260713 - [FIX][trunk] No cursor to indicate image resizing

Status: VERIFIED FIXED

(Firefox :: General, defect)

Description

[Bill Mason]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8a4) Gecko/20040920 Firefox/0.9.1+
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.8a4) Gecko/20040920 Firefox/0.9.1+

The magnifying glass cursor that normally appears over an image that has been
resized to fit the browser window (via the pref) no longer appears.

The image still toggles between full/shrunk size when clicked.  There's just no
cursor change.

Worked in the 2004091908 trunk build.
Broken in the 2004092008 trunk build.

Reproducible: Always
Steps to Reproduce:
1. Position mouse cursor over a 'shrunk to fit browser window' image.
2.
3.

Actual Results:  
No mouse cursor change occurs.

Expected Results:  
A magnifying glass mouse cursor should indicate that resizing has occurred.

This works in the 20040920 Aviary branch, and the 20040920 Seamonkey trunk. 
Only the FF trunk is affected.

Comment 1

[Bill Mason]

Mats, I'm wondering if your fix for bug 259639 ("Remove some cursors") caused
this regression.

Comment 2

[Mats Palmgren (inactive)]

I wonder too :-)  I get the same results as you - I can reproduce with FF but
not Moz. Here's the strange bit: the bug does NOT occur with a current
CVS "debug" build of Firefox on the same host (this is on Windows XP).
Then I tried an "opt" build - still no problem. Finally, I tried
with "--enable-static --disable-shared" (which is how the FF nightlies
are built) and then the bug occurs.

Furthermore, it's not just the zoom cursors, try attachment 154812 [details]
(Testcase from bug 163174) - most of the cursors on the lower part of the page
does not work (again everything works fine in the non-static builds.)

Looks like a build config/compiler/linker problem?

Comment 3

[:deanis74]

I added these cursors way back and the change hasn't been reverted nor
overwritten.  They're only defined if MOZ_STATIC_BUILD is defined.

http://bonsai.mozilla.org/cvsview2.cgi?diff_mode=context&whitespace_mode=show&file=splash.rc&branch=&root=/cvsroot&subdir=mozilla/browser/app&command=DIFF_FRAMESET&rev1=1.5&rev2=1.6

Comment 4

[Mats Palmgren (inactive)]

Attachment: Patch rev. 1

Thanks Dean, that was the problem.

I found that this file has been copied to two other places as well:
calendar/sunbird/app/splash.rc
xulrunner/app/splash.rc

I haven't tested those two changes though.
(and in the case of calendar I think the ..\\..\\ is not enough,
it probably should be ..\\..\\..\\ but I don't know so I'm leaving
it as is)

I have tested this patch on a static trunk build of Firefox on
Windows XP and it works as expected.

Comment 5

[:deanis74]

Comment on attachment 166058 [details] [diff] [review]
Patch rev. 1

Ahhh... those defines changed.	r=me, but what about mail/app/splash.rc?

Comment 6

[Mats Palmgren (inactive)]

Attachment: Patch rev. 2

Fixed mail/app/splash.rc too...
Do we need to have these values in xpfe/bootstrap/splash.rc too?
(it doesn't have any cursors at the moment)

Comment 7

[:deanis74]

It all looks good to me.  r=me if that stands for reviews in windows-specific
code in sunbird and mail.

Comment 8

[Phil Ringnalda (:philor)]

*** Bug 271309 has been marked as a duplicate of this bug. ***

Comment 9

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

Comment on attachment 166061 [details] [diff] [review]
Patch rev. 2

There should really be a better way to do this.  It seems like you could at
least get the constants with an #include, judging from widget.rc.  Could you
get the whole thing from an #include, or does that not work?

Comment 10

[Mats Palmgren (inactive)]

I checked in attachment 166061 [details] [diff] [review] (2004-11-26 11:49 PDT) to fix the regression.

I think an #include would work for the constants, but I don't know
how to fix the relative paths for the cursor files.
I have filed bug 271910 on this.

-> FIXED

Comment 11

[Bill Mason]

V 20041127 PC/WinXP

Comment 12

[:deanis74]

Comment on attachment 166061 [details] [diff] [review]
Patch rev. 2

setting r=me to get this out of my request queue