[RFC] deprecate repository_has_same_url, replace with skip_fetch.

[Robert Collins]

This is fairly urgent in getting stacking working correctly. So if I
don't hear back soon, well it will just happen as a patch rather than
discussion :).

We currently short-circuit some fetch logic, for performance. The way we
do this is by checking if two repositories have the same url. If they
do, we do nothing.

We don't need to do this now that deep ghost-filling is no longer the
default. We can instead just fetch always except where fetching implies
a disk os read lock - pre metadir all in one trees. For those, it will
break on windows when you fetch within a single repository.

Stacked branches means that two repository instances at the same URL can
have different external references (this is good for scaling), and that
means that a fetch from one branch into another in a shared repo should
sometimes fetch data.

So - I propose a skip_fetch(other) method. This will:
return True for many old repositories when the url's are the same.
return False for stackable repositories when the reference lists are the
same.

-Rob

-- 
GPG key available at: <http://www.robertcollins.net/keys.txt>.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20080225/54c7c8ca/attachment.pgp