> 
> If you feel like testing, please copy over the following python files from 
> your 4303 to 4305 installation:
> 
> python/user-defined-restraints.py
> python/coot-utils.py
> python/extension.py
> 
> fingers crossed that should help.

Thanks Paul,
However, I copied over the python files from 4303 (I don't have a 
user_defined_restraints.py) and still get the error. 

Fatal Python error: PyThreadState_Get: no current thread
/sw/bin/coot: line 6: 37821 Abort trap: 6           /sw/bin/coot-real "$@"
Regards

MGM



On 4 Sep 2012, at 12:22, Paul Emsley <[email protected]> wrote:

> On 04/09/12 10:23, Martin Montgomery wrote:
>> Hello,
>> 
>> I'm having trouble with 0.7-pre-1-4305 installed through fink on osx 10.8.1
>> I updated to fink 0.34.3 then coot 0.7-pre-1-4305.
>> Coot now crashes after the launch of the splash screen with the following 
>> message:
>> 
>> Spacegroup: P 1
>> Cell: 40.631 109.18 93.243 90 90 90
>> initalize graphics molecules...done.
>> (filter-fileselection-filenames-state)
>> (get-active-map-drag-flag)
>> (use-graphics-interface-state)
>> DEBUG:: stating pydirectory /sw/share/coot/python
>> INFO:: importing coot.py from /sw/share/coot/python/coot.py
>> Importing python module coot using command from coot import *
>> INFO:: coot.py imported
>> Fatal Python error: PyThreadState_Get: no current thread
>> /sw/bin/coot: line 6: 50934 Abort trap: 6           /sw/bin/coot-real "$@"
>> 
>> The previous version (4303) was working perfectly (and still is on my laptop 
>> with fink 0.34.2).
>> 
>> A quick google showed at least two others had found the same problem but 
>> hadn't posted to the list.
>> 
>> Is there a fix?
>> 
>> 
> 
> I didn't know about this recent problem until I (too) did some googling.
> 
> I did see a ChangeLog entry from 2004:
> 
>       * src/main.cc (main): Put PYTHON macroed code before GUILE, so
>       that python is initialized when we have --with-guile
>       --with-python, which deals with "Fatal Python error:
>       PyThreadState_Get: no current thread".
> 
> 
> which makes me believe that I have inadvertently somehow made a similar 
> mistake (that is only apparent on Mac XXX).  How I could have done that and 
> why it only fails on Mac, I don't know.
> 
> Having said that, if you have localized the problem between 4303 and 4305 (as 
> seems to be the case), Bernhard (who is now on holiday) wrote r4304 and r4305 
> - but that consists mostly of python scripting changes (the prosmart 
> interface, in particular).
> 
> If you feel like testing, please copy over the following python files from 
> your 4303 to 4305 installation:
> 
> python/user-defined-restraints.py
> python/coot-utils.py
> python/extension.py
> 
> fingers crossed that should help.
> 
> Paul.
> 
> 
> 

Reply via email to