[Dbix-class] Another SQLite 1.13 Issue?

[Marc Espie]

On Mon, Apr 09, 2007 at 08:39:31AM -0400, Christopher H. Laco wrote:
> Brandon Black wrote:
> > On 4/7/07, Christopher H. Laco <claco at chrislaco.com> wrote:
> >> Christopher H. Laco wrote:
> >> > Before I write this off as a SQLite 1.13 problem, I just wanted to get
> >> > some feedback and see if anyone else has seen this. IT all started when
> >> > perigrin tracked my down on IRC about some failing create tests in
> >> Handel.
> >> >
> >> > What id came down to was that the last passing test was testing for
> >> a PK
> >> > violation on insert. Then the next test, which inserts a record with a
> >> > unique fails with the [almost] same error about id not being unique
> >> even
> >> > though it is.
> >>
> >> Also an item of interest...
> >> If I put the pk test after the failing test, that test passed...but the
> >> next test after the pk test fails...in nother words, any create after
> >> the pk violation  test fails.
> >>
> > 
> > Is this with -current ?
> 
> Yes. And it does not happen with DBD::SQLite 1.12

Other similar frameworks (e.g., Rose) recommend to not 
use DBD::SQLite 1.13.

In fact, I downgraded the OpenBSD port of DBD::SQLite to 1.12 because
of various issues.

It also seems DBD::SQLite is poorly maintained, as this is a known issue,
but hasn't been fixed over the previous months. 

To the best of my knowledge the developer of DBD::SQLite is aware of the
problem, but has had his personal life taking over...