After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 81089 - Background window shows corrupt image
Background window shows corrupt image
Status: RESOLVED NOTABUG
Product: gdk-pixbuf
Classification: Platform
Component: general
git master
Other other
: Normal normal
: ---
Assigned To: gtk-bugs
Luis Villa
Depends on:
Blocks:
 
 
Reported: 2002-05-07 20:36 UTC by Andrew Sobala
Modified: 2010-07-10 04:06 UTC
See Also:
GNOME target: ---
GNOME version: ---

Attachments
Graphical bug description (22.48 KB, image/jpeg)
2002-05-07 20:36 UTC, Andrew Sobala 		Details
Problematic image (598.19 KB, image/jpeg)
2002-06-27 16:48 UTC, Andrew Sobala 		Details

Description Andrew Sobala 2002-05-07 20:36:03 UTC 
Version: 1.99.9

Look at the picture: http://www.e-consort.co.uk/gnome-aes/bug.jpg

This (greenness) only occurs in the stretch-preview, not on the desktop itself.
Comment 1 Andrew Sobala 2002-05-07 20:36:37 UTC 
Created attachment 8266 [details]
Graphical bug description
Comment 2 Jody Goldberg 2002-06-27 01:40:46 UTC 
1) does this still occur
2) if so can you attach the image
Comment 3 Andrew Sobala 2002-06-27 16:48:33 UTC 
Created attachment 9480 [details]
Problematic image
Comment 4 Federico Mena Quintero 2002-06-27 17:15:13 UTC 
Sigh, another bilinear scaling bug in gdk-pixbuf.
Comment 5 Federico Mena Quintero 2002-06-27 17:15:45 UTC 
This is very easy to reproduce.  Just load the attached image in EOG,
select bilinear scaling, and zoom down to a small size.
Comment 6 Matthias Clasen 2002-06-29 00:38:20 UTC 
The same shift-to-green bug appears for any interpolation method
except nearest neighbor. This will probably be fixed when we fix
the scale-down-by-large-factors inefficiency; see bug 80925.
Comment 7 Owen Taylor 2002-07-30 03:04:14 UTC 
Federico - The scaling down code for bilinear scaling
 is almost unrealted to the scaling up code.

Matthias - Won't changing over to different code just hide
 this problem? Or are you expecting some problem with
  integer overflow.
Comment 8 Matthias Clasen 2002-07-30 23:23:02 UTC 
My comment wasn't based on a deep analysis of the bug, just the fact
that it occurs for all interpolation modes which rely on creating
huge filter matrices for scaling down.