Conference logo
[ ASCILITE ] [ 2004 Proceedings Contents ]

The application of massively multiplayer online role playing games to collaborative learning and teaching practice in schools

Ken Eustace, Geoff Fellows, Allan Bytheway, Mark Lee
School of Information Studies
Charles Sturt University

Leah Irving
TAFE WA
The authors believe that the overlap between computer-supported collaborative learning (CSCL) and games design theories can now be tested and applied to high school teaching practice, due to the large uptake of Internet access in schools. Teachers and students at a high school in Perth, Western Australia, are being used in a pilot study as "proof of concept". This study will entail the development, testing and investigation of the use of a massively-multiplayer online role-playing game (MMORPG) to facilitate collaborative learning in the English and Science domains. This paper describes the background, preliminary work and findings involved in the pilot study.


Introduction

Global demand for international K-12 education (offshore and by e-learning) is expected to increase 9-fold up until 2025 in Australia and is already Australia's 8th largest export sector, according to a report by Böhm, Davis, Meares and Pearce (2002). Online delivery by the institution may place the economics of delivery over pedagogy (e-commerce before e-learning) through the use of ICT products as commercial conduit (Dean 2002), enhanced by broadband multimedia, device networks and mobile learners, connected by camera phones.

Adjacent education (Eustace, 2003) occurs as students learn and act on new knowledge in their local context. Whether on-campus or online, learning takes place adjacent to the learner, and removes the word 'distance' from use in education. In many ways the educational technology of instructional gaming ("edutainment") is another instance where learning is adjacent to the gamer.

Rieber (1996) proposed a hybrid learning environment in which the constructivist concept of a microworld was supported and with simulation and gaming characteristics. The authors take this proposal further in two ways by proposing:

  1. that massively multiplayer online role play games can be used for developing social and emotional skills that can then be applied in real life.
  2. that instructional gaming which combines collaborative learning and game design theories can be part of the professional practice of many ICT educators. (Many educators often wonder if the instructional gaming is worth the effort required by teachers and students).
The research question overarching the project is:
Can a combination of computer supported collaborative learning (CSCL) and games design theories work in secondary teaching practice with a MMORPG?
The aims and significance for all participants is novel. At the end of the project: Swan View Senior High School (Perth) teachers and students are used in the pilot study as "proof of concept" using teachers and students covering drama and computing domains. The researchers believe that the overlap between CSCL theories and games design theory can now be tested and applied to the use of instructional gaming in teaching practice, due to the large uptake of e-learning resource access in schools.

The Research into Professional Practice, Learning and Education (RIPPLE) research centre at Charles Sturt University has funded this instructional gaming project that seeks to develop, test and investigate the use of massively-multiplayer online role-playing games (MMORPGs) based on the authors' existing work with virtual learning environments. The research team was formed within the Internet Special Project Group at Charles Sturt University after an approach by Leah Irving from TAFE WA to develop, test and investigate the use of a collaborative role play game (MMORPG) with the assistance of Dianne Hobbs and Mark Weber (Technology Coordinator) at the host site at Swan View Senior High School.

The researchers believe that the overlap between computer-supported collaborative learning (CSCL) and games design theories can now be tested and applied to high school teaching practice, due to the large uptake of Internet access in schools.

This paper reports on the progress of the 2-stage project using enCore open source virtual learning environment to develop and test the use of a MMORPG for promoting instructional gaming and collaborative learning in schools.

