Re: C-x v i bug

[Dan Nicolaescu <dann@ics.uci.edu>]

Stefan Monnier <monnier@IRO.UMontreal.CA> writes:

  > >> >   > hg status test_front_end_spec.py
  > >> >   > *** failed to import extension rdiff: No module named rdiff
  > >> >   > ? test_front_end_spec.py
  > >> >   > The above warning confuses emacs.  C-x v i report 'already
  > >> >   > registered'.
  > >> > Indeed, vc-hg-status parses the result of "hg status
  > >> > test_front_end_spec.py"
  > >> > Is there are way to tell hg to ignore .hgrc ?  I.E. something like
  > >> > emacs -q ?
  > >> >   > After I fixed the above warning (removed rdiff from my .hgrc),
  > >> >   > then C-x v i worked correctly.
  > >> HGRCPATH= hg showconfig
  > > So should we run HGRCPATH= hg status by default in vc-hg-state?
  > 
  > I haven't used Hg and don't know what you can put in .hgrc, so maybe I'm
  > way off base, but at least for Bazaar and CVS I think it would be wrong
  > to ignore the user's .cvsrc and .bazaar/bazaar.conf.

At least for vc-hg-state using HGRCPATH= is justfied:
1. vc-hg-state needs to be very fast, so not having to initialize
plugins should help.
2. the return value should not depend on whatever the user puts in .hgrc
(I think, I've never written a .hgrc file, so...)