So maybe this is done to relieve the parser for doing the semantic check, and so improve query speed..
Well, if that's the reason, maybe is better that way. Thanks. And sorry about the repost, it won't happen again. Regards On Sat, Oct 16, 2010 at 9:16 AM, Drake Wilson <dr...@begriffli.ch> wrote: > Quoth Germán Herrera <germanh1...@gmail.com>, on 2010-10-16 00:10:23 > -0300: > > As you may know, both MySQL and SQL Server engines would refuse to run > > the last query, indicating an error because not all columns come from > > aggregate functions and there is no "group by" clause.. > > > > Is this left on purpose?, can this behavior be switched? (already > > searched in the Documentation, and in the list of pragmas and couldn't > > find anything). > > I doubt it has to be "left on purpose"; in fact it's more the > opposite. I would think it's more work to detect queries that use > combined aggregate and non-aggregate results in ill-specified ways, > depending on how one's query compiler is built, so it's just a matter > of an extra feature that was never implemented because there was no > need for it. > > Is there a reason you want this type of query to raise an error? Is > it just a matter of a safety net, wanting to know when you're doing > something that's not that well-defined? The query is semantically not > very good, but there are many other kinds of meaningless queries that > are valid SQL; it's not really SQLite's job to check that for you. > > (I don't know what the SQL92 standard has to say on this, FWIW.) > > ---> Drake Wilson > _______________________________________________ > sqlite-users mailing list > sqlite-users@sqlite.org > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users > _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
