Summary: | REGRESSION: 404s are not displayed | ||||||||
---|---|---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Denis Defreyne <amonre> | ||||||
Component: | WebKit Misc. | Assignee: | Trey Matteson <trey> | ||||||
Status: | RESOLVED FIXED | ||||||||
Severity: | Major | CC: | ap, trey | ||||||
Priority: | P1 | Keywords: | Regression | ||||||
Version: | 420+ | ||||||||
Hardware: | Mac | ||||||||
OS: | OS X 10.4 | ||||||||
URL: | http://www.apple.com/akjwfelwajkfwajfkewwefawegawef | ||||||||
Attachments: |
|
Description
Denis Defreyne
2006-04-01 11:51:25 PST
Caused by the fix in bug 7739 (TOT REGRESSION: Assertion failure loading acid2 test in -[WebCoreFrameBridge installInFrame:]). I will take a look. Fixed by limiting the previous change to the case of an <object> element. I also added a couple test cases for this area, and tweaked the test harness to allow Perl files to be executed as tests (so as to generate a specific 404 error). Created attachment 7472 [details]
proposed patch
Comment on attachment 7472 [details]
proposed patch
+ return element ? [DOMElement _elementWith:element] : nil;
No need for this ?: expression here. The method _elementWith: handles this case, yielding nil. I know you copied it from the _ownerElement method, which also doesn't need it!
Created attachment 7474 [details]
proposed patch
Yep, that's where I copied it from. Same patch with the extra nil checks removed from both places, and [WebCoreFrameBridge currentForm] to boot.
Comment on attachment 7474 [details]
proposed patch
r=me
I applied the patch, and it clearly does fix real 404 erors. But I can't get the new layout tests to pass, so I can't land yet. (In reply to comment #9) > But I can't get the new layout tests to pass, so I can't land yet. It was a permission problem. Fixed now. |