On Sat, Oct 10, 2009 at 07:47, Iwan Vosloo <[email protected]> wrote: >> Well that global (ie provided by default) collection, session and >> metadata was a poor design decision and this is one of the numerous >> problems it brings. At some point, I should really think things >> through and see how to get rid of them with minimal upgrade hassle. > > I'd be interested to see what you come up with...
Don't hold your breath though, it might take a while before I fix that issue, if ever. Of course, a patch in that direction would help ;-). >> > Neither do we want to be >> > dependent on when what gets imported... >> >> Well, that's the only semi-clean (IMO) option currently available. > > The problem is that we don't always have control over that. And, it is > such an unintuitive place to look for problems. Such a secondary > impact of import order is almost invisible in a way. For me the > logical place to do something about it feels like setup_all (or > something analogous to it)? > > We're doing the following now (before setup_all): > > elixir.session.configure(autoflush=True, transactional=True) > elixir.session.registry.scopefunc = custom_scope_func Well, I didn't dig that far into SQLAlchemy's code. But I thought this would work until I saw your post on SQLAlchemy's list... Note that a session is *not* thread-safe, so whatever you are trying to do needs to take that into account (and that *might* be the reason your test is failing). -- Gaëtan de Menten http://openhex.org --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "SQLElixir" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/sqlelixir?hl=en -~----------~----~----~----~------~----~------~--~---
