[MUD-Dev] Apple WWDC?

[Brian Hook]

At 02:56 AM 5/10/2002 -0700, Sasha wrote:

> Perhaps because maintaining heavily cross-platformed code can
> multiply the amount of labor it takes to do anything?

Only if you're inexperienced at it.  Writing cross-platform code can
be difficult at first because most programmers have to learn through
trial and error what features of a library, API, language, tool set,
etc. don't translate to new platforms cleanly.  Once you accrue this
knowledge, however, writing cross-platform code just isn't that
difficult unless the _majority_ of your code is system specific in
nature.  And for many applications, that's just not the case.

Case in point: my day job is writing mass market puzzle games.  Our
first game, Candy Cruncher, currently runs on OS X, Linux, and
Win32.  The next revision will also run on MacOS, Sharp Zaurus and
PocketPC.  Oh, and BeOS.

> erupt into another thread - I am picking up a little SDL here and
> there and will report if it turns into anything).

I have nothing but good things to say about SDL.  In fact, I have
nothing but good things to say for many cross-platform open source
toolkits that I'm leveraging.  zlib, libpng, ijg, Lua, etc. have all
dropped in and "just worked" every time now.  I also hear good
things about wxWindows, a cross-platform C++ GUI toolkit.

> explanations are available.  For example, that the expected return
> on investments in cross or multi platform development can be
> pretty poor.

Any intelligent game company will realize that the engineering costs
are almost trivial if you have a competent dev team that cares about
portability from the outset.  The REAL costs are maintenance and
support.

Right now I have a version of Candy Cruncher for BeOS just sitting
on my hard drive, but I can't release it because we're not in a
position to support it.  Basically we recompiled and ran it on BeOS,
and it worked, but we don't have the BeOS specific knowledge to
actually support it in case someone purchased it and had a problem.

This can be projected to larger game companies thinking about the
Mac or Linux -- the extra engineer involved in a port is
inconsequential to the cost of any major title, it's the training
required for support personnel to deal with the new system that is
going to be the headache.

I mean, Windows is a big enough support headache as is, but can you
imagine people with custom compiled kernels calling in for support
using their old version of Caldera Linux and a video card that
hasn't been manufactured in 4 years using open source video drivers?

> Even if there is a starving mac audience ready to drop their
> weight in gold on glorified graphical MUDs, the people making the
> games might not know that on authority which is sufficient to
> convince them to drop the money it takes to find out.

I think the Mac is a massive market waiting to be tapped for this
type of game.  They don't have anything except maybe WW2OL to play,
and it's not an RPG.

> What assurances would I have about dropping more to target a bunch
> of platforms?

I guess my feeling is that you're greatly overestimating the cost to
be cross-platform.  I'm a single programmer with one artist and yet
I've managed to develop small portable games that take 6 weeks to
develop now that my basic infrastructure is in place.  And since
they're games they have to deal with sound, graphics, input, threads
(for audio), disk IO, etc. which are some of the gnarlier aspects of
cross-platform development.  We also have a 3D app under development
that supports Direct3D and OpenGL and I anticipate will take less
than a week to bring up under SDL and get working on non-Windows
boxes.

Brian

_______________________________________________
MUD-Dev mailing list
MUD-Dev at kanga.nu
https://www.kanga.nu/lists/listinfo/mud-dev