postgres(at)athena(dot)frontlogic(dot)com writes:
> Goinging throught the libpq docs, I noticed that in all of the examples involving
> select statements, transaction blocks are used. I see why this is necessary for
> write operations, but I don't see the need in read operations that don't commit. Am
> I missing something? Any help appreciated.
It's done to get consistent data snapshots across several commands.
--
Peter Eisentraut peter_e(at)gmx(dot)net http://funkturm.homeip.net/~peter