Surfin' Safari

WebKit hits 98/100 on Acid3

Posted by Eric Seidel on Wednesday, March 26th, 2008 at 10:16 am

We fixed our XML processing in r31316 to stop auto-recovering from invalid byte sequences and instead treat invalid bytes as fatal errors, as required by XML 1.0. To quote the spec:

It is a fatal error if an XML entity is determined (via default, encoding declaration, or higher-level protocol) to be in a certain encoding but contains byte sequences that are not legal in that encoding

Internet Explorer also has this strictness requirement (for XML loaded via XMLHttpRequest), thus we believe this added strictness should not cause any real-world compatibility problems. (IE’s behavior can be confirmed by Hixie’s test.)

Once again, enjoy the pretty picture:
WebKit hits 98/100 on Acid3

or you can grab the latest nightly and try Acid3 yourself!

21 Responses to “WebKit hits 98/100 on Acid3”

  1. dipser Says:

    98/100 is passing, but one or two tests are taking too long.

    Test 00 passed, but took 31ms
    Test 26 passed, but took 91ms

  2. Mark Rowe Says:

    @dipser: Please see Maciej’s previous comment about the timing messages.

  3. eAi Says:

    Both are known issues: http://bugs.webkit.org/show_bug.cgi?id=17871 and http://bugs.webkit.org/show_bug.cgi?id=17510

  4. Gladiak Says:

    Amazing step !!!
    Keep it up my friends :D

  5. Roberto Says:

    if you guys wanto to win the race you must hurry up! ;) http://my.opera.com/desktopteam/blog/2008/03/26/opera-and-the-acid3-test

    and please focus also on these bugs – http://bugs.webkit.org/show_bug.cgi?id=18026 and http://bugs.webkit.org/show_bug.cgi?id=18027 (without :hover this CSS works just fine)

  6. solipsism Says:

    @ Roberto.

    I have doubts that their internal builds will make for a 100/100 pass when released to the public for their various platforms. Public betas of Opera only pass 77/100 right now. WebKit looks like it will have an easier time of creating a stable release while passing all 100 tests.

  7. solipsism Says:

    build 31323 just fixed the glyph issue in the upper right hand corner. It’s still at 98/100 but at least the image looks complete. Though, there is an issue with the border thickness, and possibly spacing, between coloured blocks that is preventing the rendered image from being pixel for pixel perfect.

  8. Thinine Says:

    I’m getting some intermittent failures with the latest Windows nightly that knock the score down to 96. And the Opera folks are just claimed a 100% score on Acid3. :(

  9. Aisys Says:

    Opera is claiming 100, but with glitches that prevent them from declaring full pass (as does WebKit). Looks like WebKit just hit 99/100 (no new build available for it yet). FF3 is 68 it seems. With IE’s new rendering engine mode switch deal, why don’t they just adopt WebKit? ;)

  10. dipser Says:

    @Mark Rowe: I see.
    Forgot to say: “Great work, so far.”

  11. rossjones Says:

    Thinine: http://my.opera.com/desktopteam/blog/2008/03/26/opera-and-the-acid3-test

    So, so close. But don’t stop now .. see if you can beat the IE team to it (j/k)

  12. xfinite Says:

    I was toying around with webkit and acid3 and I can’t really figure out what kind of error causes the top right border corner to draw incorrectly. There’s a background image which is almost shifted out of view, but the border is still overdrawn. Can anyone explain?

    Oh, and good work guys!

  13. Robert Blaut Says:

    lop3u2: “There are some remaining issues yet to be fixed, but we hope to have those sorted out shortly.
    We will release a technical preview version on labs.opera.com within the next week or so.”

    It sounds that they have some work to do.

  14. Aisys Says:

    xfinite: that was just fixed. Check out the latest build.

  15. Robert Blaut Says:

    xfinite: Take a look at a bug: http://bugs.webkit.org/show_bug.cgi?id=17086
    Dave Hyatt, fixed the problem in http://trac.webkit.org/projects/webkit/changeset/31322 So the latest Webkit r31326 has no problem with it, but only on Leopard.

  16. Robert Blaut Says:

    And Windows of cause.

  17. masterleep Says:

    The only thing better than this would be if IE8 was in the competition.

  18. macman7 Says:

    about opera passing the test I would say hold on a second and lets think about it. Not that I don’t believe the opera team but they say on their blog they will be releasing a PUBLIC version of the passing build in the next week or so… The webkit team will probably have the first public build to pass the test and safari will more than likely be the first official commercial release to pass it so my money is still with the webkit team.

  19. The Ancients Says:

    Congratulations guys :-)

    Also, this is on slashdot already: http://tech.slashdot.org/article.pl?sid=08/03/26/1938251

    Opera are claiming a 100/100, but as I commented on /. the highest I see when browsing to http://acid3.acidtests.org/ is 98/100 using Safari.

  20. Roberto Says:

    I think that WebKit can be first public, but not official release, and Opra 9.5 have chances to be first official released browser

    We’ll see… ;)

    Acid3 or not, but WebKit has better CSS3 support :D

  21. Mithrandir Says:

    My not so secret browser whih uses WebKit will be officially released as Browser v1.0 as soon as Webkit passes acid 3. It will be the first stable browser passing Acid 3 :D : D