mgetty should use autoconf and automake

• Messages sorted by:[ date ][ thread ][ subject ][ author ]

Hi,

Sverre Hvammen Johansen wrote:
> > mgetty is not, and has never been, in the Public Domain.  It's free for
> > nearly every purpose, but it's NOT Public Domain.
> 
> >From mgetty.texi:
> 
> >    This license expires, that is, five years after release date, a given
> > `mgetty+sendfax' release falls into the public domain, and everybody is
> > absolutely free to do with the code what he wants.
> 
> And you have stuff Copyrighted in 1993, which means that part of the 
> code lays in the public domain.

Well, yes.  A given *release*.  That is, all versions of mgetty up to
mgetty 0.13 (released july 26, 1993) fall under this.

This doesn't mean "any piece of code that was originally written in 1993
(but modified later)" falls into the Public Domain today.

> > Why?  This just makes scripts more clumsy, and gives much more possible
> > ways to misspell an option.
> 
>    One of the advantages of long-named options is that they can be
> consistent from program to program.  For example, users should be able
> to expect the "verbose" option of any GNU program which has one, to be
> spelled precisely `--verbose'.  To achieve this uniformity, look at the
> table of common long-option names when you choose the option names for
> your program (*note Option Table::.).

Well, yes, this is nice.  But then, what does "--verbose" do?  Sendfax has
a "-v" and "-V" switch that both do some kind of "verbosity", which one
should become "--verbose"?

> By using GNU getopt, some other extensions are included for free.

Yes, the GPL virus.

> > I will happily look at your configure script, thank you for that.  I won't
> > take all the other changes.
> 
> There are many more reasons why you also should use automake:
> 
> -- Others are used to configure-scripts made by autoconf and makefile.in
>    made by automake. The produced makefiles follows the brand standard
>    settled by the GNU project. 

mgetty is not part of the GNU project, and I do not have to follow their
"brand standard".

> Your makefile are missing some useful targets.
>    TAGS, dist, distdir, distcheck, uninstall, dvi, check, ...,

I will have to read the GNU docs what those are exactly, and implement
them, if I find them useful.  That's not too hard.

>    and your makefiles does not work with a VPATH build

Besides that, yes, VPATH support would be nice, on certain occasions.  

I haven't needed it anywhere yet, because with remote CVS, I have even
more flexibility in building the same set of sources among different
machines and architectures, without being limited to NFS mounted
disks.

> -- There is no problem to include your local stuff into the new specification.
>    I think it works by just dropping it into Makefle.am.

I will look at it.  I am NOT going to restructure everything "now", as I
said before.

> -- When using autoconf, there is some requirements that is tricky
>    to write on your own. By using autoconf, that part is generated 
>    by autoconf.

Last time I looked at autoconf (2.1), there were no "special" requirements
that weren't easy to do by hand.

gert
-- 
USENET is *not* the non-clickable part of WWW!
                                                           //www.muc.de/~gert/
Gert Doering - Munich, Germany                             gert@greenie.muc.de
fax: +49-89-35655025                        gert.doering@physik.tu-muenchen.de