Andy Green wrote: > Today we had Marco telling us specifically that 2.6.28 touchscreen > filtering arrangements made things slow for him. But, when we looked > closer, it seems the keyboard code itself can be broken and slow (I gave > a link to the thread on Community list) and we did not nail down that > there was any problem with our existing arrangements, he didn't reply > yet about tests with the python script so we can isolate the slowness to > the keyboard or the touchscreen filter action. It also turned out he's > running a unique old build of stable-tracking (not recommended anyway) > with a bunch of patches ripped out.
Sorry I didn't try yet your suggested image. Today has been very busy for me and I had to use the spare time for some webkit/efl tests, however I want to remark here that the "slowness" I saw in my tests with the keyboard aren't related to the Illume keyboard slowness in looking up the dictionary. This is a bug that is completely illume related and that doesn't affect my experience in any way (since I'm using a very small dictionary, and also because that bug doesn't cause to ignore the taps, only it makes the words-search very slow [i.e.: every tap is taken anyway, also if takes longer to be analyzed]). As add on, I'm using the same rootfs for my tests (i.e. the same illume version and configuration). I neither think that the new kernel is slower in analyzing the touchscreen events. Simply it uses too strong filters imho. -- Treviño's World - Life and Linux http://www.3v1n0.net/
