On 19 April 2010 21:41, Stéphane Ducasse <stephane.ducasse@inria.fr> wrote:
igor
do you think that we got have a svg canvas: as a merge between what you show us at Brest -- with the tiger and using the rome cnavas -- having a dedicated subclass for svg.
Well, the problem is, that the above frameworks putting us into a constraints. While, we can do things like that (http://http.developer.nvidia.com/GPUGems3/gpugems3_ch25.html) ourselves, without using 3rd party stuff. I know, it sounds like NIH syndrome, but i am concerned, that using rome or svg we will be able to render a 10 browser windows on a desktop with 25 frames/second rate. The above stuff may be fast enough, when you updating only a portions of the display surface (like currently Morphic does and optimized for that). But things like zoom-in or zoom out, obviously require a full desktop update at each frame, so we need a framework which can support a decent frame rate for this. This means, that such framework should not rely on a techniques/optimizations, which require updating only a portions of screen, but instead, its throughput should be high enough to redraw a full screen each 1/25 second.
Stef
On Apr 19, 2010, at 7:32 PM, Igor Stasenko wrote:
This is where, i think we should be heading:
And this is why i think, a graphics frameworks should be vector based, not a pixel based one.
-- Best regards, Igor Stasenko AKA sig.
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
_______________________________________________ Pharo-project mailing list Pharo-project@lists.gforge.inria.fr http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
-- Best regards, Igor Stasenko AKA sig.