[sldev] Google Summer of Code 2008 update
lenglish5 at cox.net
Wed May 7 19:53:49 PDT 2008
Mike Monkowski wrote:
> Callum Lerwick wrote:
>> Ever seen the surface editors in something like Lightwave or Maya? Guess
>> what SL is missing... :) Not to mention they support shader plugins. I
>> still think the future lies in supporting user-supplied GLSL shaders...
>> (This is the cue for the sldev peanut gallery to shout me down and
>> explain how I'm an idiot for even suggesting such a thing.)
> I don't understand why you would expect to be shouted down. I don't
> remember anyone objecting to plugins before. I don't know if there's
> been any discussion about shader plugins specifically, but shaders are
> just code. You have access to all of the GLSL code for the viewer.
> The only argument you might get is that you say "the future" and I
> would suggest it should be "the present." Why wait?
The problem with any plugin architrecture for the SL viewer is that it
requires a great deal of "buy-in" from Linden Lab to make it work. And
my own view, speaking as a documentor of SL prototocols, is that plugins
won't make much sense until the protocols are well-documented, and the
transition to a new viewer structure (which will hopefully happen as the
protocols change radically over the next 2 years) is far enough along
for people to see what makes sense as far as plugin APIs go.
As we create new viewer test harnesses to test login, inventory, gorup
IM, etc., it would make sense to create test modules for new graphics as
well. The modular-viewer VAG proposal is still sitting there if anyone
wants to forgo jumping on the libsl/openviewer/realXtend bandwagon and
design the modular/pluggable viewer of the future....
More information about the SLDev