Re: My patches, complete

[prev] [thread] [next] [Date index for 2004/08/08]

--AqsLC8rIMeq19msA
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Sun, Aug 08, 2004 at 15:38:19 +0100, Tony Bowden wrote:

> Lots of tiny little patches which are self-contained and make sense on
> their own and can be integrated in a few minutes are much more likely to
> be applied quickly.
>=20
> If that's too much work for you, then feel free to just leave it, and
> I'll get some time some time to look at this, but it will be at least a
> month or more.

Most of the patch lines are the test (4 new test files, 3 test lib
modules, normalization of 09-hasmany and Actor.pm) and doc (60 or so
new/changed lines), which work on the patch as a whole, but the
features themselves can be broken apart to bits. The tests for the
links_together thing, for example, also do some stuff with many many
that has to do with the fixes.

There's the fixes to issues I raised earlier, and there's the
links_together relationship, which are very independant. I guess the
fixes are more work for you, and they're about 60% (estimate) of the
functionality patch lines. I'm not sure how easy it would be to
break apart the fixes, because they solve joint issues, but at least
the links_together mess is unrelated.

I suggest that you start reading the diff to HasMany.pm, which I
will be more than welcome to comment excessively if that helps, and
then see the LinksTogether.pm, and the doc changes in DBI.pm.

Afterwords the tests are long, but pretty standard, and I can assure
you they pass (on darwin 5.8.1 and linux 5.8.4).. =3D)

Lastly, I'd like to note that there's a lot of chaff in the diff making it =
look
big, like:

-	Film->has_many(actors =3D> Actor =3D> 'Film', { order_by =3D> 'name' });
-	Actor->has_a(Film =3D> 'Film');
+	Actor->has_a(film =3D> 'Film');
+	Film->has_many(actors =3D> Actor =3D> 'film', { order_by =3D> 'name' });

which was necessary for some stuff to work, since the accessors/relationshi=
ps,
table declaration, and sql were not in agreement.

To conclude - I would be happy to split it into some diffs that should be
applied together, but have to do with different parts. That's not too much
work, and breaks up the task, but I don't think it would help with a logical
partitioning that much.

Sorry that this came out so long... Ciao, and thanks!

--=20
 ()  Yuval Kogman <nothingmuch@xxxxxxxx.xxx> 0xEBD27418  perl hacker &
 /\  kung foo master: /me does a karate-chop-flip: neeyah!!!!!!!!!!!!!!


--AqsLC8rIMeq19msA
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFBFkIXVCwRwOvSdBgRAm/YAKCg6WrBXa6OLCTutCzQThffJK2PlACfRoul
QdVTo6CQstri+X7aHKz5LRk=
=92rl
-----END PGP SIGNATURE-----

--AqsLC8rIMeq19msA--