Your first response to my question suggested that the 'shelve' object doesn't handle concurrent read-write access, so it would seem to me that using it as you suggest below would be problematic.
Could I suggest/recommend that you give some thought to moving the current data-access code to a new object so that folks can work on the other storage backends w/o having to mess with the core code? I can probably find some experienced Python programmers to help with this effort. We have about 80,000 Jabber users internally at EDS, and many of them really like to be able to chat with Yahoo users. Regards, Brian On Fri, 2006-10-27 at 08:46 +0200, Norman Rasmussen wrote: > On 10/26/06, Brian Safford <[EMAIL PROTECTED]> wrote: > > Having spent about 10 minutes looking at the yahoo-transport code, it > > seems to me that there might be a pretty simple way to get it to support > > multiple instances while using the dbm file methods. Here are my > > thoughts: > > Your thoughts are 100% valid, the problem is finding time for someone > to make the changes :-P > > Preferably I'd like to build multiple storage backends that all work > the way shelve expect them to. That way almost none of the existing > code has to change, and the admin gets to choose gdm/xml/sql in the > config file. > _______________________________________________ py-transports mailing list [email protected] http://www.modevia.com/cgi-bin/mailman/listinfo/py-transports
