Report - Computer Science - ANU - Australian National University
Contents
1. 17 18 44 3 10 The quality of the Scene This issue has also been repeated in much report which I will detail in future development MS HD 3 11 Documentation The existing online help is miles away from being enough and even not accurate at all At the first sight walking is just about keeping the left mouse button down no matter what direction you are heading for Users are left working out the operations like digging a plot and picking an artefact by themselves MS LD Criticism also comes from the confusion of what record context and search context mean to do MD LS 3 12 Some Other UI Problem The OK and CANCEL have the same effect close the dialog or record users input The OK Button and Cancel Button do the same work in these two dialogs so we totally believe that using one OK Button instead of those two is a better choice LS LD The most frequently used and also efficient enter to confirm is not implemented LS LD Using warning information too much LS LD All messages contain an exclamation mark icon which may not be appropriate for all situations LS LD Others 3 13 Multiplatform problem It works best on WinXP Java sdk 1 51 and java 3D 1 31 The appearance is little different on Window 2000 The lighting of terrain has become over strong On Linux the main program still can run after changing some file format and name for loading but the artefact can n2. ZANU Virtua m naar File Excavation Site Help Search Context Name Kiandra NS Plot Marks S Selected U Underway X Excavated Enter Column 4 Search Cancel Site Observation Actions Search Context Specify Plot Record Context Excavate Reset Display Grid Collect artefact when found Mes Al File Excavation Site Help irtual Excavation lt r AEF y 4 z F P n 1 Zeit Site Observation Actions Search Context Specify Plot Record Context Excavate Reset Display Grid 3282044 2 Requirements 2 1 User Requirements This exercise should be designed to simulate in three dimensions the physical characteristics of an archaeological excavation It should attempt to visually depict the ground surface and the underlying strata as revealed during the excavation process It should be capable of being used for any number of digging strategies in terms of the type of excavation test pit trench or open area the method of excavation by varying intervals of earth removed i e by spits 1 2 5 10 cm spits etc usually for test pitting or by stratigraphy usually used in open area excavation and by the type of recovery methods used to collect the artefacts bones and other materials present in the matrix by eye only dry or wet sieving sieve size etc Therefore the size of the data base should reflect the potential for all types of strategy to be practised It will contain the f
3. Medium seriousness and low difficulty high priority MS LD Low seriousness and medium difficulty low priority LS MD Low seriousness and low difficulty medium priority LS LD 3 1 Installation and initial config The installation is to put required files into two folders It can hardly be called as installation One report recommends using some simple installer and jar file to finish this process LS LD Since it has been decided that the end users are the students of ANU archaeology students VE should be less JAVA looking I doubt how many users could start the program without guidance for professionals Not to mention that you have to figure out CLASSPATH configuration which sometimes is not easy even for us What s more not a user understandable message is provided when installation goes wrong At the very least we should include all the files necessary in the VE package Maybe a Java tutorial also should be there The program should be able to be started by a double click of an executable file which includes the classpath setting and is represented as an icon It s anorm for any software nowadays and fits with the conceptional model of every user It was mentioned in all of the reports of COMP6422 students as suggestion for improvement Running an application is just a double click on icon provided There is no installation provided LS LD We can see that users do not need to modify a sett
4. 44 e Exposed artefact record The text field for you to put the description of the artefact le Excavation Site Budget Help retefact Description aretefact Description No artefacet has been recorded yel Site Observation Actions Default position Backfill C Birdview Gridon _ Highlight Off ir P Specify Plots Excavate Review Contexts Review Artifac e Review artefact This on allows you to review the artefact collected x pie Excavation Ske Budget Hop 39 40 44 Site Menu Has the functions needed for actions taken during the site observation phase e Review Contexts Allows the users to review the annotations findings on each context Enter the roe and column number or simply by clicking the plot you like to review Context field record will pop up mane filo Excavation Site Budget knip Pot hs S Selected Utinderway X Excavatod She Observabon Actors Spocity Plots Excavate Rewtow Contexts Detaut postion Bati Birdview Gridon C Highlight oer e Specify Plots Users start to pick what plots they would like to excavate by click 40 41 44 the grid in the panel and cancel by re clicking on the selected ones The selected plots will be highlighted on the scene HERBE eee File Excavation Site Budget Help Specify Plot Name Kiandra NSW Site Width m Site Length m 100 100 Plot Marks S Selected U Unde
5. OrganicMatter Artefact Budget Has a SZA 35 44 Virtual Excavation Users Guide 35 36 44 1 0 Welcome to Virtual Excavation Virtual Excavation in its current prototype version is easy to use interactive graphical software tool aimed at helping simulate some of the complexities of a modern day archaeological excavation The users of Virtual Excavation are undergraduate Students of Archaeology at the ANU School of Archaeology Main features e 100 Java e Renders some others 3D formats such as VRML wrl 3DS and DXF 2 0 About this guide This guide the Virtual Excavation User Guide is provided to help you get up and running as quickly as possible with Virtual Excavation It provides step by step instructions for installing and starting Virtual Excavation performing common tasks and learning more about Virtual Excavation 3 0 System Requirements Java SDK 1 5 1 Java 3D 1 3 01 NCSA Loader portfolio jar needs to be in the classpath VRML Loader vrml97 jar needs to be in the classpath Three buttons mouse 256Mb RAM minimum 32Mb graphic card minimum 4 0 Installation Virtual Excavation requires the existence of two directories e VA to host the java classes and the executable jar files e Excavation The excavation directory requires further split as follows e Excavation Shapes e Excavation Site 296s we B7 e244 e Excavation Site Textures The Excavation Shapes
6. any future development could take this version as a sound starting point without being bugged by usability problems of the previous stage 10 44 3 Analysis amp Problem detected The following section will discuss the problems found and mentioned by the Usability Engineering students reports and some of them are detected after using and developing this software Opinions from Users and tester including me the biggest user and testers up to now and some argument will be put after some issues that are usability arguable Some of them are just bugs The level of seriousness and priority of development ranking are given at the end of each To begin the usability report here I list some basic usability concepts that will be used repeatedly in the report We believe that a good design should have A good conceptual model The designer provides a good conceptual model for the user with consistency in the presentation of operations and results and a coherent consistent system image Good mappings Itis possible to determine the relationships between actions and result between the control and their effects and between the system and what is visible Feedback The user receives full and continuous feedback about the result of actions Visibility By looking the user can tell the state of device and the alternatives for action And the usability is traditionally associated with the following five usability attributes
