Archive for September, 2004

About M3G Converter

Wednesday, September 29th, 2004

I work hard to finish M3G-Converter. Two versions will be available:
- M3G-Converter Free, you could convert 3DS, LWO,OBJ, MS3D, XM3G to M3G. No optimizations for triangle strips and no editing of M3G files.
- M3G-Converter Pro, same as free edtion, plus editing-tweaking of M3G files and optimization of files. This edition could work with other M3G tools (MAX7 exporter, Superscape, Hi-Corp, …) to power up M3G. No price available
I will release a beta of Free version in the end of october.

MEE3D tutorials soooon!

Monday, September 27th, 2004

I put last touch on my starting serie on MEE3D for M3G/MascotV3. Tutorial 1 (open a World), 2 (creating a mesh) and 3 (loading a Milkshape3D Model) are ready and I finish documents for them. MEE3D is a small layer to abstract and to simply acess to other 3D engines. So you write once, and build for several targets… It is a scale down of my MEE3D-Pro version, but it will be release under GPL

3D mobile gaming

Thursday, September 23rd, 2004

For me there are two unknowns in 3D mobile gaming:
- the 3D market : will you have have only hardcore gamers or casual gamers. Hardcore gamers ok you need a highend tech phone…But for casual gamers? They play 3D games on consoles, and see 3D movies. So maybe they need the same experience on mobile.
- Programmers, you have 2 profiles : J2ME developers, they need a simple API maybe M3G fit for them. And Game developpers they need an API like OpenGL-ES, and they prefers to use C++… Brew or Symbian 8.0 will be very attractive.

Java & 3D : API fragmentation

Tuesday, September 14th, 2004

I have talk to several guys (Intel, Orange, Nokia, In-Fusio and Sagem, …) about potiential fragmentation in 3D j2me market, that could cause delay for a quick adoption of 3D.
Yes 3D HW will be soon available, but a large base market will not include 3D chipsets. So In-Fusio done Exen2 with their own 3D, and Hi-Corp continue to develop their own Mascot API. Their is a need for a low end 3D API for mobile including Fixed Point. For example Sagem have fine tuned their VM so that it works very fast on a simple ARM7. Imagination Technologies with their PowerVR solution is based on a fixed point oriented 3D rasterizer solution. And Intel Xscale doesn’t include floating point. The mobile market is driven by low power but also by low cost. And it will continue to use ARM7 and no floating point. Maybe before creating M3G API, the best was to create a low end fixed point engine standard like Mascot API. And migrate smoothly to a highend 3D standard…
Now the mobile market is fragmented in low engine not standardized (In-Fusio, Mascot, …) and M3G a big engine not fited for small phones like SE K700i and Siemens S65. I hope I am wrong, and fast M3G phones will dominate (Nokia 6630 will be a good one). But targetting JSR239 to high end phones like M3G seems not to be the best way.
So will j2me developers and game developer will work on all devices or only target certain devices? Or maybe wait for best days?

OrangeCodeCamp

Tuesday, September 14th, 2004

I was at OrangeCode Camp yesterday, very cool event.
I was speaker at java & 3D BOF. Thanks to Orange team ;-)
codecamp
flashy bean