Acid 3 - The Web Browsers Test

[The_Decryptor Veteran]

I doubt IE will ever fully pass Acid3, unless they change their mind on a few things.

Mainly font-face support, IE already supports that part of the spec, but only their own format, Acid3 tests font-face but uses a format IE doesn't (and apparently wont ever) support.

[megamanXplosion]

Serious?

Acid3 don't mean anythng in real world at present and near future, innovation, security, performance to name just 3 things are far more important.

These test should be for developers only, as at present its mis-guiding the ignorrent to how good browser is.

Rendering engines don't innovate. If they do, they're typically proprietary innovations and should, for that reason, be avoided. Cross-browser innovations occur in W3C standards, which build upon previous standards. Those must be well-supported before you can build upon them because you cannot build off a foundation until the foundation can support what you build. That is why the Acid3 test is so important. Innovation on the web depends on decent browser support for current standards. It's no coincidence we are still waiting for CSS 3, HTML 5, and XHTML 2, for example.

I agree with you that speed is important, and that also ties into idea of providing a foundation for innovation. The Acid 3 test reports on speed. In fact, you must render accurately and swiftly to pass the test.

As for security, that too is incredibly important. But working to pass the Acid3 test does not detract from it. The development teams for the various browsers will, generally, drop what they are doing and improve security if they perceive a need to do so. Acid 3, nor the encouragement to pass it, affect browser security.

As for the idea of the test being developer-only, I disagree completely. The only way most web developers can have a say about the progression of the web is if they can gather people together for a cause. There's a reason why random griping across the internet led to only small changes but the introduction of the Acid tests led to rapid progression ;)

[The_Decryptor Veteran]

Well they can innovate, but they do it within a certain framework that allows everybody to benefit.

The Mozilla guys are working on a way to treat any arbitrary element as a image for stuff like background-image, the Safari guys already have something similar, but just for Canvas tags. In both cases though, they've written out documents detailing how it works, then have gone around talking to other browser makers for comments.

<canvas> started off on a similar route, it's now part of HTML5.

Edit: Compared to how IE does it, when they mention in release notes they've created a new method of doing something, when there's already an incompatible spec going through the W3C and has browser implementations (like FX3, it was cut at the last minute due to changes in the draft)

[+M2Ys4U Subscriber¹]

XMLHTTP was a Trident innovation, now it's literally everywhere.

[Binary]

Safari 4.0 Dev release gets a score of 100.

But..

Failed 0 tests.

Test 26 passed, but took 44ms (less than 30fps)

Total elapsed time: 1.24s

Edited August 28, 2008 by Binary

[Buu]

I think IE8 has come A LONG WAY recently and i'm really impressed with it. I would actually consider switchign back from Firefox to it if it keeps this up. Although i think it would be a great move, expecially for the IE haters, if they passed ACID3 i think the fact they have now passed ACID2 is great news.

[Daninku]

lol I got a miserable 12/100 with IE8 Beta2.

[NateB1]

IE 8 Beta 2 crashes after a couple seconds when running this test...Just the tab, though.

[The_Decryptor Veteran]

XMLHTTP was a Trident innovation, now it's literally everywhere.

Yeah, it was odd how it happened, although strictly it wasn't IE, it was just a useful ActiveX control shipped with Windows.

IE7 brought native XHR.

[+M2Ys4U Subscriber¹]

True... but I think in non-web applications it was very much underused and most people used IE to do any of their data retrieval stuff.

If it hadn't have been embedable in a web page and used via javascript, it probably wouldn't have really be used at all.

[Ruti]

77/100 with google chrome

[splicer707]

77/100 with google chrome

I get..

[Reacon]

did I do something wrong? My firefox 3 got 71 yet everyone reports worse...

[The_Decryptor Veteran]

FX3 gets 71/100

FX3.1 gets 85/100 and only "fails" one test due to the time it takes (FX3 "fails" a whole bunch due to time)

[Crazysah]

78/100 with the latest Chrome Nightly installed! An improvement over the official release!

[Biotoxic_hazard_835]

71/100 with Firefox 3

[D!rtySh@dy]

78/100 with the latest Chrome Nightly installed! An improvement over the official release!

Where do you have it ?

[wellofsouls]

I can get anywhere from 75/100 to 79/100 in Chrome and all those Chromium builds every time I refresh, so it doesn't seem to be really improving on Acid3 yet. Never got one 80/100 yet.

And yes I got 79/100 the first time I run Acid3 in the first beta release of Chrome three days before.

[tiagosilva29]

Where do you have it ?

Link for Chrome Nightlies Thread

[torrentthief]

who wants to test firefox 3.1 alpha 2 with acidtest 3, i dont wanna install on my pc as my add-ons wont work with it.

http://developer.mozilla.org/devnews/index...e-for-download/

[iconoclast]

85/100 With Minefield 3.1b1pre

[revvo]

I'm getting tired of seeing that red kitty!

[Crazysah]

77/100 with the latest Chrome Nightly! It is getting closer!

[The_Decryptor Veteran]

who wants to test firefox 3.1 alpha 2 with acidtest 3, i dont wanna install on my pc as my add-ons wont work with it.

http://developer.mozilla.org/devnews/index...e-for-download/

...

FX3.1 gets 85/100 and only "fails" one test due to the time it takes (FX3 "fails" a whole bunch due to time)

I use the nightlies, it probably hasn't increased since then (I was using a newer build than A2 though)

[wellofsouls]

77/100 with the latest Chrome Nightly! It is getting closer!

I think it's always fluctuating between 75/100 to 79/100 after each refresh...