Interesting. I think I see what you're saying. My execution time is also the lowest in chrome but when I drag the bars around it has a choppy feeling, the bar would get hung up on certain parts. In IE the bar outline smoothly followed my mouse. (In firefox, the demo doesn't work at all for me.)
Maybe there is some other metric besides execution time which would capture the behavior you describe. On Feb 5, 2:23 pm, Stefan Blanke <[email protected]> wrote: > You will have to drag around the "windows" of my application to see > the effect. Chrome and Safari for Windows are very slow while moving > the windows. > Javascript-code execution is fine with Chrome and Safari, its just the > rendering of the positioned element (window). > > Stefan > > On 5 Feb., 19:55, Jon <[email protected]> wrote: > > > > > The number I am looking at for a difference is the milliseconds number. I > > assume that I want that to be lower. I see two things. First, I see that > > we seem to generate more global events than FireFox 3. That is probably a > > good thing. Second, I see that on my Intel based laptop FireFox seems to be > > doing things at about 12ms and Chrome at about 8ms. > > I want us to be faster and if you are seeing us being slower I want to > > understand how to reproduce the results. > > > Jon > > > On Thu, Feb 5, 2009 at 8:16 AM, Stefan Blanke <[email protected]> wrote: > > > > Hello again. > > > Sadly I am not able to reproduce the behaviour in a simple test-case. > > > Any test on reasons i can think of and reproduced didn't show the same > > > lagging. So, please take a look athttp://www.webmop.de/app. Start > > > dragging the "windows" by dragging the bar to see the effect. I think > > > it will also depend on your machine speed (mine: old athlon 64 x2 > > > 4200+, 939 Socket, GF 8600 GT) but comparisson with Firefox might > > > reveal what i am talking about. There are some statistics in the > > > topmost window. "Global events" are the number of all events fired > > > inside the application (high value, not managed yet...), called is the > > > number of how many times the move-function is called and execution > > > time (10-12 ms with Chrome on my computer). > > > > Stefan > > > > On 3 Feb., 19:08, Jon <[email protected]> wrote: > > > > I suggest you wrap all of this up into a zip file that we can use as a > > > test > > > > case. We are always looking for examples that show where we need to > > > > improve. If we have a test case that we can measure with each build we > > > can > > > > track and improve our performance. > > > > Once you have a self-contained example open a bug athttp://crbug.comand > > > > then send me the Issue ID so I can bring it to the attention of the > > > > right > > > > people. > > > > > [email protected] > > > > > On Sun, Feb 1, 2009 at 4:24 AM, Stefan Blanke <[email protected]> wrote: > > > > > > Hi! > > > > > I ran into some severe performance issues with my last project. Chrome > > > > > (and safari for windows) slow down rendering free positioned DIV > > > > > elements while number of unpositioned nested DIVs (inside a DIV with > > > > > scrollbars) increases. The javascript code takes about 2-5 ms (!) but > > > > > rendering of the positioned elements utterly stutters, this effect is > > > > > even worst in safari for windows. > > > > > IE7 (code: 15-20ms) performance is acceptable and Opera & Firefox > > > > > (code: 5-10ms) realy fine. > > > > > > Any suggestions? > > > > > > Greetings Stefan --~--~---------~--~----~------------~-------~--~----~ Chromium Discussion mailing list: [email protected] View archives, change email options, or unsubscribe: http://groups.google.com/group/chromium-discuss -~----------~----~----~----~------~----~------~--~---