7. context button intended to specify the plot to excavate The trick thing is both specify plot and search context panel have nearly the same appearance It sometime takes several seconds by clicking to find out you open the wrong one MS LD The RESET surprisingly reset the positioning of the scene rather that the status of the terrain which is of more importance It will confuse any novice of this software since we all know what RESET means Only one report does not mention this problem directly by put it the current reset the scene but not the data The others suggest that this one should be name as Reset position or realize the function of resetting the data MS LD The record context is redundant because every excavation entails a record action This one simply add unnecessary word load to the users and challenge them by working out how to associate the excavate and record Users are reported confused about what the context means hard to find the relation between digging and record context don t know when to press record context or found it trouble to change too frequently between Record context and excavation MS LD Better change the specify plot into specify plots and search context into Review contexts because the current ones don t indicate the real function of them properly This one is not in the report LS LD 3 3 The spe
8. it s one sort of over automation The system takes the control out of the users Users feel not informative and insecure as a result The solution of this is simple make things visible We now have the power to show what is really happening to provide a good complete image that matches the person s mental modal of the task thereby simplifying both understanding and performance Don t waste the power we have and don t fritter users out HS HD Here is one story during the test During the test process actually I was the first one to find the artefact in my group When other participants asked me to show the artefact to them as I did not remember the exact position where I got it I cannot show it to them Therefore I hope users could be permitted to query what artefacts have been picked up and this must bring more convenience to them Actually it is not only for anyone s companions Even that person himself will find it frustrated that you are not allowed to review what you have picked You will think whether I have picked that one or I just threw them away The effects are the same HS MD 3 7 Navigation It is unbelievable to have successful 3D software without user friendly navigation system whether it is a game or educational stuff like this The city jungle is labyrinthine enough People won t bother to get into anything like this in a virtual world to have themselves lost again Probably it is still
9. v Highlight Off e Highlight off Option for users who don t want the highlight of the plots 43 44 44 References Virtual Excavation Application Final Report Ben Davies Virtual Excavation Application Final Report Cesar Martinez Usability reports from comp6442 students Archaeological Field and Laboratory Manual 2004 ANU School of Archaeology and Anthropology Human computer interface Lecture Notes Henry Gardner ANU DCS eScience The Design of Everyday Thing Donald A Norman Java3D various sources 44
10. goal of the software Students should not be allowed to undo what has already happened in the site otherwise all the constraints are lost and rampant click and undo become the only rule Meanwhile certain kind of undo does exit at the stage of the users decision making period They are permitted to cancel what they have chosen for excavation before the action is taken Show the artefact quickly If the user is only interested in artefact it wastes time to click on all the 100 grids Just like a select all button in an email system it is suggested to add a select all plots button in the Specify Plot function That s because they misinterpret the goal of VE as software to study the artefact What makes sense is the procedure of excavation not only the artefact 19 20 44 3 15 General remarks I think Ben did a great job but due to time constraints this software does violate some of the usability rules which makes it a bit tough to use e Hide critical components make things invisible e Do not give sufficient feedback e Use unnatural mappings for the execution and evaluation side so that the system state is hard to interpret e Challenge users by not allowing some transferred skill within or without the software For novice users of 3D stuff they don t get enough assist when start to use for experienced ones they find challenged because their experience does not work here No wonder tw
11. of that is the buttons in them The one popping up when you excavate has Ok and Cancel while Previous Exit and Next for the other We will discuss them together The OK and CANCEL have the same effect which will record down what you ve written down in the description and comments text area So is the case for EXIT As a norm EXIT means quit without recording Not many users complained but it is mentioned in all reports as suggestion LS LD If you click previous in the first context dialog with warning information Warning Context 0 pops up that is apparently not understandable language for the users Also a warning pops up when pressing next in the last context Here constraints should be in place to make it easy to determine which action is possible is available at this point instead of kinda trapping users into pressing not function existent buttons It is mentioned in two groups report as users feel uncomfortable when warning information pops up ad no one understand the information on it LS LD All the users failed to record information for the 5th layer because of a bug When they clicked on the Record Context button a dialog box showed All selected plots have been excavated Once again the consistency rule is broken here One group of students found this problem as When the user used Search Context function the above information can be seen I
12. out another way rather than use the navigation we give HS HD VE forces people to use the combination of keyboard and mouse when dealing with translation and rotation which steepen the learning curve and increase the memory load Because users can not get any similar experience outside that means they need to establish temporary memory form the external reminder like helper in case some time has passed since the last time they use it That is not desirable for user centred design It is redundant and nearly useless to pop up the right click menu in VE considering the buttons available already as one pointed out no one uses the right click popup Instead like most of 3D games the right key of mouse should take the responsibility of rotating Another good thing of this is to spare the left key for the real excavation in future version of VE Also your view point moves too fast considering 3D scene as small as this one Users tend to get lost and suffer motion sickness One part of VE s purpose is to study the soil so there is no point flying over without presenting enough details to the users VE users prefer navigation be executed slowly and accurately It is not like what we do in the Doom being slow means death Relatively people won t complain even if VE is a bit too slow And it is preferable to have options open to the users if deferent scene is loaded MS LD If you decide to use the standard 2D mouse as a 3D input devic
13. popup when you like to cancel your selection MS MD The shape of the cursor is found to change unexpectedly when it moves over the 2D map LS LD 3 4 Search Context Panel Firstly the function of this panel is to review the contexts So it should be renamed as Review Contexts LS LD It is time consuming and frustrating to key in the column and row number in letter and number respectively This function should be achieved with the mouse click in consistency with what practices at specify plot panel All the reports point out this problem Nearly all the users can complete the task but complaining only one way to specify the spot and one group put it as not consistent with the specify plot panel This is especially misleading design since this panel share the image map with the specify plot panel MS MD And it does not support to view more than one context field record dialog at the same time and the panel switches off once the context field record is open which trouble people to open the panel again for any more review People spent more time than necessary to perform a task so that efficiency is not optimized here This problem is not mentioned in any report Professor Henry pointed it out when reviewing this software MS LD re 14 44 Backspace does work in the Row field LS LD 3 5 Context Field Record Dialog There are two kinds of this dialog the only difference
14. Learnability ease of learning Efficiency high speed of user task performance Memorability user retention over time Errors low user error rate Attitude subjective user satisfaction oOoo0o0o0o Benchmark of problems in this software o Most serious problems problems that compromise the basic value of this software o Medium serious problems problems that affect the major performance of this software mentioned many time by the usability reports as suggestion for improvement those reported as places where users complain many time o least serious problems the rest Because of the time limit of the project I give the priority of possible improvement based on the mechanism as below 10 11 44 We rank the priority of possible usability improvement according to the implementation difficulty and seriousness of the problem If implementation difficulty gt gt seriousness of the problem the priority is lowest in the opposite the priority is highest difficulty seriousness the priority is medium difficulty gt seriousness priority low difficulty lt seriousness priority high High seriousness and low difficulty highest priority HS LD Low seriousness and high difficulty lowest priority LS HD Medium seriousness and medium difficulty medium priority MS MD High seriousness and medium difficulty high priority HS MD Medium seriousness and high difficulty low priority MS HD
15. PNE THE AUSTRALIAN NATIONAL UNIVERSITY FACULTY OF ENGINEERING AND INFORMATION TECHNOLOGY DEPARTMENT OF COMPUTER SCIENCE COMP6703 eScience Project Virtual Excavation Final Report Scottie FENG u2503062 S044 Contents Introduction Background knowledge of this software and excavation Requirements The requirement of users client and the general one Analysis amp problem detected A complete analysis of the usability aspect of this software Design amp Implementation amp Fixing Some new features are introduced Usability Improvement Usability improvement implemented Appendices Some material related to this project References 3 44 1 Introduction 1 1 Background The ultimate purpose of this software is to simulate the complexities of a modern day archaeological excavation which will help to save part of the budget that will be spent otherwise on the real life excavation Thanks to the jobs done by former ANU students Ben Davies and Cesar Martinez the Virtual Excavation software has been developed as Version 1 0 But after being used by the eScience students last year at the Usability Engineering course and usability reports have been written Professor Henry recommended that an overhaul be made so that the department can keep this software alive for future development The current status of this Virtual Excavation software is far way from meeting the user s requirement to be major teachi
16. backfill Birdview Gridon Highlight Off 26 aes gare Figure 3 Bird View ANU Virtual Excavation DER File Excavation Site Budget Help Site Observation Actions Specify Plot Excavate Search Context Default position Budget backfill Bird view Grid on Highlight Off 4 7 Improved Excavation System Excavation is no doubt the core of this VE This one called new partly because the highlight function that has been covered already Meanwhile another new feature is to let give users the freedom to decide which one to excavate next in the cases when multiple plots have been chosen When you have more than two plots in the queue each time you press the excavation button you will be required to click on a 2D map to specify your next target It avoids the embarrassment of being throwing into the dark that is complained much by the users in the precious version Once again when you click the plot on the specify the plot to excavate next dialog the highlight system will help you to identify the one you are about to make decision on oy ae 28 44 4 8 New procedure So after several new features are added the basic process of finishing an excavation task 1S Specify a plot or several plots form Specify plot panel Choose one of them as your next excavation target Record the contexts Pick the found artefact Review the context if needed Review and file the collected artefact i
17. cify Plot Panel It should be the very first panel the users deal with but turns out to be the first challenge 12 13 44 So is the case to analyse Some eScience students put it as weird After when only one plot is picked there is no way to de select it when multiple plots are selected the last two selected ones are removed after you press cancel button and open this dialog again If you keep pressing cancel the plots of your selection rolls back one by one As far as I know there is not any other software executing the select and de select mission in this way It means users are forced to learn a brand new mechanism Apparent it is if not impossible very tough In other words the learning curve is steep because no skill could be transferred from any experience We should make it like this click once for selecting and twice for de selecting instead of popping up warning Someone suggest adding a clear all button in this diagram to accommodate people who just like to cancel all the selected plots once for all This one is specially stressed in one report as In this task Users were required to randomly specify 8 plots and to cancel all of them It was considered to be quite a simple task at the initial thought and that was why the benchmark was set to be 10 seconds However due to a serious bug no one managed to finish the task within the benchmark I really don t enjoy the