A short review of instructional gaming ideas

The structural foundations of virtual learning environments should support the learning processes at work, according to Kolb's analysis of Carl Jung's works (Kolb, 1984) as applied to the use of fantasy and game playing in learning:
the reality of the internal world of ideas and fantasy, [is] in equal or perhaps superior status with the demonstrable reality of the external world (Kolb, 1984, p.54)
Lepper and Malone (1987) identified seven theories for game design, divided into four kinds of intrinsic motivations that can be presented in any learning situation, even those involving one person. They are listed below and are based on previous theories conducted:
  1. Challenge
  2. Curiosity
  3. Control
  4. Fantasy
The next elements of intrinsic motivation were categorized as Interpersonal Motivation. These 3 theories of game design are of particular interest in the application of MMORPGs to collaborative learning and teaching practice in schools:
  1. Competition
  2. Cooperation
  3. Recognition
Computer Supported Cooperative Learning (CSCL) is a theoretical paradigm for the use of online role-playing games that focuses on the use of information and communications technology (ICT) as a mediating tool for collaborative virtual environments (CVE). It emphasises an understanding of language, culture and the social setting, founded in the social constructivism. CSCL is also worthwhile where problem solving or project-based learning is concerned (Johnson & Johnson, 1999; 2002) with achieving a common group goal. [See http://www.clcrc.com/pages/cl-methods.html] Their recent 2002 newsletter states:
Learning in cooperative groups while utilizing the tools of technology needs to occur in all grade levels and subject areas. Schools need to increasingly utilize technology-supported cooperative learning (the instructional use of technology combined with the use of cooperative learning).
The key theoretical domains exist in regard to those CSCL theories related to problem-based learning and project-based learning. A MMORPG by its nature, is both problem and project based in learning. Problem-based learning (PBL) is an effective technique for motivating students to learn about information and concepts needed to help solve a problem. While most of the theoretical framework is grounded in data, the initial theoretical perspectives come from a combination of computer supported collaborative learning (CSCL) and games design theories.

The use of game genre like MMORPGs provide a "lived experience" for effective learning, where the process of creating and transmitting knowledge leads to knowledge construction, individual skills development and responsible communications, fundamental to game playing success. Savin-Baden (2000) separates problem-based learning from problem-solving learning, which mostly seeks an answer or a solution linked to curriculum content, by suggesting that the focus of problem-based learning is about:

organising the curricular content around problem scenarios rather than subjects or disciplines. Students are not expected to acquire a predetermined series of 'right answers'.
In accord with Mason's (1991; 2002) desire for more research on the educational value of using ICT, Savin-Baden (2000) alludes to the lack of research which investigates the complex and challenging ways involved in applying a problem-based learning approach and its impact on teachers, students and their "lived experience". The conclusion is that problem-based learning nature of an MMORPG should have a core location in high school education curricula. This is where "lived experience" must examine not only an understanding of the self, but also the context and ways in which a student learns effectively across subject borders such as computing studies and English.

Mason's work (1991) set the framework for investigation into the quality of student learning that takes place in ICT environments, through examination of the participant behaviours which reveal other educational goals in the chosen areas like of problem-based learning, critical thinking and broad awareness of issues. Educational research has also shown that learners can have surface or deep learning strategies. Biggs (1987) suggested a close association of deep learning strategies with active interactive participation and social interaction in an "affective" environment. Critical thinking can be defined simply as thinking that involves analysis, synthesis and evaluation.

Participant interaction in designing, building and using the MMORPG provides specific cognitive advantages to student learning with ICT. Alternatively typical gaming communities or guilds display a similar pattern of to and fro, cooperative problem solving behaviour, according to a study of online gaming communities named Project Massive, Seay et al (2004) discovered some interesting patterns through interview and survey of 1836 informants, from the ages of 12 to 68. Table 1 below summarises some of those findings that are of interest to this study.

Table 1: Online gamer behaviour patterns
(adapted from work of Seay et al 2004)

Average time per week with preferred game15-21 hours
Percentage of gamers involved playing > 40 per week12%
Percentage of gamers who communicate "outside"
the game with other gamers
69%
(28% have none)
Most frequently used method of out of game
communication
Forums
(message boards)
Main activities for out
of game communication
coordination and scheduling57%
support and advice55%
social interaction53%

Educational value is defined as the change and positive experiences in learners involving seamless movement to and fro between moments of surface and deep learning, cooperation with others; contributions to the dialogue and feelings of self-worth. This sense of educational value of instructional gaming has motivated the research in this project. According to Rieber (2001):

Play offers a means for understanding motivation and learning in a holistic way. Serious play is not easy to achieve, but the reward is an intense and satisfying experience for both students and teachers.

The game design

A research ethics plan for low risk, low harm and privacy protection of student participants was developed in line with such policies of Charles Sturt University, Deakin University, the WA Education Department and the school community used in the trial. The MMORPG games design strategy involved Leah Irving taking charge of the MMORPG design. Leah provided the story outline by writing a script as a starting point, while the students and teachers then created the narrative. Rochester Castle in Kent (Cook, 2001) was chosen as the setting.

