Code review comment for lp:~mbp/bzr/715000-stacking

John A Meinel (jameinel) wrote :

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 2/8/2011 4:49 PM, Martin Pool wrote:
> On 9 February 2011 07:50, John Arbash Meinel <email address hidden> wrote:
>
>> transitive_fallbacks does ok with this, if you consider that the
>> function is actually a public Repository api. (Which I think it should
>> be, and should be tested as such, with a per_repository implementation
>> test.)
>
> I was going to recurse through the public API, and originally took
> that approach, but the problem was that in this case the public API
> returns a KnownGraph, and I couldn't see how to combine several of
> those objects. Perhaps there is a way?
>

Not trivially. We have code to add new entries one-by-one, but that
isn't great.

If we wanted to tease this apart more, something that returns a dict,
and then wrapping that in a KnownGraph would be enough. Or expose more
of _find_ancestor (I don't remember the exact 'take one step' function
name).

John
=:->

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAk1RzqsACgkQJdeBCYSNAAMXPACeIOmDqkT8KfkedqNEGn6Dv89u
9moAoLpnFYmQ0Nnv2rmhGrUJN+x60pkM
=YQNy
-----END PGP SIGNATURE-----

« Back to merge proposal