Re: Intel Mac-mini OSX 10.5.1 guile test results

[Neil Jerram <neil@ossau.uklinux.net>]

Roger Mc Murtrie <rogermc@iinet.net.au> writes:

> I agree "the patch didn't change that line".
> I don't think the problem was caused by your patch or any corruption
> of it as I actually implemented the patch "by hand".
> It seemed to me that m4_esyscmd wasn't working properly as it inserted
> a CR at the wrong place in the "# Generated by GNU Autoconf 2.61 for
> guile 1.8.3." line of the configure file (after the 3 as I remember so
> that the period following the 3 appeared on a line by itself causing
> ./ configure to abort at that line).
> I tried to get some info on  m4_esyscmd from the on-line autoconf
> manual but, although it appears in the manual's index, B.5 M4 Macro
> Index, which provides alink to the Redefined M4 Macros section, it is
> not mentioned in that section or anywhere else in the document as far
> as I can see.
>
> Any other ideas?

Was this the first time that you ran autoconf in the Guile 1.8.3
distribution - or that something in the ./configure magic decided to
run it?

Guile distributions (like most free software tarballs) are set up so
that you normally don't have to run autoconf yourself, so I think the
above is quite likely.

In that case, I think we can just forget this problem.  The autotools
are very sensitive to exact versions of autoconf, m4, libtool,
automake etc., and I daresay to host OS also, so if we go down this
road we'll end up debugging all of those tools on MacOS, as well as
Guile!  When the next release candidate comes out, you won't have to
run the autoconf step on your OS.

Regards,
        Neil