errors e Information display correctly on the web site but CPU manufacturer does not contain the right percentages CPU speed
146. ing Agenda and Minutes guidelines Page 116 of 194 Project FHaSS Quality Assurance Document 2011 There should always be a secretary present to take minutes of the meeting that should be edited and distributed to all participants within one workday as well as being placed in our Meeting Minutes folder Meeting topics may vary in scope but will at a minimum contain the following e Project updates on all completed work Since last meeting or work in progress e Questions from the team to the supervisor s e Work to be done until the next meeting e Topics for the next meeting e Any conflicts problems the supervisor should be aware of The meeting is always started and closed by the meeting leader the meeting leader may alternate or change depending on the meeting 15 1 3 Group communication The project group will use additional communication means such as Skype and MSN between each other for faster and easier communication however any major item or news should be announced via email to all affected parties 15 2 Information Exchange File sharing 15 2 1 Project files folder Dropbox The project group will be using a collaboration tool called Dropbox to store and share all our project related files Dropbox gives us easy access to all our files wherever we are as well as synchronizing them to the latest version Only the project members and our supervisors will be given access to our Dropbox files
147. ion is added the project leader should perhaps develop several different strategies for handling this additional problem and discussing the best solution with the project team or supervisors Possible solution R3 9 Change of Supervisor illness change of job complications or other reasons Having to change a supervisor will both be time consuming and slowing down our project due to spending time replacing and briefing the new supervisor Possible So en We could rely more on our other supervisor during a transition phase R3 10 Change of Project Manager MS The risk of having to change the current project manager and reinstate another Description project member A change of Project Manager would cost us a large amount of time to reinstate another project member It would also severely decrease the morale of the Consequence affected member as well as the other members of the group Loss of knowledge routines Clear communication with the other project members as well as the client and Prevention any supervisors so that any problem or issue may be solved through constructive criticism instead of a sudden change as the easiest solution Possible ealution The team must go together and to coordinate the PM change 61 4 Other risks R4 1 Acquisition of computer hardware Page 97 of 194 Project FHaSS Risk Analysis Document 2012 oe To run and test our system we have the need for an internal se
148. ion to a general CPU model and our system will then display all statistics as if the distinct CPU model actually was of the general CPU model it is mapped to nHibernateHelper NHibemateHelper GetHardwarelnfoL ist session ISession offset int limit int void GetCpuManufacturerList session ISession IList GetGpuManufacturerList session ISession IList RunAutoRelateDistinctToGe neralCpuModels UpdateUniverses void UpdateDistinctGpuModels void UpdateDistinctCpuModels void AutoRelateDistinctToGeneralGpuModels void AutoRelateDistinctToGeneralCpuModels void FilterGpuManufacturer vcard string gpuManufacturerList IList Gp FilterRam ram string int ParseString stringToParse string startWo ParseString stringToParse string startWo GroupRam ram int RamGroup GroupGpuModel modelName string GeneralGpuModel GroupCpuModel modelName string GeneralCpu Model Figure 23 HWParser classes Page 153 of 194 Project FHaSS Design Document 2012 The class diagram above shows the updated design of our parser including the CPU model additions and also general updates to the design that have been made while implementing the code The grouping logic have been moved to its own class and so has the update and auto relate methods for the distinct tables Y EL 1 1 1 SR is is iz JG 18 is Figure 24 Database tables Above are the tables
149. ituation like explained above 67 2 Physical and virtual CPU cores statistics On the client side of their MMORPG games Funcom is using the DirectX Diagnostic DxDiag tool that comes with Windows to retrieve hardware statistics to return to their hardware statistics server One of the problems with this approach is that the DxDiag tool does not take into account whether or not the amount of CPU cores that Windows is working on are physical cores on the CPU chip or just virtual cores As a result the statistics that FHaSS is producing about CPU number of cores is a merged statistics of both physical and virtual number of cores A possible solution for giving correct statistics in this context is to rewrite the hardware information gathering software to gather the CPU number of cores information from the msinfo32 system information tool instead of the DxDiag tool 67 3 Saved filter administration with AJAX Currently the saved filter administration functionality has got its own page in FHaSS This functionality could potentially have been put in the administration section but since all the functionality in the administration section controlled the backend part of the system it felt wrong to put frontend administration functionality like this in the same section The functionality could also have been implemented into the Filter Menu on the Home page but due to the Home page having to reload between each new HTTP request which a deletion of
150. kend of FHaSS such a solution could potentially also improve the loading times for the Home page 67 5 Dual page view Feedback from the Funcom employees said that it was wanted that the FHaSS web page interface should be capable of displaying two or several pages of hardware statistics without having to reload the page Such functionality would make it easier for the user to compare statistics calculated based on two or more different filter selections in the Filter Menu 67 6 Toggle displaying individual charts How the system works today is that all categories displayed in the statistical table have their own chart showing information These charts will be created as long as the amount of days in the time span chosen is under 600 days An option to choose which charts to be created could be added to reduce the loading time of the statistical page or to just show the relevant ones and use less space 67 7 Compare newest event log entry from database to local event log How the parser grouper and distinct updater of the system works today if an exception was to happen is that an event log about the exception is written to the database and to a local file where the windows service is stored If the database connection is down still after 4 retries the event log will only be written to the local event log and another log saying could not log event to database Today there is no functionality to check if the local file has newer log e
151. king sure to delete files with the P4 GUI client this problem is avoided 25 Subversion Subversion is an open source version control system developed as a project of the Apache Software Foundation There are different software that uses this open source project as a basis For the server we will test VisualSVN Server and for the client we test AnkhSVN because these are free and recommended by multiple people on stackoverflow com Throughout the document SVN Subversion will be used to refer to this system First we have to install the subversion server VisualSVN Server You have to set up users accounts on the server which clients can connect with Then we install AnkhSVN on the clients which is the plug in for Visual Studio The installation process is straightforward and without problems Page 128 of 194 Project FHaSS Version Control Document 2012 Setting up a project to use SVN is straightforward First create a project and then add the solution to subversion in Visual Studio Commit the changes and it is done After that adding files works very well just like with P4 you just have to create the file in Visual Studio and it will automatically be added to your pending change list Then you just have to commit the change and the file is added Checking out files is also very easy it is completely automatic just start writing code in the file and it is checked out With this plug in you can t see if there is so
152. ks we face during this project and also how to avoid or even solve them if they occurred It is this documents main intention to lower the chances of risks happening and help us handle them as quickly as possible when they occur The assessments that are made in this document are meant to guide the project group through hard choices regarding activities priorities strategic choices of solutions and project management 58 2 Definitions and acronyms The damage caused to our project Often in form of time loss Application Programming Interface Non Disclosure Agreement The Secret World Identity Crash When a program unexpectedly stops running Browser Internet browser MVC Model View Controller Net standard PM Project Manager N A Not applicable WCF Windows Communication Foundation 58 3 Document history Version Description Date 1 First version created 1 12 2011 1 5 Peer reviewed 14 12 2011 2 Supervisor reviewed 10 01 2012 Page 88 of 194 Project FHaSS Risk Analysis Document 2012 2 5 Correcting risks 20 03 2012 3 Finalized document 21 05 2012 3 5 Document reviewed 25 05 2012 59 Table of contents 58 General Document Information 88 58 1 Declaration Of intention ue tiit aiandi odes e nante ge Pati er eh ideen 88 58 2 Definitions and acronyms sess enne nennen nennen nnne senten nente sene nn nnn 88 58 3
153. l be the person responsible for that task being done and done properly 53 Requirements breakdown To get an idea of what has to be done to manage to implement the functionality that the requirements are describing we had to break down the requirements into tasks These tasks had to be specific enough for us to be able to give a good estimate of the work required to do them but also general enough so they could be used as the basis of a project plan Too specific task definitions would also be way too time consuming to produce and would probably not improve the time estimates because of the lack of experience we ve got in this specific technical area There is a document attached to this project plan called Requirements breakdown In that document the requirements breakdown lists and the time estimations for each task that was used as a basis for the project plan is given 53 1 Remake of the Requirements breakdown Closely related to the remake of the Gantt diagram see section 6 Gantt diagram is the remake of the Requirements breakdown This breakdown is now estimating the time usage for implementing the User Stories in the Product Backlog instead of the requirements in the Requirements Specification For the user stories in the Product Backlog following the Time Period Filter it is still hard to make good time estimates for We will therefore be keeping the time estimates for this portion of the Product Backlog as they were wit
154. l computer with backup systems If data is lost and cannot be regained the quickest solution would be for the project member losing the data to reproduce it as quickly as possible We also rely heavily on version verifying software Possible solution R2 3 Slow web page loading times The user experiences that his browser uses a long time loading the hardware Description crash statistics web page May be due to web page programming or slow database connection It may be undesirable for the user to even use our new system due to long Consequence e loading time Having a good understanding of ASP Net and C and good programming ethics will help reuse of code no memory leaks low memory usage and flexible yet robust code Can be prevented by optimizing system amp cleaning code after testing Possiblesolution If this problem is persistent after several iterations and there is enough time it Page 93 of 194 Prevention Project FHaSS Risk Analysis Document 2012 might be advisable to either take a closer look and restructure the architecture or optimize the code R2 4 Bad internal server performance all its received data it may slow down or even experience crashes It may hinder the admittance of new data into the database and in a worst case situation slow down or even break the database connection to the user s web page Enough system resources on the internal server to handle all the processing Should
155. layers These bug report files are then searched for keywords which define the category of the bug Idberrors clientcrashes clientasserts scripterrors bugreports silentasserts When the category is detected the file is stored in a folder named batchtmp subfolder where subfolder are the different categories WebRestarter py starts processes of BatchMaker py for each of the different categories Each process of BatchMaker py creates a list of files in the subfolder it is assigned to within the batchtmp folder Then it creates a batch file tar of all the files in that folder and moves the batch file to a subfolder of the same category within the batches folder To sum up the Batchserver receives bug reports categorizes them and creates batches of the files to be downloaded Initially we will not change anything on the Batchserver it works good as it is 24 3 The Internal Server The internal server consists of several python processes that together perform the downloading of files from the batchserver and then the parsing and storing of bug reports and player hardware information This section explains the code of some of the important files in the current system which we can use for future reference when implementing our system 24 3 1 General description of python files see attachments for full flow of events on some of the files Restarter Python Here all processes are started Some even have mu
156. ll and similarly with both systems 28 Conclusion Both systems work very well and include the features we need for our project However P4 is the optimal choice for us because of the feature to easily view which revision of a file you have compared to the server and if other users have checked out files Also the merge tool in P4 is also better and easier to use A potential problem with P4 is that the free version will only allow 1000 files when working with more than two users This is not a problem with SVN but it is unlikely that our project will contain so many files so P4 is still preferred 29 Sources e D2 4 1 IDE Document docx project FHaSS Last visited 03 01 2012 e http stackoverflow com questions 157429 what are the benefits of using perforce instead of subversion Last visited 16 12 2011 e http www perforce com Last visited 19 12 2011 e http ankhsvn open collab net Last visited 19 12 2011 e http stackoverflow com questions 453481 what subversion plugins for visual studio are there Last visited 19 12 2011 e http johnhforrest com 2010 09 how to subversion server for windows Last visited 20 12 2011 Page 130 of 194 9 HOGSKOLEN i Buskerud Funcom Hardware Statistics System Design Document Funcom Hardware Statistics System Client Funcom N V Acronym Sensors amp Supervisors Internal sensor Olaf Hallan Graven Internal supervisor Aurilla Aurelie