The MMORPG scenario had teachers Dianne Hobbs, Mark Weber and their students re-construct the castle and characters of the time. This meant students accessed information on the historical events, planned and designed how to find access to the drawbridge, cellar or dungeon. Each student had a specific role to play based on the medieval society. Students needed to access information on a particular medieval castle to construct their own model using the encore MOO environment. Extension activities in the castle involved finding the treasure or to create some robots and contribute to the future direction of the narrative.

The plan also had Leah and the two high school teachers co-construct secret rooms and give students hints or instructions on how to find it and gain more treasure, power, etc, according to games design theory. Meanwhile the CSU research team provided theoretical, technical, developmental and analytical support to the project.

Short history of Rochester Castle

During the reign of Richard the Third, Prince John takes over the throne of England, while King Richard is fighting the Crusades and then is subsequently taken hostage. Prince John allowed allies to build a series of castles to strengthen his position. The Bishop of Rochester was one such ally and he petitioned Prince John for permission to build a castle. He needs to recruit a work force to build this castle along strict rules of the time, then staff it with soldiers to defend it against attackers. Rochester Castle is then attacked and is under siege in 1215 AD. It was the longest and best documented siege of the time. (Cook, 2001).

Multimedia and game design architecture

The teachers and the research team were to create a virtual environment where Society and Environment (S&E) students are involved in the building of Rochester Castle (1087-1100 AD) in order to explore the outcomes in the Year 8 unit "The Changing World" in an interactive way. This could also help to determine whether role play games in enCore Xpress can be used for developing social and emotional skills that can then be applied in real life scenarios.

Our MMORPG is based on an encore Xpress-enhanced MOO (Multi-User domain Object-Oriented). MOOs have their origins in multi-player text adventure games (Bartle 1990). A MOO is a collection of objects. Locations (where avatars may gather), avatars (online representations of participants) and things (balls, boxes, shoes or anything else) are objects. Objects have attributes including actions they take in response to commands or messages. Some objects, called containers, can hold other objects. Locations, boxes and pockets are examples of containers. Objects are classified into a hierarchy with general characteristics inherited from higher level objects and special characteristics distinguishing objects lower down the hierarchy.

People connect using a network client to a MOO network server. Once correctly connected, people control their avatar and other objects by typing commands. An encore Xpress-enhanced MOO shows information about the current location, including contained objects and avatar icons, in one web browser frame, and dialogue in another frame. A third frame can be used to view other web pages. Many actions can be performed using a mouse.

Each person's avatar has a class or level. At the bottom are temporary avatars called guests who can only move around and chat to other avatars. A player has a more permanent identity but can still only move and chat. A builder can create new objects (except avatars) and give them attributes such as a description. A programmer can create new behaviours or actions (verbs) that can be applied to objects. Verbs are written in the MOO programming language. Wizards can create and control other avatars and change attributes such as location, password or quota. They can gag, disconnect or even re-cycle avatars. The Arch Wizard at the top has ultimate power and responsibility.

It takes about ten minutes to learn to be a guest: to read location descriptions, move to other locations and chat to other participants. Most people who can type take to it very easily. Budding writers like to be builders and people who like logic, syntax and semantics can learn to be programmers. Good leadership skills are required for wizards. It takes a very special person to learn to be a good Arch Wizard.

Rochester Castle MMORPG is hosted on the Internet Special Projects Group server at Charles Sturt University at http://ispg.csu.edu.au:7690/. The system is run using enCore Xpress 4.0.1 (enCore Open Source Project at http://lingua.utdallas.edu/encore/) under the FreeBSD operating system as used by busy sites like yahoo.com. The welcome screen is shown below in Figure 1.

Figure 1

Figure 1: Rochester Castle welcome screen at http://ispg.csu.edu.au:7690/ by Leah Irving.

Design aspects in detail

The students were responsible for designing and developing the game, along a scenario presented by the teacher, Dianne Hobbs. Students were given a plan of the castle as a fortified Manor house (Keep). The year 8 and year 9 students worked over 6 weeks in Term 3, 2004, in the computer lab and doing research and preparation for the game in the regular classroom.

MMORPG topology

The curtain wall featured as the great corridor, from which all other rooms in the castle come off. The central hall is where life takes place, and the big grassy area (Figure 2) is called the common. Mud maps and layout diagrams are used to show gamers how to move throughout the castle and the grassy area that surrounds it. Bishop Langton was appointed by King John to administer the castle after the siege of the barons.

