Code review comment for lp:~jelmer/bzr/rmbranch-colo

On 02/26/2012 11:17 PM, Martin Packman wrote:
> Review: Needs Information
>
> Substative changes all seems fine.
>
> Is this intended to go on 2.5 or dev? Either needs retargetting or release notes need moving (and conflict to resolve there anyway).
Yeah, mostly because it's a bug that people can remove active colocated
branches.

Alternatively, I could propose it against bzr.dev. colo support in 2.5
isn't fully polished anyway. That would mean fewer (risky) changes in
2.5. What do you think?
>
> -def lookup_sibling_branch(control_dir, location, possible_transports=None):
> - """Lookup sibling branch.
> -
> +def open_sibling_branch(control_dir, location, possible_transports=None):
> + """Open a branch, possibly a sibling.
>
> This looks technically like an api change if it's what's in 2.5.0 but as it's pretty new I guess it's okay to do.
I think that should be fine considering it's a fairly recent change,
indeed - and it's something that can't really be something that's used
by outsiders yet.
>
> - self.assertLength(2, self.hpss_calls)
> + self.assertLength(5, self.hpss_calls)
>
> Sneaking this up is from looking at more possible branch locations?
Yeah.

Cheers,

Jelmer