update on stack merging
robertc at robertcollins.net
Fri Jul 11 10:35:06 BST 2008
On Fri, 2008-07-11 at 19:08 +1000, Martin Pool wrote:
> OK, here's an overall patch that makes all tests pass in the top of
> the stacking branch!
> * KnitVersionedFiles.check should just check keys in its own file,
> then recurse into fallback kvfs.
> * The specialized get_parent_map and has_revisions in
> KnitPackRepository don't hook correctly into stacking, and are
Pls delete; we have a VCS :)
> This potentially causes a problem because I think now the base class
> version will be active, and that does not use the kvf index but rather
> reads each revision one by one, which may be slow. I have verified
> that method is on the instance but haven't profiled it. It should be
> fairly straightforward to fix.
Yes; thats an oversight for sure - we want to just return
keys = [(revid,) for revid in revision_ids]
(Modulo returning NULL_REVISION if johns goal of pushing that deeper
hasn't been done yet, and handling None).
Do you want this merged? Seems intrusive...(e.g. may hang PQM :))
+ import pdb; pdb.post_mortem(err)
I would rather we keep this test but invert the sense: it should work
rather than failing :).
- def test_error_mismatched_format(self):
- # this first implementation uses the internal pack list from
- # repository and thus needs the same format in both sides.
- base = self.make_repository('base', format='pack-0.92')
- repo = self.make_repository('repo', format=self.get_format())
- repo.add_fallback_repository, base)
GPG key available at: <http://www.robertcollins.net/keys.txt>.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 189 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20080711/6f5fda06/attachment.pgp
More information about the bazaar