816679 - Percentage max height ignored if parent has auto height but percentage max-height and grandparent has fixed height

Status: RESOLVED INVALID

(Core :: Layout, defect)

Description

[Scott Lindsey]

Attachment: screen shots of various browsers w/reduced test

User Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.16 (KHTML, like Gecko) Chrome/24.0.1297.0 Safari/537.16

Steps to reproduce:

Reduced case:

http://sample.scott.to/ffbug.html


Actual results:

The img seems not to respect the size of the container if that size is derived via max-height.


Expected results:

Webkit's behavior is more inline with what I expected, but hey they seem to be the only ones that feel that way.

Comment 1

[Boris Zbarsky [:bzbarsky]]

That's because the spec says to do what we do...

Comment 2

[Boris Zbarsky [:bzbarsky]]

Though it's hard to tell what behavior the testcase is expecting, exactly.  Especially given that it has no actual image...

Comment 3

[Scott Lindsey]

Arghh, looks like Firefox is correctly implementing a lousy spec.  

http://www.w3.org/TR/CSS21/visudet.html#min-max-heights

"percentage
Specifies a percentage for determining the used value. The percentage is calculated with respect to the height of the generated box's containing block. If the height of the containing block is not specified explicitly (i.e., it depends on content height), and this element is not absolutely positioned, the percentage value is treated as '0' (for 'min-height') or 'none' (for 'max-height')."


I suppose I'll go and file a bug with the google and see if that goes anywhere.

This is quite disappointing.  As a web developer, I expect max-height to more or less the same as does "height", but in this case per the spec -- it just does nothing.

Comment 4

[Scott Lindsey]

(In reply to Boris Zbarsky (:bz) from comment #2)
> Though it's hard to tell what behavior the testcase is expecting, exactly. 
> Especially given that it has no actual image...

Is the image I attached not suitable?

https://bugzilla.mozilla.org/attachment.cgi?id=686750

I was expecting the webkit behavior, but I gather that while it is more logical it is not to the standard.

Comment 5

[Boris Zbarsky [:bzbarsky]]

Ah, I missed that.  Yeah, that's just what the spec says.

Comment 6

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

(In reply to Scott Lindsey from comment #3)
> This is quite disappointing.  As a web developer, I expect max-height to
> more or less the same as does "height", but in this case per the spec -- it
> just does nothing.

How is max-height different from height here?

Comment 7

[Scott Lindsey]

Ahh, your right.  Neither height nor max-height (by percentage) on the image have any effect if the parent element has a height that was generated automatically.