----- "Gustavo Sverzut Barbieri" <[EMAIL PROTECTED]> ha scritto:
> On Thu, Sep 11, 2008 at 5:56 AM, Anders Petersson > <[EMAIL PROTECTED]> wrote: > > Hi, > > I have continued to investigate the slow performance of my edje > project. > > > > Of course you are right, there are many Evas_Objects and smart > objects. > > I updated to the 'calculate' patch that Gustavo committed. It works > > wonders! Thanks, Gustavo. I'll apply one of Gustavos fixes if they > are > > deemed ok: > > > http://www.mail-archive.com/enlightenment-devel@lists.sourceforge.net/msg18947.html > > Also, now I don't need to do edje_freeze()/edje_thaw() when using > > 'calculate'. > > Excellent. As for edje_freeze()/edje_thaw(), avoiding these kind of > guards are the reason I created this patch. I was doing some layout > objects here and "Oh, I don't want to do this AGAIN", then created > this Evas callback. > > > > I tried to optimize emitting "program,start" and "program,stop" to > > supply the Edje_Program object directly instead of requiring extra > > edje_match_programs_exec calls. Then I discovered that these signals > are > > just emitted in case anyone would be interested. I can't see that > we > > need this possibility so I removed the emitting of "program,start" > and > > "program,stop" altogether. This alone doubles the speed. > > AFAIK, these signals could be used by applications and Edje to catch > the program start or end and do some action in response. Frankly I > just use this callback to debug, connecting to a "*" "*" or > "program,*" "*" signals and check if stuff is running. > I know you can emulate the original "program,*" intent by > chaining > programs, like "pre-ACTION" { signal: SIGNAL_EMIT "program,start" > "ACTION"; after "ACTION"...}. So maybe we could remove this signals > altogether and those who need it could add in this form, but this > requires "grep 'program,'" in the current SVN to see how many use > cases exist. I connect "*" "*" for the log in edje_editor, maybe we can disable the emission by default and enable it by the calling of a function. Like edje_emit_program_signal(int enable) ? dave > > Also related, we could be more selective on which mouse events > objects > emit. Right now we emit all of them, which sucks. Most of times you > just want up, down or even just clicked, not mouse move, in or out > events. Probably we could do that without break .edc files, just have > the boolean "mouse_events" to be an integer, 2 = no move, 3 = no move > + in/out, 4= no move + in/out + up/down, just click, ... > > > > Enabling EDJE_PROGRAM_CACHE still gives some advantage it seems, > > although not that dramatic when I avoid all "program,start" > signals. > > > > For the moment I think this is all I need. In total I've got 5-10 > times > > faster execution. If further needs arise, I have gained some insight > in > > what to look at. > > wow, sometimes I'm amazed of efl performance. Man, you're totally off > the intended use case and you still have performance... or did you > change something, reducing the number of parts in Edje? > > -- > Gustavo Sverzut Barbieri > http://profusion.mobi embedded systems > -------------------------------------- > MSN: [EMAIL PROTECTED] > Skype: gsbarbieri > Mobile: +55 (19) 9225-2202 > > ------------------------------------------------------------------------- > This SF.Net email is sponsored by the Moblin Your Move Developer's > challenge > Build the coolest Linux based applications with Moblin SDK & win great > prizes > Grand prize is a trip for two to an Open Source event anywhere in the > world > http://moblin-contest.org/redirect.php?banner_id=100&url=/ > _______________________________________________ > enlightenment-devel mailing list > enlightenment-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
