From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:0.9.9+) Gecko/20020320 BuildID: in the above URL there is a <td width="70"> which is ignored. I found that this is deprecated in HTML4.01 but should be rendered by user agents for backward compatibility. Reproducible: Always Steps to Reproduce: 1. open the URL above Actual Results: the second <table> is divided 50%/50% Expected Results: the first column should have a width of 70px and the second should be the left space of the page
Confirming W2K 2002031104. OS -> All. Reassigning to HTMLTables. Severity -> major - lots of sites assume that setting the width attribute for the first table cell will affect the rest of the column. I've produced a simplified testcase, which I will attach.
Assignee: attinasi → karnaze
Severity: normal → major
Status: UNCONFIRMED → NEW
Component: Layout → HTMLTables
Ever confirmed: true
OS: Linux → All
QA Contact: petersen → amar
Created attachment 75577 [details] Testcase Here's a testcase which demonstrates this bug, plus some additional wierdness. Observations: * Setting 'width' attribute on only the first TD in a column is ignored (examples 1, 2, 7, 8). * Setting 'width' attribute for all cells in column works fine (3, 4, 5, 6). * The 'cols' attribute to TABLE changes the table layout slightly, even though 'cols' isn't a valid attribute for TABLE in HTML 4 or HTML 3.2, unless all TDs in a column have the width attribute set (1&2, 7&8 vs. 3&4, 5&6) * All the above apply in Quirks mode only (the !DOCTYPE given in the original URL isn't correct enough to engage Standards mode)
Created attachment 75606 [details] testcase with strict doctype cols are a quirk inherited from NN, and have been something which would be today implemented via table-layout:fixed instead. I dont see why this bug is major. For me the bug is either invalid or wontfix
Ah, I wondered what the cols attribute was. Anyway, the original report was for the ignoring of the width attribute on the TD in Quirks mode, and this still stands.
just a quote from the manual (NN4x): http://developer.netscape.com/docs/manuals/htmlguid/tags9.htm COLS="numOfCols" indicates how many virtual columns of equal width fit in the width of the window. Each actual column in the table occupies a virtual column. You would typically set the COLS attribute to be equal to the number of columns in the table to indicate that all the columns in the table have the same width. Navigator 3.0 If the WIDTH attribute is supplied, the COLS attribute indicates how many virtual columns fit in the specified width. If the WIDTH attribute is not supplied, the COLS attribute indicates how many virtual columns fit in the current window or frame. Each column in the table occupies one of the virtual columns. Suppose that the WIDTH attribute is "80%" and the COLS attribute is 4. In this case, each virtual column takes up 20% of the width of the window. Each actual column in the table occupies a virtual column, so it occupies 20% of the width of the window, so long as the table has from 1 to 4 columns inclusive. Note, however, that if the minimum width needed to display the contents of an actual column is graeter than the width of a virtual column, then the width of the column is expanded to fit its contents. If the table has more actual columns than the COLS value, then the columns in excess of the COLS value are displayed in the minimum width required to fit their contents, and the other columns divide the remaining space equally between them. For example, suppose the table has 4 columns, the WIDTH attribute is "80%", and the COLS value is 3. What happens here is that the table takes up 80% of the width of the window. The fourth column uses the minimum width necessary to display the contents of the column. The other 3 columns divide the remaining width of the table equally between them. That is what mozilla is doing here. So far for my claim, about marking invalid, the really interesting part is that NN4x did not implement as well the spec as mozilla now does. So it boils down to how far we would like to go in implementing NN quirks behaviour. It would require some tweaking of BasicTableLayoutStrategy.cpp, but it will definetely not happen before 1.0 -> this bug deserves only the milestone: future. If somebody would like to get this bug fixed then pelase provide a a regression tested patch.
Severity: major → normal
mass reassign to default owner
Assignee: karnaze → table
QA Contact: amar → madhur
Target Milestone: Future → ---
*** Bug 253694 has been marked as a duplicate of this bug. ***
Works for me. Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9a3pre) Gecko/20070221 Minefield/3.0a3pre
Status: NEW → RESOLVED
Last Resolved: 10 years ago
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.