[darcs-users] Re: why do identical changes have to be conflicts?

[Mark Stosberg]

On 2004-12-05, Tommy Pettersson <ptp at lysator.liu.se> wrote:
>
> I'd like to throw in an alternative view on conflicts.
> If two developers start to write the same lines of code,
> there is a conflict about who should write that code.
>
> Of course that doesn't mean there couldn't be a simple way
> to resolve the duplication in darcs.  If it really is a
> duplication and not just a hunk with some common code in two
> otherwise conflicting patches.
>
> So, inside darcs "ignore one of them" is a simple resolution,
> but outside of darcs the fact that the same file or the
> same lines of code comes from two different sources, could
> very well mean something much more important than "ignore
> one of them".  Or, it could just be a hunk with some common
> code in two otherwise not conflicting patches.  It's hard
> for darcs to know.

I think I understand this viewpoint and agree with it. I think it
implies that:

1. darcs should not consider two identical patches conflicts. Either
can be accepted as the correct merger.

2. darcs should report the event, in case it is of interest to the developers:
 "NOTICE: The following two patches were found to contain identical changes:
    * Susie applied diff from Zooko
    * John  applied patch from Zooko

Whether this logic can be applied to particular lines and not whole
'hunks' is not something I'll try to guess.

    Mark

-- 
http://mark.stosberg.com/