Hi, Did you try deleting all .pyc and .pyo files in /home/prologic/work/circuits recursively and then trying?
Since it looks like python 2.6 will create the pyc files in the master process and then the popen to python 2.5 barfs on them. As Holger pointed out PYTHONDOWNWRITEBYTECODE should turn off the master python 2.6 creating pyc files. But if you've still got old ones lying around then the slave python 2.5 will no doubt barf. Another solution to this maybe to give each test process it's own area. So change the working dir of each tx and specify the rsync option to get the proper code to each test area (I assume the rsync doesn't push pyc files). Also that way if your code relies on it's own resources in the dir tree (writing out to a log file for example) then they will be separate and not clash. :) On 8 July 2010 22:44, James Mills <prolo...@shortcircuit.net.au> wrote: > On Thu, Jul 8, 2010 at 10:40 PM, holger krekel <hol...@merlinux.eu> wrote: >> does it help if you set the environment variable >> PYTHONDOWNWRITEBYTECODE=1 > > No, this didn't seem to affect anything. > > cheers > James > > > -- > -- James Mills > -- > -- "Problems are solved by method" > _______________________________________________ > py-dev mailing list > py-dev@codespeak.net > http://codespeak.net/mailman/listinfo/py-dev > _______________________________________________ py-dev mailing list py-dev@codespeak.net http://codespeak.net/mailman/listinfo/py-dev
