[Zope3-Users] IRelationshipMapper

[David Pratt]

Hi Pete. Isn't this what relationship does or could do?

http://svn.zope.org/zc.relationship/trunk/src/zc/relationship

Regards,
David

Pete Taylor wrote:
> Hi all,
> The other day I had an interesting idea for managing complex
> inter-object relationships in Zope3 in a way I haven't seen done
> before.  I threw a blog up about it so I wouldn't forget it when the
> idea struck me (things I don't write down, I tend to forget, no matter
> how seeming important at the time).
> 
> I wasn't going to mention it on the list immediately, because as with
> many ideas I get enthused about, after a day or two of thinking, I
> realize one of many different things about the idea.  The realizations
> regularly include, but are not limited to, a) that it has been done
> already, b) that it doesn't solve a problem I'm likely to have more
> than once, or c) that it introduces more complexity into a given
> system than the original problem space contained.
> 
> So, I waited on this one.  And a few weeks later, it still seems like
> an interesting idea.
> 
> But what I'd like is some commentary and feedback from other people in
> the Zope community, especially from anyone who may have pondered (or
> solved!) this kind of thing before.
> 
> The post is located here:
> http://baldtrol.blogspot.com/2006/06/irelationshipmapper.html
> 
> It is basically a statement of a way to manage non-containment
> relationships between ZODB persisted objects in a way that should not
> necessitate object-intrusive code.  However, There may be complexities
> that I've neglected, or other extant solutions for this same kind of
> problem in Zope3 that I'm just unaware of.
> 
> Any suggestions, comments, or thoughts on how to improve on this
> design would be more than welcome...  I would like to start working on
> an implementation of this idea...  I've done something like this on a
> smaller scale already, but would be very interested in hearing
> thoughts on this before I undertake the larger project.
> 
> Thanks all!
> Pete
>