IE8 passes the acid test

James Bannan
30 December 2007, 1:59 AM


Microsoft has recently announced that the latest build of Internet Explorer 8 has passed the Acid2 web standards test. However there are, as always, some ifs, buts and maybes - read on to find out what the future holds for IE8.


Dean Hachamovitch, General Manager of Microsoft’s Internet Explorer team, recently announced that the latest build of IE8 has passed the Acid2 test.

Dean HachamovitchDean Hachamovitch


Acid2 is a web page rendering test, designed to measure the capabilities of web browsers to render code correctly as well as handle incorrect or invalid code, focussing specifically on CSS, alpha transparency, object elements and hovering effects. The test was released by the Web Standards Project (WaSP) on April 12th 2005, and updated with a bugfix on April 23rd.

The test is purposefully not written to W3C CSS specifications – WaSP isn’t a body which defines web standards, but rather an advocacy group for web interoperability and feature support. As such, the Acid2 test isn’t a compliance test in the true sense, but rather a benchmark for determining a browser’s rendering capabilities. As such, passing the test isn’t the be-all and end-all of determining what makes a good browser, but it’s definitely a desireable goal.

Acid2 on IE8Acid2 on IE8


The first officially-released browser to pass the Acid2 test was Safari 2.02, released with Mac OS X 10.4.3. Most popular browsers (and even some less popular ones) pass the Acid2 test, like Opera, Konqueror and iCab. The most notable exceptions are Mozilla Firefox and Internet Explorer. Firefox’s inability to render the Acid2 test is actually due to the underlying Gecko engine, and as such Firefox, Camino and SeaMonkey all fail the test. However Gecko 1.9 will support the necessary rendering functions, and Firefox 3 Beta 2 passes the Acid2 test.

Acid2 on IE7Acid2 on IE7


IE8’s new-found standards compliance does come with a “but”. IE8 has a feature called “IE8 standards mode” which has been introduced to bridge the gap between web pages coded to expect IE6-compatible behaviour and those anticipating the new standards. It seems that Microsoft has determined it best that IE8’s default behaviour should be backwards-compatibility, and so standards mode is not the default rendering engine. Web developers wishing to make use of IE8’s enhanced rendering support need to insert an optional flag which will instruct the browser to enable standards mode.

This situation is nothing new – Microsoft always has a responsibility to support the environment created by prolonged use of its own products, but it does raise an interesting (albeit highly pedantic) technical point. As standards mode isn’t enabled by default, IE8 won’t actually render the Acid2 page properly as it doesn’t contain the flag to enable the compliant rendering engine, and therefore an argument could be made that it actually fails the test. If IE8 hits a page which it could render properly but which doesn’t contain the flag, it will render it using IE6-compatible behaviour. This might seem a very minor point, but previous builds of iCab and Konqueror have in the past technically failed the test on equally minor grounds.

There’s a very interesting interview with Dean Hachamovitch and Chris Wilson, Platform Architect on the IE Platform team, available here.

 


Post your comment



Comments

RSS feed Email alert

Anonymuos:

More testing hell now (IE6 & IE7 (quirks and standards mode) and IE8 "standards mode") for developers which they were already ranting about on the IE blog. Guess, it’s better in the long run, otherwise they couldn’t bring IE up-to-date without breaking the web. Given MS’s slow pace of IE releases, I’m guessing IE8 won’t come till 2009, or Q4 2008 at the earliest.

Firefox 3.0 FTW for its addons!!

29 February 2008, 8:33 PM (5 years ago)report abuse Send to a friend reply

tin:

So MS were fine with a band-aid ripping switch in interface for MSOffice, but the same type of switch from non-standards to standards is not good enough for them?

Why not just switch it and tell website developers to fix their non-standard pages instead? That would be a better choice for all involved.

29 February 2008, 8:33 PM (5 years ago)report abuse Send to a friend reply

Vico:

It would have to be more then just past a "acid2" test for me to go back to using IE as my default browser

29 February 2008, 8:33 PM (5 years ago)report abuse Send to a friend reply

Lawrence D'Oliveiro:

IE8 is Vapourware

What's the difference between IE8 and Firefox 3? Answer: while the latter is still in beta, it is nevertheless available for anyone to download, and verify for themselves that it passes Acid2, while the former exists purely in "internal builds" accessible only to a select few.

In other words, IE8 is still vapourware. It doesn't "pass Acid2" (present tense), it should be it "will pass" or "hopes to pass in future"--assuming Microsoft doesn't start dropping crucial features to salvage a slipping ship date, like they did with Vista.

29 February 2008, 8:33 PM (5 years ago)report abuse Send to a friend reply

Anonymous78987:

Yes, very good point, Lawrence. I agree entirely with you. Also, I only use Firefox and the IE Tab plugin whenever i need to use IE.

29 February 2008, 8:49 PM (5 years ago)report abuse Send to a friend reply

SimonW:

The use of "an optional flag" in the article, as if it was some IE8-specific flag that web developers need to specifically code for, is rather misleading -- the "optionaly flag" is in fact simply a standard HTML DOCTYPE declaration. The assumption is that anyone who knows enough HTML to put in a proper DOCTYPE declaration in probably knows enough HTML to be able to assume that they've coded properly the rest of the time; so IE8 can use 'standards' mode.

Incidentally, for anyone who thinks interpreting DOCTYPE declarations and changing rendering mode based on them is an IE-only thing, Firefox (i.e. Gecko) does the same thing.

29 February 2008, 8:33 PM (5 years ago)report abuse Send to a friend reply

Thomas:

seems more like a khtml copatiblity test than anything

29 February 2008, 8:33 PM (5 years ago)report abuse Send to a friend reply

anonymous user Anonymous user