157. ltiple instances os spawnvp Page 29 of 194 Project FHaSS Current System Description 2011 Batchdownloader Python attachment 1 The process receives arguments which sets the folder directory it will traverse At this point it will start downloading batch files from this path and move the files in the folder to the batch server s outgoing folder and download it to the incoming folder on the internal server It deletes original files after successful transfer The important thing to noticehere is the connection to the batch server with user name and password over SFTP If we are going to replace the current system completely we need to download the files from the batch server with a similar method Initially however we can develop our system to co exist with the old one Reportdownloader Python attachment 2 This file does similar work as the batch downloader but moves files from the complete folder to the outgoing folder on the batch server and then downloads from the outgoing folder to the complete folder on the internal server It deletes original files after successful transfer Also here the connection over SFTP is an important thing to notice see explanation in previous section XMLBugReport Python attachment 3 This file calls the processing of the XML files when they are downloaded and extracted It uses a method to parse the XML file including the DxDiag file It also m
158. m the test data set Test criteria e Database and website is up and running e Test dataset has been made and inserted into database Expected e The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Kim Richard Johansen 23 12 2011 Errors e None Outcome Test data set loaded into the db and then went through the parser and grouper functionality The information displayed on the website corresponded to the manually calculated results for video card model VRAM and Primary resolution Multi monitor resolution has been changed to an extension of the system and therefore not a part of this test Test failure 55 5 Story 6 7 Hardware type management and grouping Hardware type management and grouping est ID T R 4 1 ire Plasa A F 5 ES Eddie Test to see if the hardware from the bug reports is grouped correctly under general hardware types and that it shows the correct percentage for each of them based on the total Page 170 of 194 Project FHaSS Test log 2012 This test will also contain e Test if registered but unrecognized hardware is grouped in the unrecognized category and shows the right percentage Test date 22 05 2012 Tester Sverre C Thune WSS ie paljierere Function test black box Test approach 3 Get the statistics page Print out write down et
159. manual calculated results in the test data set Correct amount and percentages Test Failure Display hardware software statistics Direct X information est ID T R 1 5 REG Mee A F 1 and B F 1 EE Test to see if tables are created and contain accurate information about the different Direct X information stored in the database It should be possible to toggle between number of entries and COS Show the Direct X from all systems in db and group them under different groups Data to be shown e Direct X grouped under Direct X 11 1 Direct X 11 0 Direct X 10 1 Direct X 10 0 Direct X 9 0c Unknown Test date 27 04 2012 Tester Kim Richard Johansen USS ial galuereres Function test black box Test approach 5 Putin the time period for which the statistical data shall be produced with basis upon 6 Choose to get to see statistical information for GPU 7 Wait for the web page containing the information to load 8 Compare the resulting percentages to the percentages that have been manually calculated from the test data set Test criteria e Database and website is up and running e Test data set has been made and inserted into database Expected e The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Kim Richard Johansen 16 03 2011 Errors e None Outco
160. me Loaded test data set containing DirectX into the database and ran the parser and Page 173 of 194 Project FHaSS Test log 2012 grouper system Loaded the website and checked if the information displayed corresponded to the manually calculated results It did Test Failure 55 7 Story 8 Memory information Display hardware software statistics RAM information est ID T R 1 3 AEN nA D A F 1 and B F 1 ES dG Edda Test to see if tables are created and contains accurate information about the different RAM stored in the database It should be possible to toggle between number of entries and COS Show the RAM from all systems in db and group them under different range groups Data to be shown e RAM grouped under Less than 512 MB 512 MB to 999 MB 1GB 2 GB 3 GB 4 GB 5GB and higher Unknown Test date 13 03 2012 Kim Richard Johansen ESRA Function test black box Test approach 5 Putin the time period for which the statistical data shall be produced with basis upon 6 Choose to get to see statistical information for GPU 7 Wait for the web page containing the information to load 8 Compare the resulting percentages to the percentages that have been manually calculated from the test data set Test criteria e Database and website is up and running e Test data set has been made and inserted into database Expected e
161. meone else who have checked out a file Also you cannot see if there is a newer revision of the file on the server you have to manually check if you can get a new revision Checking in files is also simple just like with P4 Every file you change is automatically added to a pending list Just click commit and the files are checked in The merging tool works well enough with AnkhSVN The new code from two users is merged automatically but some extra lines are added that show which user that part of the code was from these extra lines you will need to remove manually before committing the change There are also not simultaneous views to compare the different revisions that you want to merge Comparing changes between different revisions in the history of a file is simple Right click and choose view history and then choose two revisions to view the changes It is simple and easy to use There is no problem reverting a file to an older revision 26 Other Systems GIT and Mercurial are two other version control systems that could be used They include similar features to P4 and SVN but the server set up is more complicated than just installing and running it command line seems to be unavoidable Because the setup is not simple enough these two alternatives are discarded and will not be compared or taken into account in the conclusion 27 Comparison The installation process for P4 and SVN are both quite easy You only need to install o
162. multimedia like for instance game programming Hard Disk Drive Solid State Disk A type of HDD Operating System Condition of Satisfaction An error in the programming code Integrated development environment 52 3 Document History Version Description Date 1 First version created 24 05 2012 1 5 Updates after document review 28 05 2012 Page 164 of 194 Project FHaSS Test log 2012 53 Table of contents 52 General Document Information 164 52 1 Declaration of intentio rettet re eret reet reet tei vete ve eR ente tee vie avus 164 52 2 Definitions and Acronyms sssssesesseeeeeeeen nennen enne nnne nent eE ieai iiaii 164 KE RR le elle Tele 164 53 Table Of contents etre etre etie irent eee ee intr kaken eda fine 165 54 List of validation tests completed rrrrnnrrrnrnnnrvrrnonarvrnnnnnrvrnnnnnrvrnnnnarvrnnnnavnnsnnnarsesnnnanvesnnnarsnsennnns 166 55 Logs for validation tests performed sss nnne nennen 166 55 1 Story 3 Game TEs cdessesaccessesicccsszcnceesscdachessasacecssedactestaaaces E A kes 167 55 22 Sto Senere 167 55 3 Story 5 CPU prlottmatlobt ees ien ere Zeg Ze uege e degen Eege deed e eee REEL eo UNES 168 55 4 Story 6 Graphics card information eese esent nennt nns nnns 170 55 5 Story 6 7 Hardware type management and grouping rrranrranvrrnvrrnnrrnnvrnnvrnnvnranrrrnsnrnsrnnsrnnn 170 55 6
163. n hard to work with and far from intuitive and we had to make many workarounds to make this work All the project members had previous experience with writing SQL database queries and handling the results of such The transition to writing reading and thinking in the way of NHibernate Criteria Queries was nothing but painful And so was the adaptation of having to write out the mapping database table to object mapping XML documents A stream of errors that felt endless was hard to debug and even harder to fix was awaiting the project members when we started out on the development phase of the project 62 Filter saving To save some time on inputting commonly used filter combinations we created default saved filters for The Secret world Age of Conan and a filter combination of both games Aswell as access to these filters the user can also save their own combinations of filter parameters for a quicker and easier hardware monitoring experience The filters are saved on the database so that they will show up for all users accessing the system This was implemented purely for the reason that we want this to be a quick to use system where the users can get specified information in no time In addition to this we wanted to implement a dual view functionality e g to view a crash comparison of The Secret World and The Age of Conan Hyborian Adventure This was a low priority extension however so sadly this did not get implemented the way we
164. n solution A workaround solution would be to use a specific browser desired by the user 61 3 Human risks Page 94 of 194 Project FHaSS Risk Analysis Document 2012 R3 1 Broken NDA Leaked data Probability If one of our project group s members breaks our written NDA by disclosing Description information about FunCom s current system and their upcoming game The Secret World Project member will be unable to work for FunCom anymore not wanted on the Consequence project group for breaking our trust Keeping both our physical and digital information away from prying eyes by using Prevention password protected systems and common sense Not discussing or leak information about TSW to outsiders Possible We may be able to change the projects scope in order to accommodate the solution missing group member R3 2 Miscommunication Probability pe Problems often arise from misunderstandings and miscommunication between P people we would like to minimize these The consequence may vary depending on the matter and the misunderstanding Consequence The impact ranges from missing a workday and undesired design to extra work All resulting in lost time The user requirements and our product backlog prevent any misunderstandings towards the project requirements Minutes from meetings helps us minimize the GE chance of misinterpreting the new information Skype chatting with both our client and internal supe
165. n the intervals as the y axis value As we improved the fetching of data from the database we wanted the graphs to yield more precise information thus we made an x axis point for each single day of the filter period Although this yields better information over short periods of time 1 4 months it gets a bit overpopulated when they request data spanning over vast time periods However the visual information given by the graphs are not particularly affected by this If the user requests a date in the future a mathematical formula will calculate the predicted entries based on the past entries We didn t get time to implement a way to indicate the transition to a future date in the graphs since there was a problem extracting the point positions after the graph was rendered A fix for this is possible but alas we did not have the time to fully implement this since other known issues had a higher priority The tooltip also has a minor problem where it won t show the dates if the x point is an empty string this is solely because HighCharts require that the series data and the x axis categories arrays are of same length so we had to print empty strings in order not to overlap the dates printed below the graphs In the end the graphs turned out quite nicely after a bit of trouble getting it dynamic and finely tuned to fit into the rest of the design and layout of the FHaSS system Page 190 of 194 Project FHaSS Reflection Document 2012
166. nd predict the future evolution Test date 22 05 2012 Kim Richard Johansen WSS ie palsjuerere Function test black box Test approach 1 Put in the amount of time into the future the system is to predict the hardware evolution for 2 Load the web page containing one of the possible charts containing future hardware prediction 3 Toggle the chart to display the future hardware prediction chart lines if toggled off 4 Check that the hardware evolution prediction lines are proportional 5 Compare the updated percentages in proximity to the chart to the manually calculated percentages This stage corresponds to testing the last time point on the chart to the manually calculated percentages for the whole test data set for the whole time period put in 1 Test criteria e Database and website is up and running e Test data set has been made and inserted into database Expected results e That the hardware evolution prediction lines are proportional e The percentages the system produces and displays on the web page based on the test data set are similar to those calculated manually Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors e Encountered errors Outcome What happened Page 71 of 194 Project FHaSS Test Specification 2012 47 3 Verification tests 47 3 1 Code Review All developed code shall be reviewed by an
167. ndre Bjerkerud Sverre Christoffer Thune Kent Brian Dreyer Kim Richard Johansen Project FHaSS Test Strategy 2012 Test Strategy 28 General Document Information Deliverable nr D1 5 2 Deliverable type Report Release Public Work package 1 Responsible Kim Richard Johansen 28 1 Declaration of intention The intention of this document is to give the reader insight into how the group is planning to test the system they are to produce It contains testing strategy different testing methods and types we are going to use and how we plan to document our performed tests 28 2 Definitions and Acronyms Bug An error in the programming code API Application Programming Interface 28 3 Document History Version Description Date 1 First version created 15 12 2011 2 Document template added 04 01 2012 Information reviewed and updated 2 5 Updates after document review 29 05 2012 Page 36 of 194 Project FHaSS Test Strategy 2012 29 Table of contents 28 General Document Information rnannrononnvnnvnrnnnnrrnnnnrnnrnvnnnnrrnennrnnennnnserresnnranennensenvnesnravennrasenneeeenn 36 28 1 D clarationof intention ite ie eene Re t Peste tee dere tie ive ese Ova Ene ge 36 28 2 Definitions and Acronyms sesssssssseeeeeee nnne enne nnns nn ennt nasa siis esse tia assa sss senseri inan nns 36 pr WE sdnsabseas cM EUER 36 29 Table of co
168. ne 83 ELEM UI ICnIrmee EEEERRR 83 54 1 Milestones accomplisbment rietara deanei iaa 84 55 Gantt diagBEalflc suiit ioter eani AO vied deat TENO RE saker RA KR RENE ETE pRRKUE 84 55 1 Remake of the Gantt diagram arorrrrrnnnnnnnnnrnrvrrrrrrnnnnnnrnrrnnrrrsnrnnnnrnnnrnssrsnnnnnnrnnnrsssrannnnnrsnnsnee 84 creen 85 1 ME de E 85 56 1 1 Meetings with internal supervisor ennemis nnn 85 56 1 2 Meetings with external supervlsgor eene enne ener nnn enn 85 56 1 2 1 Sprint planning ENT 85 56 1 2 2 Sprint review meeting ENEE EEN 85 56 1 2 3 Sprint retrospective meeting 85 56 2 Updates to the project plan i rte ient ertt ee aia se tke rea cL cere ak ene Pan dee de aie deo 85 5O63 E Tele ee EE 86 57 JAtt enments vamser neser senere 86 Page 77 of 194 Project FHaSS Project Plan 2012 51 Resources In the following are the different resources that are needed to complete this project As you will see there are no resources needed for this project that aren t available at all times that is all times that people are working on the project Due to this fact there is no need for a resource plan defining when certain time critical resources can be used by whom what and so forth 51 1 Personnel The project team is the main resource needed to get any project work done at all The team members will be working eight 8 hours a day three days a week until Easter and
