Karsten Hilbert wrote: >>If we have to maintain pypgsql ourselves there seems little downside to > > ... using something else. One alternative of which is yours. > Actually, I see only two viable alternatives: "yours" and > psycopg.
Um, one thing in favour of pyPgSQl is that it has a reasonable set of unit tests - about 100 from memory. Personally I wouldn't use a database adaptor that doesn't have a minimal set of unit tests, in the same way that I wouldn't use a database which doesn't have unit tests (PostgreSQL has over 1000 I think). Tim C _______________________________________________ Gnumed-devel mailing list [email protected] http://lists.gnu.org/mailman/listinfo/gnumed-devel
