Answering myself... It may be feasible to [with little code] catch schema changes that would invalidate DB triggers before committing them. I'm not yet sure how to do that without re-entrance, but that may be OK too (hey, OP_ParseSchema does it). At least I've an idea that might be worth pursing.
Also, in general, I find it an odd feature that SQLite3 does not warn about schema changes that leave triggers broken. It is a very useful feature to not have to drop triggers just because one drops and re-creates some table that the trigger bodies refer to, but it is also very surprising! For now I'm settling for the pragma, since I needed one anyways. _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
