[Grok-dev] Re: STORM howto
faassen at startifact.com
Fri Mar 14 06:38:10 EDT 2008
Martin Aspeli wrote:
> These are good points. I don't have a particular preference one way or
> the other since I'm largely ignorant about Storm, except to say that I
> think SQLAlchemy is a marvellous project.
> I do worry a bit that if Grok tries to lessen the Zope "ghettoisation"
> (as Chris McDonough is so fond of saying), going with a Zope-focused SQL
> ORM library may be counter-productive. If people who write TurboGears or
> Pylons code have existing DB mappings, the idea of being able to port
> them to Grok and get object publishing "for free" is attractive.
Yeah, I understand. On the other hand, there's nothing particularly Zope
focused about Storm. As an interesting observation, when we have
discussions about what AJAX system to use, people make convincing
arguments for using the Zope community originated KSS.
Grok tries to reduce the ghettoisation of Zope primarily by other ways,
such as advocacy, documentation and making things easier to use. I
believe Zope 3 technology *is* mostly successful in reducing this, we
just have to take a last step. :)
Using what code everybody else is using may not always be the best
choice, as we cannot capitalize on the strengths of innovation our
community also possesses.
Anyway, that's the counter argument to this. The "reduce ghettoisation"
argument, which I support, should be kept in mind but cannot be
I'm on the fence between SQLAlchemy and Storm; SQLAlchemy does have some
features that would be very interesting for Grok.
More information about the Grok-dev