[PATCH] RFC: localrepo: look at .hg/hgrc.d/*.rc after .hg/hgrc

[John Coomes]

Peter Arrenbrecht (peter.arrenbrecht at gmail.com) wrote:
> On Thu, Mar 6, 2008 at 12:51 AM, John Coomes <John.Coomes at sun.com> wrote:
> > Bryan O'Sullivan (bos at serpentine.com) wrote:
> >  > Peter Arrenbrecht wrote:
> >  > > This is an attempt to solve the problem of configuring sets of
> >  > > related server repos such that common config is kept central.
> >  > > Instead of adding an include directive to .hg/hgrc, I simply
> >  > > mimick /etc/mercurial/hgrc.d/*.rc for .hg/hgrc.d/*.rc. This allows
> >  > > me to use symlinks to factor out common config:
> >  >
> >  > Nice idea.  Please polish it up and submit a real version.
> >
> >  The major goal mentioned on
> >  http://www.selenic.com/mercurial/wiki/index.cgi/SpecificHgrcForSetsOfRepos
> >  was to
> >         Do this extra configuration in such a way that new clones are
> >         automatically configured correctly.
> >  I don't see how symlinks in .hg/hgrc.d/* accomplish it, since they're
> >  not cloned (unless I'm missing something).
> 
> Absolutely correct. Sadly, we never got anywhere with that. And it was
> focused on handling _client_ clones. I see there's a new proposal by
> Jonathan Shapiro in this direction, which looks promising. But I
> haven't looked at it in detail yet.
> 
> Meanwhile, I am trying to solve a problem on the _server_ right now.
> ...

Aha.  Didn't realize this was addressing a different problem.  I'd
also love to have a solution for "client-side" hgrc files.

> ...
> Anyhow, I think this change is nothing radically new (see
> /etc/mercurial/hgrc.d/*.rc) and easy and safe to implement. And I'd
> like to see _something_ that helps in this direction in hg 1.0.

Yes, I agree; it's reasonable and consistent with the handling of
system-wide hgrc files.

-John