Paul Ramsey-3 wrote: > > there is nothing actually being > *done* to the database, so there is no need to COMMIT. If we didn't > need to BEGIN to get our hands on a binary cursor, we wouldn't do it. > The connection close in the end is our implicit COMMIT. >
Ok, but is it maybe just considered bad form to have a BEGIN without a COMMIT? I don't know enough about SQL to say whether or not this is the case. Would it hurt things at all to include the COMMIT? At the very least, maybe a comment in mappostgis.c clarifying why a COMMIT is not necessary might clarify things a bit. The comment about why the BEGIN is needed is indeed helpful. (Although only located at 1 of the 5 BEGIN instances, but perhaps that's sufficient.) - Rich -- View this message in context: http://www.nabble.com/postgres-transaction-warnings-tp15749333p15765216.html Sent from the Mapserver - User mailing list archive at Nabble.com. _______________________________________________ mapserver-users mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/mapserver-users
