[darcs-users] Darcs in kernel-like development model

[Tristan Seligmann]

* Miklos Vajna <vmiklos at frugalware.org> [2008-02-08 11:51:36 +0100]:

> On Tue, Feb 05, 2008 at 09:47:39AM -0600, John Goerzen <jgoerzen at complete.org> wrote:
> > People use various tools to do this.  Patch queue management tools
> > like quilt or Mercurial's mq are common.  Others use git with frequent
> > use of its rebase feature, or Mercurial with transplant.
> 
> the rebase feature is needed because in git every commit has one or more
> parent. in darcs, this is not necessary, if you create patch A and then
> patch B, then you *can* push patch B only, without pusing patch A. so no
> need to have a rebase-like feature.

This is only true if patch B does not depend on patch A. If I understand
correctly, it should theoretically be possible to attempt to commute A
and B, resulting in a conflict, but darcs does not have any way to
perform such an operation. In addition, to resolve the conflict, you
would either need to record another patch, or amend-record B, which
produces a new patch B' that now conflicts with B (!!).
-- 
mithrandi, i Ainil en-Balandor, a faer Ambar
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: Digital signature
Url : http://lists.osuosl.org/pipermail/darcs-users/attachments/20080209/e39f725e/attachment.pgp