Re: Journal activity profiling

2008-08-14 Thread riccardo
On Wed, 2008-08-13 at 18:02 +0200, Tomeu Vizoso wrote: On Wed, Aug 13, 2008 at 5:20 PM, riccardo [EMAIL PROTECTED] wrote: On Wed, 2008-08-13 at 12:13 +0200, Tomeu Vizoso wrote: On Tue, Aug 12, 2008 at 12:10 PM, riccardo [EMAIL PROTECTED] wrote: ... I find interesting to note that

Re: Journal activity profiling

2008-08-14 Thread Michael Stone
On Wed, Aug 13, 2008 at 01:11:04PM +0200, riccardo wrote: Do you have any idea for redrawing less often? what do you mean ? Think of the way that terrain based games use level-of-detail computations (LoD) to figure out which portions of the screen need to be updated. Think of the use of

Re: Journal activity profiling

2008-08-14 Thread Michael Stone
On Wed, Aug 13, 2008 at 06:02:49PM +0200, Tomeu Vizoso wrote: On Wed, Aug 13, 2008 at 5:20 PM, riccardo [EMAIL PROTECTED] wrote: On Wed, 2008-08-13 at 12:13 +0200, Tomeu Vizoso wrote: On Tue, Aug 12, 2008 at 12:10 PM, riccardo [EMAIL PROTECTED] wrote: Testcase: Fill the journal with hundreds

Re: Journal activity profiling

2008-08-14 Thread Michael Stone
On Wed, Aug 13, 2008 at 12:13:14PM +0200, Tomeu Vizoso wrote: Yeah, no C-level function takes an absurd amount of time. It's python code that makes too much work than is really needed. It's not python code proper; it's calling any sort of code and performing any sort of I/O UNNECESSARILY. Think

Re: Journal activity profiling

2008-08-14 Thread Tomeu Vizoso
On Thu, Aug 14, 2008 at 12:32 AM, Michael Stone [EMAIL PROTECTED] wrote: On Wed, Aug 13, 2008 at 12:13:14PM +0200, Tomeu Vizoso wrote: Yeah, no C-level function takes an absurd amount of time. It's python code that makes too much work than is really needed. It's not python code proper; it's

Re: Journal activity profiling

2008-08-13 Thread riccardo
On Wed, 2008-08-13 at 13:11 +0200, riccardo wrote: On Wed, 2008-08-13 at 12:13 +0200, Tomeu Vizoso wrote: On Tue, Aug 12, 2008 at 12:10 PM, riccardo [EMAIL PROTECTED] wrote: ... Yeah, we should be much smarter about that. Right now, we cache three screens worth of DS entries so we

Re: Journal activity profiling

2008-08-13 Thread Eben Eliason
On Wed, Aug 13, 2008 at 12:02 PM, Tomeu Vizoso [EMAIL PROTECTED] wrote: On Wed, Aug 13, 2008 at 5:20 PM, riccardo [EMAIL PROTECTED] wrote: On Wed, 2008-08-13 at 12:13 +0200, Tomeu Vizoso wrote: On Tue, Aug 12, 2008 at 12:10 PM, riccardo [EMAIL PROTECTED] wrote: Hi! Testcase: Fill the

Re: Journal activity profiling

2008-08-13 Thread Tomeu Vizoso
On Wed, Aug 13, 2008 at 6:07 PM, Eben Eliason [EMAIL PROTECTED] wrote: On Wed, Aug 13, 2008 at 12:02 PM, Tomeu Vizoso [EMAIL PROTECTED] wrote: On Wed, Aug 13, 2008 at 5:20 PM, riccardo [EMAIL PROTECTED] wrote: On Wed, 2008-08-13 at 12:13 +0200, Tomeu Vizoso wrote: On Tue, Aug 12, 2008 at 12:10

Journal activity profiling

2008-08-12 Thread riccardo
Hi! Testcase: Fill the journal with hundreds of entries and automate the GtkScrollWindow in the journal to scroll one entry at time on a timeout. Build: joyride-2281 + Macro investigation of cpu usage A picker graph taken during the test can be found at: