I think i disagree with the general motivation here.

The ability to arbitrarily map models and properties over data sources is a
vital tool when pulling data out of legacy stores.

What we need to figure out a way to do is separate the notion of
repositories as scopes/contexts (which i use regularly) from the ability to
give fine grained control over how a model is defined across stores.
 Basically when it comes down to it, i don't like the idea of coupling
models to specific repositories.  Models are mutable for a reason.  That's
one of the things that impressed me so much about DM when i first started
using it.  You just write your description of what data you want access to,
and boom, you've got querying over your data.

I think also that part of the lack of clarity may arise out of the fact that
once your data is in ruby land (i.e. you've got an instance) you've lost all
sense of how that data was mapped into ruby.  Perhaps if there were better
ways to interrogate repositories, and/or drill down on how results were
gathered from repositories, some of clarity issues would be alleviated.

Regarding the identity map Tony, that seems to me like something that needs
to be fixed in the ID map.  But again maybe making DM smarter about
repositories would make that easier.

-- 
You received this message because you are subscribed to the Google Groups 
"DataMapper" 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/datamapper?hl=en.

Reply via email to