169. ne plug in for each of the version control systems but by also installing the P4 GUI client you get some extra features For example being able to see if there are newer revisions of files on the server just by pressing refresh F5 is one of these features that are very useful With SVN this is not available Adding new files checking out and checking in files in a project are easy with both systems within Visual Studio it is done almost identically You can also use the P4 GUI client Both systems have tools to merge a file edited by two different users The merge tool in P4 works better than the merge tool in SVN In P4 you can simultaneously view the current revision on the server the two new revisions from the users and the automatically merged version Changes are easily done and saved if needed The tool with SVN works alright but not as good as P4 The new code is merged however some extra lines are added that show which user that part of the code was from these you need to remove manually before committing the change With SVN you also don t have the nice simultaneous view of all the revisions Page 129 of 194 Project FHaSS Version Control Document 2012 Viewing the changes between revisions of files is easy in both systems but P4 does have better graphical representation of the changes The lines highlighting the changes are linked between the views for the different revisions Reverting to an older revision works we
170. neous Python files assesses eee nnne 31 24 3 8 Short rundown on remaining python files esses eene 32 244 The GUI server tree rates ter retten es recte EE 32 SEELEN 32 26 Sources amp Referellces ceteri en exta aaae K aia aaae KAER raaa e eiaa Taa aae aaa REEE anae Keia E 32 PE EM uro apice 32 Page 28 of 194 Project FHaSS Current System Description 2011 24 The current system 24 1 Introduction Funcoms current bug report system consists of three servers the batchserver the internal server and the GUI server The first two are important for us to understand because we will need to integrate our own system to co exist with these The third we will completely replace This document contains descriptions of the current system which will be a good reference for ourselves when we are designing and implementing our system Therefore it is quite technical We describe the current code with and then take notice of some of the important things we will need to take into account 24 2 The Batchserver The Batchserver is what we call the server which receives the bugs from the players It is located outside of the firewall of Funcom s network It is responsible for creating batches a collection of bug files which can then be downloaded to a server on the internal Funcom network A php file named bugsubmit receives the bug report files the XML files from the the p
171. ng The actual filtering criteria can be given as user input through the web page interface And the resulting filtered statistics returned and shown through the same interface For the users to be able to filter based on both geographical location time period game title and game server each of these criteria must have their own possible input values 20 Benefits To both get rid of unknown hardware types and group almost identical hardware types as a single type will make the statistics a lot more informative and complete which in turn gives the employees more and better hardware statistics of players of Funcom games In other words the system is enhanced with regard to what was the purpose of the system in the first place Improved GUI will in the first place of course give the users a more pleasant system to work with The statistics charts will put more information into less space and graphs can also be read fast and efficient without having to go into a very detailed level The detailed level is what the tables are for With the new filtering functionality an employee can get more accurate and precise information as well as the more general information that the current system delivers He she can now for instance get statistical information about the hardware of Korean players of Age of Conan in 2010 playing ona specific server as well as see the general statistics of all players ever reported a crash to the system This functionality
172. ng for the report is that many of the tests are a part of the same module The reports therefore create a good overview over all the tests and summarize them into modules Currently there is just one module but as we work on the project extensions e g more requirements may be added and from there more modules can be added Modules e Web functionality 35 Responsibility There are three roles to be filled when it comes to testing e Test manager e Test analyst e Tester In smaller projects it s most common for the roles Test manager and Test analyst to be filled by the same person The Tester role is common to be filled by the whole group This is the case in our project Page 41 of 194 Project FHaSS Test Strategy 2012 35 1 Test manager A test manager has the main responsibility for testing e Ensure that tests are planned and the requirements are arranged e Ensure that all test reports are made e Keep all test documentation up to date and organized e Define what test strategies to be used e Verify test results with the rest of the group and the employer 35 2 Test analyst The main responsibility for a test analyst is e identify test elements e Describe test progression e Evaluate the quality of a unit tested 35 3 Tester The tester role is a role every project member can take The tester s tasks are e Follow the test procedure to execute the test e Write a test log with results e Separ
173. ng the market also increase the different amount of components available Take for instance the evolution when it comes to graphics processors and graphics cards on the PC market There are two main producers of graphical processors today nVidia and AMD 2 with their graphics processor series GeForce and Radeon respectively These processors are then bought by graphics cards manufacturers who build them into their own type of graphics card There are far more graphics cards manufacturers than the two graphics processor producers and many of the manufacturers are making graphics card versions based on both of the processor technologies In addition some manufacturers are even making several graphics card types based on the same processor This all sums up in that there is a very large amount of different graphics cards being distributed on the user market An almost similar explanation can be given for other types of computer hardware such as processors main boards and memory as well Taking into account the evolution of the technology this results in a market that is full of different computer components for the end user to buy Because a regular PC consists of several different hardware types this again results in an enormous amount of different combinations of hardware a user potentially can have in his computer Funcom as a software developer has to take this fact into account when developing their games Either specific hardware types or sp
174. none assistance from the project team Some of these deployments has happened after the set milestone date of 19 of May but the project team consider their last major deployment as the final deployment and therefore consider this milestone as accomplished The sixth 6 and last milestone date is yet to come but the project team is on schedule wioth regard to the preparations for the 3 and final project presentation 55 Gantt diagram In the attachment Gantt diagram of this document the Gantt diagram for the project periodis displayed The Gantt diagram lists the requirements to be implemented and when the work on them should start and be finished All the presentations are added to this diagram as well The milestones High Priority Requirements and Middle Priority Requirements are displayed only as the end of their corresponding work packages 55 1 Remake of the Gantt diagram The first version of the Gantt diagram was based on the requirements in the Requirements Specification Because some of those requirements include the functionality described in several user stories we found it hard to decide whether or not a specific requirement was completed or not We decided of this reason to remake the Gantt diagram 31 01 2012 to instead be based on the User Page 84 of 194 Project FHaSS Project Plan 2012 Stories in the Product Backlog This decision and remake has proven itself very useful as it is now a lot easie
175. ns and Acronyms The damage caused to our project Application Programming Interface Non Disclosure Agreement The Secret World Identity Integrated development environment When a program unexpectedly stops running Perforce version control software Subversion version control software Revision A version of a file in the files history Check out A user checks out a copy of a file from the server which he can edit Check in A user checks in an edited copy of a file so the server so the server has the latest revision Merge Combine the changes from two users on the same file User A developer in the team 20 3 Document History Version Description Date 1 First version created 16 12 2011 1 1 Supervisor suggestions integrated 03 01 2012 1 5 Updates after document review 29 05 2012 Page 125 of 194 Project FHaSS Version Control Document 2012 21 Table of contents 20 General Document Information 125 20 1 Declaration of Intention ertet temer tetendit ur Petite ene vae tentent eer 125 20 2 Definitions and Acronvms sssini eied aaea aE E 125 203 reels c 125 21 Table of contents eerte ime epe Ee e Y eed etr on ve ee Eee E ee ee eese Ree pea te ge pvs 126 Puerum 127 23 Iis TUE 127 NM IIR 127 25 SUBVERSION im 128 26 Other
176. nt 2012 La 1 1 GetHardwarelnfoList zs 1 2 GetColumnEntries For each Hardwarelnfo entry Ll 1 2 1 FilterCPUManufacturer 1 2 2 FilterCPUSpeed P 1 2 3 FilterCPUCores 1 2 4 FilterGPUManufacturer 1 2 5 FilterVRam la 2 InsertintoDB Figure 16 GPU methods added to HWParser sequence The sequence diagram above figure 12 shows how the GPU methods in the sequence for the HWParser and also some changes that were done to the design during implementation of the previous user story The best way to actually transfer the GPU information to the new tables will be to delete all the previous entries and run the parser through fully once more 449 N Figure 17 GPU info added to database Page 147 of 194 Project FHaSS Design Document 2012 Above figure 13 displays the additions to the database to store the GPU information There is a new table that will contain the existing GPU manufacturers and the hardware table have a new foreign key and vram column ReadData Service GetGamesList string D GetGameForUniverse string GetUniversesForGame g ameTitle string string D SetGameForUniverse universe string gameTitle string void UpdateUniverses void FetchFromDb universes string regions string dateFrom int dateTo int FetchResult I GroupCpuS peeds hardwarelnfoList IListsHardwarelnfo gt FetchResultPart GroupCpuCores hardwa
177. ntents e bett ree e lekrere 37 30 Requirements for LTE 38 31 T eStstrateBy uode eere pe PE iim cases stems Pe E 38 32 Testing Categorles 5 etse esca teet tod ter ote seek sje Enden ee Eee e RENE caches vaeataea Eve den aaRS 39 32 1 Verification and validation sess enne enne nnne nennen nennt nnne eaea AEE 39 32 2 Functional and non functional testing ccccccccccessssssssseeececessesecseseeeeecessesssaeaeeeeeeessesseaeees 39 32 3 Static and dynamictesting e nettes o is tese tact e cor aequos e tene diete buena 39 MEME le siis 39 33 1 ge TE 39 SS LT Bl ck boxtesting sasoian aiir aara E TETA OE TEOTEO RT 39 33 12 WMhite boxXx testirig ode oe etes eost see onte Eed dote eo nua edet s 39 33 2 Meth dS risau 40 33 21 delt Ir LOS T iie teet rode O O OA OAO OTO 40 3322 Code Compilations iei ii octo aiios e de iei dede esc eaae nee edet 40 3323 TE 40 33 24 TEE 40 33 2 5 USEF lnterfaCe ee erret eire aai 40 33 2 6 P LECCE 40 3327 JStralni test oue ee ERE Rn A nd i e UE 40 33 2 8 R gressjontestuu jur erre ER Ere isdans 40 33 29 Performance testari cas2eeened idaen 40 34 Testdocumentatioh wiar ends ees Reese eeu E Ge ane dede Dv ea eg edu e eek E Uns sandete 41 34 1 Test SPECIFICATION EE 41 342 EE 41 343 En
178. ntil the whole string list is filtered The parser will run in a console application containing a Main method with a loop that starts RunParser after a certain amount of time has passed This console application will therefore need a timer to keep track of how much time has passed so it will know if the RunParser method should be executed 34 2 1 The Parser interface Since the parsing and grouping of tables happen automatically as data is entered into the system or each time an administrator changes the grouping rules there is the need for an interface where it is possible to check the parser and grouping threads status due to their complex nature and long runtime Due to the nature of the current database structure and size we need to create a user friendly interface to handle the separate parsing thread The interface will provide basic functions to the user of the system so that they can control the parser and view parsing statistics without having access to the underlying systems Page 140 of 194 Project FHaSS Design Document 2012 To achieve this the parser will be controlled by the help of an admin web page where the user can start and stop the parser while reviewing the parsing status The parser is design to run its own thread and terminate gracefully when the user so chooses The only communication between the parser and the web page is by the model and the model will call methods in the parser to achieve its goals The MV
179. ntries than the database This could be implemented so the user could be given a message of this when viewing the event log on the parser interface page under administration without having to check this manually 67 8 Recommended hardware software to look into these need attention The group has come up with an extension of the system that could be helpful for Funcom employees using the FHaSS system Another list could be added to the statistical table showing the most frequent hardware software found in the database This could identify the source of most recent crashes This extension should also have the possibility to remove entries in the list if the hardware software is not the cause of the bug but is just often used in computers that experience crashes or that the specific hardware software issue has been resolved Page 194 of 194
