Margin inconsistently applied with "writing-mode: vertical-lr"
Categories
(Core :: Layout, defect)
Tracking
()
People
(Reporter: zdanevich.vitaly, Unassigned)
Details
Attachments
(2 files, 1 obsolete file)
I recorded a video about it https://youtu.be/ShzU7cVqr9s
Updated•5 years ago
|
Updated•5 years ago
|
Comment 1•5 years ago
|
||
I'm having an hard time reproducing what is going wrong here.
I'm trying out of a simple example and don't see any difference with chromium:
data:text/html,<div style="margin: 10px; background: blue; width: 10px; height:10px;"></div>
Could you describe more precisely what is wrong or even better have a simple example where the issue happens?
| Reporter | ||
Comment 2•5 years ago
|
||
I can give you the account to this closed system where you can get real DOM.
Comment 3•5 years ago
•
|
||
I can reproduce on a small test case. Will upload.
I don't think it's purely devtools, it might rather be a layout issue.
Comment 4•5 years ago
|
||
STRs:
- opened attached page
- Note: here there's already a rendering inconsistency between FF and Chrome.
- open devtools
- select the div
- uncheck the margin property
- check the margin property
ER: The rendering should be the same as before we unchecked margin
AR: The box for the div is now bigger than it was before we unchecked margin. The rendering is now the same as on Chrome.
When closing DevTools the rendering goes back to its initial state.
Comment 5•5 years ago
|
||
Alternate test case without devtools involved.
STRs:
- open the page
- (layout is different from chrome, the
this not expanded by its child div) - click on the button
ER: After clicking on the button the margin of the div is restored to its initial value, so the rendering should be the same as before clicking on the button
AR: Rendering is different, the th is expanded to contain the div fully.
This seems 100% linked to using writing-mode: vertical-lr (meta Bug 145503)
Updated•5 years ago
|
Comment 6•5 years ago
|
||
(uploaded the wrong file!)
Comment 7•5 years ago
|
||
I think this is basically the same root issue described in bug 1332555. Marking as a duplicate.
Description
•