18. contains the files with the artefacts to be rendered in the virtual scene Valid file types are VRML wrl 3DS and DXF Texture can be applied on these objects for that the pictures can be stored in this directory The Excavation Site subdirectory contains the following files e jnitial txt file with the instructions of how the virtual scene is going to be build an e terrain txt with the x y z coordinates double values The Excavation Site textures contains the texture to apply to the terrain representation as well as to the digging action 5 0 The initial txt file Readme file structure e One line per instruction e Double forward slash for comments e Instruction followed by semicolon followed by data Spaces are not allowed A typical readme file Site name SITE_NAME Kiandra NSW Description Text describing the site description SITE_DESCRIPTION The city was the main location for gold miners of the 19th century Site Dimensions x y z are integers pos follow by comma no spaces SITE_DIMENSIONS 100 100 100 Site Texture picture to be used as texture Number of pictures name of picture picture format SITE_TEXTURE 25 grid jpg SITE_GRID_Y_POSITION SITE_GRID_Y_POSITION 10 0 Avatar Image to be used as Avatar on the scene file name texture name scale double AVATAR MagnifyingGlass 3DS MagnifyingGlass jpg 0 125 Artefact object to be loaded name texture scale double position x y z do
19. e I suggest you review the literature and study the 3D applications that your target audience is likely to be familiar with 3D games are the favourites in this field Emulate the paradigms that your users are experienced with wherever possible it is unlikely that your application is so unique that no one has previously encountered your UI problems 3 8 Artefact Panel 16 ara yee The panel facilitates students investigating the artefact found This panel could be totally ignored by some novice users Most of the new users don t realize that the Artefact Panel is there after they picked an artefact MS LD The button Reset Scene only works correctly with the 6 buttons So if users use mouse to drag the object they cannot use this function to go back previous status Any position after mouse drag is being taken as a default position for the Reset function Users are cheated again One group found it as Unfortunately the function Reset Scene only works correctly with the 6 buttons So if users use mouse to drag the object they cannot use this function to go back previous status LS MD And they are also confused by guessing what is Z or Z It s not their language LS LD The system lacks the Zoom In Zoom Out function As an observing system it is important to have such kind of function This can help users to investigate the 3D object with more details or have a general imp
20. eeded in the future 24 ssi 44 With auto locating and highlight colour present in almost every your action in his site you will get all the feedback necessary for what are you doing and what you are about to do U Virtual Exc T File Excavation Site Budget Help Specify Plot Kiandra NSW Plot Marks S Selected U Underway X Excavated OK CLEAR ALL Site Observation Actions Specify Plot Excavate Search Context Default position Budget backfill C Birdview Gridon Highlight Off Figure 1 Highlight the plot 4 5 Navigation System The former keyboard corporate with mouse method is abandoned Not to get lost in this 3D scene the modified Navigation System also contributes The former keyboard plus mouse method receive lots of criticism from the trial users That is not the fault of Ben since in 2001 three key mouse is not that popular as nowadays In this version it is adopted to fully get use of three keys of mouse No corporation is needed between mouse and keyboard The workload of users is decreased Both memory workload and physical workload To be in alignment with other standard 3D software like SimCity serials the middle key is designed for scaling and rotation with right key The popup menu from right click is redundant and it has got a latency problem In this way it fits with users mental model from experience of other similar softwares Also some
21. esson for usability testing of VE The far end should be ignored Another defect is also hard to believe You can pick the artefact without get it exposed to air Once more it is a dangerous criminal at large It is understandable that few people will try to use pick Ctrl right click on any soil where no artefact is situated But just in case someone full of whim like me at that moment comes across the idea of having a try the software is suddenly regarded as a joke These two problems are lurking well The chances that users would encounter them are not big but the result is devastating 5 2 Some important fixing or Improvement An executable file is provided to launch the program and a real readme will give some very brief idea of this software before they get into it They are import because the bugs they are dealing with are of big changes that users will be bothered by them Nearly all the important functions are realized through the buttons aligned at the lower side of the VE interface They should be put in order with the sequences of the actions normally the user will take instead of letting users to jump among buttons Specify 290s 30 44 Excavation Research Review This is probably the most common problem that bothered users in the precious version Related to the buttons order the Record Context disappears in this version because it is just indispensable to each excavation Actual the goal of e
22. f needed Backfill the site before finish Iga 2590 44 5 Usability improvement This part is about the usability improvement through debugging the previous version 5 1 Two critical fixings They are critical because the bugs they are dealing with are fatal ones What s more no one has ever detected them before But once they show up the quality or the basic value of this software are compromised severely Both of them have meaning for future usability testing The furthermost row row 10 is not selectable Each time you select them the VE responses with the row 1 selected What makes it worse the sitemap shows the one you like to excavate but the excavation happens at different spot It has been too late you find you ve digged wrong plot In this case maybe the undo system is more in need there It won t be exaggerate if we say 1 10 of this software is malfunctioning This defect has been fixed with one method that parses the string of the plot to int index So big the problem is but no one even mentioned a word The funny thing here is after one month of developing it is then spotted It is probably because the users habit when playing around 3D environment they tend to put more attention in the area where is near themselves but much less in the far end That is why nearly all the 3D games nowadays define something like area of interest the area out of which is rendered with less polygons It is a l
23. h of the article Then you can review the artefact through the menu item and record detailed context after you study the plot Not any excavation could be called finished unless you backfill the site fully This is one of the most underlined procedures in the excavation guideline Users of this software can leave the site by two ways exit or backfill The difference is the by exit you ee o a4 leave without a trace but by backfilling you will get a briefing of what you have achieved and your action in this site is a complete one A feature related to backfill is that only after you backfill the site you are allowed to use the New to open a new session without exiting the program 4 3 Mini map Actually these two functions were proposed by eScience students and have been carefully thought about but finally abandoned The probably biggest issue of 3D is how to locate the target and walk through the scene The mini map which will facilitate both of them is desirable in this sense But it doesn t apply here It makes sense to categorize 3D software into DOOM like and SimCity like The former one people is a human being who is supposed to explore the 3D scene and whose behaviour is human like murdering or firing But in SimCity you are actually the god though you are told to play a role of mayor instead because obviously no one can rotate and fly over the scene at will The scene is more in your control
