29th May 2001
When I started thinking about how I could re-jig my first C adventure adv to work with multiple users, I quickly realised that the source was totally unsuitable, and that I'd do better to start again.
What I came up with was MU (which stood simply for ``Multi User'') - a server/client system in which a single server maintained the game state and communicated with an arbitrary number of clients, handling their requests to interact with that state, and multiplexing the responses back out to the relevant other clients.
Through one long night of hacking from the ground up - 18th April 1988, for anyone who cares - I put together the first, simple version. It was what we would now describe as a ``chat'' system, similar in spirit to irc or to Napster's chat facilities. But back in those days, we'd never heard of such a thing.
In fact, it's possible that this very early MUNDI was the world's first multiplexing chat system. Certainly the author of irc states in his own history at www.irc.org/history_docs/jarkko.html that he started work on it in late August of the same year, but he refers to an earlier program called MUT (MultiUser Talk). I welcome confirmation or denial of the conjecture that MUNDI was the first multiplexing internet chat system.
Over the next few days, weeks and months, more and more features began to pile into MU, which was renamed MUNDI. This is the Latin word for ``world'', plus it's an acronym for ``Multi-User Networked Dungeon Interaction'', or perhaps ``Mike's Unfinished Nebulous Daydream Idea''.
The initial set of locations and objects appeared very quickly - basically limbo, the clearing, and the dozen or so locations surrounding it - along with player actions for manipulating the objects and each other. I remember being particularly pleased at getting recursive containment working:
You can also see: A huge wooden box which is open and contains: A number of frogs and toads. A little glass case which is closed and contains: A herring. A velvet glove. A gratuitous item of treasure.
At this stage, I had a playable, if incomplete, MUD that ran over TCP/IP. What's more, since I was at a University full of midnight-oil burners, I had an enthusiastic user community who lost no time in proposing all sorts of enhancements, many of which went into the game.
Another historical note here. Hans Persson's compilation of MUD information at www.lysator.liu.se/adventure/mud.txt (which, by the way, also mentions Causes of Chaos) says:"It wasn't sold commerically [sic], but got very widely played at Warwick University. I believe it was the first multi-user adventure game that worked across the internet."Sure enough, my perfunctory investigations have not turned up anything older that this: earlier MUDs such as AberMud and the Trubshaw/Bartle original were restricted to running all on one machine. Again, I welcome confirmation or denial of the conjecture that MUNDI was the first internet MUD.
Kimberly M. Antell's MUD information, found on the Early MUD History document at www.ludd.luth.se/mud/aber/mud-history.html further says that:
MUNDI was written by some MUD players at Warwick University and never distributed. It's mentioned here because it was probably the first British MUD written to be efficient and networkable.Which is nice.
After MUNDI was up and running at Warwick, other programmers - notably Harvey ``Pizza DeathMatch'' Thompson and Martin ``Midas'' Senior - added more features, including NPCs/mobiles and inter-player combat with weapons. By that point, though, I'd started to think about Herring (and might have left Warwick - I don't remember clearly.)