Figure 2

Figure 2: Rochester Castle layout and navigation map as seen from the grassy area.

Playing the game

The game starts when the Archbishop of Rochester petitions the King John for a Cantenation. This is granted and the Bishop must now recruit his building force. The guilds involved are Stonemasons, Blacksmiths, Farmers guilds, etc, and student groups will take these on allocated by ballot. Students may trade their guilds at the start of the game. Students are rewarded for participating in the game by acquiring barter tokens and gold. Any disputes are settled first by the Bishop of Rochdale or in the extreme by King John. Daily law enforcement is carried out by the Bailiff. The castle is built by programmers creating rooms in the MOO. Students apply to the wizard for permission to create rooms. Students may be allowed to form their own kingdoms and may even be given permission to be a programmer. To build anything, they need to get permission from the Archbishop of Rochester to be allocated permission as a builder (or programmer).

Another class taught by Mark Weber was the attacking force. They research the engines of war for medieval times and needed to organise themselves into teams, then attack the castle. They then laid siege to the castle.

Development ideas for Rochester Castle

Developing a 'Turing Bot' to act as a guard on the drawbridge provided some challenges for programmer, Allan Bytheway. The idea was that the 'bot' (affectionately named Rod) would interact with the school students involved in the project and challenge them to provide a password. This password, if correct, then allowed them entrance to the Castle. If they did not have the correct password they would be directed to move to the 'field' surrounding the castle. The speech patterns to be used by the 'bot', like other characters, were to be based on Ye Olde English. This raised some security issues, so the idea of using a 'key' was developed.

Develop a 'key' in place of the guard 'bot'

To solve the security issue of keeping the players to their appropriate parts of the game, a 'key' object was created. This key was based on a generic thing object and a 'turn' verb was written which moved the players to the 'Big Grassy Area' inside the walls of the castle which surrounded the castle keep. The key was located in the 'Rochester Castle drawbridge' room, which was the default login position for all players. A list of players authorised to use the 'key' was coded into a property of the 'key'. If an unauthorised player used the key, they were told to go to the surrounding field. Code for 'turn' verb of the 'key' object:

1: if (is_member(player.name, this.valid_workers))
2:   player:tell("Ye open the drawbridge door to enter the Castle.");
3:   player.location:announce(player.name, " enters the Castle");
4:   player:tell("Ye see the big grassy area surrounding the castle keep.");
5:   move(player, #170);
6: else
7:   player:tell("Ye are not allowed to enter the Castle, be gone with ye!");
8:   move(player, #173);
9: endif

Develop a 'stair' object to move the players inside the castle

The real Rochester castle had only a north stairwell and a south stairwell which lead to 6 different galley levels. To simulate moving between the six different levels of the castle, a separate 'stair' object (based on a generic_thing' object) was created which moved players to each of the other levels. This required the writing of a verb which moved players to the appropriate location. Two versions were written, a 'climb' verb and a 'descend' verb, to be used as appropriate i.e. if a player wanted to go to an upper level then the 'stair' object contained a 'climb' verb, conversely if a player wanted to go down to a lower level then the 'stair' object contained a 'descend' verb. The code for the verb to be used for each stair object is based on the following examples and was modified to suit each location.

Code for 'climb' verb of the 'stair' object:

1: player:tell("Ye start to climb up the stairs, as ye climb ye look out the 
   window that is located half way up and ye see the forces of King John 
   working on building siege engines. After a few minutes ride ye arrive 
   in gallery 2");
2: player.location:announce(player.name, " takes the stairs");
3: move(player, #542);
4: player.location:announce(player.name, " arrives in the gallery");

Code for 'descend' verb of the 'stair' object:

1: player:tell("Ye start to climb down the stairs, as ye climb ye look out the 
   window that is located half way up and ye see the forces of King John 
   working on building siege engines.  After a few minutes walking ye arrive 
   in the castle keep.");
2: player.location:announce(player.name, " climbs down the stairs!");
3: move(player, #697);
4: player.location:announce(player.name, " arrives in the keep!");

Fighting during the game

The issue of simulating the fighting required for the game it was suggested that players 'page' each other with carefully crafted, aggressive but appropriate messages. Development of a 'foundation' object was made, to move attacking siege players inside the walls of the castle.