24. hallenge to the excavation This components include time restraints budgetary limits information overload i e excavating so many items that it is impossible to keep track of them all and data recording The exercise has to be designed to enable students to observe the whole excavation site in particular its surface From the evidence obtained from the surface search students will have to decide where to excavate This requires that the student selects an enei appropriate method from the range of research excavation and recovery strategies available Here the simulation has to model certain physical characteristics of an archaeological excavation in three dimension space such as excavation by stratigraphy or by spits discrete layers Here the users need to be able to inspect the pits being excavated and make decisions on how the digging proceeds The user will have to decide when to quit the dig and return to the lab Here the user should examine the finds and make the best assumptions as to what the data means or represents This should be done as an analysis report that is to be produced entirely by the student The purpose from the student s perspective will be to provide them with a simulation which will enable them not only to get experience of the logistics of an archaeological excavation but they will have to analyse their results data and write up an excavation report This will enable supervisors to monitor student s
25. ing file directly in many well designed software package However according to the current VA System it is inconvenient for users to changer maps and set initial parameters because they have to exit the software first then modify the readme txt file and restart the program again Design a setting form for the VA System so that user can change the map dimensions and other initial parameters about the software when the program running LS MD ll aiis 3 2 Main interface Namely this is the interface that the users interact with and surely the most important one The site observation panel contains all the buttons and checkbox important for your action in this site The most imposing problem here is the sequence of buttons form left to right does reflect the steps that in normal case any user will take In a typical scenario people specify excavate research And users mental model from experience is that the steps of a process should start from left to right as we do for reading As a sequence of this misarrange people will keep pressing the leftmost one search context when they would like to excavate till their conceptional model has been changed This will affect the learnability efficiency and users attitude towards VE negatively Two of the reports mentioned this problem as it is strange that the excavation button goes before the specify one Also I experienced many time I click the leftmost search
26. ion factor is modified so that the artefact won t move out of the scene so quick as it does in previous version The main interface e The RESTE is renamed as Default Position e The specify plot is rename as specify plot e The Search Context is renamed Review Contexts Some warning popup have been replaced with information ones The confusing readme file is renamed initial safe ne 30a 5 4 Future Development It s time to shun the responsibility Future Development seems to be tactful tone of something I really like to do but didn t The possibility is indefinite and many ideas come across my mind during the process of development As mentioned in Ben s report lab analysis is an option I would try to make all the stuff in this site touchable or live That means you can really behave and interact rather than release your command through the buttons It means every grid will be pickable and selectable Wether this software is really valuable or not also depends on the complexity of terrain it can present to the users since to ultimate goal is to test the students ability to make wise decision The material or the reasons that their decisions are based on must be provided Also difference digging method should have different effects on the terrain which also triggers many related issue like the budget It needs much help from archaeology professionals I
27. n other words it will be a team effort So is the case for lab analysis which could be more specialized It will be very desirable if VE can load 2D map with extra statistics of Y axis In that way the functionality of this software will be explosively more powerful than what it is now An efficient loader needs to be built The other direction of this software is to move it onto wedge which provides a more immersive environment Any person is a real presence in the scene These two directions could be a good combination The wedge one should be more of software for presentation and specified for archaeology department This one could be more public and of general purpose I can not agree with Cesar who says this software as powerful But it does have the potential to be so 390i s lt 2 33 44 6 Appendices Context Diagram of the System Student Plays simulation Virtual Excavation Supervisor system Prepares modifies dig data Virtual Excavation Domain Chart Application Domain Virtual Excavation Are Architectural Domain al Domain t Domain Domain Java Excavation Data oe en Gee Graphical Information Model contains specifies has a has Site SiteD atalviodel data is stored in Determines plot for User Data Location specified by a generates generates Stone Artefact E BoneArtefact T g D a are added to z ShellArtefact Carbon Artefact
28. ng assistant software However the basic modelling has been done and the possibility of extension is there 1 2 Archaeological Excavation Archaeology is a science that is very concerned with precise measurement and recording An archaeological excavation is examined and documented primarily using handwritten measurements notes and photography which are often irreplaceable during the on site excavation During the excavation three dimensional spacing is an important factor to be visualized Research on archaeological excavations methods in general and the ones used by the ANU School of Archaeology in particular reveals stages in the process of digging an archaeological site typical tasks include in each stage are A Site Preparation o Site identification o Sites may be identified using a range of methods such as studying relevant documents e g historical texts maps and drawings Topographical plan The height of each point in the grid is recorded Excavation unit definition Grid layout O O OO B Excavation o Remove the turf layer first layer o Remove the each strategic unit till the culturally sterile sediments are reached C Sieving nee ee o The scraped sediment is then placed in a plastic bucket and taken to the spoil heap and sieved D Artefact collection o During excavation all artefacts that are seen are carefully exposed recorded on the context sheet and then bagged They should not be removed as soon a
