Re: Vertex submission (was Re: [Plib-devel] Re: 3DNow! acceleration)

[Sam S. <sa...@sp...>]

-----Original Message-----
From: Steve Baker <sjb...@ai...>
To: pli...@li... <pli...@li...>
Date: 04 April 2000 00:18
Subject: Re: Vertex submission (was Re: [Plib-devel] Re: 3DNow!
acceleration)


>Sam Stickland wrote:
>
>> How about how the vertexes are submitted?  Am I right in thinking that
>> currently ssgVTable simply does glBegin's and glEnd's on the vertex
>> information?  I assume there could be some big wins here if we started
using
>> CVA's (or even just vertex arrarys?) ?.
>>
>> Or is the idea for people to create subclasses of VTable to fix the
>> situations as neccessary?
>
>I intend to write a CVA or VA class at some time in the future - but I'm
>just so short of free time right now that it's not going to happen for a
>while.


Well I have some time :)

>I don't think it's hard to do though - if you'd like to contribute a
>suitable class, I'd be VERY happy to accept it.
>
>Since CVA's are only an extension to OpenGL, we'd have to conditionally
>fall back on VA's on platforms that don't support them.  That being the
>case, we probably only need one class - I don't think there are many
>cases where people would want simple VA's if CVA's are available.


Makes sense.

To get the best speed out of CVA's won't it need some adjustment to the
scene graph traversal?  Say we had this structure:


             Transform Node                    Transform Node
                           |                                            |
                           +------- CVA Node ---------+

Then you'd want to lock the arrarys, draw the CVA Node in both positions and
then unlock it - rather than locking it twice.

I guess this might start to get a little to complex, particulary if you
start to factor in different states.

Sam