180. o database Expected e That the produced filtered percentages based on the test data set are results similar to the manually calculated percentages for the tested filter combination Test creator Kim Richard Johansen 24 04 2012 Errors e None Outcome Combinations tested Each country and combination of countries from different regions Each region and combination of different regions Mapping interface All worked successfully Test failure Page 175 of 194 Project FHaSS Test log 2012 55 9 Story 11 Time period filter Filter statistics time period filter est ID T R 3 5 ire Pias A F 3 Uae Ed aielk Test to see if the time filter combinations are working correctly e Test if the data shown on the website corresponds to the manually calculated results after setting a time filter Test date 27 04 2012 Kim Richard Johansen WSS ie gleereles Function test black box Test approach 4 Setthe filter combination to be tested 5 Wait for the web page containing the filtered statistics to load 6 Compare the produced filtered percentage statistics to the manually calculated percentages Test criteria e Database and website is up and running e Test dataset has been made and inserted into database Expected e That the produced filtered percentages based on the test data set are results similar to the manually calculated percentages for the tested filter
181. of testing 33 2 3 Debugging This is a testing method done with the API to find out what part of the code that fails when a code compilation fails When a verification test fails and a code compilation is successful debugging can be used to go through the code step by step to indentify the error by examining the values of variables 33 2 4 Code Review A person just reading the code is a form of static testing He should not be the same person that wrote the code 33 2 5 User Interface This type of testing is about confirming that the user get the correct response based on his actions 33 2 6 Ad Hoc test It s a test commonly done after large system implementations and late in the project where the tester goal is to break the system by trying different system functionality This test is done without any formal test plan This test can also include negative testing 33 2 7 Strain test The purpose of the strain test is to give the system variable amounts of data e g bug reports over a lengthy time to see how the system handles it While the variable data is given to the system we will use the website to load show results from the database The point behind this test is to see how the system handles operates during what is looked at as normal phase e g after bugs is fixed and software crashes has reduced and a high pressure phase e g after large software implementations game extensions where they may be a
182. on Deliverable nr D1 4 1 1 DAMIEN ien Reference Release Public Workpackage 1 Responsible Kent B Dreyer Sverre C Thune 22 1 Declaration of intention The purpose of this document is to help us understand how the current bug report system at Funcom works This will help us integrate our own system into the environment at Funcom 22 2 Definitions and Acronyms Secure File Transfer Protocol Graphical User Interface Tape Archive File DirectX Diagnostics A collection of files 22 3 Document History Version Description Date 1 First version created 07 12 2011 1 5 Renamed document from 10 01 2012 Current System Research to Current System Description Updated structure 2 Updates after document review 29 05 2012 Page 27 of 194 Project FHaSS Current System Description 2011 23 Table of contents 22 General Document Information 27 221 D claration of intention reete ie eee teneo t Pese tree tie vester ren Eng 27 22 2 Definitions and Acronvms eene nennen enne a intense ai a 27 22 33 Document History avsette keen 27 KC SET E 28 24 The current Syste Messeniens NENN NEES EAE 29 24 1 e de DE 29 24 2 The B tehs erveruvsarseurae eter E EAEE EEE REEE Eege 29 24 3 The lnteral Senger E 29 24 3 1 General description of python les 29 24 3 2 Miscella
183. or removing EEN them from the group permanently grown up constructing way and have teambuilding sessions If an internal conflict occurs it is important to know that you can still keep things at a professional level between co workers and not necessarily dragging personal feelings into the workplace Possible solution R3 7 Disencouragement and workload It is easy to lose morale and be discouraged if you put a lot of work on yourself for an extended period of time this may cause burnout But on the other hand it is quite easy to deliver uncompleted work and miss deadlines if you take your work too lightly If a member of our group experiences burnout during the course of this project and renders himself discouraged from working we as a group need to use time Consequence and energy nursing our member into working normally By taking our work too lightly we risk missing deadlines and delivering sub par products Page 96 of 194 Description Project FHaSS Risk Analysis Document 2012 Discussing your workload with the project leader teambuilding and celebrating Prevention e achievements R3 8 Major project changes additions The consequences may vary depending on the addition but in large cases where Consequence new systems should be implemented it may mean an extra set of documentation omme in addition to the actual system development and supervisors to get an insight to their future plans If a major addit
184. or subset of them in the next iteration release Throughout the iteration sprint the stakeholders can adjust the plan by adding removing or editing requirements XP is known for its practice of Pair Programming where two people is programming on the same computer together That is one who is writing the actual code and the other who is ensuring correctness of the code and both taking part in the discussion of how to code XP is also known for its frequent and in depth testing throughout the development period 6 Conclusion There are mainly two forces pulling on us in a direction of which project model to use so the decision is not actually one of the group to make H gskolen I Buskerud HiBU requires of us to deliver certain documents at certain points in the project period Many of the documents are typical deliverables in the standard Unified Process pre face Inception and Elaboration phases but are documents that not usually are produced when Page 105 of 194 Project FHaSS Project Model Document 2012 following the agile models Scrum or Extreme Programming The agile models are a lot more straight to the point of development whereas UP is starting up with a thorough pre face before the actual development starts So in one way our school requires of us to more or less follow the practices of UP which isn t a factor we can influence On the other side we ve got our client Funcom Game development companies in general
185. ory caching 61 1 3 1 Single SQL query Up until this point the WCF web service interacting with the database had done one database query per hardware category per day As a result a lot of queries had to be done for each user request By following the test and fail procedure the project team managed to put together more sophisticated SQL queries that retrieved all the information for all days for one hardware category The load time improvement the system saw after this enhancement was enormous compared to the amount of work hours that had been used for the implementation of it 61 1 3 2 Memory cache There was a lot of back and forth regarding the load issue but just needed the users to accept the fact that we are dealing with a lot of information here this is a heavy system So fetching all the entries in the database 4 000 000 entries takes about 40 seconds which is far from the original requirement but thinking about it how often does the users need to view this long term information This is a quick analysis system to check on how the latest patches are working out and what types of hardware that causes the most significant crashes that s why we came up with the caching solution The caching was implemented by fetching every single entry for the last month and storing them in the program memory as a static variable and because the entries were fetched using nHibernate it automatically created objects and groups of that fetch for us
186. osoft Network Messenger Workpackage Exempli gratia for example FunCom Hardware Statistics System 12 3 Document History Version Description Date 1 First version created 20 12 2011 1 5 Peer reviewed 15 03 2012 2 Updated 19 03 2012 3 Document finalization 29 05 2012 3 5 Document reviewed 28 05 2012 Page 113 of 194 Project FHaSS Quality Assurance Document 2011 13 Table of contents 12 General Document Information 113 12 1 Declaration of intentiOh edo rente ten Lr ebrei rene vet eee ke aeu e see eene ee pee venue 113 12 2 Definitions and Acronvms ssiseirecssoecerinedresiir decoderi derininde iae daiane da aiiai aaee 113 123 Document HISLOTFV incite eoe iio Eee docs giants Pee eea p aeae Ea ee sien totes eet spams eo ed Erei 113 13 Table e LEE 114 14 reete 116 15 Communication and Information Exchange 116 15 1 Communication oerte ote teer e cere ote Rte orbe duces sedcdeeageatececeeacbeseataiee 116 15 13 E mail guidelitiges idee Seege EENS 116 15 1 2 Regular supervisor meetings eese esee nennen enne enne nennen nennen nnns 116 15 13 Group Communication 117 15 2 Information Exchange File sharing 117 15 21 Project files folder Dropbox nette mee cete reos ee 117 16 Deliverable g ideliries 2 teneor onere tene tna ete rane deeg Eet een ie Deren excea
187. other person than the one who developed it The organization of code reviews will be as follows The code developer notifies the Test Manager Kim R Johansen when the code development has been completed It is then up to the Test Manager to delegate the task of reviewing the code to another team member The review shall result in a code review feedback document for the code developer to examine The feedback document shall consist of comments and suggestions for improvements of the overall code or parts of the code Code parts shall be identified using line numbers It is up to the code developer to consider whether or not the comments and or suggestions shall be taken into account implemented Both part can and shall request a short meeting discussion when disagreements arise about the developed code A third part can be brought into the meeting discussion to further investigate the code 47 3 2 Code Compilation It is up to each code developer to ensure that the code he has produced compiles without errors or warnings It is natural that compilation will be attempted regularly throughout the code development process and that the source of eventual errors warnings is corrected before resuming the code development This is to ensure that faults in the code are fixed as early in the development process as possible which in turn will save development time because of the exponential growth of time needed to fix errors as development continues
188. oves the files to an archive when it is finished with them See attachment 3 for a more detailed flow of events fcBugMain Python The function ProcessFile runs the bugParser function which returns a parsed bug report object This object is processed further and certain properties are updated if some conditions are true Our client informs that this bug report system functions well thus should initially not be changed After the bug report object has been processed fcBugDXDiag function is called and the DxDiag file which belongs to the bug report is parsed After the DxDiag file is parsed the bug report object is sendt to Bugzilla which is a third party bug tracking server software However if the object has the type DxDiag it is not sent The type of the bug report object is the category of the bug originally parsed from the XML file Under certain conditions extra emails are sent to specific developers so that they get an alert when a certain bug occur When we are going to develop our own parser for the DxDiag file changes will be needed here We will want to keep the bugParsing but use our own DXDiag parser We might even use the already exsisting DxDiag parser but then parse again to get better control on and more accurate information Page 30 of 194 Project FHaSS Current System Description 2011 fcBugParser Python attachment 4 Here we have the code responsible for parsing the XML files and creating bug
189. p www cio com article 197152 PHP s Enterprise Strengths and Weaknesses Take 2 pa e 3 amp taxonomyld 3038 Last visited 15 12 2011 e http www cio com article 176250 You Used PHP to Write WHAT Last visited 12 12 2011 e http www creativewebmall com cwm php strengths and weaknesses 444 Last visited 15 12 2011 e http coding smashingmagazine com 2009 02 11 the big php ides test why use oneand which to choose Last visited 14 12 2011 Page 14 of 194 Project FHaSS Programming Language Document 2011 13 2 C ASP NET e http en wikipedia org wiki C Sharp programming language Last visited 14 12 2011 e http www mono project com Main Page Last visited 14 12 2011 e http msdn microsoft com en us library z1zx9t92 v vs 80 aspx Last visited 14 12 2011 e http en wikipedia org wiki Common Language Runtime Last visited 15 12 2011 e http en wikipedia org wiki ASP NET Last visited 15 12 2011 e http en wikipedia org wiki List of CLI languages Last visited 15 12 2011 e http wiki answers com Q What are the advantages and disadvantages of aspnet Last visited 15 12 2011 e http msdn microsoft com en us library k4cbh4dh aspxHY10154 Last visited 14 12 2011 13 3 Python e Mhttp wiki python org moin PythonVsPhp Last visited 11 12 2011 e http python org doc Last visited 11 12 2011 e http en wikipedia org wiki Python programming language Last visited 11 12 2011 Page 15 of 194 9
190. percentages for the tested filter Page 67 of 194 Project FHaSS Test Specification 2012 Test creator Errors Outcome combination Kim Richard Johansen 31 01 2012 e Encountered errors What happened Requirement ID Test description Test date Testing method Test approach Test criteria Expected results Test creator Errors Outcome Filter statistics time period filter est ID T R 3 5 A F 3 Test to see if the time filter combinations are working correctly e Test if the data shown on the website corresponds to the manually calculated results after setting a time filter 27 04 2012 Kim Richard Johansen Function test black box 1 Set the filter combination to be tested 2 Wait for the web page containing the filtered statistics to load 3 Compare the produced filtered percentage statistics to the manually calculated percentages e Database and website is up and running e Test dataset has been made and inserted into database e That the produced filtered percentages based on the test data set are similar to the manually calculated percentages for the tested filter combination Kim Richard Johansen 24 04 2012 e Encountered errors What happened Requirement ID Test description Filter statistics geographical location filter est ID T R 3 6 A F 3 Test to see if
191. project every file that belong to the project is stored on this server A user a developer in the team can then connect to the server and check out the files he is going to work on When the user is finished working on the file he needs to check in the files he edited so that the new revision is stored on the server When a file is checked in it becomes a part of the current revision of the project Before it is checked in it is just an unfinished temporary copy of the file There are some criteria for checking in files that should be followed The most important is that the project should always be able to build on the client before checking in updated files If you check ina file that makes the project unable to build this can lead to everyone working on the project to be not able to build Never check in files that does this Also every check in should have a comment attached that shortly describes what the changes are 24 Perforce Perforce P4 is a version control system developed by Perforce Software You can use P4 for free with either two users and five workspaces or unlimited users but up to 1000 files Installing the P4 server is very easy you only need to download the installation file and run it You need to make sure that the clients can connect through the firewall on the server if there is one The P4 GUI client is also as easy to install When the client is installed you just run it and connect to the server with the I
