The UAD MUD

1
The UAD MUD

• Michael Heron

2
The Problem

• We have numerous problems when teaching students
  how to code.
• Its a very difficult occupation!
• One of the problems is the bag of tools
  problem.
• We give a bag of tools, but few problems to
  solve.
• Often this is because of the restricted range of
  potential tasks.

3
The Problem

• Those with some experience of programming often
  have a library of pet projects.
• When learning a new language, one of these can
  serve as a platform for development.
• This library of projects comes with experience.
• It is developed through encountering problems
  that we know how to solve with some code, if only
  we had the time to sit down and do it.

4
The Problem

• Most students lack this experience, and so have
  little opportunity to develop skills outside of
  the university.
• This is disadvantageous, since the only way to
  learn how to code is to sit down and write.
• The tasks we set in first and second year are
  largely proof of concept.
• Useful from a technical point of view but often
  without long-term merit.

5
The Problem

• Providing larger projects in the past has helped
  utilise the talent of stronger students.
• Names Removed to protect the guil er innocent
• This is difficult to sustain on a long-term
  basis, and excludes the weaker students.
• What would be useful is a self-sustaining project
  that can accommodate all skill levels in varying
  roles.

6
A Solution?

• Enter the UAD MUD!
• A multi-user textual game environment written in
  an event-driven, object oriented programming
  language.
• It is self-sustaining.
• Most of the development will be driven by the
  students, with only minimal staff involvement.
• It is long-term.
• Depending on the theme, development can continue
  indefinitely.

7
UAD MUD

• For my sins, I am a developer on a MUD based on
  the works of Terry Pratchett.
• 200(ish) coders.
• 4000(ish) players.
• The codebase for this MUD is open-source and
  freely available for download.
• The codebase is broken up into two parts
• The driver
• The mudlib

8
The Driver

• The driver is written in good, old-fashioned C.
• The code is very complex most developers never
  touch the driver framework.
• The driver is essentially the operating system
  for a MUD.
• It is very similar to the Virtual Machine for
  Java code is written and then interpreted by
  the MUD driver.

9
The Mudlib

• The mudlib consists of all of the code written to
  run on the driver.
• The mudlib is written in a language called LPC,
  which is an object oriented C variant.
• Its not C.
• The language has a very high degree of
  transferability with Java and other C-type OO
  languages.
• The two main differences are that LPC provides
  multiple inheritance and function pointers.

10
What do we do with it?

• The main thrust of this project would be to
  develop an environment where students of all
  skill levels can contribute.
• Some MUDs require large amounts of technical
  know-how before any real contributions can be
  made.
• The Discworld Mudlib is flexible enough to allow
  for development on a wide range of projects.

11
What do we do with it?

• The first step is decide on a theme for
  development.
• An appealing idea is to develop a simulation of
  the University of Abertay and its surrounds.
• Firstly, an environment must be developed.
• In Discworld jargon, this is called building, and
  requires only a little technical skill but still
  allows for practise with coding fundamentals.

12
A Very Simple Room

• inherit "/std/room"
• void setup()
• set_short("main lecture room")
• set_long("This is the main lecture room
• for the university. There is a
• stage at the front of the
• room and hundreds of chairs
• stretching off into the distance.\n")
• add_property("determinate", "the ")

13
Building

• Building involves mostly descriptive writing.
• But its writing for an audience.
• It also demonstrates the use of
• Methods
• Parameter passing
• Inheritance
• As far as technical skills go, building is the
  simplest thing that can be done on the MUD.

14
NPCs

• The next step up is to develop NPCs, or Non
  Player Characters.
• These are little computer controlled characters
  that function independently of any players.
• NPCs range from the very simple, to the extremely
  complex.
• There is much scope for developing interesting AI
  routines of all levels of complexity.
• At the simplest end of the scale, it is only a
  little more complex than building rooms.

15
Items

• Of a similar level of complexity is the
  development of actual items that can be
  manipulated during the game.
• These can range from weapons and clothes to
  complex multi-functional devices.
• A rich environment comes from a combination of
  rooms, NPCs and items.

16
Quests

• Quests are puzzles that can be solved during the
  game for rewards or prestige.
• They can be very simple
• Take this thingy to the ogre on the hill
• But they can also be extremely complex, requiring
  thousands of lines of code to handle the logic.
• Writing quests is a task that is suitable for
  students of all skill levels.

17
Moving On Up

• Rooms, items and NPCs can accommodate students of
  low and intermediate skill levels.
• There exist many other areas supported by the MUD
  architecture in which stronger students can make
  more fundamental contributions.
• Inheritables
• Handlers
• Game logic
• Database routines
• Web programming

18
Moving On Up

• Examples of some more complex objects that I have
  written
• A web-based project management system
• A text based monopoly game
• A derby day style horse-racing game.
• A mafia family infrastructure that provides
  suitably interested players with a system for
  earning money.
• A warhammer fantasy battle style wargame based
  on Discworld.
• Trainable pets, in the style of the creatures in
  Black and White.

19
How Do We Do It?

• First of all, we need to set up an administrative
  structure suitable for supporting development.
• The model suggested here is based on the
  Discworld structure, and involves the creation of
  largely independent domains.
• Each domain has the responsibility for the
  development of a particular part of the MUD.

20
How Do We Do It?

• Each of the domains is headed by a domain leader,
  who is an experienced coder familiar with the
  processes involved in development.
• The domain leader is responsible for
• Designing the plan of development
• Subdividing the plan into projects
• Hiring coders
• Training coders
• Assigning work to coders
• Ensuring the quality of submitted work.
• Usually a significant chunk of this is delegated.

21
Domains

• Each coder is attached to one or more domains.
• Ensuring the co-ordination of all domains is done
  by a super domain called admin.
• Members of the admin domain are the ones who
  appoint coders as domain leaders, and handle the
  creation of domains when required.

22
Domains

• Some of the domains in Discworld
• Guilds
• Ramtops
• Ankh-Morpork
• Learning
• Each domain has a clearly defined area of
  responsibility.
• There also exist service domains
• Playtesters
• Learning
• Liaison

23
Domains

• The admin domain would be populated by interested
  members of staff.
• The rest of the domains can be populated by
  students.
• The important thing is to start off small and to
  expand as the student interest grows.
• To begin with for example, perhaps only an admin
  domain, a learning domain and a main building
  domain.

24
Domains

• This structure allows for a degree of stability
  in the short-term, and for a gentle transference
  of responsibility in the long-term as students
  graduate/quit.
• It also gives interested students a chance to
  practise project management skills while involved
  in the administration of a particular domain.
• So, assuming a main building domain, how do we
  get started?

25
Step One The Environment

• Rooms need to be written for each location in the
  university.
• The precise granularity is up to the domain
  leader, but this can be changed as time goes by.
• NPCs need to be written to represent interested
  members of staff.
• Ideally, well write our own.
• Items need to be written to simulate common
  objects in the university.
• All interested students will be able to
  participate in this.

26
Step Two Stuff To Do

• The environment itself is pretty stale, since
  theres little interactivity.
• We need to provide a range of activities to
  interest people.
• On one level, the MUD itself will be
  interactivity enough.
• A social representation of the university.
• On another level, we need things to entice people
  to play.

27
Step Two Stuff To Do

• Much of the code I have written for Discworld can
  be adapted in one way or another.
• Students should be encouraged to submit ideas for
  things to do
• Quests
• Games
• Toys
• There is very little limit on what is possible.

28
Step Three Develop a Community

• This is perhaps the hardest part.
• There is a level of critical mass required to
  ensure a steady stream of interested students who
  want to contribute.
• Previous experience with the VMS River system in
  the university, and the current usage of MSN
  Messenger in the labs suggests that a MUD as a
  social environment could be successful in
  generating interest.

29
Step Four Goto Step One

• As domains are developed, fleshed out and
  expanded, we can add new domains.
• This is directly related to the number of
  students playing and contributing.
• Once the main building is done, we can move onto
  the old college.
• Then Bell Street.
• Then the rest of Dundee.
• Then the world.
• (Bwah-ha-ha-ha!)
• We can continue to expand indefinitely.

30
What Does the MUD teach?

• On one level, it teaches coding.
• On another, it teaches the things that we can
  only emphasise as important, but rarely
  demonstrate within the limits of proof of
  concept assessments
• Change management
• Project management
• Maintenance
• Revision control
• Documentation
• Efficiency
• Literacy

31
What are the benefits?

• There are a number of benefits to be gained by
  running our own MUD
• Low maintenance most of the work will be done
  by students.
• Low cost a moderately low spec PC can handle
  even fairly substantial player loads.
• Collaborative development There is a sense of
  community within domains.
• Peer reviewed feedback Better coders can
  provide useful feedback, and players can likewise
  comment on the MUD from their perspective.
• All skill levels are provided for No one is
  left behind.

32
What Are The Problems?

• Its entirely text-based.
• A bit old-fashioned perhaps for the Quake
  Generation.
• It is very much dependant on getting people
  interested.
• Many MUDs die simply because there are no
  players.

33
How Do We Know If It Works?

• My experience of the MUD has been very positive
  with regards to programming.
• The experience of many Discworld coders is
  similar.
• However, we should have some metric by which we
  determine if the MUD is genuinely improving
  student performance.
• In the short term, we have pass rates for
  programming modules.

34
How Do We Know If It Works?

• It is the long term however that provides the
  most interesting results.
• Many of the skills the MUD teaches are never
  assessed.
• Part of the benefit as a coder comes in providing
  a development environment for flexing programming
  skills.
• Part comes from the confidence that this
  experience can bring.

35
How Do We Know If It Works?

• Such things are difficult to quantify short of
  self-reflective evaluation.
• This doesnt even take into account the ancillary
  benefits that may be observed as a side-effect of
  providing a UAD MUD.
• These benefits are primarily social.
• So, the answer is I dont know.

36
Conclusion

• A UAD MUD offers a chance to implement a
  self-sustaining, student-driven environment with
  only a little overhead in terms of cost.
• The benefits of such an environment are many and
  varied, although actually quantifying them is
  somewhat more difficult.
• The MUD can cater for all levels of experience
  and LPC provides a large degree of language
  transferability that allows for concepts explored
  to be used with more common languages.