148784 - display offsets incorrect under 100dpi X server

Status: RESOLVED WORKSFORME

(Core :: Layout, defect, P5)

Description

[Jason Lunz]

This bug summarizes a recent thread on the debian-user about a mozilla layout
bug and gives a workaround for it, in hopes that the workaround can be made the
default until the actual bug is fixed. I'm filing this (perhaps) overly-detailed
bug report because it took me many months to figure out what caused this problem
and gave me a bad impression of mozilla.

Using either mozilla or galeon under an X server running at 100dpi (a setting
that is often the default under debian woody, but not on many other linux
distributions), the default setting for user_pref
"browser.display.screen_resolution" causes the layout engine to have display
corruption. Sample images are attached, and descriptions of the buggy behavior
are in the thread linked below.

Using the mozilla UI to add this setting:
    user_pref("browser.display.screen_resolution", 96);

to prefs.js fixes the problem for both mozilla and galeon. This is the thread
where the bug and its workaround are mentioned:
    http://lists.debian.org/debian-user/2002/debian-user-200205/msg04163.html

This all happened under debian woody, with debian sid versions of galeon and
mozilla:

    xserver-xfree86 4.1.0-16
    Mozilla 1.0 Release Candidate 2
    Mozilla/5.0 Galeon/1.2.1 (X11; Linux i686; U;) Gecko/20020421
    Debian/1.2.1-2

Comment 1

[Jason Lunz]

Attachment: display corruption example #1

Comment 2

[Jason Lunz]

Attachment: display corruption example #2

Comment 3

[Jason Lunz]

Attachment: display corruption example #3

Comment 4

[Tuukka Tolvanen (sp3000)]

See also bug 136028, bug 136295

Comment 5

[robertojimenoca]

This bug should be fixed. Please close it.

Comment 6

[Jason Lunz]

I certainly haven't seen this in ages. I'm still running gecko in galeon (where
I first saw this), I'm still using debian X servers at 100 dpi, and I have
browser.display.screen_resolution set to 0 in my prefs.js. That combination used
to cause it, so it's fixed for me.

Comment 7

[Jason Bassford]

No specific bug / patch specified as the fix.

-> WORKSFORME

Comment 8

[Loic Minier]

Hi,

Andrew Moise confirms this bug is still present for him and provided screenshots
of the bug in the Debian report.

His report is at <http://bugs.debian.org/148916>, and the screenshots are
<http://bugs.debian.org/cgi-bin/bugreport.cgi/corruption-0.png?bug=148916&msg=20&att=1>
and
<http://bugs.debian.org/cgi-bin/bugreport.cgi/corruption-1.png?bug=148916&msg=20&att=2>.

Please reopen this bug.

Comment 9

[Loic Minier]

(Might be Galeon specific although I don't see how.)