Where Rev could be going...3D?
jbv.silences at club-internet.fr
Thu Nov 23 05:21:33 EST 2006
> Hey, hey
> I provoked the 3D stuff not because to force someone starting developing Rev
> 3D module from scratch...
> The intention was to draw your attention to the fact that OpenGL 3D engine
> for the Revolution ALREADY EXISTS and sits on the shell exposed by iGame3D
> with seemingly no one being interested in the ready-to-use thingy.
> Any ideas on how can we "take" it? Because if we do not manage to "take" it
> somehow now, it will likely be gone in a few months... and then somebody
> (definitely not me ;-)) will have to code the whole thing from scratch again
Don't worry, we all do understand your concern.
But at the same time, please allow me to consider the problem from the other
way round : have you ever wondered why some MC / Rev users (like me) have
been waiting for years for some vector graphics to be available in Rev, or why
this 3D API has been sitting on the shelf for so long ?
IMHO this situation is an inheritance from the old HC days when the only way
to add features to the engine was through externals. Which means you have a
proprietary engine (a bit like a black box), and the possibility to add limited
functions to it via the limited externals API.
But times are changing, and now open source is much more different from what
it could have been in the late 80's. I'm not pleading for complete open-ness of the
Rev engine code, but rather for some general framework that would allow skilled
coders to jump in and start coding to enhance the engine by adding primitives to
the Transcript language itself. Again I'm convinced that some "heavy" enhancements
can't be added to Rev in any other way. And since such a framework isn't available,
Rev users have to wait and wait for the skilled but more-than-busy Rev crew to
achieve and debug any improvement...
Those coders would first build these new primitives for their own use (because they
need it for a specific project - it was my case in 2003 when I needed some openGL
interface for a specific client - and most important, it would prevent them to switch
to another development tool), but then nothing would hinder them to make the new
primitives (or the code itself) available for others, just like in open source.
Eventually, others could tweak the code to meet their own needs, and finally a more
or less universal and final version could be included in the engine by RR.
BTW this would re-inforce the deep natural english-like nature of xTalk, by copying
natural language in its improvement & evolution process...
Please correct me if I'm wrong, but I'm convinced that more "universal" developments
could be achieved that way. 3D is a good example : I understand that opening Rev to the
development of 3D games is a good thing, but why limit the use of 3D to games ?
A powerful set of primitives for 3D creation and control would open more horizons :
industrial process simulation, education, scientific imagery, etc.
And who knows what kind of new developments would arise by simple interaction
with other existing features (I for one have a few ideas that I don't want to disclose
right now, but hope I'll find the time to code in a more or less near future) ?
And last but not least, here's another innocent question : please correct me if I'm wrong,
but as far as I understand, the improve-rev list is limited to those with an Enterprise
licence. How come someone (like me) who has spent thousands of hours since 1987
coding with the successive versions of HC / SC / OMO / MC / Rev is less qualified to
discuss possible Rev improvements than anyone who discovered Rev and xTalk last
week, but with enough $$$ to afford an Enterprise licence ?
More information about the Use-livecode