167787 - Erroneous image-clipping on mouseovers.

Status: RESOLVED DUPLICATE of bug 79315

(Core :: Layout: Positioned, defect)

Description

[Adam D. Moss]

This page includes the following style rule:

a:hover {color:#990000; position:relative; top:1; left:1; text-decoration:none; }

I'm guessing that the author intends this to move image-links
down and right by one pixel when they're mouseovered.

On mouseovering the thumbnail images his does indeed seem to be happening
(unfortunately the table row has to expand to accommodate the new position of the
images which causes unpleasant table-wibbling but I think that's just poorly-
conceived style rules).

However, one undesirable effect which I believe to be a mozilla layout
bug is that on being mouseovered the thumbnail images also get their
top ~5 pixels cropped out to the background colour.  I don't believe
that this should be happening.

This is with mozilla-trunk at time of writing, Linux/gtkfe, x86.

Comment 1

[Adam D. Moss]

CC:ing dbaron based on a vague and unreliable recollection that
he's a guru of styling and tables.

Comment 2

[Adam D. Moss]

That's interesting, the 'back' icon also has the mis-clipping
problem and it's not in a table, so the table thing is likely a
red herring (removing from summary).

Comment 3

[Adam D. Moss]

Attachment: Minimal test-case.

This is the (very) minimal test-case.  It doesn't matter that
the image doesn't exist, the problem is still demonstrated.

Comment 4

[Adam D. Moss]

CC:ing pavlov, since the bad clipping only seems
to apply to image elements.

Comment 5

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

Perhaps a duplicate of bug 79315?  (I'm not sure.  I didn't look very closely.)

Comment 6

[Adam D. Moss]

That does look quite likely!  I'm not confident that I'm qualified to judge, though.

Comment 7

[HARUNAGA Hirotoshi]

this is a dup of bug 79315.

Comment 8

[Boris Zbarsky [:bzbarsky]]

.

Comment 9

[Christopher Hoess (gone)]

->R&A Pos

Comment 10

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

The URL is now 404, but the testcase looks exactly like one of the ones on bug
79315 that was fixed by the patch I just checked in.

*** This bug has been marked as a duplicate of 79315 ***