Slashdot Mirror
Opera Lays Down Acid2 Challenge
sebFlyte writes "The CTO of Opera has proposed a new version of the acid test for browser compatibility, and has challenged Microsoft to make IE7 a browser worth having that will do the Web good. He's asked to help from Web designers the world over to build a new page for Microsoft to test IE7 with to make sure it does everything Web designers want it to. "
http://www.w3.org/MarkUp/Test/?
I think the number of people who care about IE are just under 90% of the browser market, and anyone who might fix their computer (another 6-7%).
So maybe a few people.
Please stop stalking me, bro.
Or keep your stable release and install SlashFix.
Actually they make quite a bit of money selling Opera. It's used a lot as an embedded app in cell phones, PDAs and such. As I understand it, Opera runs on more devices than any other browser. They also are more than willing to make custom versions for just about any platform or purpose.
Any money they make from people sitting in front of desktops is just bonus.
Any man who afflicts the human race with ideas must be prepared to see them misunderstood. -- H. L. Mencken
http://validator.w3.org
6_16-BF-01 Frame Target Names (part C/D/F(puts it in the D window)
7_4_2-BF-02 The TITLE element (fails miserably - but does any browser actually PASS that one?)
11_2_6-BF-02 Table cells: The TH and TD elements (on columns, colspan = 0 is set to 1, not all)
13_3-BF-01 Generic inclusion: the OBJECT element (no image displays)
13_3_1-BF-01 Rules for rendering objects (renders the object anyway, even though it shouldn't be able to)
13_5-BF-01 Notes on embedded documents (text didn't show)
13_6_1-BF-02 Client-side image maps: the MAP and AREA elements (fails - some parts are clickable)
16_2_2-BF-02 The FRAME Element (fails - cells are not vertically resizable if in the same row as nonresizable cells, even if noresize isn't set)
I too suspect it is in some way related to the iFrames - when I run Firefox without AdBlock I get the errors very regularily (a rough guestimate would be something in the region of every 10-15 page views) but with AdBlock blocking the OSDN ads the problem seems to go away (or at least it occurs very rarely).
There's mischief and malarkies but no queers or yids or darkies within this bastard's carnival, this vicious cabaret.
What's funny is that Slashdot gives an HTTP 403 to validator.w3.org
I saved the source of this comment page and fed it to the validator. Made 117 errors, among them a fairly serious one:
Line 2007, column 7: end tag for "TABLE" omitted, but its declaration does not permit this
It also has _tons_ of unclosed LI tags. These obviously can mess up the display quite a lot. Except for that, the errors are mainly cosmetic - & instead of & and some spurious attributes which aren't in the 3.2 standard (nobr, iframe height etc.).
The missing </table> is probably the most serious issue.
Roses are #FF0000, violets are #0000FF, all my base are belong to you
When Slashdot makes their code sane, then we can blame the browser.
No, when Firefox devs admit it's a bug in their code, we can blame the browser. When a valid HTML testcase accompanies the bug report in Bugzilla, we can blame the browser. When nightlies with the bugfix in render Slashdot appropriately, we can blame the browser.
How many times does this need to be explained to people? Believe it or not, Firefox does have bugs, and your pathological need to point the finger elsewhere is just stupid.
XMLHTTPRequest is not specified in any standard. It's more Microsoft extension nonsense which Mozilla foolishly embraced. Then again, the Mozilla guys tend to make poor decisions (hello IDN!)
Slashdot declares HTML version 3.2, and the w3c validator is actually much stricter than the written standard. For example, you don't need to close LI elements, as they are implicitly closed by the next LI element. Same with non-standard attributes and even tags - the spec explicity defines behavior for undefined attributes and tags, which is a no-op. It should be a warning that you may be relying on non-standard behavior, but the standard does not prohibit them. Even the missing table end tag may be permitted under the standard, although I'd have to do some digging to check. There's something about block level elements and implicit closing by a new block element that I'd have to get the exact wording on. It's why you don't need a close tag for P elements in HTML 3.2, for example.
MS has never shown the initiative to make things compliant
Internet Explorer was the first browser with CSS support. It was the first browser with XSLT support. It was the first (mainstream) browser with XML support. It's the only browser with P3P support.
Maybe Slashdot should just put it on the front page.t ension.html
http://hardgrok.org/blog/item/slashfix-firefox-ex
This is a firefox extension that fixes the strange rendering that Slashdot's broken html creates.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
Yes, you're the only one. :) In fact, one programmer decided to do something about it: Slashfix. Been using it for a while now and it works just great.
Which is more painful? Going to work or gouging your eye out with a spoon? Find out!
http://www.workorspoon.com
Browsers shouldn't render broken HTML.
Compilers shouldn't compile broken code
Compilers shouldn't, because their user is a developer and he can (and should) fix the problems. But browsers should indeed render broken code, or make a best effort thereof. It's called graceful failure and it's a very important characteristic of production-strength software.
The user of the browser isn't the web page developer, and he isn't interested in the minuties of CSS and W3C standards. The user wants to see the page, and the browser should show it to him instead of crapping out with weird and useless errors.
We could argue that a resilient browser encourages bad code development habits. That may be true, for bad or amateur web developers, but it's still not the fault of the users. There are tools and validators that allow good devs to check the correctness of their code and fix it before deployment.
edxwelch's comment, take 3: There is a standard available (XForms), but no one has yet deployed it.