Changes between Initial Version and Version 1 of Ticket #12342, comment 9

Nov 23, 2012, 2:43:01 AM (9 years ago)
Colin Snover


  • Ticket #12342, comment 9

    initial v1  
    22* The test for 10095 makes no sense to me, and I suspect the fact that it is failing doesn’t mean a whole lot. Supporting what that test code appears to be trying to do is stupid (delete an element from an iframe from the parent frame, then set the iframe src back to itself, then check to see if the element doesn’t exist anymore? this seems like an incredibly indirect inference to see if iframe.src = iframe.src with a javascript: uri reloads an iframe and nothing more), so I’m not even going to bother spending my time looking into it any more. If someone else wants to do it, feel free. As far as I am concerned this test should be deleted from the test suite.
    33* Same thing with the quirks mode tests. Quirks mode is called **quirks** mode for a reason, and it’s a waste of time to create tests to make sure things are “consistent” when it is being used. It’s 2012, use a bloody doctype. Also, these tests pass just fine when outside the test runner. When they fail it appears to be the issue where border gets folded into the element margin calculation, and there is some code to address that, but obviously the dumb combination of quirks mode iframe in strict mode document doesn’t work great in IE9. Stop using quirks mode!
     4* The testSpeed failure hits all versions of IE and is #11200.
    56As far as I am concerned the tests that are still there and failing probably should not be there at all because supporting anti-patterns like these is dragging the toolkit down, bloating our code, and wasting effort that needs to be focused on leading the industry. I look forward to people telling me I am wrong, but only so long as those people fix the tests first.