Closed
Bug 368898
Opened 18 years ago
Closed 17 years ago
Testcase from bug 368715 shows room at the bottom where there shouldn't
Categories
(Core :: Layout: Tables, defect)
Tracking
()
RESOLVED
DUPLICATE
of bug 370525
People
(Reporter: martijn.martijn, Unassigned)
References
()
Details
(Keywords: regression, testcase)
Attachments
(2 files)
This is a testcase that Ryan made, it's from bug 368715, in current trunk builds it shows a white space at the bottom, which doesn't happen in older trunk builds/branch builds. This regressed between 2006-12-07 and 2006-12-08, so a regression from the reflow branch landing.
Updated•18 years ago
|
Flags: blocking1.9?
Flags: blocking1.9? → blocking1.9+
Comment 1•17 years ago
|
||
This started to work correctly between 2007-06-25-01 and 2007-06-26-01 builds. I would guess that the patch for bug 370525 fixed this, since it changed when we do a special height reflow here. Does modifying the testcase to still trigger that special height reflow cause this bug to reappear?
Depends on: 370525
Flags: in-testsuite?
Comment 2•17 years ago
|
||
Yup -- changing the testcase to trigger a special-height reflow (by adding "height=800px" on the intermediate table) causes the buggy behavior to reappear.
Posting a modification of the original testcase, with that change incorporated, for the record. (original testcase is attachment 253425 [details])
So, I'm pretty sure this bug is fixed.
Comment 3•17 years ago
|
||
So the whitespace in your test case is expected and desired?
Comment 4•17 years ago
|
||
This version of ryan's original testcase shows that the innermost div actually falls outside of the innermost table, because there are no "td" tags around the div. More explanation to follow.
Updated•17 years ago
|
Attachment #270771 -
Attachment description: ryan's orig testcase, with borders div/table behavior → ryan's orig testcase, with borders to show div/table behavior
Comment 5•17 years ago
|
||
(In reply to comment #3) > So the whitespace in your test case is expected and desired? Yes, it's expected and desired. Sorry, I wasn't very clear about that. Reasons why: -- It matches FF2 behavior :) but that's not a very good explanation. -- Better explanation for what's going on: First of all (this took me a bit to realize) the 800px-high div isn't rendered inside the innermost table, because it **is not inside of a cell**. Notice, in the html: <table style="height: 100%"> <tr> <div ....>...</div> </tr> Since there are no "td" tags there, then for the purposes of this testcase, this is *actually* equivalent to: <div...>...</div> <table style="height: 100%"> <tr> </tr> This is demonstrated in attachment 270771 [details]. (Div has blue border, and innermost table has red dotted border -- it's the tiny red block in the lower-left corner.) So. Given this, the intermediate 800px-high table actually has two children in its cell: 1. an 800px-high div 2. a 100%-high table, which will expand to its parent-table's specified height, 800px. 800+800 = 1600, which is higher than the intermediate table's 800 px height, so this causes overflow and the resulting whitespace. In Ryan's original testcase, the difference is that the intermediate table has no specified height, so the innermost table doesn't expand to fill it and is actually empty (as shown by the tiny red box attachment 270771 [details]).
Updated•17 years ago
|
Status: NEW → RESOLVED
Closed: 17 years ago
Resolution: --- → FIXED
Updated•17 years ago
|
Resolution: FIXED → DUPLICATE
Updated•17 years ago
|
Flags: blocking1.9+
Updated•17 years ago
|
Flags: blocking1.9+
You need to log in
before you can comment on or make changes to this bug.
Description
•