Medyk: The testcases that IE6 supports are the so called 'control' cases.
Take for example :first-child. There is a testcase that checks if the first child is matched by the selector, but there is also a control test case that makes sure the other children are not matched.
The control test cases will usually pass, even when the selector is not supported. This doesn't mean anything. It is only significant if the selector is supported and one of these control test cases fails. This is an indication that there is a bug in the implementation of the selector. Posted to
" CSS Selectors test " by
Niels LeenheerOctober 10th, 2006
I agree that passing the CSS selector test isn't a guarantee that the implementation is bug free. The test does not include all selectors and does not test for all possible bugs. For example ::selection isn't part of this test because of difficulties with automated testing. The same applies to :hover.
With these limitation Konqueror was the first to pass the test. Opera the second. Perhaps Opera has more complete support and has less bugs than Konqueror... but that is not within the scope of this test.
That being said, the test is a great indication of how much attention is spend on implementing this feature and that there is at least a common baseline between browsers.
Btw, the selector test does include some tests for updating the selection when the DOM is dynamicly updated. Posted to
" Webkit passes the CSS Selector Test " by
Niels LeenheerFebruary 11th, 2008