192. quirements It is the responsibility of the Product Owner in our case this is the external supervisor to add so called user stories to the PB User stories are nothing more than user requirements formulated in a specific way We have converted the user stories from the PB into ordinary requirements for this document Most of the requirements in the PB are the same as those captured in the meetings as described in the section above 39 3 Technical Research A Technical Research with a Technical Document as the result has been used to confirm that some of the user requirements actually are the best possible choices for the given system 39 4 Courses with Lorenzo Aurea Lorenzo Aurea is an employee in Funcom and also one of the group s unofficial supervisors in this project In the start of the development implementation period of the project after New Year Lorenzo organized two courses for the group to ease the many problems we most probably would face with the technology we were to start using In these courses some of the already defined requirements in the PB were further explained in a bit more detail and on a more technical level and some new tweaks to them were introduced Those tweaks has effected some of the requirements defined in this document 40 Changes in requirements The Scrum project model handles the problem of fluid requirements by letting the Product Backlog PB be fluid It is up to the Product Owner to do chan
193. r ede e REIS Sagra berede FCR RE ao ONE UR 75 50 Table of Contents orte Ree eee evene e iv vem ete Re dei aar ve xi a 76 D1 ROSON ES nos up e OR RT n Eu Oo eee 78 51 1 Personer 78 512 HardWare saa 78 A Wee 78 51 4 eg uge Te 78 Meetings amp Planning i cett anos tetti rasan eed xe be da One pl xe V eaten Rana 78 Documentation te tette tede cena eeneg a aor es 79 LEE CECI UEM STER S 79 e implementatie rarasan iTA EEA EE AVANE AEA EE ees 79 H DET EE 79 LEMEEI S 79 LEE en NEE 79 EE 79 51 5 Actualr source Usage dete coe ce To reet e epe eec Ce e o Ee ER ere d se Dru 80 51 5 1 en EE 80 51 5 1 1 September October and November 80 51 5 1 2 December EE EE ME M 80 51 5 1 3 HE UC Le E 80 51 5 1 4 Feasa 80 51 5 1 5 Marcussen 81 BCE dp LE 81 51 5 1 7 MD OO 81 51 5 1 8 JUNG 81 51 5 2 For the whole project period coerente n eee n rhon deae aa Ewa e Yan 81 Page 76 of 194 Project FHaSS Project Plan 2012 52 RESPOnsibility Sraa RO EOE OOO ENNEN EENS 82 52 1 Fields of responsibility ce title iege Eaa eae Ee eere unt dee od eheu eene 82 52 2 Scrum team responsibilities esseesses sese eee nnns enne 82 53 Requirements breakdown ie ene tee chari ena rn e nane aav aede aae ENNEN Ed 83 53 1 Remake of the Requirements breakdown rrrnrannrrvnrannrvvnrensrvenssnsrrenssnsrrvnssnsvrenssssrrenssnnnnvnss
194. r system Steam Hardware Survey in terms of functionality and appearance and from that a few key elements were quickly in place like the evolution charts hardware statistics table header and footer Our distinguished feature is the filter menu which can in some situations take a lot of room while it still needs to be readily available on the page In addition to this we need to display meta data information about the selected entries to the user as well as figure out how the menu charts and table should behave from user input Page 143 of 194 Project FHaSS Design Document 2012 Games The Secret World 7 Tswserver1 Tswserver2 Tswserver3 Age of Conan H Aocserver1 Aocserver2 Sort by O A Z Windows 7 64 bit H Ei Windows 7 32 bit 1Ghz to 2Ghz 2Ghz to 3Ghz 3Ghz to 4Ghz ipsum dolor sit amet consectetur adipisicing elit sed do eiusmod tempor N incididunt ut labore et dolore magna aliqua Ut enim ad minim veniam quis CO nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur Excepteur sint occaecat cupidatat non Figure 13 Early page design In the earliest design it became clear that the filtermenu felt most natural on the left hand side of the page it also seemed natural that the meta information about the entries where the data comes from in terms
195. r to gauge the progress of the project with regard to the diagram 56 Conduction 56 1 Meetings 56 1 1 Meetings with internal supervisor Meetings between the project team and the internal supervisor AurillaAurelieArntzen will occur on a regular basis throughout the project period Each week on the same day and time of the week a follow up meeting will be held between the two parts A follow up document is to be delivered to the internal supervisor at last 24 hours before the start of this meeting The contents of these meetings will be a review of the follow up document and other project related subjects that are of interest to that should be discussed with the internal supervisor 56 1 2 Meetings with external supervisor Following the Scrum project model the project team and the Product Owner shall arrange meetings before and after each sprint All of the three following types of meetings will generally be held as one single meeting separated into three parts at the start of a new sprint That means there will only be a sprint planning meeting at the start of the first sprint 56 1 2 1 Sprint planning meeting The purpose of this meeting is to decide on which user stories requirements or which parts of them the project team shall be working on in the following sprint It is mainly up to the Product Owner to decide which user stories requirements the team shall focus on but it is solely up to the team themselves to accept what to
196. re notified of these bugs and when resolved they were removed from the list 57 6 Performance test Has been done regularly to keep track of loading times for the different web pages of our system and the time our parser and grouper uses average per workload All in all a big help for improving the performance of the code and finding new ways to do the same task but with a shorter loading time Page 182 of 194 H GSKOLEN i Buskerud Funcom Hardware Statistics System Project Reflection Document Funcom Hardware Statistics System Client Funcom N V Acronym Sensors amp Supervisors Internal sensor Olaf Hallan Graven 5 29 2012 All members Internal supervisor Aurilla AurelieArntzen External supervisor Rui M Monteiro Casais Sondre Bjerkerud Dag Hem Sverre Christoffer Thune Kent Brian Dreyer Kim Richard Johansen Project FHaSS Reflection Document Reflection Document 58 General Document Information 2012 Deliverable nr D8 3 1 DEM Er CHE Report Release Public Workpackage W8 Responsible All members 58 1 Declaration of intention The purpose of this document is to reflect on our work during the development period of our product and present features with intentions as well as what could have been done differently 58 2 Definitions and Acronyms FHaSS JPEG PNG nHibernate NET
197. relnfoList IListsHardwarelnfo gt FetchResultPart GroupCpuManufacturers hardwarelnfoList IListsHardwarelnfo gt FetchResultPart GroupGpuManufacturers hardwarelnfoList I List lt Hardwarelnfo gt FetchResultPart ene GroupV Ram hardwarelnfoList IList lt Hardwarelnfo gt FetchResultPart ieee I gpuManufacturerlD int universe string V ReadDataService svc cs cpuSpeed int nHibernateHelper NHibemateHelper zint H zin ReadDataService Mm ES CpuCores Group G eneral CpuCoresGrouplD int number int ResultParts IList lt FetchResultPart gt sessionFactory ISessionFactory puSp TotalCount int NHibernateHelper E GetSession ISession CloseSessionFactory void Groups IList lt string gt Counts IList lt int gt Figure 18 ReadDataService additions The ReadDataService and the MVC also needs to be updated to include the GPU information figure 14 above First of all there is the mapped classes for the additional tables needs to be created Then we have the new grouping methods for the GPU manufacturer and the Vram which will be similar to the previous grouping methods In the end the MVC also needs to be updated to display the GPU information the same way as for CPU information Also new MVCs needs to be added for administrating the GPU manufacturer and VRam groups so that a user can manually change these 38 Memory and OS user stories For the memory user story we
198. rement A NF 6 38 Table of contents 37 General Document Information 44 37 1 Declaration Of Intentlofis eise e rent teen en entrego Yves ent ia e eye RV aee enne 44 37 2 Definitions and ACrONyMS ssssessseisesssresseeessssressrroivensrrsasseesssisoasrrearsnsrrsi veesi srona Vrsi ren irrar serrare 44 343 Document History e IR 44 38 T bleof contents or reete treo enger teur eu ew oyente rn y ve ene pue ben Ev eye Rv as ete os eR UM 45 39 Requirements Capturing 5 cie tette pee Ne ERE e e ER RSN NAE SERE e HRS NER SERE XN ERROR AUS EECHER duds 47 39 1 Meetings with the external supervisor sesssesseseseseeeren enne ennemi nns 47 39 2 The Product Backlog cditio recense ud den endene 47 89 3 Technical Research eire Dette Pini deci obige cra detinet eee 47 39 4 Courses with Lorenzo Aurea ener enne nnn entente nennen nennen 47 40 Changes irrequiremients d eonun REENEN SEENEN REESEN ES 47 Ad Requiremierits Sto Ae ege dee en tne eagle endr 48 41 1 AR quirements ciere tssceee aceite ct reed ed ied Eder aba zd Ead ee raw en ERE e RENE DEN 48 NN DNR GE Tele Oe IT EE 48 41 12 Non F nctioral eere tete eere a edet ner Vere pene tue x enn nea Quot 48 41 2 B Requitemehts erected ree NOE vedaene Pa need eU p Nee pego eee eee RARE 48 4121 E lee Oe e M 48 41 22 Non Funetional suussgerusaum gjerde demanding 48 41 3 CReq lrements cesi Dto eene rp tni ipee
199. rement specific identification e Date The date the requirement was captured In case the requirement has been captured several times the date of the first capture will be given e Origin From what source was the requirement captured The possibilities are the three mentioned capturing sources described in the section above e Description Self explanatory A F 1 Display hardware software statistics A A F 1 27 10 2011 Meetings amp Product Backlog Display hardware information for players of Funcom games over time in table format in a web page The table shall contain the percentage distribution of the different possible hardware software that players are using The different hardware software categories to be displayed statistics for are a CPU o Speed In MHz o Number of cores o Model b Graphics card o GPU model o VRAM size amount o Primary Display Resolution o Multi Monitor Resolution c RAM size amount d Windows OS In groups by version with 32 and 64 bit versions being separate groups Page 49 of 194 Project FHaSS Requirements Specification 2012 The grouping criteria that is the boundary values that separate on group from another is to be defined put in by the administrator s of the system The time period of hardware data for which the calculations are based upon shall be put in by the user of the system A F 2 Display hardware software evolution cha
200. riche 49 41 3 4 Functional een eR a a Ric ee trusa 49 41 3 2 NoncF nctiorial iori orici ne He tarte see raden desea berikende 49 KE ee ET 49 A F 1 Display hardware software Statistics ccccccccssccsssseessececssecesseeesseeeeseeceessecsseeeesaeeeesaeceeees 49 A F 2 Display hardware software evolution charts rsrrnnrrnnrrnnvnrnvrrnvrrnnrrnnvannrrnnrrnnsrnnnrrnnrsnnnsnnne 50 A F 3 Filter Statistids eegener enee Eeer 50 A F 4 Hardware type management 50 A F 5 Hardware grouping c cccccssccccessececesseeeeceeseneeceseeeeecesceesecseeeeaeceseeesecsseeeeecsseenaeeseeeeaeeseeenaeens 51 A NF 1 Programming language esses nnne nnns sten nnn esee rn nisse nnns nennen 51 Page 45 of 194 Project FHaSS Requirements Specification 2012 A NF 2 Model View Controller ssssessssssssssssssssssrsssrsssrsssrsssesssessserssesssessstsssessstesstesstesstesstesstesstene EN A NF 3 Database EE 51 A NF 4 Database Connection 51 A NF 5 Database access sunne EEGENEN eam ane 51 A NF 6 Web page loading me 52 A NF 7 Wiki pages documentation esses eene enne ennt 52 A NF 8 BugZilla ue ET 52 B F 1 Display hardware network statistics ccscccscesscessssesssesssesescsescsescseecsuecauecsuecsaeseasseaeseaeees 52 B F 2 Display software statletics 52 B NF 1 Windows service riore tien tiere ated RED a wine HEAD eee RENE eee 53 C F 1 Future hardware prediction esses sesenta 53 C F
201. rs should be aware of such as schema changes or getting acquainted with the new functionality After submitting a new build of the project it is advised that at least one other project member immediately test this new version to uncover any bugs that might appear due to discrepancies between their versions if any 18 3 Plugins Each student in the project group can and should take use of two additional plugins at their own leisure to help them conform to the set standards Each student is responsible for their own licenses 18 3 1 StyleCop StyleCop is an open source code analysis tool that helps us use the correct standards when it comes to code naming readability ordering spacing maintainability and layout amongst others ReSharper helps us maintain our code standards and thereby keep the code as easy to read and structured as possible without hampering implementation We have implemented Funcom s StyleCop rules into our own add ons to make sure our codefollows our client s standard Each student using StyleCop is advised to take use of the configuration filelocated in Other Settings StyleCop 18 3 2 ReSharper ReSharper is a refactoring and productivity extension used to point out bad programming practices and better solutions to our code It is a tool that helps us with hints errors and warnings to make sure our code is as quick and clean as possible We have implemented Funcom s ReSharper rules into our own
202. rsion 3 Wait for the chart to load 4 Doas exact a comparison between the manually calculated data and the chart The chart shall be compared to the manually calculated data at five 5 times points with equal distance between them The first and the last times points are the same values that was put in at stage 1 of this test which will be the start and end of the chart lines 5 The percentages displayed in proximity to the chart are compared to the manually calculated percentages for the last time point the chart is describing Test criteria e Database and website is up and running e Test data set has been made and inserted into database Expected e That the chart seems to display similar values to those who have been results calculated manually for each of the five points times of the chart e That the percentages displayed in proximity to the chart based on the test data set are similar to those calculated manually Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors e Encountered errors Outcome What happened Filter statistics est ID T R 3 1 creases A F 3 EEE Test to see if the different filter combinations are working correctly e Test each filter by itself That is for instance putting in values for the time period filter only leaving the other filters with null value the Page 65 of 194 Project FHaSS
