Bug 20682
| Summary: | Page renders inadequately | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Dan Sanduleac <sanduleac.dan> |
| Component: | Layout and Rendering | Assignee: | Nobody <webkit-unassigned> |
| Status: | RESOLVED INVALID | ||
| Severity: | Normal | CC: | sanduleac.dan |
| Priority: | P2 | Keywords: | NeedsReduction |
| Version: | 528+ (Nightly build) | ||
| Hardware: | PC | ||
| OS: | Windows Vista | ||
| URL: | http://b08.oi.edu.pl | ||
Dan Sanduleac
I think tables are misrendered but I didn't have the time to look into the source of the page to see the problem exactly. A table seems to be utterly thin when rendered with webkit. In Firefox 3 latest or Opera 9.5 latest, it displays correctly, or more precisely "readable", how I think it should look, without first checking the page source for html bugs.
The problem manifests itself in all of 525.21 (Safari 3.1.2), nightly build WebKit-36012 (the one precompiled for windows and run through Safari), and also in Google Chrome/0.2.149.27 (build 1583).
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Dan Sanduleac
Using Chrome's (version stated before) inspector component I see that div#main_container is automatically set a width of 0, as I anticipated. In Firefox, through FireBug, the dimensions are set to 670x424 regardless of browser's window size, and this looks normal. I'm at a loss whether to place the bug under Layout and Rendering, or under CSS.
Dave Hyatt
I see negative margins. Maybe something tripping up there. Firefox renders the page correctly.
Dan Sanduleac
After examining the source code, I found that there is a html source error involved, the left side menu div having sort of a glitch - one of the menu items has a 'submenu' div, and some of the submenus are commented out, but they mistakingly commented out the </div> for closing the 'submenu' div as well, thus leaving it open [line 203].
I'm resolving the bug as INVALID, yet i'm still at a loss whether this is a bug or not, since IE7 latest, Firefox 3 and Opera 9.5 all seem to ignore or identify the error and work around it.