Re: Installing gnulib from git

[Jim Meyering]

Bruno Haible wrote:

> Reuben Thomas wrote:
>> It didn't occur to me to read the manual, as I hadn't figured out
>> how to build it. I looked in Makefile, but that didn't seem to do much.
>
> This can be improved:
>
> 2009-04-01  Bruno Haible  <address@hidden>
>
>       * Makefile (info, html, dvi, pdf): New targets.
>       Reported by Reuben Thomas <address@hidden>.
>
> --- Makefile.orig     2009-04-02 03:09:05.000000000 +0200
> +++ Makefile  2009-04-02 03:08:58.000000000 +0200
> @@ -1,19 +1,31 @@
>  # Makefile for gnulib central.
> -# Copyright (C) 2006 Free Software Foundation, Inc.
> +# Copyright (C) 2006, 2009 Free Software Foundation, Inc.
>  #
>  # Copying and distribution of this file, with or without modification,
>  # in any medium, are permitted without royalty provided the copyright
>  # notice and this notice are preserved.
>
> -# Produce some files that are not stored in CVS.
> +# Produce some files that are not stored in the repository.
>  all:
>
> +# Produce the documentation in readable form.
> +info:
> +     cd doc && $(MAKE) info
> +html:
> +     cd doc && $(MAKE) html
> +dvi:
> +     cd doc && $(MAKE) dvi && $(MAKE) mostlyclean
> +pdf:
> +     cd doc && $(MAKE) pdf && $(MAKE) mostlyclean

Hi Bruno,

Do you mind if I factor that?

    html info:
            cd doc && $(MAKE) $@

    dvi pdf:
            cd doc && $(MAKE) $@ && $(MAKE) mostlyclean

or even (with or without the mostlyclean):

    dvi html info pdf:
            cd doc && $(MAKE) $@

But I'd leave out the mostlyclean or rename it to something
less suspicious sounding.  In automake-generated Makefiles,
mostlyclean removes far more than the hand-coded rule in
gnulib's doc/Makefile does.

A doc-*creation* rule like "make info" that automatically runs any
"clean"-like rule would seem to violate the principle of least surprise.