1. open inkscape 2. draw a box 3. set background type gradient 4. click edit gradient 5. switch to the secound color -> crash Happens every time Version: 1.2.1 Terminal output: Gtk-CRITICAL **: IA__gtk_paint_layout: assertion `style->depth == gdk_drawable_get_depth (window)' failed The program 'inkscape' received an X Window System error. This probably reflects a bug in the program. The error was 'BadMatch (invalid parameter attributes)'. (Details: serial 59544 error_code 8 request_code 140 minor_code 3) (Note to programmers: normally, X errors are reported asynchronously; that is, you will receive the error a while after causing it. To debug your program, run it with the --sync command line option to change this behavior. You can then get a meaningful backtrace from your debugger if you break on the gdk_x_error() function.) when I switch to gtk theme raleigh inkscape will not crash so i reported it here.
Can't reproduce with Inkscape 0.47pre4. What's your version?
Can't reproduce with inkscape 0.48.1 either. Please specify version (and version of oxygen-gtk)
inkscape --version Inkscape 0.48.3.1 r9886 (Mar 6 2012) on Launchpad I found something similar: https://bugs.launchpad.net/ubuntu/+source/inkscape/+bug/948178
So ... it seems like a "regression" in inkscape. I'd recommend to file a bug directly to them. possibly adding a pointer to this page. Doesn't mean that we can't fix it, but it might be easier on their side. Besides, I will only be able to try fix once my inkscape version is updated ... Hugo
PS: also, version 0.48.3 is apparently still tagged "unstable" by inkscape devs. Stable would be 0.48.2. Can you try downgrade inkscape and tell me if the crash still happens ?
inkscape --version Inkscape 0.48.2 r9819 (Jul 11 2011) same The error was 'BadMatch (invalid parameter attributes)'. (Details: serial 49281 error_code 8 request_code 140 minor_code 3) (Note to programmers: normally, X errors are reported asynchronously; that is, you will receive the error a while after causing it. To debug your program, run it with the --sync command line option to change this behavior. You can then get a meaningful backtrace from your debugger if you break on the gdk_x_error() function.) will report that to the inkscape project too
... interestingly, I just compiled and installed version 0.48.3.1, and still can't reproduce. Just to make sure that we are talking about the same thing: http://wstaw.org/m/2012/03/12/plasma-desktopzP3275.png That's the dialog that crashes for you (when you click on the drop down list) . Correct ? Also: what oxygen-gtk version are you using ?
(In reply to comment #7) > That's the dialog that crashes for you (when you click on the drop down > list) . Correct ? Yep: for better understanding here a video: http://www.youtube.com/watch?v=LXbEOu6wn1E > Also: what oxygen-gtk version are you using ? Oxygen-gtk is V1.2.1
Thanks for the video. Sadly enough, that's exactly what I did, and with no crash. Can you also provide your gtk+ version (running out of ideas here I must say)
I really dont know how to get the gtk version. my packagemanager calls them "2.24.10-0ubuntu5"
ok ! I have 2.24.9 I guess I should update that too :)
... still no luck with gtk+-2.24.10 (compiled manually) .... seems inkscape is rock solid here ...
ok. On my "other" computer, I am able to reproduce. Investigating.
ok. I 'think' I was able to track it down to an issue with latest glib. Using glib-2.31.20 -> crash Using glib-2.31.18 -> no Crash @Ruslan: The crash here is actually related to the issue you have in gimp when (image->scale_canvas), change unit. using glib-2.31.20 -> new unit does not appear glib-2.31.18 -> unit does appear. So, with you seing the gimp bug you should also be able to reproduce the inkscape crash (can you ?) In both cases they are due to depth mismatch. In which case. Can you reproduce my observation when downgrading glib ? If yes 1/ that would be good news 2/ I might even be able to track the guilty change in glib
ok. I have 1/ a dirty fix: trigger on size-request event rather than parent-set event to re-attach style (that fixes the depth mismatch), cause the later is apparently not working any more, but the former does. This is very innefficient 2/ a possible clean (and even cleaner than before) fix, to use a Signal rather that a Hook, on parent-set event, attached to all GtkMenuItem's children. Still working on 2/, and I'll push it to master as soon as done, for testing.
Git commit 8710e03009c987db65b27a877b8e682bfc3101fd by Hugo Pereira Da Costa. Committed on 19/03/2012 at 11:37. Pushed by hpereiradacosta into branch 'master'. moved style re-attachement from oxygenargbhelper to dedicated menuItem engine. M +1 -0 src/CMakeLists.txt M +1 -0 src/animations/oxygenanimations.cpp M +8 -0 src/animations/oxygenanimations.h A +93 -0 src/animations/oxygenmenuitemdata.cpp [License: LGPL (v2)] C +20 -24 src/animations/oxygenmenuitemdata.h [from: src/oxygenargbhelper.h - 059% similarity] A +58 -0 src/animations/oxygenmenuitemengine.h [License: LGPL (v2)] M +2 -66 src/oxygenargbhelper.cpp M +0 -9 src/oxygenargbhelper.h M +6 -0 src/oxygenstylewrapper.cpp http://commits.kde.org/oxygen-gtk/8710e03009c987db65b27a877b8e682bfc3101fd
... crap. This fixes inkscape, but not gimp. Committed too soon.
my bad. Wrong compilation. It actually does fix both inkscape and gimp. So closing. Ruslan: please double check (at least for gimp) I'll merge to 1.2 and gtk3 in a couple of days.
PS: thanks for reporting !
Yes, fixed for both inkscape and gimp (I was able to reproduce this bug with glib-2.31.20).