203. rsonal Computer Problem Sprint recognized Sprint nr 6 amp 8 Severity EE At the first occurrence a laptop charger broke Second time a team member lost Description their charger Consequence In both occurrences the affected member had to work from home a small Solution amount of time before fixing acquiring new chargers R2 6 System is not viewed the same on different browsers Sprint recognized Segen Severity Discovered some minor visual glitches on the page when using different Description browsers Changed some deprecated css usage while agreeing with the client on to optimize towards a specific browser while maintaining functionality in other browsers Consequence Solution R2 1 Bad API Documentation for nHibernate Sprint recognized n a Severity experience to use and understand it correctly Consequence More time spent learning the software than anticipated Relied heavily on 62 3 Human R3 3 Lack of knowledge Sprint recognized ma Seerty Moderate Page 99 of 194 Project FHaSS Risk Analysis Document 2012 The nature of the project required us to take use of some programming Description languages and standards we were inexperienced with such as C the Net MVC WCF services and nHibernate These challenges did create some small issues during the course of our project In the early stage our inexperience made our analysis and designs somewhat
204. rts A A F 2 27 10 2011 Meetings amp Product Backlog Some of the hardware software information that are displayed in tables A F 1 shall also be displayed as evolution charts The horizontal axis of the charts shall describe time and the vertical axis shall describe the percentage of players using the corresponding hardware software at the given time The different charts to be displayed are a CPU Split between AMD and Intel b GPU Split between nVidia Intel AMD and others c CPU cores Split between 1 2 4 and other number of cores d Windows OS version and DirectX version Split between DX11 GPU amp Windows 7 DX10 GPU amp Windows 7 and DX10 DX11 and Windows XP The current percentage distribution shall also be displayed in proximity to the chart The time period that the charts are showing the evolution for is to be put in by the user of the system A F 3 Filter statistics A A F 2 27 10 2011 Meetings Filter the hardware and software statistics It must be possible to filter by a combination of options from the different criteria The criteria and their options are a Geographical location All the different continents and countries that are registered in the database The countries shall be grouped by their respective continent Time period Date from and to Funcom game The Secret World Age of Conan Pets Vs Monsters d Gameserver The list of servers respecti
205. rver hosting our Description software If there are no servers available to us we will be unable to test the system in a Consequence tar proper and realistic environment Research if either the school or our client can offer lend us the necessary Prevention hardware Possible We could test the system exclusively on our own systems test the system at solution Funcoms location or acquire our own personal hardware R4 2 Relocation of offices hosting Probability Description If we for some reason have to relocate our equipment Relocating would cause both a time loss and other possible Consequence e software network firewall issues S A thorough discussion with the possible parties that can accommodate us such Prevention as HiBu or Funcom to establish whether or not they can meet our requirements If we are unable to find a suitable office we may have to work at one of the Possible project member s house If we are unable to find a suitable hosting solution for solution our internal sever we might be able to set it up in one of the project member s apartments 62 Occurred risks 62 1 Requirement related R1 1 A single A Requirement BugZilla not implemented Sprint recognized N A Severity ven TEC TANE It was an early request made by our client that our system should use their P current Bugzilla to post error reports Consequence i EE With agreement with our client we decided
206. rvisor can help clear up misunderstandings quickly Weekly meeting with our internal supervisor as well as daily internal meetings within the group R3 3 Lack of knowledge By lack of knowledge it is meant both your personal coding skill and your general understanding of the system we are creating Description A lack of knowledge may hamper our teams work pace considerably This is considered a likely probability due to the fact that we will use ASP Net and C as none of us is familiar with it Beyond the slow work pace when working with an unfamiliar language you may also experience a morale decrease Another consequence is our project plans Consequence W may be written on the wrong basis if we have a shallow knowledge of the current system It is each person s own responsibility to keep their coding skill at an acceptable S level To avoid misunderstandings around the system we should use both our Prevention internal and external supervisors to help us gain a better understanding of what we re doing R3 4 Group member is unavailable for an extended amount of time Probability E There are outside forces that may affect our ability to work such as long term Description F opus illness and jail time amongst others Page 95 of 194 Project FHaSS Risk Analysis Document 2012 This will leave the rest of the group with more work while we may not be able to Consequence shorten down the project scope T
207. s2 Chart raw ET 53 eee 53 C E 4 Chart ZOOImiz i aeter rte bete beidet Gade EE 53 EC NF 1 Filter HE 53 43 e deu 54 Page 46 of 194 Project FHaSS Requirements Specification 2012 39 Requirements capturing The requirements has been captured from the following four 4 sources 39 1 Meetings with the external supervisor There have been carried out two major meetings where the assignment has been discussed the first meeting at the Funcom office in Oslo in September and a user requirements capturing meeting in October In the first meeting a brief explanation of the assignment was given for the group to be able to consider accepting it whereas a more in depth requirements capturing interview was done in the second The second meeting was held via Skype and the interview part lasted for about half an hour The group asked a series of open questions regarding details about the assignment and eventual extensions of it Throughout the project period there has before and after each Scrum Sprint been held a set of three meetings with the external supervisor Product Owner In the latter of the three the Sprint Planning meeting there has several times come up completely new requirements or changes or extensions to already existing requirements 39 2 The Product Backlog The Scrum project model is using a feature called the Product Backlog PB for the task of editing organizing and prioritizing re
208. sprint and potentially be shown off for the stakeholders or others In each team there is one ScrumMaster The ScrumMaster s primary role is to address all problems that arise that can influence the optimal progress of the team and potentially result in the team not reaching their sprint goal s The ScrumMaster is not supposed to be a developer 5 Extreme Programming Extreme programming XP as the name hints is a project model that takes programming to the extreme That is to take all the beneficial elements of traditional software development to the extreme on the theory that more is better XP like Scrum is a project model of the agile family and there are of this reason many similarities between the two XP is arranged around four basic activities Listening Designing Implementing and Testing Listening is the activity of listening to the customer and produce a list of requirements which in turn the Designing Implementing and Testing activities are supposed to accomplish With XP the team is working in short iterations typically one week and like Scrum also defines certain releases that is a set of iterations and where the result is a more polished version of the system developed so far At the start of an iteration and or a release a so called Planning Game is done by the team accompanied by the client At this meeting user requirements in the form of user stories are developed and the team is committing to complete a set
209. t date 13 03 2012 Kim Richard Johansen WSS ie galjierele Function test black box Test approach 1 Putin the time period for which the statistical data shall be produced with basis upon 2 Choose to get to see statistical information for GPU 3 Wait for the web page containing the information to load 4 Compare the resulting percentages to the percentages that have been manually calculated from the test data set Test criteria e Database and website is up and running e Test dataset has been made and inserted into database Expected e The percentages the system produces and displays on the web page results based on the test data set are similar to those calculated manually Test creator Kim Richard Johansen 13 03 2011 Errors e Encountered errors Outcome What happened Display hardware software statistics COS information est ID T R 1 4 REG Meueu ES A F 1 and B F 1 EEE Test to see if tables are created and contains accurate information about the different OS information stored in the database It should be possible to toggle between number of entries and COS Show the OS from all systems in the DB and group them under different groups Data to be shown e OS grouped under Windows 7 64 bit Windows 7 32 bit Windows Vista 64 bit Windows Vista 32 bit Windows XP 64 bit Windows XP 32 bit Page 63 of 194 Proj
210. t of the final system completely in each iteration Scrum represents requirements in terms of user stories User stories are small stories that describes a specific thing that the system should do for a specific user and if found necessary why this is User stories are added and prioritized throughout the project period by the Product Owner that is the person that the stakeholders has decided that will represent them towards the Project Team and are stored in a Product Backlog Page 104 of 194 Project FHaSS Project Model Document 2012 A project team working by the Scrum model is divided into several smaller teams consisting of about ten people at max Teams work in time boxed iterations of one to four weeks called sprints where each team is supposed to completely develop one vertical slice of the final project Before every sprint each team picks user stories from the Product Backlog that they commit to fulfill in the upcoming sprint in a sprint planning meeting After each sprint the workflow and problem areas are discussed in a sprint evaluation meeting meant to improve the cooperation and effectiveness of the team A so called daily scrum meeting is a short meeting held each day in a sprint to ensure that the team is on track against their sprint goal and to let potential problems be discovered Several sprints can be gathered and called a release In the end of a release the developed software will be polished even more than after each
211. t rep aeea e e evene esee er voee do ee bl ee venue 102 1 2 Definitions and Acronvms sinici ada a aE 102 1 3 Document TE ee 102 2 T ble of contents eerte eme epe knakende 103 3 Uhified ProC SS EA 104 3 1 CEPTION iss EIE E 104 3 2 lee ET 104 3 3 Veloen EE 104 3 4 Mecum EST 104 Ed la EET 104 5 Extreme ProgrammiFig 2er serere terree aa oue nee ee unen ev e vae un sve eene a Eee Ne reos ee eee Sepe see UR A 105 6 CONCIUSIO 105 Page 103 of 194 Project FHaSS Project Model Document 2012 3 Unified Process The Unified Process UP is an iterative and incremental development process UP is supposed to work in close collaboration with the Unified Modeling Language UML A Use Case Diagram is used for requirements capturing In each iteration the project team is developing a set of use cases entirely that is all the way from requirements capturing to deployment When using UP the team is supposed to address all possible risks as early as possible both in the overall project period and in each iteration The Unified Process consists of four major phases Inception Elaboration Construction and Transition Within each phase especially in the three latter ones it is common to work in iterations 3 1 Inception In this phase the project team is focusing on making major decisions for
212. t this is more practical In the start of the main project period we will only be working three days a week on the project and one full week of work 5 days a week almost equals two weeks of three days a week Therefore sprints will start out to be of two weeks length As a sum up we can say that we will be following a combination of all three of the described project models the preface in accordance to the Unified Process the day to day workflow and practices in accordance to Scrum but incorporating the short iterations common for XP Page 106 of 194 9 HOGSKOLEN i Buskerud Funcom Hardware Statistics System IDE Document ooo Projetname ooo Project name Funcom Hardware Statistics System Client Funcom N V Acronym Sensors amp Supervisors Internal sensor Olaf Hallan Graven Internal supervisor Aurilla Aurelie Arntzen External supervisor Rui M MonteiroCasais 10 01 2012 Author Kent Brian Dreyer Sondre Bjerkerud Sverre Christoffer Thune Kent Brian Dreyer Kim Richard Johansen Project FHaSS IDE document 2011 IDE document 7 General Document Information Deliverable nr D2 4 1 DEM ETER Research Release Public Workpackage 2 Responsible Kent Brian Dreyer 7 1 Declaration of intention The purpose of this document is to review available IDE options and clearly define which IDE will be most beneficial for us to use for this
213. tage 1 of this test which will be the start and end of the chart lines 10 The percentages displayed in proximity to the chart are compared to the manually calculated percentages for the last time point the chart is describing Test criteria e Database and website is up and running e Test dataset has been made and inserted into database Expected e That the chart seems to display similar values to those who have been results calculated manually for each of the five points times of the chart e That the percentages displayed in proximity to the chart based on the test data set are similar to those calculated manually Test creator Sondre Bjerkerud Sverre C Thune 21 12 2011 Updated by Kim Richard Johansen 23 12 2011 Errors e None Outcome The manually calculated results had data from 07 01 2012 to 12 31 2012 set with a time span resolution of 6 The same time span and resolution was chosen in the filtermenu and the site refreshed The charts displayed correctly and the values for each category and time resolution corresponded to the manually calculated results Test Failure Other Filter statistics est ID T R 3 1 ire Pias A F 3 ES Eddie Test to see if the different filter combinations are working correctly e Test each filter by itself That is for instance putting in values for the time period filter only leaving the other filters with null value the standard value e
214. tant The basic elements of the system will have the highest priority and will be tested first Regression tests will be done after code alteration of previous tested functions or when new functions have been added to the system This is to ensure that previously tested functions still work When we are getting closer to a full release and most of our grade A requirements have been completed and tested we will run an ad hoc test This is to find functionality that may or may not be intended that can break the system When this had been completed a strain test will be performed This test simulates different work load for the system and will ensure us that our system may handle a high pressure of crash reports and that no data gets corrupted due to a high workload for the system Page 38 of 194 Project FHaSS Test Strategy 2012 32 Testing categories In the following are short descriptions of different testing categories our tests will be categorized under 32 1 Verification and validation Verification tests of the system will answer if the system is built correctly based on the requirement specification Validation tests will answer if the software created is what the customer wants 32 2 Functional and non functional testing Functional testing is testing that verifies a specific functionality of the system These are usually directly related to the requirement specification Non functional testing tests aspects of
