Hans-Christian Andersen
hans.andersen at phulse.com
Sat Feb 2 13:01:43 CST 2013
> I don't care that much how impartial the tests are - after all they made for Mozilla and I've tested IE and Chrome - may I assume the tests are impartial? I suppose I may.... > > And it doesn't matter for me that IE 10 runs better than Chrome 24 (or Safari) or not - AFAIS IE 10 is not bad at all for everyday tasks - am I missing something? It matters because many browsers are implemented differently, but they are all trying to work well enough with the same standard. If a test is designed to work best with your browser (and another one happens to work ok with, but not all of them), then you are testing a non-standard feature. The only thing that has been proven is that Internet Explorer 10 is more compatible with Mozilla Firefox, if you catch my drift. :) > Out of just sport interests I'd be curious to run really impartial Javascript tests on IE10 and compare them with Chrome and Safari - what such tests could be IYO? These days, performance is pretty decent across all browsers, so it really doesn't matter at all. The only ones who care are the companies/teams behind the browsers, so when searching for any decent javascript benchmarking sites, you will come across one from microsoft, one from mozilla, one from google, etc. If you want an impartial test, I might suggest this one: http://peacekeeper.futuremark.com/ Futuremark is a company which builds benchmarking software (originally for 3D graphics cards). I would trust them a lot more than I would trust a benchmark from any of the above companies. Another one is: http://www.speed-battle.com/ SunSpider is another, although it is a little questionable about how impartial it is, at least it's not as biased as the others you pointed out and can be included in the data. - Hans On 2013-02-02, at 9:55 AM, Salakhetdinov Shamil <mcp2004 at mail.ru> wrote: > Hi Hans -- > > I don't care that much how impartial the tests are - after all they made for Mozilla and I've tested IE and Chrome - may I assume the tests are impartial? I suppose I may.... > > And it doesn't matter for me that IE 10 runs better than Chrome 24 (or Safari) or not - AFAIS IE 10 is not bad at all for everyday tasks - am I missing something? > > Out of just sport interests I'd be curious to run really impartial Javascript tests on IE10 and compare them with Chrome and Safari - what such tests could be IYO? > > AFAIU to test memory leakage issues, which was the case with IE (in the past) the tests should run unattended for quite some time - I can use Hyper-V VM for that long running tests - what tests should I run? > > Thank you. > > -- Shamil > > > Суббота, 2 февраля 2013, 0:43 -08:00 от Hans-Christian Andersen <hans.andersen at phulse.com>: >> Just two issues that i want to point out >> >> Firstly, Dromaeo is not an impartial test. It's made by Mozilla. Shouldn't it be impartial and also from multiple impartial sources? >> >> The other issue I have with Dromaeo is that it crashed the Safari browser on my iPad as well, which says to me that the tests are not purely JavaScript tests, as Safari and Chrome have radically different JavaScript engines. >> >> Best regards, >> Hans-Christian Andersen >> >> >> On 1 Feb 2013, at 16:21, Salakhetdinov Shamil < mcp2004 at mail.ru > wrote: >> >>> Hi All -- >>> >>> I have just tried the subj - http://dromaeo.com/?jslib - for IE10 and Chrome 24 on Win8 Prof - here are my results: >>> >>> Test title / IE 10 result / Google Chrome 24 result >>> >>> DOM Attributes (Prototype): 400 / 677 >>> DOM Attributes (Prototype): 278 / 890 >>> DOM Events (Prototype): 468 / 197 >>> DOM Events (jQuery): 28 / 25 >>> DOM Modifications (Prototype): 86 / 73 >>> DOM Modifications (jQuery): 264 / 1359 => Stopp >>> DOM Style (Prototype): 270 / - >>> DOM Style (jQuery): 124 / - >>> DOM Traversal (Prototype): 172 / - >>> DOM Traversal (jQuery): 31 / - >>> >>> IE 10 used not more than 180MB of RAM while running the tests. >>> Google Chrome 24 - got at 900MB memory consumption for 'DOM Modifications (jQuery)' test then displayed "Aw, Snap! Something went wrong while displaying this webpage. To continue, reload or go to another page.". I tried to run this test twice for Chrome and it failed both times. I could have been unlucky with it this time. >>> http://dromaeo.com/?dromaeo >>> Arrays: 1741/2051 >>> Base 64 Encoding and Decoding: 851 / 1764 >>> Code Evaluations: 1754 / 539 >>> Regular Expressions: 168 / 474 >>> Rotating 3D Cube: 960 / 1505 >>> Strings: 1848 / 3321 >>> IE: http://dromaeo.com/?id=189589 (50MB) >>> Chrome: http://dromaeo.com/?id=189590 (33MB) >>> >>> http://dromaeo.com/?dom >>> DOM Attributes: 196 / 525 >>> DOM Modifications: 68 / 255 >>> DOM Query: 2137 / 11009 >>> DOM Traversal: 29 / 316 >>> IE: http://dromaeo.com/?id=189591 (230 MB) >>> Chrome: http://dromaeo.com/?id=189592 (102MB) >>> Thank you. >>> >>> -- Shamil >> > > _______________________________________________ > dba-Tech mailing list > dba-Tech at databaseadvisors.com > http://databaseadvisors.com/mailman/listinfo/dba-tech > Website: http://www.databaseadvisors.com