[MUD-Dev] MMORPG/MMOG P2P design

Crosbie Fitch crosbie at cyberspaceengineers.org
Sat Feb 22 11:05:10 New Zealand Daylight Time 2003

From: lynx at lynx.purrsia.com

> Well, the question is, how do you structure an MMOG around the
> fact that you don't know how much you can trust your peers, and
> you certainly can't count on them to be up at all times?

You measure 'trustability' in the same way that humans do with the
people they deal with, i.e. you keep records of how reliable certain
PCs have been in the time you've known them. The longer they've been
reliable the more you can trust them. Moreover, there must be a
benefit to being trusted (a non-malicious benefit), i.e. improved
simulation quality. Someone breaks that trust... goodbye. Some
damage possibly, but we'll get over it.

> It seems to me that you have to invert the design: instead of
> having characters attempting to gain levels and precious items,
> start with the presumption that your characters already have all
> the levels and items they could possibly want.  What's left to
> want-- but the attention of your peers?

Interesting idea, but I wouldn't have thought the nature of the
system's design should that visibly affect the game design.

> So the challenge is to design a piece of the game world, over
> which you are the God, to be so attractive that other players will
> flock to it.  They take up the task of mirroring part of your
> world, feeding out such things as conversations and battles and
> what might be seen in some particular corner, but your server
> remains authoritative in case there is any question over what is
> where.

There is no server. Or on the other hand, every node is a server.

Authority is distributed.

> Perhaps if you sufficiently trust others, you might enter into an
> agreement that they can be considered authoritative as well, but
> then you run into data consistency issues.  (there is a Holy Grail
> kept in a glass case.  You've allowed servers A and B to know
> about it.  You go down, server B is down, server A has someone
> break into the case and steal the Grail.  Server A then goes down
> and you and server B come up, and assure people with calm
> confidence that the Holy Grail is still in the case.

Think in terms of 2 million PCs, not just 2. Massive redundancy.

However, yes, the thought experiment you describe is valid, but
there is no solution in such a contrived situation. Even so, it is
possible to ensure that there is a reliable way of arbitrating over
such cases (should they ever happen) and thus ensure the system
doesn't get its knickers in a twist.

> Could make an interesting design problem.

Yeah, I reckon.

Some more discussion here:


MUD-Dev mailing list
MUD-Dev at kanga.nu

More information about the MUD-Dev mailing list