On Sat, Dec 12, 2009 at 1:35 PM, Vladimir Mosgalin <mosgalin at vm10124.spb.edu > wrote: > Hi Carl Eugen Hoyos! > > On 2009.12.12 at 11:19:19 +0000, Carl Eugen Hoyos wrote next: > > > http://www.phoronix.com/vr.php?view=14340 > > http://www.phoronix.com/forums/showthread.php?t=19983 > > Note especially the comment from an AMD employee who explains that XvBA > was > > never released, that no SDK is available (for you and me) and that XvBA's > goal > > is not adoption by open source players (pages 23 and 24). > > I don't think original poster is concerned about xvba or amount of > bridges between fglrx and mplayer, if HW acceleration is possible this > way, why not? > > you're both correct, my interest is mainly having the hw accel available, therefore i could be content with the multiple bridges solution and/or a standard api (that's what we could expect to be "the most convenient way" for every hw accel solution in a short time, if va-api is enough well written) but on the other hand... the problem is betting money on a hardware that i plan to buy almost exactly to be able to play videos, the idea of finding out that my working system got screwed up because of _undocumented_ api changes - and the api is not thought to be used even by open source developers - _is_ a little scary. i'll think about it, but i think i'm going to choose a nvidia platform if things don't improve steadily. thank you both for the very insightful explaination :)
