Re: [PyQt] Very strange matplotlib behavior

Serge Montagnac Wed, 14 Oct 2009 02:15:18 -0700

Hans-Peter Jansen wrote:

Am Dienstag 13 Oktober 2009 schrieben Sie:

I recompiled present snapshot of sip with debug options :


#0  0xb7391840 in ??
()
#1  0xb75ebfd0 in PyDict_GetItem (op=0xb4200390,
key=0xb08d1f00)
    at
Objects/dictobject.c:658

#2  0xb5dc0131 in sip_api_is_py_method (gil=0xbfa38b3c, pymc=0x9c2b526
"",
    sipSelf=0xb4208eec, cname=0x0, mname=0xb70c5d26 "event") at
siplib.c:6053

Program terminated with signal SIGSEGV, Segmentation fault.
The program no longer exists.


#2 points to a problem, that might be fixed with tonight's snapshot. 

Pete

A lot of thanks to you Pete and Phil,

The sip-20091013 snapshot resolved the problems in embedding qt4agg backends
and widget + their use in Eric4 as mplwidget ...!
The pyqt-20091013 compiled nicely (with g++.conf fix) but I still had to complete
the QtCore/sipQtCorecmodule.cpp attribute bug (qDebug(a0,"%s") of gcc-4.4.1 as
said by Phil.

Great gurus, I promise next time to apply immediately the full debugging procedure ;-)
It is a good lesson of a bug track for all.

Best regards,
Serge.

-- 
"We all know Linux is great...it does infinite loops in 5 seconds. 
                                                 Linus Torvalds"
--------------------------
     Serge Montagnac
  http://www.obs-psr.com
    [email protected]
--------------------------

_______________________________________________
PyQt mailing list    [email protected]
http://www.riverbankcomputing.com/mailman/listinfo/pyqt

Re: [PyQt] Very strange matplotlib behavior

Reply via email to