Hi! > After a few bounces (why doesn't PyDS respond to ^c and stop itself?)
The 0.7.2 still doesn't hook up signal handlers when running in the foreground. So you should stop it with pyds-stop in another terminal window. > trying to fix the calendar layout, I found that the upstreaming and > events databases were corrupted. Trying to access the upstreaming The upstreaming database can be deleted, but you will need to reregister with the Python community server, as that data is stored in the upstream database. Use the same email address and password you used with your current installation and you just get your old usernumber back on registration. The events database can just be deleted. > I am not sure whether this is PyDS threading issues or Metakit bugs. > In any case, Metakit should not crash while attempting to read data > from a database! PyDS threading shouldn't have any influence on Metakit, as it only uses the Python metakit interface that itself says it's python-threading-aware. And all write access to Metakit files is secured by locks in PyDS anyway, so only one thread at a time accesses the metakit database. > I tried updating to core Metakit CVS (+ my Python bindings) from > 2.4.9.2, but that did not have any effect. The corruption occured > under stock 2.4.9.2. Weird. There _are_ some situations where people reported corrupted metakit databases, but I was never able to reproduce them, except one situation where my PyCS settings.dat file was broken. It showed the same symptoms like your database corruption showed, so it might be the same problem. In my case a dying hard disc was the reason for the corruption, though. I never had corruptions with Metakit 2.4.9.2 during normal use, even though I quite often need to kill my PyDS installations with kill -9 due to my development versions I am running. So I think it must be rather weird and special situations that lead to this corruption issue. > If anyone (Jean-Claude?) wants to see one/more of the databases, I can > send them. Yep, that would be good, as the database file I have for PyCS is very big - over 25 MB - and so not really usefull for debugging. bye, Georg _____________________________________________ Metakit mailing list - [EMAIL PROTECTED] http://www.equi4.com/mailman/listinfo/metakit