29. o reports give its satisfactory rates concerning its HCI as 54 and 40 Most reports rate the memorability less than 60 benchmark as moderate One report even pointed out it is not desirable for a small software like this need 0 5 hr to learn 20 an DT eA4 The Current Usability Status TD Problems deee Beriousnessibificulty Priority Solved estan and al eg aie ae Not easy to config the initial status Medium a f il ES The arrangement of buttons is not intuitive Low er hard to find the relation between digging and record Low ian as context i yes 08 10 Te misname of the reset Low Low e e e 09 Can not cancel what is selected correctl edium Medium Medium 10 e e e Misname of specify plot and search context s ow s s The specify Plot Panel O Z O Z o o o O s Low Search Context Panel O Z o Z o o y O Ee e Medium Low Low medium Context Field Record Dialog O O O Z J O yY O aaa Low Low Warning information Warning Context 0 not readable Low Medium Low s The context name is not noticeable enough Low C E E Not any feedback for digging High Artefact not really collectable Medium e M The corporation between mouse and keyboard is tough High to use M Medium y Low Artefact panel CT ae ie Users don t know the panel is there after artefact is Medium Low high yes picked Reset button can not re
30. ollowing sets of information provided by the archaeological supervisor 1 A three dimensional plan of the ground surface of the site and where possible appropriate photographs and a plan of the distribution and type of surface features of interest to the archaeologist 2 A three dimensional description of the various sediment layers and their physical colour texture etc and chemical pH characteristics including any structural or other man made features such as trenches pits post holes man made walls and footings hearths graves 3 A three dimensional description of the artefacts raw material and its matrix dimensions length width thickness weight shape decoration type function and evidence of use wear or resharpening bones by species type preservation plant debris type condition ash and carbon particles giving particular attention to the size and position of them Where possible images will be provided of a range of finds 4 A three dimensional description of the type of datable materials and the appropriate radiocarbon or other results obtained The exercise is to be designed to enable students to observe the site and its surface decide where to excavate and present them with a range of research excavation and recovery strategies As material is recovered from the excavation it should be briefly examined i e some basic evidence provided and then bagged archived to be accessed later after the e
31. ot display in artefact panel due to compatibility problem of loading 3DS on Linux No work has been put on this part LS HD 218 19 44 3 14 Some Disagreements All the groups did very detailed and inspiring usability reports within a very short time But their reports are constrained by three factors e Time constraints which hinder them from going through more possible cases Two fatal errors are ignored that I will detail later e They are not supposed to read the source code and Archaeological Field and Laboratory Manual 2004 carefully In other words most of testers test the software as an eScience students who focus more the software technically rather than from the viewpoint of any real users the archaeology students Only one group recruited the archaeology students as testers In his case their opinions are too critical i e don t know what is context But some are ignored like z in artefact panel because they are armed with graphics knowledge Once again as the biggest tester and user of VE I have a big say on it Some issues have been put forward repeatedly can not be taken as possible improvement here The undo system It has been found clearly that the software package has not a function for back step or cancellation of excavation so if users make a mistake during the excavation process it needs to restart the whole program again to avoid errors The function will go against the basic
32. progress and understanding of the concepts and techniques in archaeology 1 4 The basic scenario of V1 0 To perform the excavation task users are required to go through at least these steps as followed Specify a plot to excavate through specify plot panel File Excavation Site Help Specify Plot Kiandra NSV 100 100 Plot Marks S Selected U Underway X Excavated Site Observation Actions Search Context Specify Plot Record Context Excavate Reset _ Display Grid s644 Click the excavation button or select it from menuitem to execute excavation By irtuc CE Pee J File Excavation Site Help Excavation Site Observation Actions Search Context Specify Plot Record Context Excavate Reset Display Grid Record the context for every layer ANU Virtual File Excavation Site Help Context Field Record Site Name Kiandra N Square B7 Context Context 0 Depth m 0 5 Brick Rock Metal Charcoal Glass Ceramic Description of Context and Comments Context 0 Artefacts Discard Pott idle butten ey Ry y wa gt a Wt 4 7 s Poxi ne A Z E B wa 7 Cancel k A lt 4 fe n x Site Observation Actions Search Context Specify Plot Record Context Excavate Reset Display Grid sae Search for a specific context in the Search Context panel
33. ression of the object No one will notice that the right key implement the translation and middle key for scaling I found it after using the software and reading the source code for one week Probably people will think about using the wheel or amp to resize the article If the design is not to be changed users deserve reminder like some popup dialog or at least it should be included in the online help LS MD 3 9 Display Grid All users use the grid to identify the plot at the first time but they found it non useful It indicates users need grid but just not the one we provided People can read the coordination but with much afford They distort at several positions An associated problem with this is how to get a bird s view easily which will help users map the grid to the terrain Also the corresponding Display Grid should have a tick indication when it is on Maybe the colours of the letters have to be changed as well Probably billboard behaviour will automatically rotate the sign such that the face of the sign is perpendicular to each viewer s line of sight Alternatively aligned geometry could also be used such as Raster or OrientedShape3D if the Billboard behaviour is too computational intensive MS HD Another problem reported is that most when you choose the menuitem display grid the check box is not ticked correspondingly which cause confusion Every group reports this problem LS LD
34. rway X Excavated Site Observation Actions Specify Plots Excavate Review Contexts Default position Backfill Bird view Grid on Highlight Off e Default position Show the default view of the scene ANU Virtual Excavation Default postion Backfiat Bird view Gridon C Highlight Off 4 isd 20d e Site Description Pops up a dialog that contains a short text description of the site This provides the basic overview of the site amp Site Description Kiandra NSW he city was the main location for gold miners of the 19th century el Budget Menu e Status Shows your budget status Help Menu e Help Presents a quick reference to the navigation keys e Typical scenario Gives a typical procedure for you excavation task e About Version information Button Bar Site Observation Actions Some description of the following buttons is mentioned in the Site Menu Specify Plots Excavate Review contexts Review Artefact Default position Backfill iid mcd The final step before completing your task in this site e Bird view Gives a bird view of the site e Display Grid Clicking this checkbox draws a grid over the site 10 of the width or length of the site This is to make estimation easier and help locate plots ANU Virtual Excavation File Excavation Site Budget Help Site Observation Actions Specify Plots Excavate Review Contexts Default position Backfill Birdview v Gridon
