Working together for standards The Web Standards Project

What’s the best test for Acid3?

By Kimberly Blessing | January 16th, 2008 | Filed in Acid2, Acid3, Browsers, Web Standards (general)

Now that all the major browsers (and many minor ones) have pledged support for Acid2, Ian Hickson has moved on to preparing Acid3 — and you can help!

Skip to comment form

Perhaps you expected this announcement, given the news that IE8 passes Acid2. Or maybe it already seems like old news, if you caught the recent Slashdot post. In either case, the news is true — Acid3 will be coming soon!

Whereas Acid2 tested static HTML and CSS, Acid3 will focus on ECMAScript and DOM — the dynamic side of the Web. While it will include a few rendering tests to make things look interesting, the core of the test will be 100 scripting subtests.

And that is where you can contribute! Ian has compiled 84 of the 100 scripts but is seeking 16 additional tests. If you’re up to the challenge, read the criteria, create your test, and submit it — within the next five days!

As with Acid2, the Web Standards Project will play host to Acid3… so stay tuned for the final test, and the play-by-play on the browsers’ progress!

Your Replies

#1 On January 16th, 2008 7:19 am Vectorpedia (Rick) replied:

I applaud Ian for taking on the completion of the Acid3 project. I’m looking forward to seeing the final results – and the competitive heat that’s been applied to CSS-spec implementors being applied to ECMAScript implementors. This new test will put more pressure on the browsers to attempt to adhere to further standards with JavaScript. Hopefully they will all pass it soon with their push towards adopting web standards as they are written.

#2 On January 16th, 2008 7:25 am Leyton Jay replied:

I can’t wait to Acid3, it will be very interesting.

Its a shame that Acid2 does not yet work in more browsers though. I had heard that Firefox 3 beta and IE 8 beta have passed it – not sure if I truly believe that though. Maybe on a programmers desk it works in theory, but it’s not official in my mind until it works on every PC with every version of Windows at every screen res.

#3 On January 16th, 2008 7:53 am bruce replied:

It’s still not completely certain whether IE8 passes Acid 2 yet; although it does an excellent job of rendering the smiley face, the test requires that “you should use the default settings of the browser”.

#4 On January 16th, 2008 9:44 am The Web Standards Project prepara el Acid3 : Planeta Reixa replied:

[...] La nueva versión del test de compatibilidad con los estándares de los navegadores, se centrará en ECMAScript y DOM. Tienen listos 84 de los 100 tests que usará, pero piden ayuda para los 16 restantes. [...]

#5 On January 16th, 2008 11:40 am Gérard Talbot replied:


My guess is that IE8 (in its alpha and – eventually in its final releases) passes acid2 test but only with a special -tag which will trigger IE8 into pure/full or complete standards compliant rendering mode. Something like




or …

In which case that would mean that IE8 does not pass acid2 test without such



#6 On January 16th, 2008 11:41 am Gérard Talbot replied:


My guess is that IE8 (in its alpha and – eventually in its final releases) passes acid2 test but only with a special <meta>-tag which will trigger IE8 into pure/full or complete standards compliant rendering mode. Something like

<meta name=”full-standards-mode” content=”true”>
<meta name=”full-standards-mode” content=”yes”>
<meta name=”standards-mode” content=”full”>
<meta name=”rendering-mode” content=”web-standards”>
or …

In which case that would mean that IE8 does not pass acid2 test without such



#7 On January 16th, 2008 3:36 pm Rick replied:


You can test Firefox beta2 yourself and see that it does pass the Acid2 test. It is available for download. I did, tried it, removed it, and got my firefox 2 back with no problems, no data loss.

As for IE8, I won’t believe anyone from the evil empire. Let me try it myself. Which will never happen.


IE already has the hack you’re looking for. Remember doctype switching??

#8 On January 16th, 2008 6:52 pm Gérard Talbot replied:


I do not trust Microsoft a lot either; I, too, want to be able to test this myself.

Just about a couple of days ago, a well known web developer testing IE7 told me that he had doubts about IE8 fixing all the bugs in IE7 because there were just so many of them. My exact thought too.

> Remember doctype switching??

