[MUD-Dev] Re: PDMud thread summary

[Chris Gray]

[Darrin Hyrup:]

 >I also believe that an O-O approach to the design from the inset would be
 >the best... in both the implementation of the server, as well as in the
 >design of the internal language.  Ideally, the internal language should
 >support inheritance (at least single, although multiple would be nice in
 >some situations.)

I believe there are two aspects to the OO-ness of such a system. The
world the system models will contain lots of objects (swords, books,
furniture, decorations, etc. etc.), so an OO approach seems desireable.
The question of single versus multiple inheritance there will be quite
contentious, I believe. I would vote for single inheritance (because
of its higher efficiency), along with interfaces (like the Java ones).

The OO-ness of the core language is quite a different question. There,
you are more into the programming preferences of those involved. Also,
if a module is programmed in, say, C++, then a C++ object, with its
inheritances, will likely have nothing much to do with the objects in
the world that are stored in the database and manipulated by the in-MUD
programming language. Quite different levels.

Now, if the entire system is one cohesive C++ system, then you *can*
make the two object systems one and the same, but you don't have to.
I would argue that even then you don't want to. The inheritance rules,
data representation, etc. of C++ objects is not necessarily what is
best for representing the world in a dynamic MUD system. LPC tends
to do things mostly that way (from what little I've seen), but other
systems don't. These are all details that have to be ironed out. All
I'm saying here is don't jump to early conclusions.

--
Chris Gray     cg at ami-cg.GraySage.Edmonton.AB.CA