[Ocaml-biz] the game market

[Brandon J. Van Every]

Ok, I brainstorm some thoughts on possible markets for showcase
projects.  No promise that they are good thoughts.  Please do the same,
as I can only comment on my (limited) areas of knowledge about the OCaml
landscape.

I'm a Game Designer and game technologist.  My personal goal is to
replace C++ as the dominant language in the games industry.  I believe
OCaml can be that 'successor language', with sufficient legwork.  I
concentrate on the Windows side of OCaml problems because commercial
computer game development is mainly a Windows thing.  Macs are an
ancilliary market and Linux isn't a relevant market at all.

Consoles are relevant, but I don't see OCaml getting ported to this
generation of consoles.  They aren't specced high enough, and they are
not open source friendly anyways.  I made a decision to forego consoles
some time ago, knowing that tool choices on consoles will be regressive
for the next few years at least.  If I strike it rich with a PC game,
I'll use that war chest to get into the console space.

The best open source 3D engine out there is The Nebula Device.
http://nebuladevice.cubik.org  It is Windows-centric but Linux builds
are starting to appear now.  It is MIT licensed.  Radon Labs
http://www.radonlabs.de is a commercial game company.  They contributed
the 1st release of the engine and still add new chunks from time to
time.  Their business model is to give away the engine but charge 464
EURO for their Maya export plugin, if such you want.  No high quality
open source exporter plugins currently exist for Nebula2.  Some
half-baked ones are available.  Nebula1 was commercially proven in
several titles, most notably Project Nomads.
http://www.project-nomads.de/  Nebula2 isn't commercially proven yet.  I
hope some Nebula2 titles ship this Christmas so that we can actually see
it *doing* something.  Right now, there are no compelling eye candy
demos for Nebula2.

Nebula2 is written in C++, so I have contemplated wrapping it with OCaml
using SWIG.  http://www.swig.org  As of April 2004, SWIG's OCaml support
was used only by the author of the support.  Thus, nobody could possibly
consider it ready for prime time.  In the best possible case, it
couldn't be ready for prime time for at least 1 year.  That's assuming
SWIG got used immediately for a high profile project, the project was
widely disseminated, and it got lotsa testing.  Average case is 2..3
years, worst case is nobody ever uses it and it never goes anywhere.

An OCaml 3D engine has enormous 'kewl' value for techies, and is a great
basis for doing all kinds of bells 'n' whistles marketing demos.  It
could definitely be a 'killer app' for OCaml.  This is, however, not an
easy project.  Either one has to seriously improve SWIG to make wrapping
Nebula2 worth it, or one needs to start from scratch with a C, OpenGL
approach.  Nebula2 isn't a 'best fit tool' or somehow ready to go.  I
bring it up because I'm fully committed to solving the 3D engine problem
somehow anyways.  If I had allies, it would be easier.


Cheers,                     www.indiegamedesign.com
Brandon Van Every           Seattle, WA

"The pioneer is the one with the arrows in his back."
                          - anonymous entrepreneur