On Sun, 30 Sep 2012 08:45:55 -0700 Bill Spitzak <spit...@gmail.com> wrote:
> On 09/29/2012 01:28 AM, John Kåre Alsaker wrote: > > >> Wayland also does scaling and rotations but the difference between linear > >> and srgb is much smaller for these. I think the real solution for this is > >> to > >> allow clients to know the actual transform of their surfaces to the screen > >> and be allowed to draw the transformed image, since this will avoid > >> resampling and filtering, as well as letting the clients choose the color > >> space. > > Part of the point of Wayland is that clients shouldn't know about how > > they are placed or transformed, even if they did, the compositor would > > still have to compose them and know about the gamma encoding they use. > > Clients certainly should not *have* to know about the transformation, > and the shell will transform for them. > > However, even thought there is a deliberate attempt to make clients not > know it in current Wayland, I think client awareness of the transform is > going to be needed eventually. Lots of clients could draw their > transformed windows enormously faster and higher quality than then if > they drew the non-transformed one and the shell then transformed the image. And you would like to kill all innovative uses like: http://www.youtube.com/watch?v=_FjuPn7MXMs right? And spherical VR displays, too? No, this is not yet the reason to expose transformations to clients. You are only suggesting to make the protocol a huge deal more complex and restrictive, to optimize a tiny detail prematurely, again. On a regular desktop, the transformed case is very rare, and even more rare on cases where the image quality would actually matter that much. - pq _______________________________________________ wayland-devel mailing list wayland-devel@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/wayland-devel