[sldev] Nvidia 3D Goggles?

[Lawson English]

Argent Stonecutter wrote:
> On 2008-09-18, at 10:23, Dahlia Trimble wrote:
>> On Thu, Sep 18, 2008 at 8:17 AM, Lawson English <lenglish5 at cox.net> 
>> wrote:
>>> Seriously, with capability-based data transfer, it should be 
>>> possible to request optional streams of data for specialized viewers 
>>> that support these optional streams (e.g. different camera view), 
>>> just as limited clients will be able to NOT receive specific data 
>>> streams as the OGP matures and is incorporated into the SL main grid.
>
>> Sounds to me like it's something that would be implemented entirely 
>> client side
>
> Nod.
>
> I don't see why either of these would need any server-side work, nor 
> what OGP would have to do with it. The former was almost possible in 
> First Look 57575 using a reflective HUD (completely non-scripted, the 
> 'rear view' was done entirely in the viewer), and the latter would be 
> a user-interface element for the binocular client 
> (Advanced->Character->head size ^^ (or maybe ^_^ or ^__^)).
>
>
well, I wasn't thinking very well concerning the camera view since a 
3-4" difference in perspective doesn't require any change in scene graph 
info. As far as OGP is concerned, same deal. If you can generate the 
proper geometry for the second perspective from the info the client 
currently gets, you obviously don't need to ask for more info from the 
server.

Though, it is something to keep in mind anyway: would a separate 
pipeline for rendering info make things more efficient generically, not 
just for stereoscopic glasses? They are already experimenting with 
sending textures via their own cap. Would sending other info separately 
make more sense as well?

Lawson