[prev in list] [next in list] [prev in thread] [next in thread] 

List:       darcs-conflicts
Subject:    Re: [darcs-conflicts] Problem with 4 non-commuting patches
From:       David Roundy <droundy () darcs ! net>
Date:       2006-01-09 13:59:55
Message-ID: 20060109135953.GB12301 () abridgegame ! org
[Download RAW message or body]

On Sat, Jan 07, 2006 at 06:09:09PM +0000, Ian Lynagh wrote:
> On Fri, Jan 06, 2006 at 08:22:02AM -0500, David Roundy wrote:
> > 
> > But it's not just the middle patches that are too limited.  If I try (as I
> > did at first) to commute A and C to the right of D, by all reasoning D
> > ought to be a left-conflictor, but our left-conflictor also isn't
> > sufficiently flexible: the effect ought to be AC, but that's not possible.
> 
> Hmm, right, given the commute of 2 conflicting patches, by symmetry we
> must have:
> 
> ABCD <-AB-> o[ A <| <@| -< B ]a
>             a[ A >- |@> |> B ]b
>             bCc
>             cDd
>      <-CD-> o[ A <| <@| -< B ]a
>             a[ A >- |@> |> B ]b
>             b[ C <| <@| -< D ]c
>             c[ C >- |@> |> D ]d
>      <-AD-> o[ A <| <@| -< B ]a
>             a[[B, D]]b
>             b[[C, A]]c
>             c[ C >- |@> |> D ]d
> 
> where the [[B, D]] has conflicted with A and C, and the [[C, A]] with B
> and D.

I agree.

I've updated http://darcs.net/DarcsWiki/Permutivity to include this
permutation (with a few question marks), and to have a few more effects
written in for other permutations.

> > representation, and we'll have to look at alternatives (e.g. Arjan's
> > idea).
> 
> IIRC Arjan hasn't solved the problem of knowing where the bits of effect
> should go when commuting conflictors, which seems to be the root of the
> whole conflictor problem.

True, as I understand it, his idea is more based around the idea of not
using commutation on conflictors.  I'd really rather not go that way, but
something has to be done.  Right now I've feeling a bit more optimistic
about conflictors that I was when I wrote that.  But I'm also terribly busy
right now.  :(
-- 
David Roundy
http://www.darcs.net

_______________________________________________
darcs-conflicts mailing list
darcs-conflicts@darcs.net
http://www.abridgegame.org/cgi-bin/mailman/listinfo/darcs-conflicts
[prev in list] [next in list] [prev in thread] [next in thread]