[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Linux GameDev IRC meeting (fwd)



Tristan Wibberley <twibberley@llamacom.com> writes:

> Jeff wrote:
> > > Notes:
> > > * The commercial people will be interested most in *portable* APIs/SDKs/...
> > > Itīs easier to convince them to use some new library if this lib
> > > supports Win32 *and* Playstation *and* MacOS *and* Linux in a portable
> > > way instead of *only* Linux.
> > 
> > it is folly to try to support cross-platform to the PSX...I am a
> > professional game developer (most of what I do is on the PSX, been
> > doing that platform for about 2.5 years now) and it is so proprietary
> > that it would be a large waste of memory and speed (which is sorely
> > needed on a 2 meg 33MHz machine) to try to put a layer in between the
> > game and the hardware API (which, given, sucks, but it's more in how
> > it's laid out more than it's speed).  Concentrate more on going
> > linux/win32/mac/un*x than trying to port to a console as well (never
> > mind the NDA issues)...
> 
> Because there is a single standard hardware api on a console, much of
> the internal libggi fluff can be left out for cross compiles to those
> platforms. And because ggi is so modularised, you only need to load into
> memory the bits you need. This means that it is likely that only a few
> extra KB per extension will be used by libggi.

this is true, however most places that develop cross-platform
(pc/console) tend to write their games very system independant
anyways, only having the core drawing/io functions separate, this is
what we do.  The system/console where GGI will come in, and where *I*
personally feel the GGI people should be looking (instead of GGI) is
the Dreamcast (formerly the Katana), it will be running on a quasi-MS
operating system that supports directX, etc.  So really, the porting
to it would be easy in the first place, but perhaps even easier using
GGI.  The problem with using GGI on consoles is you have a LOT of
system level hardware/register tweaking to do to squeeze that little
bit extra out, depending on what type of world drawing algorithm you
have, what type of animation system you are using, etc.  

Basically my personal opinion as an experienced next-gen console
programmer is that trying to do a port of GGI to the PSX (without even
taking into consideration sony's utter unwillingness to participate in
something of that nature...their NDA system is quite harsh) isn't in
GGI's best interests.


j