> Over the last few weeks there's been growing interest in using the > support for HTML5's Canvas present in Browse.xo. A very quick test > felt fairly slow, but Alex and his colleagues are interested in > studying, profiling and working towards a solution. > > So here's some notes I've collected for them -- it's not my area of > expertise anyway, so it'd be great to hear from others. > > Cairo and 565 surfaces. The XO display doesn't use 8-bits per RBG > channel (aka 888). Instead, it downsamples colours to 565 (is that > still true?). If the whole stack is using 565, then we avoid costly > downsampling, and generally hit the fast paths... > http://lists.laptop.org/pipermail/sugar/2007-April/002247.html
You might want to look at what mozilla is doing with fennec, their mobile browser, in regards to speed. Their weekly meeting notes [1] cover quite a few bits performance stuff for small devices. They have a bug open for 565 perf stuff [2]. But if you look back through the meeting notes there's a number of bugs that would help perf on small machine for everything from sqllite to rendering etc... there are some incomplete notes here [3] Cheers, Peter [1] https://wiki.mozilla.org/Mobile/Notes [2] https://bugzilla.mozilla.org/show_bug.cgi?id=485426 [3] https://wiki.mozilla.org/Mobile/Performance _______________________________________________ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel