Re: [pysqlite] Possible Regression in pysqlite 2.5.3

Markus Gritsch wrote:
> [...]
> The remaining question is if SQLObject is doing something obviously
> wrong, which lead to the error in 2.5.3 which should be fixed in its
> implementation?  In this case I would try to bring it to attention in
> the SQLObject mailing list.

As Edzard found out, it was really a problem caused by a change in 
pysqlite 2.5.3. Basically, isolation_level=None in the connect() call 
did not work. This has been fixed in 2.5.4 released today, though.

-- Gerhard