According to persistent rumors and several sources, including Opera’s CTO, it may be more than just doctype switching but another opt-in method on top of the doctype switching, which could be a <meta>-tag, something like

<meta name=”standards-mode” content=”full”>
<meta name=”rendering-mode” content=”web-standards”>

Regards, Gérard

#9 On January 16th, 2008 7:23 pm Gérard Talbot replied:


“It seems that IE8 will not display Acid2 correctly by default. Instead of following established conventions for how to switch between quirks and standards mode, it seems that Microsoft plans to introduce a new opt-in scheme based on the <meta> tag.”
taken from Acid2 in IE8!
and that assumption/deduction would be consistent with other sources and also with how IE8 can/will render webpages with document.compatMode == “CSS1Compat” differently than with IE7.

In my opinion, there are now well over 750 known bugs in Internet Explorer 7 when triggered into standards rendering mode with specific doctype declarations, all reproducible, all with testcase-ed. And about 500 properties, attributes and methods in HTML 4, CSS 2.1, DOM 1 & 2 currently not supported in IE7. If IE dev. team implements all these and correct all the bugs, then the rendering of webpages would be immensely different between IE7 and IE8. Thus the need for another specific opt-in method.. for full compliance or to trigger document.compatMode == “CSS2Compat”.

Regards, Gérard

#10 On January 16th, 2008 8:01 pm Neal replied:

It is obvious that JavaScript has a long way to go for compliance amongst various browsers but I feel that Acid3 shouldn’t focus solely on JavaScript alone. Perhaps two separate test should be created, one for JavaScript and one for CSS.

The fact that a test even exists is a small amount of motivation for a browser vendor to make their browser more compliant. If we do not continue to push browsers to be more compliant they may loose sight of our goal.

Perhaps a similar test can be created for the upcoming versions of HTML as well.

#11 On January 17th, 2008 10:50 am Lars Gunther replied:

I have been in a conversation with Ian H about the possibility of including E4X in the test. However, major browsers vendors besides MS, I was told, do not want E4X at all, and it won’t go into the test. Shame on you webkit! (And everyone else with this attitude.)

E4X is also too new, as the standard must be from 2004 or more.

It would be easy to write test where current implementations of ECMAScript 3 fail, but where this has been rectified by the upcoming ES 4 spec. I trust Ian won’t consider such entries, though.

Now I will get working on my contribution: media queries!

Lars Gunther

#12 On January 17th, 2008 10:52 am Lars Gunther replied:


(X)HTML 5 will have something better: A formal testsuite.

BTW, there are many test for CSS 3 as well, both formal and informal.

#13 On January 17th, 2008 1:58 pm Lars Gunther replied:

Aargh! Reread Ian’s tests carefully. Media queries are in! Which is a good thing, BTW, but I’ve got to find another niche.

This will put some pressure on Mozilla to fix bug156716

Mozillas progress on ACID 3 can be seen at and

Lars Gunther

#14 On January 18th, 2008 10:47 am Daniel replied:

@Gérard Talbot: According to Chris Wilson, the new trigger for full standards mode will probably a meta-tag. However it is not yet finalized.

So we’ll have to wait before we know what they’re up to. But I’m pretty sure it’ll be some standrd conformant way.

Besides. This could be a way into a really great web.

#15 On January 18th, 2008 12:06 pm Dominic Shiells replied:

It would be smart if it tested all the current W3C recommendations such as
CSS 2.1 or CSS 3 (Font Embedding), this would change the face of the web
ECMA script (javascript)

SVG would be good in all browsers as it would add much more room for creativity.
To me the basic aim of the Acid3 test would be to test current W3 recommendations. So that we can program.
Also it would be smart to specify an aim for the acid test is it just to test css or is it to test more widely than that.
To me all I care about is as a web developer is not to worry about different browsers when typing in the same code

#16 On January 19th, 2008 6:21 pm Jeff Walden replied:

Lars Gunther:

A little knowledge is a dangerous thing.

In truth, E4X is a horrible specification which breaks JavaScript in numerous ways not immediately obvious if you aren’t very familiar with the specification. If you are familiar with it, you know that it’s horribly buggy. No attempt seems to have been made to prove important properties about it, such as that you can only produce well-formed XML or that serializations will always be valid XML. It has had type errors which have shown up only during actual implementation testing and which have been security problems in Mozilla. The spec is a buggy copy of a buggy implementation in Rhino, and it shows. If we knew what we know now, we would not have implemented E4X in SpiderMonkey (the Mozilla JS engine).

