Re: [PATCHES] Various

[Ludovic Courtès]

address@hidden (Taylan Ulrich "Bayırlı/Kammer") skribis:

> address@hidden (Ludovic Courtès) writes:
>
>> address@hidden (Taylan Ulrich "Bayırlı/Kammer") skribis:
>>
>>> address@hidden (Ludovic Courtès) writes:
>>>
>>>>> +       (uri (string-append "http://www.leptonica.com/source/leptonica-";
>>>>> +                           version ".tar.gz"))
>>>>
>>>> Really, they didn’t use ‘make dist’?
>>>
>>> I don't understand; what does make dist do differently?
>>
>> A tarball generated by ‘make dist’ (like most tarballs of GNU packages)
>> already has a ‘configure’ script and Makefile templates; thus it doesn’t
>> depend on Autoconf, Automake, etc.
>>
>> However, this package recipe explicitly adds Autoconf, Automake, and
>> Libtool as inputs, plus it runs ‘autoreconf’.  So I suspect this is
>> because upstream’s tarball isn’t self-contained, right?
>>
>> Ludo’.
>
> Oh, I did that because otherwise I get the following at start of the
> build phase:
>
> (CDPATH="${ZSH_VERSION+.}:" && cd . && 
> /gnu/store/nq6idcqwqc9x6z7g9jxq11a58jqx6w8x-bash-4.3.33/bin/bash 
> /tmp/nix-build-leptonica-1.71.drv-1/leptonica-1.71/config/missing --run 
> autoheader)
> aclocal.m4:17: warning: this file was generated for autoconf 2.68.

This may be because a .ac, .am, or .m4 file is newer than ‘configure’,
for instance.  Does any of these get patched, which would explain the
newer mtime?  Can it be worked around by changing that file’s mtime
using ‘set-file-time’ or similar?

> You have another version of autoconf.  It may work, but is not guaranteed to.
> If you have problems, you may need to regenerate the build system entirely.
> To do so, use the procedure documented by the package, typically `autoreconf'.
>
> However, it seems to build fine in the end.  Should I remove the
> autoreconf phase?

If it’s not needed then yes, definitely.  The important thing would be
to remove autoconf, automake, and libtool from the inputs though.

Thanks,
Ludo’.