Re: The PerfectMUD? Just an opinion

From: Project Leader McCoy (
Date: 10/31/96

On Thu, 31 Oct 1996, Michael Scott wrote:

> 	Hrm, what is the PerfectMUD?  And if EVERYONE had it, would it be 
> perfect anymore?

Well, I rather dislike getting involved in something this off-topic,

> 	While this may be a nice thought, putting together a 'better' starting 
> base ... IMHO, Circle (stock as it is) is one of the BEST coding bases, 
> that I've seen, and HALF of that is because of its simplicity, and 
> standardness.  

I agree.  As far as DikuMUDs go, Circle is by far the best.  Although
Merc, Rom, and so on have more features, they're so nonstandard,
undocumented, and buggy that it isn't worth it.

> By adding anything to it, you are going to see 100's of 
> muds with those same additions.  

And that's OK.  However, the fact that about 90% of the CircleMUDs out
there have the easy-color patch applied would seem to indicate something
about what people want in the stock code.

> And for a mud where people still debate 
> over which olc is best, there's NOT that much I would think of adding to 
> the base.  

Adding a standard OLC base would be good.  Colors would be.  In fact,
terminal autodetection would be good, too.  There's a long list of
features that are mostly under the hood, so to speak, that I feel should
be added to stock Circle code.  (But I can save the list for Circle 4.0.)

> In fact, I would almost say that there's too much of a world 
> to circle.   

I wouldn't almost say it.  I'd just say it right out.  There's too much of
a world to Circle.  A single zone should be available to help new admins
learn by example, but definitely no more than that.

> That encourages some guy who knows nothing about c or 
> running a decent mud, to pick up the code and run AS IS.  

OK, I have two major problems with this.  First, how do you expect him to
learn how to run a decent MUD without doing it?  We all have to start
somewhere.  I know that I knew absolutely nothing when I started playing
with CircleMUD from an admin perspective, and now I think I know quite a
bit.  Second major problem: why do you feel it's so vitally important to
limit the CircleMUD admin base to only people who know C?  Most people who
are extremely creative don't take to the boundaries and limits of computer
programming very well.  A good MUD world has little to do with computer
knowledge, and nothing to do with an understanding of C programming.

> I think it 
> should come with 1 room. (grin)   What about races and all this 
> 'standard' stuff that everyone should have when they start out with a 
> mud?  

I'd agree that Circle should come class-less and race-less if it weren't
for the fact that it's a chore to add classes and races.  Sure, it's
basically a no-brainer, but it involves lots of grep time, ten minutes of
mindless drudgery, and half an hour of debugging since it's almost
impossible to get everything the first time.  In other words: it's

A better system, which I hope will be in place soon, would allow for much
simpler modification of existing and addition of new races (or classes, or
guilds, or clans, or whatever).  This will also cut out about 150 lines of
useless code (in particular, the level crap) and make the whole thing
much, much easier to use.

> Remember the more complicated you make the base, the harder it is 
> for people like me to change it 

I disagree.  The mark of good work is that it *seems* simple without
*being* simple.  I've written some very small amounts of code like this,
so I know it can be done - a function or program that can handle any
possible contingency and do anything you want it to, but is still easy to
understand and (more importantly) easy to use.

Here, look at it this way.  I feel that CircleMUD should really be
something more like CircleAPI - we get a set of common functions and use
that to build our dream MUD.  So it doesn't *come* with races, but it
comes with a set of functions that will set it up for us.

Adding races should be this simple:

racelist.add("elf", "elves", ...);

Unfortunately, this requires C++, but something very similar and only
slightly more complex could be done in C.

> (I'm re-doing even the 'taken for 
> granted' things like how eq is handled, the D&D base itself, and what 
> is a dang stat thing called move, and how does that relate to anything?
> I'm tossing em in the trash..)

I think that a lot of these are hold-outs from DikuMUD.

> 	I think that the very 'stock'-ness of circle encourages the right 
> kinds of behaviors.   It allows good solid coders who want a challenge to 
> make something new.  And, if you get in-depth enough into it, most 
> well worked, custom crafted circlemud's DO NOT look alike when you get done
> with em.   Each can be a work of art on a canvas that just happened to be 
> supplied by the circle crew.

I think that the very difficulty of easily modifying CircleMUD to your
standards encourages the *wrong* kinds of behaviors.  If you can get a
patch from a friend that does roughly what you want, why not use it?
Especially if the alternative is spending an hour or two reinventing the

Code interchangeability is a big feature of the entire programming world.
In fact, #include files were developed almost exclusively so that programs
could use bits of other programs with a minimum of fuss (or so I've
heard).  C++ does the same in an even bigger way.  The UNIX world is
almost totally free, and everyone shares things from each other.  I really
don't see how this is a problem.

I agree - customizing a MUD is a requirement.  But let's not carry this
too far and say that everyone has to write his own OLC!  If I visit a MUD
that has flashy ANSI "graphics," lots of spells, OLC, etc etc, that won't
convince me to stay there.  What makes me stay on is the quality of the
writing.  If it reads like a book, I'm STAYING!  But if it reads like a
Choose-Your-Own-Adventure "book," I'm getting the hell out of there, even
if the damn thing plays music and dances for me.

I think you're overestimating the value of the interface.  But then, I've
always been more interested in substance than appearance - which is why I
don't use Windoze 95 and don't plan to.

> 	While I agree that patches can be USEFUL things, for additions 
> such as olc's, improvements and/or some great insight into the basics of
> the mud, I also believe that handing out too much of code like "here's 
> how to add x spell into your mud" and "here's the races code" can cause 
> the viscious strains of genericism, for which circle is ALSO known for.

I think that the primary reason CircleMUD is known for being generic is
because of all the areas distributed as "stock Circle."  My suggestion is
to make them available for download, but don't distribute them with the
regular CircleMUD - that way, they're there if you need them, but you have
to go out of your way a bit.

Here's my list of BAD things in a MUD:

1. Poor writing
2. An admin who can't spell
3. Stock areas
4. Little or no storyline
5. Only slightly modified areas (like "Olympus" for "Valhalla")
6. More admins than players (doh!)
7. Being overrun by about 30 shouted messages on logon
8. Having to go through a 15-minute questionnaire procedure
9. Having to be validated by e-mail by a person
10. Players with names like "Eidolon who is (bright red) REALLY PISSED at
(bright white) Trejak now because he's (bright green) SUCH a
(yellow) JERK!!!!!!!!!!"

Well... just some things to consider.

>                       Michael Scott -- "Living in the Eye of the Hurricane."
> 		      FLAMES/COMMENTS to
> +-----------------------------------------------------------+
> | Ensure that you have read the CircleMUD Mailing List FAQ: |
> |   |
> +-----------------------------------------------------------+

                          "The human brain is like an enormous fish --    it is flat and slimy and has gills through
                           which it can see." - Monty Python

| Ensure that you have read the CircleMUD Mailing List FAQ: |
|   |

This archive was generated by hypermail 2b30 : 12/18/00 PST