Re: weird AC_REQUIRE expansion issue

[Ralf Wildenhues]

Hi Stepan,

Again I'll only refer to part of your well-written mail, not because of
ignorance but because this issue is more important to me at the moment,
and I will come back later to the other one:

* Stepan Kasal wrote on Wed, Aug 31, 2005 at 04:11:08PM CEST:
> 
> On Wed, Aug 31, 2005 at 01:53:07PM +0200, Ralf Wildenhues wrote:
> > But a question for now: Can I assume the second parameters of both
> > AC_REQUIRE and m4_require to be public interface?
> 
> I seems I have just decided to change it!  Lucky it wasn't documented!

Well, for CVS HEAD Libtool, it would be very nice to be able to say

# LT_INIT([OPTIONS])
# ------------------
AC_DEFUN([LT_INIT],
[m4_require([_LT_SET_OPTIONS], [_LT_SET_OPTIONS([$1])])[]dnl
# ...
])

# _LT_SET_OPTIONS([OPTIONS])
# --------------------------
m4_defun([_LT_SET_OPTIONS],
[m4_foreach([_LT_Option], m4_split(m4_normalize([$1])),
        [...])
# ...
])

Would this be allowed?  If not, what can we do instead
(_LT_SET_OPTIONS will only be called once)?

> The fact that Autoconf users squeeze me asking what is documented and
> what isn't is two-edged: I'm no longer ashamed to change things which
> are not documented.  ;-)

This is a very dangerous "conclusion".  Most of the questions I ask are
motivated by the desire to mimimize Libtool maintenance costs.

I can tell you about one other project I am looking at a bit at the
moment[1], and trying to work around Autoconf-2.59 within a configure.in
written for 2.13 causes major pain.  Luckily most of that project has
been converted to use the most recent stable autotools.

> I'll make a patch implementing these ideas, eventually.

See above, such a thing would be "very nice to have".

Cheers,
Ralf

[1] http://www.open-mpi.org/svn/