35. s the Budget Status check when necessary It is not surprising you will be required to leave the site if budget runs out The detailed breakdown of the budget is listed as followed There are two constraints in VE basically the maximum time and the maximum attempts of digging allowed The money concept is integrated into these two factors For any plot in this scene the first digging that includes vegetation removal in real life normally will cost two attempts compared with one attempts for the rest of digging The maximum attempts limit is set mainly because that will ask the students to make wise decision instead of clicking all around The maximum time as another limit seems to have more realistic meaning Digging context recording context searching reviews and file artefact will take time The backfill action will take out some time depending on the plots excavated since it is the last but indispensable procedure before finishing excavation at this site All the value of this part should be entered by site supervisor while the default values are presented 4 2 File artefact and backfill The artefact found should be filed and labelled for future lab analysis And the context of each should be carefully recorded In this version every time an artefact is found and picked the programme will switch to the artefact panel automatically after reminding you using the middle and right key of the mouse where you can have a close watc
36. s they are found but should be left and progressively uncovered so that their context can be carefully examined and recorded o Bagging and labelling All finds must be appropriately bagged and labelled The descriptor provided on the label must correspond with the descriptor on the context field record sheet E Completion of a Context o At the end of a context the context record sheet should be completed describing methods find and soils and so on F Completion of the Excavation Phase o All excavation is backfilled from spoil heap as soon as possible after the excavation and recording has been completed G Laboratory Analysis o Artefact analysis o Artefact drawing o Data entry In version 1 0 the excavation and context recording are implemented and so is part of the artefact collection 1 3 Functionality of complete Virtual Excavation Software The proposed software application essentially has to be developed in a way that provided a virtual environment in which a simulated excavation could take place Students have to be able to work through the simulation making a series of trade offs and decisions throughout the process This is like a real excavation in which decisions constantly have to be made Such decisions will ultimately effect the quality and quantity of the data retrieved as per a real dig Various logistical components require modelling for inclusion in the simulation These are included to add realism and a c
37. set the mouse behaviour Medium 29 The x six button are not understandable to the Low yes users Medi 30_ No close button provided Low Not any reminder for the mouse usage especial the Low Low middle key TET ow Display Grid High box is not ticked correspondingly The quality of the Seene oo o y O ee a High Documentation gt o o yY O bees ah Low ai AA 38 Not any typical scenario o O Medium The user manual are not detailed enough Low higan medium Others No enter listeners for most dialog Medium 44 The artefact can be picked without digging High Medium medium Some field like depth In context record should be not editable low 48 The initial config file should not be named as readme low____ low Multiplatform problem io 49 Can not load 3ds file under Linux Medium _ Ca A en O o medium medium ium a aa Pa 90 s 2223 44 4 Analysis and Implementation All the function about to discuss is new features or at least modified ones in this version 4 1 Budget Plan The simulation will have a budget plan that affects all phases of the dig which contains a financial limit It will be set by the site supervisor once before any the virtual session begins and made unavailable to anyone else unchanged during one session The students will be informed to check the budget when it reaches some low threshold or pres
38. site button in the context dialog can help the users to relocate the plot again in case they have transferred focus on other places Also the users are allowed to view the artefact when reviewing the plot context from where that article was exposed After picking the artefact this version avoids the problem of program crashing when you happen to click the artefact site once more A modified users guide 5 3 Some Other Fixings These are something called bugs Not big but the usability of this VE is compromised if not fixed 30 ore eee a In the context field record at least several bugs as follows are fixed e Remove the uneditable artefact table whose mission is now taken by the review artefact e Change the depth to 0 05 and make the context and depth not editable to avoid confusion e Give default PLS enter the description in description text area e Set grey the pre and next button when needed e Fix the problem that at layer 5 users can not record e Notify users if one plot s recording excavation is finished In the research context dialog e Fix the problem the backspace doesn t work at row field In the artefact panel e All the button names have been changed to the users language For example rotate clockwise instead of Y e The default position is implemented to default the mouse behaviour as well e The translat
39. t is clear that Context 5 is meant to be recorded otherwise it will not be searched Unfortunately there is no way to record any findings for layer 5 which is a serious bug MS LD Depth field apparent should be not editable because it is modelled so but so users still seems to be able to change the value only to find out what they key in is not recorded finally The artefact table in the panel not to mention that no one knows on earth what it does is not recordable either These distract users from their normal behaviours and increase the error rate that will cause users frustration It is described in one group s report because one student found it strange he can not find what he has entered in depth field when reviewing the context LS LD The Site Name is redundant because users can only open one site once LS LD 14 15 44 3 6 Excavation Users are in the dark of what happened when they keep clicking this one after their selecting several plots What is to be happening Where is it What did I do No Feedback Surely they can search around the site to find the happenings but that s the extra work load they are not supposed to burden Probably the sores problem of this software Several times in every report During the excavation users can not get the information about how deep has been excavated No information is provided about anything has been excavated In the other hand