215. tasks However these files are not essential to understand in detail for our first task creating the new GUI therefore we will not go into the details regarding the source code of these files at this time Below is a list of these files with a general description of their responsibility TT Library designed for sending emails automatically regarding bugs or crashes to specific developers fcBugzilla Library for communicating with Bugzilla fcPreforce Contains functions for version control with Preforce This class retrieves and represents data from the database in HTML BugStats This works as the graphical representation of the system on the internal server fcBugStats Library for manipulating the MySQL database fcBugSymbols Library for processing files e g callstack for debug symbols Page 31 of 194 Project FHaSS Current System Description 2011 24 3 3 Short rundown on remaining python files DumpRevision Simply calls upon fcPerforce dumpRevisionList 63101 63102 createTables Simply calls upon the create_tables from the fcBugStats class Contionously prints the number of files in the category subfolders to Reporter the console e Collects and manages map files and sends it to fcBugSymbols for mapFilelmporter parsing 24 4 The GUI server This is the server which runs the website GUI in form of a website for looking at the hardware information extracted from the DxDiag files Hardware information is sele
216. tchResultPart GroupCpuManufacturers hardwarelnfoList IListsHardwarelnfo gt FetchResultPart NHibernateHelp sessionFactory SessionFactory NHibernateHelper GetSession Session CloseSessionFactory void V ReadDataService svc cs nHibernateHelper NHibemateHelper ReadDataService General Figure 12 ReadDataService class diagram After feedback from a Funcom employee we also decided to make the NhibernateHelper non static and instead initiate an instance of it in the ReadDataService svc cs class Since the UpdateUniverses method will be run in its own thread we concluded to open a new session from the NhibernateHelper for each time our system is talking with the database 36 Designing the page layout The web page design and layout have so far primarily been designed through internal discussion within the team as we add functionality to the page such as the hardware statistics table Both we and our client have the need of a layout mockup and CSS design so that we have something steadfast to iterate on while we also make sure that our vision of the system matches with our clients Because the system is still in its early stages there is a lot of unknown functionality that we cannot account for or it simply isn t implemented yet but we can still create a general layout for the page and CSS styles from that The early designs were created from the idea that our system should replicate a simila
217. ter just the geographical filter and just the universe filter All data corresponded to the test data set Tried combinations of time with both geographical and universe filter and all data corresponded to the test data set Zero errors Test Failure 56 Validation tests that were not performed and why This is meant as a short summary for each of the validation tests that were not performed during our project period also with an explanation for why 56 1 List of validation tests that were not performed Name of test Test id Display hardware software T R 1 statistics Filter statistics user interface T R 3 2 Display software statistics T R 5 Future hardware prediction T R 6 56 2 T R 1 Display hardware software statistics This test ended up being too vague due to covering all the information to be showed in the statistical table without going into details about each category The test was therefore split into 5 new tests T R 1 1 T R 1 2 T R 1 3 T R 1 4 and T R 1 5 NET Display hardware software statistics est ID T R 1 REG Mee DH A F 1 and B F 1 Page 178 of 194 Project FHaSS Test log 2012 EEE Test to see if tables are created and contains accurate information about the different hardware software stored in the database WSS eres Kim Richard Johansen WSS ie paljuareles Function test black box Test approach 1 Putin
218. ter 73 47 3 7 Regression test soe e esee eere dediserte osea ones e naues 73 47 3 8 Performance test oic cec a saken Eee s aD ceu ee c ERR d Beg tee ront n dx ees 73 AS TOI ri 73 Page 57 of 194 Project FHaSS Test Specification 2012 46 Test layout 46 1 Test data set To test the system we will be creating a test data set that will be used in most of the tests The type of tests is explained under test criteria for each of the tests We will manually calculate values for the different categories beforehand of the tests and compare them to the results the website produces The data set will consist of fictional bug reports that will be used to test each part of the system such as testing the system web functionality i e sorting and grouping algorithms calculation algorithms and testing the result by sending complete bug reports that ventures through the different modules of the system i e parser database and web functionality i A complete fictional bug report will contain this information e CPU o Speed o Number of cores o Manufacturer o Model f Graphics card VRAM size o Primary Display Resolution o Manufacturer o Model g RAM size h Windows OS version i DirectX version j Geographical location may be in the form of IP k Date and time Funcom game o Gameserver Our system will group most of this information e g CPU speed cores GPU Model
219. test that will be done when all our A grade requirement are completed and successfully tested We will insert variable amounts of bug reports over a length time to ensure that the system can handle the pressure and still be operational by our client under the two second load time they wanted for their website The bug reports will be created beforehand of the test and the results will be manually calculated to check up against the data from the website after the test This is to ensure that the data from the bug reports did not get corrupted due to high work load the system encounters 47 3 7 Regression test Regression tests will be done after code alteration of previous tested functions or when new functions have been added to the system This test is to ensure that previously successful tests still give the same results 47 3 8 Performance test This test is about measuring the performance of the system e g the time it takes for the website to load show results as charts or tables with information from the database This test will be done regularly to ensure that a new functionality or code alteration keeps the loading times under two seconds as the client wants 48 Sources e http en wikipedia org wiki Test plan Last visited 15 12 2011 e Mhttp en wikipedia org wiki Test case Last visited 15 12 2011 e RIVET test specification HIBU Kongsberg library e Dragonfly test specification HIBU Kongsberg library Page 73 o
220. the geographical location filter combinations are working correctly It should be possible to choose between continent and country as a filtering option COS That the data shown after choosing a continent or country corresponds to the manually calculated results Continents to be able to choose from North America South America Antarctica Africa Europe Asia Page 68 of 194 Project FHaSS Test Specification 2012 Australia Test date 27 04 2012 Kim Richard Johansen WSS ie palsereles Function test black box Test approach 1 Set the filter combination to be tested 2 Wait for the web page containing the filtered statistics to load 3 Compare the produced filtered percentage statistics to the manually calculated percentages Test criteria e Database and website is up and running e Test dataset has been made and inserted into database Expected e That the produced filtered percentages based on the test data set are results similar to the manually calculated percentages for the tested filter combination Test creator Kim Richard Johansen 24 04 2012 Errors e Encountered errors Outcome What happened Hardware type management and grouping est ID T R 4 1 ire Pias A F 5 ES Eddie Test to see if the hardware from the bug reports is grouped correctly under general hardware types and that it shows the correct percentage for
221. the work to come in the later phases The scope of the project is established and the use cases are outlined as well as one or several potential system architectures It is also in this phase that the major risk analysis is done and the project management is setting up a preliminary project plan as well as a cost estimate 3 2 Elaboration In the elaboration phase the work done in the inception phase is elaborated upon with the primary goal being to address the risk factors and to establish and validate the system architecture The architecture must be validated to ensure that it is possible to implement the major system requirements which are also captured in this phase A final deliverable is supposed to be a plan for the construction phase 3 3 Construction The project team will in this phase be building the complete system upon the foundation that came out of the Elaboration phase This phase is the largest of the four and will be divided into a series of time boxed iterations 3 4 Transition In this phase the system is deployed at the user s location and feedback is gathered for a future system update User training and conversion is also included in this phase 4 Scrum Scrum is one of several so called agile software development methodologies which aims to be flexible throughout the whole project period regarding changes in requirements They also aim to reduce the risk of the stakeholders by developing a par
222. ther work in the field is adequate e Following the document template e Correcting any typographical errors such as spelling and sentence structure e Using correct cross references if you are referencing something within the same document In addition the process is also intended to ensure that every contributor to the deliverable ensures that their submission adequately reflects their intention Page 118 of 194 Project FHaSS Quality Assurance Document 2011 16 2 1 Document creation and Structure Any new deliverable is to be created from the document template foundin Temporary This template contains the modified styles that are changed according to this projects color and layout such as numbered headers different theme colors and general structure examples A newly created deliverable or document that is yet to be completed should be placed in either the Temporary folder or any personal folder Most deliverables should contain the most basic of content to keep the wanted document structure e A front page e Ageneral document information page e A table of contents e Anintroduction relative to the technical field the deliverable is about e Asummary or conclusion Where applicable e Alistof references and sources Where applicable Beyond this any deliverable should be neatly structured for easy navigation using the correct headers and sub headers 16 2 2 Timelines A document should be completed by its ori
223. this area was not tested as frequently and thorough as it probably should have and as well there should have been created one or several tests in the test specification for testing this area specifically 67 FHaSS extension suggestions In the following will several ideas to extension features that either the project team has thought of themselves or that Funcom gave as feedback after the first successful deploy of the system be discussed and given proposals to design solutions of Page 192 of 194 Project FHaSS Reflection Document 2012 67 1 CPU cores groups with range The grouper is currently grouping hardware statistics entries for the CPU cores hardware category based on groups representing distinct number of cores values for instance four 4 cores Say in five years the computer hardware market will have changed much and the CPU evolution trend of more and more CPU cores onto a single CPU chip will most probably just continue Using FHaSS Funcom will then have to have one CPU cores group for each possible CPU cores number while it is at this point probably more informative to get statistics for how many CPUs there are in this and this range of number of CPU cores like FHaSS has for CPU speed for instance The current system does not take this into account because it is not a demand of the current computer hardware market A range solution would both work with the current market situation and an eventual future s
224. to display the raw data that the charts defined in requirement A F 2 C F 3 Textual help C C F 3 24 04 12 Meetings Functionalities in the web page interface that is not self explanatory to a completely new user must be explained through some form of textual help in the interface It must be possible for the user to see the help text in the current page he has loaded In other words he she shall not have to enter another page to view the help text C F 4 Chart zoom C C F 4 4 05 12 Meetings It must be possible for the user to choose to get displayed a zoomed enlarged version of one of the charts defined in requirement A F 2 C NF 1 Filter URLs C C NF 1 4 05 12 Meetings When the web interface is displaying hardware statistics as defined in requirement A F 1 and B F 1 based on the filters defined in requirement A F 3 the web browser must display an URL that when accessed at a later point in Page 53 of 194 Project FHaSS Requirements Specification 2012 time will make the system reproduce the same statistics filter selection 43 Attachments e Product Backlog Page 54 of 194 9 HOGSKOLEN i Buskerud Funcom Hardware Statistics System Test Specification Funcom Hardware Statistics System Client Funcom N V Acronym Sensors amp Supervisors
225. tween the software code and the hardware Bug An issue with the software application causing it to crash or behave in ways it isn t supposed to A minor software update that usually fix specific issues do small changes to or add minor functionality 14 3 Document History Version Description Date 1 First version created General 15 10 2011 outline structure 1 5 Most text rewritten 13 12 2011 2 Further detailing of text 21 12 2011 2 5 Project document template 10 01 2012 adopted 3 Updates after document review 29 05 2012 Page 17 of 194 Project FHaSS Vision Document 2012 15 Table of contents 14 General Document Information 17 143 Declaration of tatentionn irae rere ee tee erento Paene ie ne uw ee eege reueg 17 14 2 Definitions and Acronyms esses enne nennen nennen asoa aaa aaaeaii 17 14 3 Document HIStOFV ose atra Sees e pea S Geert 17 15 Table of contents irte eret Ere HER e Ive eres 18 ene 19 17 Context ofthe stud asset 19 17 1 Computer technology evolution esses esses ener nnne nennen nnns 19 17 2 Shortening of production me 20 17 3 Fault tolerance eoe ettet e eerte ceres dee eee este LE hie 21 174 Hardwaretrends eerie ciento cxtra cada one ek iod ag A Eo D Reda ade ado dusin adu 21 EMEN I c 22 17 6 Usage of the system ice cine te haare enar l sec ena pe
