114430 - margin inside table cell is not correct

Status: RESOLVED FIXED

(Core :: Layout: Tables, defect)

Description

[karnaze (gone)]

In the test case to follow, the cell block ends up with a negative y origin, but 
I have stepped through the table cell code and it doesn't appear to be setting 
it to that value (in FinishReflowChild).

                            TableCell(td)(1)@0172F2F0 next=02D94764 {30,0,3060,1
410} [state=70040004] [content=02D72C88] [sc=0172F220]<
                              Block(td)(1)@0172F350 {30,-195,3000,1815} [state=0
1c40004] sc=0172F3D0(i=1,b=0)<

Comment 1

[karnaze (gone)]

Attachment: testcase (3rd attachment from bug 111872)

Comment 2

[Julien Cegarra]

WFM (BuildID: 2002072204)

Could we close that one now ?

/jc

Comment 3

[Christopher Hoess (gone)]

->Tables

Comment 4

[Mats Palmgren (inactive)]

Attachment: Screenshot of testcase

Bug still occurs in a current CVS trunk build:

TableCell(td)(1)@0x8280f88 next=0x8291f74 {32,0,3264,1504} [state=30040004] ...

  Block(td)(1)@0x828102c {32,-208,3200,1936} [state=00c40014] ...

Add lines to the text area until the scrollbar appears and then the fun begins.

Comment 5

[Vedran Miletic]

retargeting

Comment 6

[Boris Zbarsky [:bzbarsky]]

Mats, what is the bug here, exactly?  The fact that the lower border of the
<textarea> is not painted?

Comment 7

[Mats Palmgren (inactive)]

I'll have to admit I can't remember exactly, my comment above isn't very clear
on what the *rendering problem* was.  It is clear regarding the negative
y-offset though (which is what karnaze reported as a problem in comment 0).

Comment 8

[Boris Zbarsky [:bzbarsky]]

Oh, I see.  So this bug is about the frame model not being what's expected...
gotcha.

Comment 9

[Bernd]

one needs to verify that this does not happen in
http://lxr.mozilla.org/seamonkey/source/layout/html/table/src/nsTableCellFrame.cpp#588

Comment 10

[Bernd]

Attachment: patch for the margin

mozilla up to now centers a overflowing block with respect to the table cell.
With the patch we would not create a negative offset (this makes the initial
rendering compatible to IE and opera)

Comment 11

[Bernd]

taking

Comment 12

[Bernd]

Attachment: testcase with different valign

Comment 13

[Bernd]

Attachment: clip ytop

Comment 14

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 158901 [details] [diff] [review]
clip ytop

r+sr=bzbarsky

Comment 15

[Bernd]

fix checked in