segfaults when running graphical reports

[John Ralls]

On Apr 25, 2014, at 9:46 AM, Geert Janssens <janssens-geert at telenet.be> wrote:

> On Friday 25 April 2014 09:20:07 John Ralls wrote:
>> On Apr 25, 2014, at 6:22 AM, Geert Janssens <janssens-
> geert at telenet.be> wrote:
>>> Yet there must be another factor here because I can't reproduce this
>>> at all. We may still have to dig a bit deeper to find the real root
>>> cause. But I have no idea where to start as it works on my
>>> machine...
>> These Python bugs might help clarify why it works on some systems and
>> blocks on others: http://bugs.python.org/issue14892
>> http://bugs.python.org/msg173501
>> 
>> Regards,
>> John Ralls
> 
> Thanks for the pointers. So there also the suggested workaround is to 
> inhibit the SIGTTOU signal. Except the links propose to add it right 
> before calling import readline in the code. I like that better than 
> calling it in the gnucash binary.
> 
> Should be re-enable the signal after readline is imported ? I'm not too 
> comfortable with blocking signals for no good reason...

If it works, sure. But maybe we should leave this to Mike, since he can actually get the hang and we can’t.

Regards,
John Ralls