Hi list,
I'm having a serious issue with inkscape on FreeBSD -current. I'm using the port available for FreeBSD and even a forced reinstall of it doesn't fix the problem. The maintainer of the port suggested to report the problem through this list.
The issue is quite simple to express: adding a layer to any opened file freezes inkscape. I can open any svg file without a trouble, and a file can contain layers, they are perfectly displayed. I can also hide any of the layers of a document, or rename a layer, change the order of layers, without problem. Only creating a new layer (or duplicating the current layer) is impossible, and leads to a freeze.
Tracing the process (using truss on FreeBSD) doesn't show anything once the process is frozen. If there is anything else I could do please let me know.
Does such an issue speak to anyone here? Merry Christmas all gb
As not many of us have access to a FreeBSD machine, it would be helpful if you could specify the exact version you're using (e.g. the output of inkscape --version ).
Thanks,
pjrm.
Peter Moulder (Peter.Moulder@...38...) on 26/12/2009 at 22:52 wrote:
Hi
As not many of us have access to a FreeBSD machine, it would be helpful if you could specify the exact version you're using (e.g. the output of inkscape --version ).
Oh sorry, should have tought about that:
~# inkscape --version Inkscape 0.47 r22583 (Dec 25 2009)
That's the version I get from the port. I also tried with the current svn revision:
~/inkscape# svnversion 22629
same issue with layers addition.
For Linux/Gnu users, this bug can be seen by doing
_INKSCAPE_GC=disable valgrind ./inkscape --verb=LayerDuplicate
The provisional suggestion is to modify document-subset.cpp by moving the "records.erase(obj);" line to just before the "removed_signal.emit(obj);" line.
However, this unfortunately still leaves me with a
"Conditional jump or move depends on uninitialised value(s)"
error from valgrind when I exit inkscape.
Anyway, Guy, see if this change is enough that layer manipulation works for you under FreeBSD.
pjrm.
On jan 01, Peter Moulder wrote:
Hi,
The provisional suggestion is to modify document-subset.cpp by moving the "records.erase(obj);" line to just before the "removed_signal.emit(obj);" line.
However, this unfortunately still leaves me with a
"Conditional jump or move depends on uninitialised value(s)"
error from valgrind when I exit inkscape.
Anyway, Guy, see if this change is enough that layer manipulation works for you under FreeBSD.
This change fixes the layer issue I was facing. Thank you very much, Peter! Happy new year all
gb
On Fri, Jan 1, 2010 at 12:23 PM, Guy Brand <gb@...2297...> wrote:
The provisional suggestion is to modify document-subset.cpp by moving the "records.erase(obj);" line to just before the "removed_signal.emit(obj);" line.
This change fixes the layer issue I was facing.
so should we commit this? any potential issues?
On Jan 1, 2010, at 7:45 PM, bulia byak wrote:
On Fri, Jan 1, 2010 at 12:23 PM, Guy Brand <gb@...2297...> wrote:
The provisional suggestion is to modify document-subset.cpp by moving the "records.erase(obj);" line to just before the "removed_signal.emit(obj);" line.
This change fixes the layer issue I was facing.
so should we commit this? any potential issues?
Yes. I'd actually thought it was committed, otherwise I would have.
At the moment I'm hooking in a filter bit to prevent "inkscape:*" attributes from sneaking into our plain SVG output. Once I finish that I was going to check and commit for pjrm.
participants (4)
-
bulia byak -
Guy Brand -
Jon Cruz -
Peter Moulder