244686 - Too small table cell with defined table cell with and nested tables inside it.

Status: RESOLVED FIXED

(Core :: Layout: Tables, defect)

Description

[Martijn Wargers (dead)]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8a1) Gecko/20040520 Firefox/0.8.0+
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8a1) Gecko/20040520 Firefox/0.8.0+

I will upload a testcase, which describes the problem.

Basically in this testcase, the first table cell has style="width:205px" but is
getting rendered smaller (computed width 39px).
The second table cell has style="width:335px", but has a computed width of
572px. That's because of two divs which are nested inside a table. The first div
has style="width:333px;" (computed also). The second div has
style="width:225px;" (computed also).
The parent table has style="width:760px;" so the two table cells should at least
cover that width (which they don't, that's why you see the red gap).


All in all, Mozilla should behave more like IE and Opera, which render the first
table cell at least as wide as 205px and cover the whole 760px wide table with
it's two table cells. 

Reproducible: Always
Steps to Reproduce:
1.Load testcase
2.
3.

Actual Results:  
First table cell too small (39px).
Red block at the right side of the table.

Expected Results:  
First table cell at least 205px wide.
No red block should be seen, only a red border.

Comment 1

[Martijn Wargers (dead)]

Attachment: Testcase to show the bug

Comment 2

[Martijn Wargers (dead)]

Attachment: Testcase2

Well, I've managed to simplify the testcase some further, no nested table is
needed (so the summary isn't right, I guess).
I've used width: 568px; for the div. This renders wrongly, but when I change
the width to 567px, then it renders right.

Comment 3

[Bernd]

that looks like a layout strategy issue

Comment 4

[Martijn Wargers (dead)]

This got fixed between 2006-12-07 and 2006-12-08, so fixed by the reflow branch landing.

Comment 5

[Martijn Wargers (dead)]

Attachment: reftest

Comment 6

[Martijn Wargers (dead)]

Attachment: reference rendering

Comment 7

[Martijn Wargers (dead)]

Attachment: reftest

Comment 8

[Martijn Wargers (dead)]

Attachment: reftest diff

The only difference between the two is the width: 300px/width: 200px.

Comment 9

[Martijn Wargers (dead)]

Comment on attachment 261193 [details] [diff] [review]
reftest diff

Oops, I forgot about this test.

Comment 10

[Daniel Holbert [:dholbert]]

Two notes on attachment 261193 [details] [diff] [review]:
 - The testcase and reference case currently have different endline characters -- test uses Dos-style, and ref uses Unix-style.  Pick one or the other -- preferably Unix-style. :)  
(Currently, "diff" thinks the two files are completely different because of the endline difference.)

 - In which historical build(s) does this reftest fail?  I tested out both FF2 and Trunk from pre-reflow-branch (20061201), and in those builds, the testcase and reference case are *both* broken in the same way -- both show the red block at the right edge.  So those builds would still pass this reftest.

Comment 11

[Daniel Holbert [:dholbert]]

Sorry, I should've been more specific -- both of those notes refer to this pair of files from the reftest:
  244686-1.html
  244686-1-ref.html

Comment 12

[Daniel Holbert [:dholbert]]

(In reply to comment #11)
> Sorry, I should've been more specific -- both of those notes refer to this pair
> of files from the reftest:
>   244686-1.html
>   244686-1-ref.html
> 

oops, guess that's obvious -- I got confused 'cause I was also looking at attachment 293877 [details] [diff] [review] in another tab, which has these same reftests accidentally included in a larger reftest patch.

Comment 13

[Daniel Holbert [:dholbert]]

Comment on attachment 261193 [details] [diff] [review]
reftest diff

Marking reftest patch r-, per comment 10.