On Tue, 28 Feb 2017 08:48:02 +0000 Hick Gunter <h...@scigames.at> wrote:
> "If they are not the primary key, then the parent key columns must > be collectively subject to a UNIQUE constraint or have a UNIQUE > index." Thank you, Hick. I have always thought af a foreign key as an existence test. Looking around, I see that other implementations also often require the referent to be unique. I'm not convinced that's justified theoretically, but at the moment I can't check against my usual resources. I now see how to solve the conundrum I faced that motivated my complaint. It requires more columns and UNIQUE constraints than I think are strictly necessary, but it can be made to work. --jkl _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