40. than being waited to be explored The VE belongs to the latter When you take a bird view of the site or when you scale the map change your view point along y axis as you like Mini map is simply not applicable 4 4 The highlight system This is the feature that this version boasts most System doesn t necessarily mean being complicated This system has two branches the locating and appearance Whenever you specify any plot through the site map the plot will be put in the middle of the screen with green colour With the option highlight off it is up to the users whether or not to have the highlight colour on Basically the colour will help the users to locate the spots of interest But if you like to have a close look of the surface before giving a go propagable you should leave the option highlight off on Also the highlight colour of any specific spot will disappear when the excavation begins there The program will also locate a plot if any new excavation is happening there No locating for further digging because the users in most cases will adjust to their preferred view after that In case that you still like a more apparent indication a red colour appear and disappear with duration long enough to notice will flash on the plot you are excavating to let you know that something hot is happening there Check it out By the way it is very easy to modify this one into being an undo system if it is really n
41. the trickiest problem up to for any 3D software Even sophisticated 3D games have yet tackled it perfectly Follow the way that most users are used to instead of inventing something new is good strategy Apparently VE is not of the same calibre with SimCity which is a PC game legend for more than ten years SimCity is a classic game in which the player will take the role of a mayor whose responsibility is to build a city from scratch The game has been put on 3D 15 16 44 platform since the third version SimCity3000 The latest version is SimCity 4 SimCity and many so called SimCity like games are followed by tens of millions of players that are arguably to most popular 3D stuff up to now Left key translate middle key scaling right key rotation are most commonly used Of course it is used in SimCity Even most 3D stuff provides keyboard navigation but unfortunately most users don t like it except the direction arrows So mouse plus not combining with direction arrows are the un texted standard That is what this software should implement The current navigation was criticized by everyone All users use the help documents to search the mouse operation Any task related is hindered for its low efficiency One extreme example is no one can directly find the specified plot in the map and travel to that plot Most of users use Excavation to identify the specified plot People even bother to work
42. translation and rotation factors are reset to fit in with this kind of small VE The desirable keyboard navigation is implemented by adapting default java 3D keyboard navigation behaviour This makes a good additional way to walk through the scene and locate the plot 2254 26 44 An avatar that is represented by a torch light is added It answers the questions where are you and where are you looking at It also helps enlighten the area that the viewer is looking at and is very useful when the context of soil is observed closely 4 6 Improved Grid Function This is dilemma during the development since the precious spent quite a lot coding on this but the result is dismal But it seems that the function could help the users to establish a good mapping between 2D map and 3D scene potentially The function has been improved by adding the bird view button and adjusting the default position of the scene From these two perspectives with the Display Grid function on you can have a clear idea how each plot is mapped to names like A10 However this function is much more auxiliary in this version since other function like highlight will take most of the job assisting the users to map between 2D and 3D Figure 2 Default view with grid ANU Virtual Excavation File Excavation Site Budget Help Site Observation Actions SpecifyPlot Excavate Search Context Default position Budget
43. uble double double ARTEFACT ancientbowl 3ds travertn jpg 0 5 26 0 2 5 3 50 Viewer Position x y z int int int VIEWER_POSITION 0 12 75 Background Image to be used as the scene main background file jpg BACKGROUND stars j pg Images tobe used to simulate the act of digging Number of pictures name of picture picture format SITE_DIG_TEXTURE 5 dig jpg whe sa 58a cad 6 0 Menus File Menu e Exit Closes the application e NEW New a session which is enabled only after you do the backfill to complete Excavation Menu e Excavate Whenever you have chosen more than one plots you will start excavation by selecting this one A context field record panel will pop up for you to record information of interest ANU Virtual Excavation File Excavation Site Budget Help Context Field Record Kiandra NSW F5 Context 1 0 05 Brick Charcoal Metal Glass Ceramic Description of Context and Comments pls enter you description Site Observation Actions Specify Plots Excavate Review Contexts Review Artifact Default position Backfill Bird view Grid on Highlight Off In case you have more than one plots selected a specify the plot to excavate next panel will pop up Also by clicking the plot you can make a decision what to do next 38 39
44. xcavation It is integrated into excavation button The context record pops up every time you perform excavating Another beauty of this design is it gives another feedback to users action excavation in this case In the previous version nearly all the dialog has got a cancel button even in the help and about Most of them are though to be redundant and have been removed now For users habit the enter listener is now implemented in most of dialogs and panels Users can OK a dialog by a simple hit of Enter as they usually do in all the other softwares Tooltips are added to several places where something may need to be clarified like Display Grid It helps the users to understand what is about happen by using that component A typical scenario is given in this Typical Scenario to help novice to have a brief idea of how the whole process is And the online help is modified to provide clear usage information of keyboard and mouse The specify plot dialog has implemented the undo you can cancel your selection simply by clicking the plot selected again or once for all using clear all As mentioned everything is still reversible before your real action For the desirable consistency the Review Contexts dialog is also made available for mouse selection The previous key in selection is kept but obliviously it will be much less used predictably The go to the
45. xcavation has finished as part of the student driven analysis of the finds It should have the facility to accommodate a budget which is used up as the simulation proceeds and is designed in such a way that the amount of earth removed becomes a function of budgetary constraints The purpose of the exercise for the archaeological student s perspective will be to provide them with a simulation which will enable them not only to get experience of the 9 44 archaeological excavation but they will have to analyse their results and write up an excavation report 2 2 Client s Requirements Professor Henry requires three main places to work at e Acomplete and detailed usability report regarding the current status of VA should be delivered e All the problems contained in that report should be solved properly and the usability improvement related should be apparent and well explained e Several extensional functions like budget and filing artefact should be developed 2 3 General Requirements After this round of development this software should be much more user friendly and bug free It should give students a reasonable environment to gain some basic knowledge of archaeology excavation Some constraints like budget should be added for realism The students are expected be able to go through the main process of excavation and get visual ideas of concepts related Also it will give any future development a sound starting point Also
Download Pdf Manuals
Related Search