[Haskell-cafe] Re: [ANNOUNCE] git-darcs-import 0.1

[Peter Hercek]

Aaron Denney wrote:
> On 2008-06-04, apfelmus <apfelmus at quantentunnel.de> wrote:
<-- cut -->
>> Or the  next->topic  path relies on features from  next  that are not 
>> present in  master . But then, you're screwed anyway
> 
> Yep.

Well not really, depends what kind the dependency is, this kind of rebase
  is useful when "topic" depends only syntactically (as you pointed later)
  on "next" or when the semantic dependency is only on a small part of "next".
  Git rebase allows you get the syntax or the small part of semantics to the
  rebased "topic" by asking you for (manual) conflict resolution. This would
  correspond to commuting darcs patches which depend on each other (again
  possible by providing manual conflict resolution).
Of course this happens only when it was anticipated that upstream merge
  of "next" happens before "topic", but then the upstream maintainers
  decided that "topic" should go upstream first. So, not often.


>> and should merge some parts from next into master so as to advance the
>> point where master and next fork.
> 
> That's one solution.  Of course, darcs doesn't have semantic dependency,
> but syntactic dependency.  (You can add extra dependencies to
> model semantic dependencies, but you can't take away the syntactic
> dependencies.)  Another solution, if there's syntactic,
> but not semantic dependencies, is to manually use patch and diff to get
> 90% there, and then cleanup and record.

OK, so I think this is what I expected for such a case.

Thanks for the explanation of the meaning of "merging patches prior head".

Peter.