[MUD-Dev] Virtual machine design

Alex Stewart riche at crl.com
Mon Apr 19 12:19:15 CEST 1999


> In other words, at least the players I have seen, can barely write
> decent english (or any other language) or hold a thought long enough
> to write an entire zone.  A few simple scripts have been attempted by
> others, but nothing like conditional branches or anything...

Boy, we are just full of condescention toward our users, aren't we?

Sheesh..  Had it ever occurred to you that maybe the reason your users are all
stupid is because your system is one that only attracts stupid users (possibly
because of its lack of support for anything more sophisticated)?

Having worked for many years with programmable MUD systems and their users, I
can personally say that there is a _huge_ number of users out there who are
capable and interested in learning such languages and using them to do, in some
cases, truly amazing things.  I've worked on systems whose entire code base has
been continually evolved over long perioods of time by literally hundreds of
contributors, most of whom were simply average users who decided they wanted a
feature, and so coded it.

> I doubt that anyone will put forth the effort to learn a new language
> to write mud code for your server.  If they are the kind of person
> that will, then they really want to write the server instead :)

You seem to be really missing the point here.  If the server is defined by its
internal code, then by writing code in the internal system, they _are_ writing
the server.  Moreover, with proper security constraints, average users can do
server hacking on many aspects of the system, without even having to be
the system admin, without needing to work with tedious details the system
can handle better anyway, and can see their results immediately (and share them
with a pre-existing community which can then extend them in other ways).

> I'll offer an anecedote:  I used to play a Diku (JediMUD).  I got
> tintin++
> and learned how to use it's 'programming' language (ie a bunch of
> pattern
> matching and scripting).

If this is your definition of "programming" then I'm beginning to understand
your myopia.

> the grail that led me to code.  I don't think there is anything short of
> an OS that could hold my attention, and the possibilities I wanted to
> code.  That is why I doubt that anyone would really want to use such
> an in-game coding language.  It is either too restrictive, or you've
> re-written UNIX.

You've obviously never looked at MOO, LPMUD, or ColdMUDs (there are several
others, but these are the most notable).

> Be careful here.  It may be more efficient to run a few scripts and have
> your normal code be native, than put almost everything in interpreted
> code.

Oh, yes, it would definitely be more efficient to put things in native code,
but sometimes efficiency isn't everything.

> Those java folks weren't stupid,

Actually, I have to disagree with this.  Some of the design decisions in Java
were, in my opinion, actually rather stupid (or at least poorly thought out),
but that's a side issue...

> and they had extreme ammounts of manpower and time...

This isn't really true either.. in actuality most of Java was done by only a
few guys and then had its goals changed several times until it was rushed into
production by Sun when they saw an opportunity to grab a market.

> I wonder how much easier it will be to add your own code at run time,
> versus
> writing it in c/c++/java/?? and promoting into the [CVS??] load.  Either
> way you have security issues that will be hard to beat.

These security issues have been beaten quite effectively in many systems
already.  Look around a bit.

-alex
-------------------------------------------------------------------------------
     Alex Stewart - riche at crl.com - Richelieu @ Diversity University MOO
                         http://www.crl.com/~riche
           "For the world is hollow, and I have touched the sky."


_______________________________________________
MUD-Dev maillist  -  MUD-Dev at kanga.nu
http://www.kanga.nu/lists/listinfo/mud-dev




More information about the mud-dev-archive mailing list