> For the sake of laziness, I want local database to be sqlite. Bad idea. You need to be testing on your production database from the beginning.
For example, the Postgres driver will throw an error when it sees a query like "SELECT foo FROM bar WHERE foo.id NOT IN ()..." Sqlite will accept the query same query. You can create such queries with a statement like "~(db.foo.id.belongs(foobar))" if foobar is an empty list. Likewise the collating orders for different databases are different as are the upper/lower case boolean matching rules. You can encounter surprises in your sorting and searching widgets if your development and unit test database is different from your acceptance test/ production database. LIKE clauses in MySQL are case insensitive, whereas in Postgres they are case sensitive. On Jan 9, 10:04 pm, seongjoo <seongjoo....@gmail.com> wrote: > Thank you for letting me know a wonderful technique. Web2py rocks! > > For the sake of laziness, I want local database to be sqlite. Then, I see a > little problem that the deployed code will also create its sqlite on the > web server too. I hope the deployed code stores its data to production db > only and not to sqlite of the web server. Can there be a programmable > solution of this too?