The game required the foundations of the castle to be breached, by the players laying siege to the castle, by burning pig fat in tunnels dug down the foundations. To simulate this situation, a 'foundation' object (based on a 'generic_thing' object) was created containing a 'crumble' verb. This object was located in a room object called 'tunnel' linked to the 'Countryside'. This verb moved the player using it inside the castle walls.

Code for 'crumble' verb of the 'foundation' object is below:

1: player:tell("Ye see that the pig fat fire has burnt the foundations and 
   it starts to crumble, ye then are able to enter the castle");
2: player.location:announce(player.name, " moves through the hole in the 
   foundation");
3: move(player, #640);
4: player.location:announce(player.name, " arrives in the east tower");

Interim Analysis of online gamer behaviour

Interim analysis of gaming participation showed that 53 student online gamers used Rochester Castle for a total of approximately 223 student hours. Table 2 below records the highest and lowest number of hours as well as the total average time spent playing the game over all sessions. Figure 3 reveals the pattern of time spent playing the game by the students.

Table 2: Student participation statistics for Rochester Castle


Gaming Hours
Highest participation20.55
Lowest participation0.1 (5.7 min)
Average no. hours5.62

Figure 3

Figure 3: Rank order of the time spent (hours) gaming by all students at Rochester castle.

The student player called "Charles the Squire" had almost double the average playing time of the next 5 highest in the list of top 25 gamers by time spent at Rochester Castle. The gradient of the curve gets significantly steeper after the player called "Simon the Steward". An ethnographic study of the Rochester Castle game community (Eustace, 1998) will make an interesting follow up analysis.

Table 3: List of top 25 gamers by hours spent

CharacterHours
Charles the Squire20.55
Sir Eramus13.27
Reginald the Squire12.60
Claramunda10.95
Andrew10.46
Simon the Steward8.27
Oliver the Don7.66
Ralph the Squire7.17
Bryan the Squire6.88
Valentine the Cotter6.83
Agnes6.53
William the Squire6.07
Warren5.88
Isabelle5.88
Walter de Hugh5.48
Griffin5.47
Zacheus5.25
Apothecary4.91
Sir Bartholemew4.71
Lord Nicholas4.68
Geoffrey the Bailiff4.50
Lord Richard4.01
Maxwell3.87
Sir James3.84
Avis3.73

Table 4: Summary of non-student gamers participation

Type of participantNamesHoursSessions
Research teamOphelia87.8771
Castledog25.8644
Bailiff Bill7.7833
Total121.51
SWSHS TeachersLangton10.3320
King John8.4026
Total18.74

Conclusion

The development, testing and investigation of the use of a massively-multiplayer online role-playing game (MMORPG) to facilitate collaborative learning in the English and Science domains has been a rewarding challenge and learning journey for the research team, staff and students at Swan View Senior High School.

The average time spent by teachers supporting the game (about 9 hours) is about double the average time spent by students and seems to be typical of the normal lesson preparation time. This interim analysis suggests that instructional gaming with Rochester Castle and similar interactive MMORPGs is about the same workload as conventional teaching where each hour in the classroom is matched by an hour of preparation beforehand, as well as time spent mastering the ICT, games design and research skills needed.

The authors are hopeful that further analysis of data will confirm the educational value of instructional gaming with an EnCore Xpress MMORPG. Further analysis will examine:

The authors plan to conclude analysis and produce a final report by May 2005.

Acknowledgements

The authors wish to thank and acknowledge the support given by those people and organisations on the following list:

References

Bartle, R. (1990). Interactive Multi-User Computer Games. [viewed 18 Nov 2004] http://mirrors.ccs.neu.edu/MOO/papers/mudreport.txt

Biggs, J. B. (1987). Student approaches to studying and learning. ACER, Melbourne.

Böhm, Davis, Meares and Pearce, (2002). Global students mobility 2025: Forecasts of the global demand for international higher education. Media briefing, IDP Education Australia. [viewed 24 Feb 2004, verified 23 Nov 2004] http://www.idp.com/marketingandresearch/research/internationaleducationstatistics/GlobalStudentMobility2025.pdf

Cook, M. W. (2001). Rochester Castle.Castles, Abbeys and Medieval Buildings. http://www.castles-abbeys.co.uk/Rochester-Castle.html [viewed 18 Nov 2004]

Dean, A. (2002). Telelearning: Invention, innovation, implications: Towards a manifesto. e-Journal of Instructional Science and Technology, 5(2) [viewed 12 May 2003] http://www.usq.edu.au/electpub/e-jist/docs/Vol5_No2/dean_frame.html

Eustace, K. (1998). Ethnographic study of a virtual learning community. Internet Research: Electronic Networking Applications and Policy, 8(1), 83-84.

