On Jul 4, 2008, at 00:45, Jure Vrscaj wrote:
> Could this be fixed in 1.0? It breaks compatibility, I know.
Although I'm not in any kind of position of making such a decision, I
can make an educated guess: no. The reason is that most developers
currently focus on getting a Django 1.0 out the door, and so there'll
be little to no time for a core committer to actually look at the
patch and commit it, adding tests and that.
That said, you should *definitely* file a bug report, and attach a
patch that fixes the misbehavior and also add a regression test. Doubt
there'll be any docs needed for this.
Also, what compatibility does it break? I can't see how this is a
common use-case.
Regards,
Ludvig "toxik" Ericson
ludvig....@gmail.com
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.8 (Darwin)
iEYEARECAAYFAkhtZp8ACgkQXnZ94Kd6KaettQCfbC6DfVjwbKj+1kcK4lXbnHs0
ar4AniFUpLSjQpe7PBjFnkluGqkSoJVS
=o0QH
-----END PGP SIGNATURE-----
We're always going to look at things like this. It's 10 minutes of
effort. :-)
Certainly file a ticket so that the information doesn't get lost, Jure.
The inconsistency looks a little wrong, although I'd have to think a bit
about which return type is the "right" one, since there are arguments
both ways.
Malcolm