On Thu, Jun 23, 2016 at 10:23 AM, Mark Brand <mabr...@mabrand.nl> wrote:
> > Thanks. I'm curious about the thinking behind this change. What are the > benefits that offset the increased risk that a particular database will > seem to have been created correctly only to fail at runtime? I'm not > complaining, just curious. > IIRC, this change was added because with older versions of SQLite if you do "CREATE TABLE t1; CREATE VIEW v1 AS SELECT t1 ...; DROP TABLE t1; CREATE TABLE t1;" then the next time you open the database, the schema will fail to parse because the CREATE VIEW occurs out of order. Version 3.9.0 fixes that. The other changes to VIEWs in 3.9.0 adds the ability to specify column names in a VIEW which is consistent with common table expressions and the behavior of PostgreSQL. -- D. Richard Hipp d...@sqlite.org _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
