To summarize, the list feels that this is an incorrect model
BEGIN;
... first set of commands
ROLLBACK;
... second set of commands
END;
whereas this is how things are meant to work:
BEGIN;
... first set of commands
ROLLBACK;
BEGIN;
... second set of commands
END;
and that since ROLLBACK ends a transaction, it releases locks. The above is
correct for all journalling models except for OFF, where the effect of ROLLBACK
is undefined. (I'm ignoring SAVEPOINTs for now.)
Simon.
_______________________________________________
sqlite-users mailing list
sqlite-users@mailinglists.sqlite.org
http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
