bug#9026: [PATCH] aclocal: handle ACLOCAL_PATH environment variable

[Stefano Lattarini]

[CC:ing automake bug 9026, which I should have done right away]

On Monday 19 September 2011, Paolo Bonzini wrote:
> On 09/19/2011 06:05 PM, Stefano Lattarini wrote:
> > Resurrecting the old thread "Add ACLOCAL_PATH to aclocal"; references:
> >   <http://lists.gnu.org/archive/html/automake-patches/2010-11/msg00089.html>
> >   <http://thread.gmane.org/gmane.comp.sysutils.automake.patches/4972>
> >
> > I really want the ACLOCAL_PATH feature to end up in 1.11.2, which I hope we
> > will be able to release in a few weeks.  The test cases I've added should
> > IMHO ough to show that the new feature is solid enough and interacts nicely
> > with the pre-existing aclocal features.  The documentation could probably
> > be improved, but that can easily be done at a later moment.
> >
> > I'll push the attached patch in 72 hours if there is no review by then.
> >
> > Paolo, since it's you who have written the previous version of this patch,
> > from which I've drawn heavily, are you ok to be listed as the main author
> > in the ChangeLog entry, and in the "Author" field of the git commit?
> 
> Just to be sure, you are not including 
> http://lists.gnu.org/archive/html/automake-patches/2010-11/msg00091.html 
> are you?
>
No, not yet at least.

> If so, why not?
>
Because, if I'm not reading it wrong (which I might be doing in fact, since
I've only taken a perfunctory look at it so far), it introduces an
incompatibility between the "system acdirs" added with ACLOCAL_PATH (which
can be used to override automake macros) and those added with dirlist
(which on the contrary cannot be used to do so).  I'm not saying this
is necessarily a bad thing, but is something to think through before
proceeding.  In any case, feel free to re-base and polish the patch if
you want, so we can take a "fresh" look at it; I'd love to have some more
feedback and contribution on this topic!
 
Thanks,
  Stefano