Eustace, K. (2003). Educational value of e-learning in conventional and complementary computing education. Proceedings of the 16th Annual Conference of the National Advisory Committee on Computing Qualifications. Palmerston North, New Zealand. 6 July 2003.

Johnson, D. W. & Johnson, R. T. (1999). Learning together and alone: Cooperative, competitive, and individualistic learning (5th ed.). Boston: Allyn and Bacon.

Johnson, D. W., Johnson, R. T. & Stanne, M. B. (2002). Cooperative learning methods: A meta-analysis. [viewed 1 Aug 2003 at http://www.clcrc.com/pages/cl-methods.html, verified 23 Nov 2004 at http://www.co-operation.org/pages/cl-methods.html]

Kemmis, S. and McTaggart, R. (1988). The Action Research Planner, 3rd edn, Geelong: Deakin University.

Kolb, D. (1984). Experiential learning: Experience as the source of learning and development. New Jersey: Prentice-Hall.

Lepper, M. R. & Malone, T. W. (1987). Intrinsic motivation and instructional effectiveness in computer-based education. In R. E. Snow & M. J. Farr (Eds.), Aptitude, learning and instruction. Volume 3: Cognitive and affective process analysis. Hillsdale, NJ: Erlbaum

Mason, R. (1991). Evaluation methodologies for computer conferencing applications. In A. R. Kaye (Ed), Collaborative Learning through Computer Conferencing. Springer-Verlag, Berlin.

Mason, R. (2002). The Global Classroom. In H. Adelsberger, B. Collis and J. Pawlowski (Eds), Handbook on Information Technologies for Education and Training. Springer, Berlin.

Rieber, L. P. (1996). Seriously considering play: Designing interactive learning environments based on the blending of microworlds, simulations, and games. Educational Technology Research & Development, 44(2), 43-58.

Rieber, L. P. (2001). Designing learning environments that excite serious play. Proceedings 18th ASCILITE Conference. Melbourne, 9-12 December. http://www.ascilite.org.au/conferences/melbourne01/pdf/papers/rieberl.pdf

Savin-Baden, M. (2000). Problem-based Learning in Higher Education: Untold Stories. Buckingham. Open University Press.

Seay, A. F., Jerome, W. J., Lee, K. S. & Kraut, R. E. (2004). Project Massive: A study of online gaming communities. Proceedings of CHI 2004. April 24-29 2004, Vienna, Austria. [verified 23 Nov 2004] http://www-courses.cs.uiuc.edu/~cs598kgk/papers/p1421-seay.pdf

Authors: keustace@csu.edu.au, malee@csu.edu.au, gfellows@csu.edu.au, abytheway@csu.edu.au, gilriv@iinet.net.au

Please cite as: Eustace, K., Fellows, G., Bytheway, A., Lee, M. & Irving, L. (2004). The application of massively multiplayer online role playing games to collaborative learning and teaching practice in schools. In R. Atkinson, C. McBeath, D. Jonas-Dwyer & R. Phillips (Eds), Beyond the comfort zone: Proceedings of the 21st ASCILITE Conference (pp. 263). Perth, 5-8 December. http://www.ascilite.org.au/conferences/perth04/procs/eustace-poster.html

© 204 Ken Eustace, Mark Lee, Geoff Fellows, Allan Bytheway and Leah Irving
The authors assign to ASCILITE and educational non-profit institutions a non-exclusive licence to use this document for personal use and in courses of instruction provided that the article is used in full and this copyright statement is reproduced. The authors also grant a non-exclusive licence to ASCILITE to publish this document on the ASCILITE web site (including any mirror or archival sites that may be developed) and in printed form within the ASCILITE 2004 Conference Proceedings. Any other usage is prohibited without the express permission of the authors.


[ ASCILITE ] [ 2004 Proceedings Contents ]
This URL: http://www.ascilite.org.au/conferences/perth04/procs/eustace-poster.html
HTML created 23 Nov 2004. Last revision: 23 Nov 2004.