165703 - Background-image positioning on items broken over a line is incorrect

Status: RESOLVED DUPLICATE of bug 14777

(Core :: CSS Parsing and Computation, defect)

Description

[Stuart Langridge]

If an inline element is broken across a line by wrapping (on hitting the right
edge of the viewport), a CSS-defined background-image styled as "right
no-repeat" is displayed on the right-hand-side of the first part of the element
(before the line break), rather than at the right-hand-side of the whole element
(i.e., the part after the line break), as in the example URL.

Comment 1

[Phil Miller]

I'll confirm that I see this on Win98 build 2002-08-30, but I'm not sure if what
you state is the proper behaviour, need someone with such knowledge to confirm that.

Comment 2

[Boris Zbarsky [:bzbarsky]]

Oh, yay.  That inline containing block stuff again...

Ian, David, what's the right rendering here?  It seems to me that the expected
rendering from comment 0 is correct, so we should be looking at our
continuations when doing this.... 

One interesting question is what happens for a background image styled as "right
repeat-y" for an inline that wraps... do we get:

  ____________________________
__|________________________img|
|__________________img|

or do we get:

  ____________________________
__|________________img________|
|__________________img|

?

Comment 3

[Eric A. Meyer (dead account)]

Boris (comment 2): I believe the answer is neither.  Given my understanding of
inline layout, we should get:
   _____________________________
 __|____________________________    <-- note I removed your end-line caps
 ___________________img|

Inline objects are constructed as if they were laid out as one long block and
then broken across lines.  Thus your example would start as:
 ___________________________________________________
 |______________________________________________img|

...and then be snapped into however many pieces were necessary for display
across multiple lines.  That's why borders on lines broken across lines don't
cap on the ends.  Outlines are a different story, but then they always have been.

Of course, if the WG decided something different recently and I missed it, David
or Ian will be able to correct me.

Comment 4

[Hixie (not reading bugmail)]

Well, what you wrote is what I proposed in my CSS 2.1 comments, so I agree with
you (Eric). The spec, however, is silent on the matter at the moment. Indeed,
according to CSS 2, 'background-position' doesn't apply to inline elements.

Comment 5

[Eric A. Meyer (dead account)]

Er-- how'd I miss that?  You're right, and even in CSS2.1 and CSS3 
'background-position' does not apply to inline elements.  Damn, there goes a
trick from one of my recent books.  Is there any likelihood of changing that
behavior for either 2.1 or 3, or has that already been discussed and shot down?

Comment 6

[Hixie (not reading bugmail)]

It's in both my CSS 2.1 last call comments and my CSS3 last call comments and at
a recent meeting I _think_ we resolved to do it (check the wednesday afternoon
minutes to be sure).

Comment 7

[Hixie (not reading bugmail)]

*** This bug has been marked as a duplicate of 14777 ***