Re: CDBI::Relationship::Glue ?

[prev] [thread] [next] [Date index for 2004/06/22]

--xHFwDpU9dbj6ez1V
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Wed, Jun 23, 2004 at 21:57:30 +0100, Tony Bowden wrote:

> I'd actually argue that the meta_info bit is important, rather than
> cruft. One of the main points behind meta_info is that application
> writers can interrogate it for information about what relationships
> exist. Yes, there'll still be the info on each of the has_a() and
> has_many() relationships, but most applications will care more that
> there's actually a many-many in place. Providing that for them is a very
> useful thing...

My objection to this is that using define_link will no longer be the
same as using the 4 relationship declaration it does for you.

If it becomes common enough, people might rely on it. As long as it's
not really going to be the interface when CDBI supports many to many
with multiple foreign keys, and stuff like that, I don't think it's a
good thing.

Currently this module still fakes many to many, because CDBI still deals
with many to many relationship as a programatic and recursive inflation
of several relationships more than a type of relationship. Is this a
good time to introduce meta data for something we don't yet offer?

Am I inventing a problem where there shouldn't be one?

--=20
 ()  Yuval Kogman <nothingmuch@xxxxxxxx.xxx> 0xEBD27418  perl hacker &
 /\  kung foo master: /methinks long and hard, and runs away: neeyah!!!


--xHFwDpU9dbj6ez1V
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (Darwin)

iD8DBQFA2L6MVCwRwOvSdBgRAoPcAJ9XbyAiU4Vux/9hJRikp2BLD2MA2gCcCLI7
Vezwsu/DCLGk/xnNNwOohuA=
=jhw7
-----END PGP SIGNATURE-----

--xHFwDpU9dbj6ez1V--