I’m not Brendan; I didn’t write the E4X implementation in Mozilla. However, I hacked on it a fair amount for awhile fixing bugs in it, and after fixing bugs, reading the spec, and watching incoming reported bugs, I agree. It’s all too common that when an E4X bug is filed, it’s a bug in the specification (I’d guess upwards of 60-70%).

I would not encourage other browsers or JS engines to implement E4X. It’s too buggy and incorrect to be usable.

#17 On January 20th, 2008 4:41 pm La domo de karotoj » Acidotesto tria replied:

[...] En 1997 oni kreis la skatolan acidoteston por konstati, ĉu TTT-legiloj ĝuste faris skatolojn laŭ la normo HTML4. En 2005 WaSP (projekto pri TTT-normoj) proponis novan teston (acidoteston duan) por testi pli novajn trajtojn, kiujn TTT-legiloj ne tute regis (precipe CSS, sed ankaŭ PNG-bildoj, datum-URL-oj k.t.p.). Nun, ĉiuj grandaj kaj multaj malgrandaj TTT-legiloj subtenas (aŭ subtenos) acidoteston duan (eĉ Esplorilo oka kondiĉe subtenos acidoteston duan), do WaSP volas, ke ni donu proponojn por novaj testoj por ke oni povu krei acidoteston trian. [...]

#18 On January 21st, 2008 4:45 pm En preparación el test Acid3 para navegadores replied:

[...] En preparación el test Acid3 para por fgmunoz hace pocos segundos [...]

#19 On January 23rd, 2008 5:54 am tomek-k replied:


My proposal for Acid3 test is a rendering test for pages served as XHTML. I’ve noticed that some browsers may render XHTML pages differently depending on if they are served as XHTML or as HTML. For example, Firefox (up till version 3b2):
if a web site’s body has a non-zero margin and some non-white background color, then the margin is colored using the bg color in “HTML mode” and it stays white in “XHTML mode”.
In any other browser that accepts XHTML, the margin is always colored using the bg color (either in “HTML mode” or in “XHTML mode”)

I don’t know if there are any other similar inconsistencies, but I think that it would be good to test browsers for such behavior.

The other thing worth considering is, I think, testing for SVG support – all in all it is a W3C recomendation.

#20 On January 25th, 2008 2:02 am Wielkimi krokami nadchodzi test ACID 3! | Ptaszor i jego blog! Lol... replied:

[...] Rozpoczęto prace nad przygotowaniem testu ACID 3, który już wkrótce będzie siał spustoszenie :). ACID 2 opierał się na statycznym języku HTML oraz CSS, natomiast wersja trzecia opierała się będzie głownie na ECMAScript oraz DOM – dynamicznej stronie. Silnikiem testu będzie 100 skryptów, które będą zawierały ‘pod testy’. [...]

#21 On January 26th, 2008 1:03 am Come_one_Micro replied:

All I have to say is:
SVG!!! (in IE)

#22 On January 27th, 2008 5:41 am Freyjkell replied:

I hope so much that Acid3 will be testing application/xhtml+xml (in some or in the main document) and SVG. Also it would be nice to have MathML and XMLHtttpRequest (without fallback to MS ActiveX) test.

#23 On January 27th, 2008 5:42 am Freyjkell replied:

I hope so much that Acid3 will be testing application/xhtml+xml (in some <object src=”data:application/xhtml+xml,…”> or in the main document) and SVG. Also it would be nice to have MathML and XMLHtttpRequest (without fallback to MS ActiveX) test.

#24 On February 5th, 2008 7:24 am Wczasy replied:

Some browsers didn’t pass Acid2 tests but I count on my favorite IE8 will pass Acid3. I’m loking for tools who tests CMS aplications too

#25 On February 9th, 2008 5:22 am ethana2 replied:


Lay on the heat! For Open Standards!
…let’s try to push page embedded .odt ^_^

#26 On February 21st, 2008 7:08 am Helem23487 replied:

FjzYXJ qpoertwyr534nn

#27 On February 27th, 2008 4:17 am Richard Sherman replied:

I can say from experience that the Firefox3 Beta passes Acid2 without any problems. I can also say that at the current stage of Acid3 FF3 Beta gets 59/100, which is good considering it’s still i the beta stage. FF2 gets 50/100, which is very impressive considering that IE7 only gets 12/100, very disappointing. I’ve tested many browsers including Opera9, and FireFox is the one that passes the Acid test most convincingly. I recommend getting FF3 as soon as the full version is released. There are some minor bugs in the beta but none that damage your computer. Mozilla beats Microsoft in my books any day.


