Comments on: Testing CSS Performance (pt 2)

By: lillbra » Blog Archive » Tänkvärda tips om prestanda

lillbra » Blog Archive » Tänkvärda tips om prestanda — Tue, 27 May 2008 05:25:58 +0000

[...] Sykes testar rendreringshastigheten hos olika css-selektorer i artikeln CSS child selector performance (via Ajaxian). Testerna visar att man där möjligt bör använda så korta selektorer som [...]

By: CSS Performance | Binary Tendencies

CSS Performance | Binary Tendencies — Thu, 15 May 2008 07:31:27 +0000

[...] read an interesting article over at jpsykes.com; testing css performance (pt 2). His conclusion is that using child selectors are slower than using direct class name [...]

By: More CSS Performance Testing (pt 3) by jpsykes

More CSS Performance Testing (pt 3) by jpsykes — Thu, 15 May 2008 04:38:03 +0000

[...] using a variant of the 20,000 item loop HTML page to test that I used in my last set of CSS tests. Basically, I have two loops one for 20 columns and one for 1000 rows. However last time round [...]

By: Testing CSS Performance | Candes Projects | Cristian Neagu - Web designer, developer

Wed, 14 May 2008 14:51:57 +0000

[...] I took a new approach which does seem to return some valid and rather interesting findings, particularly regarding Safari and Firefox 3 and how they react to child selectors and [...]

By: jay7

jay7 — Tue, 13 May 2008 07:43:53 +0000

oh sorry for my messup, besser means better :)

By: JP

JP — Mon, 12 May 2008 23:47:21 +0000

@jay, totally correct, my mistake for using the wrong term in the post, I’m going to go back and edit my references to that, thanks for the correction :-)

@jake, Jim and I are already compiling a number of other tests to run, I’ll add that to it, I’m also going to take Jay’s point above and test child selectors. :-)

By: Jay

Jay — Mon, 12 May 2008 22:13:34 +0000

Child selectors would be table>tr>td
table tr td are descendant selectors.

By: Jake Archibald

Jake Archibald — Mon, 12 May 2008 22:02:42 +0000

@JP: A td must always be within a tr which must always be within a table to be valid. If a td is placed outside a table does Firefox 3 implicitly create the table & tr in the DOM (like most browsers do with tbody)?

Firefox 3 may be using this rule to skip assertions.

I’d be interested in seeing the same test but using unrelated tags, such as “div p a.whatever”

Jake.

By: Performance web » Archive du blog » Gains de performance des sélecteurs CSS

Mon, 12 May 2008 21:40:45 +0000

[...] a été dans l’impossibilité de distinguer une quelconque différence. Il lui a fallu faire un second essai avec tableau de 20.000 cellules bourré d’identifiants et de classes, avec autant de [...]

By: JP

JP — Mon, 12 May 2008 20:15:29 +0000

@Chris, don’t get me wrong I’m not saying descendant selectors are bad and shouldn’t be used, more that they do seem to come with a price and it’s worth being aware of it if you need to be thinking about optimization of your code and speed of rendering.