I think you're right. The SQL testing was done earlier with less data in the tables. I think at that time my Table A and B had the 1:1 relationship I planned for. Somewhere in the intervening data, the relationship changed and Postgres slapped me upside the head for doing something wrong. I'm using all the columns now and although it's painfully slow it works. The painfulness can likely be reduced by choosing the right indexes for my tables.
Thanks for your help Niphlod. You are Massimo are constant sources of great information. -- Joe On Monday, July 15, 2013 11:07:29 PM UTC-7, Niphlod wrote: > > that psycopg error is a postgresql error. I find very strange that a query > run through psycopg gives you an error while the "other tool" doesn't. > However, as stated earlier, in a query with a group by you "must" (read, > should, even if some tool outsmarts the relationships) fetch either > aggregates or the columns in the group by clause. That's T-SQL. > > >>>> -- --- You received this message because you are subscribed to the Google Groups "web2py-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to web2py+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
