Re: Gaming Engines

From: "Enno Rehling" <schick.mir.spam@gmx.de>

Date: Tue, 24 Oct 2000 00:00:00 +0000

"MindShift Design" <teammsd@mindshiftdesign.com> schrieb im Newsbeitrag
news:8t4coo$bdc$1@nnrp1.deja.com...
> I posted this to the PBMDesign email list on Egroups as well and
> thought I'd post it here to start some discussion (no shortage of
> debaters in here <g>)...

I think everyone who's done what we're doing has thought about this to some
extent. There are pros and cons to building an engine, I think. While the
Pros should be obvious, there are several cons that are easily overlooked,
I'll concentrate on those. I'll take Atlantis-/Olympia-/Eressea-Style Games
as an example, because that's where I come from and what IMHO most people
tend to play.

Seperation always means defining an interface between engine and rules. If
you keep the engine small, the interface will probably be small as well, but
you're not really gaining much in terms of reuse.
If what you consider your engine is bigger (I'm aware that this could be
multiple layers of an engine) you're specializing the game. Once I put a
combat system in the engine and only leave weapons, hitpoints, etc outside,
I'm pretty much focused on a special type of game. However, making things
like a combat system into a pluggable module would mean that we're talking a
small, well-defined interface for that as well. However, the smaller the
interface, the less variation iside the module. In a rich game, magic and
combat are closely interwoven. It's not only really hard to design them as
separate modules; making an interface always means cutting back on some
really cool special effects that are only possible because both modules know
of each other.

You can get a pretty high degree of what I'd like to call interaction
between the models if the objects shared between them (in this case the
weapons, the spells, the longer-lasting magical effects on combatants or the
buildings they are fighting in, etc.) are complex objects (in our case,
since we're doing C, we have a growing number of function pointers in the
descriptive structures of these obejcts). However, you're making the
development of such a module more complex, because what you're really doing
is widening the interface.

If building an "engine" for PBEMs is just about turns, orders or reports -
well, that's the smallest part of any game, and while you can do this well
or pretty dumb, there's only so much work in that. The real work is the
intermediate parts - Trade, Combat, Magic, Movement, Overall Economy, and
balancing them. A game engine that helps me do this well has a lot of game
logics in it, and takes away a lot of the design freedom.

As for an engine that pumps out reports in PDF, HTML, etc - I don't think
that a) you can write this general enough to support more than a very
limited type of game (e.g. none that does Galaxy, Atlantis and United). And
b) it's not the way to go for a mass market. People are demanding more and
more graphical clients. Our game doesn't have one, but a format that's
easily read by a computer program (These days I'd use XML, honestly) and two
or three really good clients written by players. I'm sure that without them,
we wouldn't have half the population that we do.

Enno.

I've been absent from this group for a while, so I'll give a bit of my
background: I'm not hosting a commercial game. I'm one of three developers
of Eressea, probably the biggest german-language PBEM. We have a world more
than 4 years old with 1200+ players, and the game is free, with about 50
people joining every week. We're putting about 30 hours of combined work
into the game, and developing it from one turn to the next - still adding
stuff. We've been thinking about going commercial, which is why I pursued
the move towards a clearer seperation of game engine and rules -- it's far
easier to set up tailored worlds that you can still extend or change around
when all you're affecting is the code for that one world, not all of them.
We've also added localization, so it's technically possible to play one game
of Eressea with people from all over the world. So yes, I've got some
experience :-)
--
Enno Rehling         enno [at] eressea-pbem.de
Eressea PBeM         http://eressea-pbem.de/

References

Gaming Engines
MindShift Design <teammsd@mindshiftdesign.com>
Tue, 24 Oct 2000 00:00:00 +0000

Referenced By

Gaming Engines
MindShift Design <teammsd@mindshiftdesign.com>
Tue, 24 Oct 2000 00:00:00 +0000