#28 On February 29th, 2008 8:54 pm Carlos replied:

Opera 9.5 gets 65/100 which is in beta stage too, and ie7 just gets 5/100.

#29 On March 4th, 2008 3:08 am Les replied:

Looks like IE8 decided on “standard by default”, so when a Beta finally appears, it will be Acid2 compliant. Yeay! Now if they would only fix the zillion things missing with their ECMAScript support.

#30 On March 5th, 2008 9:25 pm Geoff Crook replied:

Q: I removed IE8 Beta – now the problem is how do I get IE7 to function?
I reinstalled IE7 and it will still not reload.
All help is very much appreciated…/gc

#31 On March 6th, 2008 3:22 pm Eraserhead replied:

@ Richard, clearly you haven’t tried the latest builds of Webkit, they get near 90%.

#32 On March 6th, 2008 3:25 pm Alexander Maier replied:

Mac OS X 10.5 / Leopard under Safari / Webkit r30790 does 90/100 – I think that’s the record so far! Is there anything better?

#33 On March 10th, 2008 9:17 pm Josh replied:

IE 8– 17/100 Acid 3 (*sort of* after clicking a few pop-ups & freezing up a few times) & it does pass the Acid 2 test (as far as I could tell) (tested on Windows XP SP2)

Firefox 3 Beta 3– 61/100 Acid 3 & it does pass the Acid 2 test (tested on Windows XP SP2 & Linux)

Opera 9.5 Beta 1– 60/100 Acid 3 (sometimes crashes browser) & it does pass the Acid 2 test (tested on Linux)

Webkit Development– 90/100 Acid 3 & it does pass the Acid 2 test (tested on OS X)

Safari 3.04 (latest shipping version)– 39/100 Acid 3 & it does pass the Acid 2 Test (tested on Windows XP SP2 & OS X)

#34 On March 11th, 2008 9:30 am Josh replied:

Firefox 3 Beta 4- 67/100 Acid 3 & it does pass the Acid 2 test

#35 On March 11th, 2008 7:40 pm Kyle replied:

Yup Firefox 3 is quickly improving everything that was wrong with FF2

#36 On March 14th, 2008 4:31 pm James C. replied:

Firefox 3 Pre-beta 5 was 69/100 and then with an update today it became 70/100.

#37 On March 18th, 2008 11:10 am Josh replied:

Safari 3.1 (latest shipping version)- 75/100 Acid 3 & it does pass the Acid 2 test (tested on Windows XP SP2)

#38 On March 18th, 2008 11:33 am Alex replied:

Webkit on Leopard now does 91%

#39 On March 18th, 2008 6:16 pm Roger D. Parish replied:

Safari 3.1 passes the Acid2 test only on initial load; it fails on a reload of the page.

#40 On March 26th, 2008 12:22 am alcatholic replied:

Latest Webkit Nightly Build (r31307) now gets 96/100

They’ve also just introduced an optimization for pageloading that is very sweet on script heavy pages.

“The latest WebKit nightlies contain some new optimizations to reduce the impact of network latency. When script loading halts the main parser, we start up a side parser that goes through the rest of the HTML source to find more resources to load. We also prioritize resources so that scripts and stylesheets load before images. The overall effect is that we are now able to load more resources in parallel with scripts, including other scripts.”

#41 On March 26th, 2008 12:29 pm Someone replied:

Firefox 3 beta 5 pre i currently am using is getting 71/100 solidly.

#42 On April 2nd, 2008 10:08 pm Josh replied:

Firefox 3 beta 5- 71/100 Acid 3 & it does pass the Acid 2 test (tested on Ubuntu Linux)

Return to top

Post a Reply

Comments are closed.

All of the entries posted in WaSP Buzz express the opinions of their individual authors. They do not necessarily reflect the plans or positions of the Web Standards Project as a group.

This site is valid XHTML 1.0 Strict, CSS | Get Buzz via RSS or Atom | Colophon | Legal