226. ually fixed or not by trying to force the same situation that created the issue in the first place The game developers in general use the collected information and read its hardware trends in order to be able to foresee what type of hardware will be mainstream when the product they are working on at the moment will be released This will enhance hardware compatibility and make visible potential future problems A typical crash scenario A player is playing one of Funcom s games The game crashes A report of the crash including hardware specification and the part of the source code where the crash occurred is sent to a Funcom server The report is stored in a database Employees access the database through a web interface to get crash specific and hardware information 18 What to improve 18 1 Crash report parsing When a crash occurs at a players computer a crash report is produced This report consists of several parts e P ofthe client date time username game server etc e Where in and the actual part of the code where the crash was produced the callstack e A so called DxDiag file containing hardware information of the client computer in pure text format This report is sent to one of Funcom s web servers where it is stored and extracted A parser is then starting to extract information from the components of the report The information is stored in a database for Funcom employees to later access Page 22 of 194 Proj
227. ue For example one day the memory will contain data from 10 04 2012 to 10 05 2012 and the next day the memory will contain data from 11 04 to 11 05 A logical check on the dates is needed in the ReadDataService to decide whether or not to use the cached memory the input parameters like universe and country and the output the result from the service remains the same Page 161 of 194 Project FHaSS Design Document 2012 New methods need to be written to iterate through the cached entries for sorting counting and grouping before returning the result in the same fashion a normal database query would 51 Conclusion The system has evolved quite a lot throughout the project period Some of the requirements we have implemented have made us realize better ways to solve specific problems which have led us to rewrite parts of the code This is especially true for the parser where the design and implementation has been changed quite a few times It started out just as a simple console application that later was being handled by a WCF Service but in the end we ended up designing and implementing it as a Windows Service On the MVC i e the user interface side after our mock ups were done of the system not much have been changed other than implementing new features over time as we got to designing and implementing the requirements The WCF Service have changed quite a lot From only doing database queries with nHibernate to usin
228. ument 5 General Document Information Deliverable nr D1 2 2 PENE Er MEN Research Release Public Workpackage 1 Responsible Kent B Dreyer Dag C Hem Kim R Johansen 5 1 Declaration of intention The purpose of this document is to clearly define the computer language s we will use for the development of our given task 5 2 Definitions and Acronyms ASP Active Server Pages Debugger Tool for locating errors in software CLR Common Language Runtime IDE Integrated development environment 5 3 Document History Version Description Date 1 First version created 15 12 2011 2 Document template added 10 01 2012 Information reviewed and updated Page 8 of 194 Project FHaSS Programming Language Document 2011 6 Table of contents 5 General Document Information AANEREN 8 5 1 Declaration Of intentioDhz iicet rettet te ete ele estet envie tee Or eee thes 8 5 2 Definitions and Acronyms cccccccccsssssessccececsessnsseesececesseuseasecseceeeusesseasaeeeceseessassasseeecetsensaaes 8 5 3 Document HIStory E 8 6 Table Of contents tre teet te Aeddi 9 ABE occur mc a 10 WE el EE 10 8 1 Is it possible to use this programming language to solve our bachelor task 10 8 2 Does it work well with other planned programming languages cssccecessseeeesesteeeessaaes 10 gt e 11 10 C ATP Ig 12
229. ument to latest QA 24 05 2012 standards Updated information on tests 4 5 Document reviewed 29 05 2012 Page 56 of 194 Project FHaSS Test Specification 2012 45 Table of contents 44 General Document Information 56 44 1 Declaration of intentions eter e rrt a Even Ine ee vv eae daten Sa Even vene 56 44 2 Definitions and Acronvms esses enne nennen nina sa nasse sense ti ada s assess sean 56 Lr E MEM PIeeiiuidaluec MEER 56 45 Table of contents ere terere tire teenie cauudssiyeveieus Er conse eee geen 57 46 IMLUe M HO 58 46 1 Test EE 58 46 2 TESt types EET D UU ILLIS 59 46 3 Identifying different tests and reouirements enne 59 Mzegiuide c m 59 Test ET 59 EMI 60 ALA T sttemplate 5 tette eneste hdd ete 60 47 2 Validation tests oce ette Cem ente oen Lene d e teke dead 60 47 2 1 Listof validation tests tci etae rien dent t qn e Pede Ee AIR FRE CARE RH RR Tav deua 60 OG MEER LCD EE EE 61 473 MernifiCatlOn tests oiie ekee on en ane EXER EE te x V Rene rent venu iet eni Re uu ey eene duet iege 72 47 3 1 Code nU REDEEM 72 47 3 2 Code eer El EE 72 urs MEMINI CIDCEEHRTL 72 47 3 4 User Interface Testing cccssccsssccsssecesssecseseecseececsaeesessecseneecsaeesssaessesseceeeeessaessesaeesnees 72 LEA SE GREC 72 4 3 6 Stralli test svet
230. ured in a simple yet expansive fashion The UML diagram below shows a simple version of the structure and shows the parts we used to generate the charts for our system The Highcharts js file contains the main graph rendering tools where as we customize it further with the use of the themes class to make it match our system s color scheme and the export class simply to export the charts as image or PDF files There are many ways to define the options of the charts either via an options instance or directly in the charts instance right before rendering We went for the latter one giving us complete control to customize the individual parts of the charts from our dynamically generated hardware lists 44 Geographical Location Because of the nature of our system is to filter information according to what the user needs it is imperative to be able to filter the database entries by their geographical location Being able to filter entries by location would help the users to easily see statistical differences in hardware between countries or entire regions In order to specify the origin of a bug report we would first need its IP address but because of a limitation within Funcom s current python parser it does not currently provide this information There were also several other pieces of information that our new system needed such as OS bit resolution and DirectX information and after we brought this to Funcom s attention they did t
231. ve to the game selected c If no game is selected it should not be possible to pick game server o o All the filter criteria are to be put in by the user of the system A F 4 Hardware type management A A F 2 27 10 2011 Meetings It must be possible for an administrator of the system to add edit and delete so called general hardware types The system must be able to recognize if new types of hardware has been registered When this happens it must be possible for an administrator of the Page 50 of 194 Project FHaSS Requirements Specification 2012 system to map the newly recognized hardware type onto a general hardware type The Hardware Statistics displayed in tables A F 1 and in charts A F 2 will only be displaying information for the general hardware types The different categories of general hardware types that can be registered are e CPU Model e GPU Model A F 5 Hardware grouping A A F 2 27 10 2011 Meetings The system must be able to recognize almost identical hardware types and group them to be shown as only one more general type in the statistics The general hardware types are those of requirement A F 2 The user shall be given information about currently registered but not recognized hardware types The user will then manually have to decide what hardware type the different unrecognized hardware types is to be grouped under
232. volves around two main problems what are the chances of us running into this risk and how much threat does this risk pose to our project These two questions can be summarized in two words probability and impact and the product of these two become the total severity involved with a given risk The following graph shows the connection between probability and impact Impact Probability Matrix Impact c LSAN COM Probability Fig 1 Risk probability matrix The hard part about risk management is to detect large risks that may occur during the development of a project and identify these properly If a risk can be foreseen and understood before it has the possibility to occur we stand a much better chance at making the right decision when we know the gravity and seriousness of our choice If a new risk is found and it is believed to be a viable threat to our project its will be brought to the risk manager for further research If a risk is deemed dangerous enough the risk manager will analyze the risk in an effort to learn more about its nature these are the most important subjects e The consequence of the risk Time setback data loss or other e How to prevent the risk If any measures can and should be taken to avoid this risk e Finding a possible solution If the risk have occurred and if it has any simple solutions 60 2 Risk table template As risks are found and identified they are put in a basic risk table and categoriz
233. w work was not documented but instead orally communication between the project members was used and worked without problems 57 2 Code compilation When compiling produced code we removed most of the possibilities for a group member to upload a code to the source control with issues but since more than one member could be working on the same code fileuser related issues still did occur 57 3 Debugging Debugging has been done regularly by each group member when developing functionality on the system The built in debugger in our IDE Visual Studio was quite powerful and enable us to have a full overview of the system when we executed the code developed Debugging has therefore been a big help when trying to indentify bugs issues that occurred 57 4 Strain test This test ended up being unnecessary due to change in our project requirement Our system will parse newly added entries in Funcoms database when theirown parser has parsed the bug Page 181 of 194 Project FHaSS Test log 2012 reportssupplied by players The amount of entries to parse group in our system is set to only deal with a fixed amount at a time so therefore the strain test was unnecessary 57 5 Regression test Ad hoc test These tests have been performed by the test manager at regularly intervals to detect issues bugs on the system If something was detected the issue bug has been added to a list Group members that were responsible for theissues bugs we
234. wanted but the filter save functionality will somewhat fill the dual view s place and give them the specified information needed 63 Read Data Service Our system uses two WCF services in order to separate two largely different information services where one handles everything related to administration of the system the AdminService while the other has the sole role of delivering the actual statistical data the ReadDataService The ReadDataService uses just as the AdminService the nHibernate API to fetch and insert information to and from the database which has caused some issues during development The main issue with nHibernate was a lack of experience from our side and the lack of documentation that came with nHibernate this mainly caused minor problems where we were unsure about how we should use nHibernate to get the correct information back as simple as possible nHibernate also caused some larger issues when we were unsure about how it worked and even what we could use it for this is explained more in depth elsewhere This WCF service slowly changed over the course of the project period due to a number of optimization efforts and the increasing the demand for more and more statistical information this Page 189 of 194 Project FHaSS Reflection Document 2012 way it has become a part that has been completely rewritten and gone through numerous iterations since we first created it in the beginning of the project The
235. xchange section handles the management of digital files such as deliverables amp documents source files and other project files 15 1 Communication 15 1 1 E mail guidelines E mail is our main communication channel to both our supervisors and our client and will mainly be used for e Announcing completed documents that are ready for comment and review e Announcing up coming deadlines to our project members e Administrating meetings More information below e Notifications about large project changes or changes to the product backlog While we use E mail to announce documents that are to be reviewed or that the review is completed it is encouraged that comments and changes is added to the document using the built in office tool See point16 2 4Reviewing a deliverable If any large changes are to be made it should be announced via email to the parties involved 15 1 2 Regular supervisor meetings The project group will hold regular meetings with one or more of our supervisors approximately once a week This meeting may be held in personor via teleconference The participants should be given approximately 1 day notification email ahead of the meeting to give them the opportunity to reschedule The meeting agenda should also be sent to all participants at least one day before the meeting as well as being placed in our Meetings Notices and Agendas folder See section15 2 1Project files folder Dropbox and section17 1Notice of meet
236. y Several new specific windows service methods needs to be added like for starting and stopping the service Other than the new methods that are used to manage the application as a windows service the code of the parser is the same lt lt derive gt gt Figure 29 Illustrating the parser as a windows serivce 46 General model relation search manufacturer In the general CPU and GPU models section of the administration page it is possible to add new general models to the database similar to the other administration pages Specifically it works like the universe to game administration in the way that a general model will have a database relation to multiple distinct models There are however a lot more distinct models to relate to a general model than there is universes for a game therefore an easier way to relate many of the models automatically is wanted so that the user can avoid relating one distinct model at a time The automatic relation should be done when the model is first created A simple way to do this will be that the user provides a string that will occur in the distinct model names One will then retrieve a list of all the distinct models that do not already have a relation iterate through it and check if the names contains the provided string If it does the relation will be set to the newly created general model Another feature that is wanted is to easily be able to sort the models after manufacturer and also

Funcom Hardware Statistics System Vision Document

Contents

Download Pdf Manuals

Related Search

Related Contents