[RFC] bzr.jrydberg.versionedfile

[Robert Collins]

On Wed, 2005-12-21 at 12:32 +0100, Johan Rydberg wrote:
> Robert Collins <robertc at robertcollins.net> writes:


> KnitFetcher violates these rules right now, assuming that the
> transaction can be roll back changes to a known state, which is not
> really true as Robert pointed out.  The reason this was done in the
> first place was to gain performance while fetching large amounts of
> revisions.

I don't understand this. I've not said anything about the algorithm by
which the things to be fetched is determined, only about the way things
are written to disk. The *maximum* cost you should be paying for this is
a memory queue of things-to-write-to-disk-later.

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/20051223/8551cf8a/attachment.pgp