[PATCH 0 of 5] histedit: fix some problems

[FUJIWARA Katsunori]

At Wed, 4 Sep 2013 09:46:46 -0400,
Augie Fackler wrote:
> 
> On Wed, Sep 4, 2013 at 9:17 AM, FUJIWARA Katsunori
> <foozy at lares.dti.ne.jp> wrote:
> >
> > At Mon, 26 Aug 2013 09:50:44 -0400,
> > Augie Fackler wrote:
> >>
> >> On Mon, Aug 26, 2013 at 04:41:47PM +0900, FUJIWARA Katsunori wrote:
> >> > This patch series fixes some problems around histedit extension.
> >> >
> >> > For patch #5, I chose editing only outgoings which are ancestors of
> >> > working directory, when there are some roots in outgoings. But should
> >> > I choose aborting in such situation ?
> >>
> >> Not sure what you're asking here, can you rephrase and/or provide a diagram?
> >
> > I'm just afraid that aborting "hg histedit" was suitable for Mercurial
> > philosophy than automatic choice of target revisions, if there are
> > multiple roots in outgoing revisions.
> >
> > For example, if it could be assumed that users often misunderstand
> > about outgoing (and histedit target) revisions when there are multiple
> > roots in them, convenience of target revision choice shouldn't be so
> > important. "hg histedit" should be aborted to avoid confusion.
> 
> 
> If I understand this, you're asking if 'hg histedit --outgoing' should
> be the same as 'hg histedit outgoing()' or 'hg histedit "outgoing()
> and ::."'. The answer is that it should be 'hg histedit outgoing()',
> and that if there are multiple outgoing roots, we should abort.

Thank you, Augie. Comparison between 'hg histedit outgoing()' and 'hg
histedit "outgoing() and ::."' explains my fear well.

I'll send the revised patch which aborts histedit if there are
multiple outgoing roots.

----------------------------------------------------------------------
[FUJIWARA Katsunori]                             foozy at lares.dti.ne.jp