Out of the way

What is Acid3?

It's a web page that includes a collection of tests put together in such a way that you can pretty much see at a glance whether you pass them all (by comparing to the required final rendering), or some of them. Following the Acid and Acid 2 tests, it's a bunch of complex things including some SVG support, lots of javascript and CSS stuff, and some odd ideas. There is an article about Acid3 on Wikipedia that is reasonably accurate and moderately clear if you want to know more.

Why is it important?

It's important because it tests a bunch of useful things, in combinations that didn't work in any browser when it was published. In other words, it is a device to push browser makers to solve interoperability problems.

It isn't critically important, since it doesn't test everything. More importantly, it changes. About the time we announced we had got to 100/100 the Webkit developers convinced Ian Hickson (the maintainer of the test) that he should change it to match their interpretation of the SVG specification, so he changed it. Today there was another request for change, so they could remove the hack they introduced to resolve a problem they had interacting with their operating system - I presume that change will be made too.

With shifting goalposts like this the real value of the test is not actually in who comes first (although getting it done faster is an indicator of a good platform to build on, and good development). The point is that as more browsers pass the test, developers are able to rely on more things working that help them to build better applications, instead of trying to work around the poor standards compliance that used to be the order of the day.

Does anyone actually pass the test?

It's hard to say, actually. The test itself requires getting to 100/100, and pixel-perfect rendering, but it also includes the performance requirement that it "run smoothly". In particular, test 26 just requires a mountain of processing, and test 69 requires loading stuff from the network - so if you have an old computer or a device with lower power or hit a slow patch on the network then the only way to make it smooth would be to slow down the rest of the browser to make it smooth - something that in my opinion is not actually useful to anybody.

In any case, both Opera and Webkit have said that they plan to work on improving performance (well, no surprises there), which should smooth out the animation - but I do not think it is so important to do it just for the test - overall performance is rather more complicated if the goal is to deliver a better browser to real users.

So what does it all mean?

• WebKit and Opera have both resolved bugs and implemented features to the standard needed for Acid 3.
• There is another desktop browser engine that lets you use SVG SMIL animation - something already in Opera (since version 8) as well as all current mobile-phone-based SVG players, but not yet in Mozilla/Firefox. (Webkit's implementation leaves work to be done, but presumably this was part of the ongoing effort to improve the browsers that are based on webkit and that work will be done. This is for me one of the most exciting results from the whole race).
• There are serious performance improvements in both Opera and Webkit - we are pushing each other to be even faster and more efficient, instead of sitting on our laurels. The fastest browser on earth is getting faster, as the others have started seriously trying to race us.

The real point, as noted by in the blog entry announcing that IE 8 now passes Acid *2*, is interoperability. Getting the advanced things right means that developers will be more easily able to build cooler applications, instead of spending their time focusing on dealing with cross-browser problems.

Anyway, I am very proud of what the core engineers in Opera did to get us here. We produce a browser with more features on more platforms than anyone else, but with a sensible architecture and good development we are still at the front on a very tough test of a set of features that only covers the narrower scope common to most browsers. This is not exactly a sideshow, but it is just one part of the process of developing the next generation of Opera. There's lots more cool stuff to come, and now we can return to the dungeons and concentrate on getting it out...