Maciej Stachowiak wrote up a nice piece on the struggles to pass the ACID3 test. We too had to fix a number of things to get the test 79, submitted by Cameron McCormack, to work, though the other SVG tests were of course easy for us since Opera has supported most of SVG 1.1 since Opera 9, and SMIL Animation since Opera 8.5.
The tests I wrote for ACID3 are probably easier to pass than the test Cameron submitted, and I agree with Maciej that they are indeed not as stringent. It would have been possible to create a far more advanced SMIL test. Time was in short supply though, that's the only excuse for the somewhat simplistic SMIL test. Really I would have liked to have a few visual tests for SVG there instead, but the rules of the contest weren't such that it allowed that unless it had verifiable results using scripting. There could have been one or more tests for SVG filters for example. The SVG 1.1 testsuite has a number of tests for all of this though, so I would recommend using that to accurately judge how a particular implementation fares with regards to SVG.
Eric Meyer writes:
...what’s really needed right now is exhaustive test suites for specifications– XHTML, CSS, DOM, SVG, you name it.
For SVG there are no less than two quite extensive public testsuites, one for SVG 1.1 and one for SVG 1.2 Tiny. These are still being actively worked on, with or without the help of ACID tests. The CDF WG also has a testsuite covering integration aspects of XHTML and SVG. Of course, exhaustive test suites take time to create, which is why the W3C is researching how to allow people to submit tests more easily. In general it seems the W3C is changing to become more open to the public, which IMHO is a good thing.
In closing, congrats to Webkit, and congrats to all the people involved at Opera for releasing a public Opera build that gets a pixel-perfect 100/100 pass rate on ACID3. Umm, what's that? An SVG-only Acid test? Naah, that's just preposterous.