commit-performance and iter_changes

Robert Collins robertc at robertcollins.net
Mon Nov 17 02:52:33 GMT 2008


On Tue, 2008-11-11 at 10:24 -0500, Aaron Bentley wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Robert Collins wrote:
> > So, we've wanted to use iter_changes for commit for some time
> 
> Actually, I thought we'd given up hope on that long ago.

It had seemed unapproachable, but I certainly hadn't given up on the
desire to reduce duplication and improve commit more.

> > I see it working something like this:
> > 
> d>  - tree to commit builder interface is a delta - iter_changes basically
> >  - on merge commits the commit builder is responsible for comparing the
> >    merged-in-tree with the basis to determine what nodes need
> >    last-changed fields updated
> 
> Seems like a reasonable plan.  It seems intuitively obvious that commit
> should only care about files that have been changed.  I think we just
> wanted to avoid having a different codepath for merge commits.

Yah - commit being such a key component, bugs in the graph generation
etc are painful and avoiding those was a key consideration in weighing
the cost of a separate codepath.

-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/20081117/471507fa/attachment.pgp 


More information about the bazaar mailing list