[MERGE] Status should not special case conflicts

[Robert Collins]

On Mon, 2006-07-17 at 20:26 +1000, Martin Pool wrote:
> On 14 Jul 2006, John Arbash Meinel <john at arbash-meinel.com> wrote:
> > The fact that 'Tree' implements conflicts() and just returns an empty
> > list, same with unknowns(). Seems like this is perfectly reasonable, and
> > simplifies the code.
> 
> Yes, that sounds fine to me too.
> 
> Including the conflicts in the result of the comparison is an
> interesting idea.  I don't see any logical reason why it couldn't be
> there, though it does stretch the usual concept of a delta a bit.
> Perhaps there should be an extended interface that gives you a delta
> including conflict information.

One significant reason to want all the data available from the delta
mapping is to be able to get everything we need from dirstate in a
single pass. If we have to build a full inventory to map conflicts that
refer to file ids back to paths, we'll lose all the performance work in
those cases.

I dont actually see how it stretches the delta concept myself, given
that we all agree that conflicts are a part of a trees state. It seems
*unusual* to me that deltas do not currently show:
 - conflicts
 - changes in parents

and possibly more things as I look deeper into this part of the code.

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: 191 bytes
Desc: This is a digitally signed message part
Url : https://lists.ubuntu.com/archives/bazaar/attachments/20060717/0fbbba1e/attachment.pgp