On Mon, Jun 13, 2005 at 03:07:09PM +0300, Jani Monoses wrote:
> 
> >>
> >>Thinking more about it I came to the conclusion that this is not actually
> >>a push variant rather a 'darcs mirror', since selective pushes might mean
> >>an inconsistent remote repo, since skipping the apply step means skipping
> >>validity check.  But for a darcs mirror operation I think it would be
> >>just OK.
> >
> >Actually, I think darcs is sufficiently well bug-tested that I wouldn't
> >worry about the consistency checking.  Leaving out the working directory
> 
> I was under the impression that unless a patch application is tried to a 
> working copy there is no way of finding out if the resulting repo is 
> consistent. Using manual copying of patches  under _darcs of a remote could 
> lead to a situation where the repo gets inconsistent. But I suppose that 
> with darcs push, the client darcs wouldn't let you push patches unless 
> their dependencies were pushed as well so I guess you're right.

No, there shouldn't be any situation where the lack of a working directory
would cause inconsistency.  It could cause you to not notice if there is an
inconsistency (caused perhaps by someone editing patches their repository
by hand), but shouldn't *cause* any inconsistencies.

> >would be quite a reasonable feature to add, and would actually make the
> >apply code significantly faster in comparison with --no-pristine.  It just
> >takes someone who wants to put in the effort to support this.
> 
> What is the new interface/option you have in mind? Adding a 
> --no-working-dir option to apply? I'll try doing it one of the following 
> weekends unless someone else steps up.

I hadn't worked out the interface.  Your suggestion sounds good to me.  I'd
probably add the flag both to apply and to pull, since the two are
symmetric.  A darcs revert -a would allow you to put the working directory
back if you ever want it... and I suppose we could also add the flag to
darcs get/put.
-- 
David Roundy
http://www.darcs.net

_______________________________________________
darcs-users mailing list
[email protected]
http://www.abridgegame.org/mailman/listinfo/darcs-users

Reply via email to