Re: Re: preview-latex 0.9 released...

David Kastrup wrote:
> > Agreed. It was based on a "this seems to work" test. libdir was
> > there for xemacs sake, and is no more. Does this work on xemacs?
>=20
> Did here.  Where do you think the XEmacs RPM was coming from?  But of
> course I have had not much feedback to rely on, and given the
> resonance on the group in the last month or so, it would have been
> very optimistic to hope that this would improve without a bad release.
>=20
> I did not hope it to be bad, but I now we are at least getting
> feedback.

Lunch hour. I hope I get some food too.=20

If $prefix is NONE, it is set to $ac_default_prefix now. May I suggest

AC_DEFUN(EMACS_SET_PREFIX,[
  if test "${prefix}" =3D NONE -o -z "${prefix}"; then
    AC_MSG_CHECKING([for prefix of $EMACS_FLAVOR])
    EMACS_LISP(prefix,[(expand-file-name \"..\" invocation-directory)])
    AC_MSG_RESULT([$prefix])
  fi
])

We might as well set prefix to something emacs-relative while we're at
it. Then the packagedir test works here too. I think there appeared
../../.. in the file-relative-name output before.

> > What worries me is that the current code seems to pick out the first
> > entry in load-path that matches "site-lisp" or "site-packages". I'm
> > not sure this is very robust either.
>=20
> Sure, so we have to sort our choice better.  Sorry for that.

Well, you do check that it is the last in the path string. What
worries me is users who have modified load-path.... but then again you
enforce that $prefix or $datadir is in the start of the path string.
Perhaps that is enough.

> If it is not under prefix...  The first thing I want is that we get a
> robust installation fitting a site-wide installation, regardless of
> what the user has configured for himself.  Everything else is asking
> for mysteriously failing package builds on some platforms.
>=20
> Once we have that in place, we can think about automated fallbacks.

Agreed. The EMACS_SET_PREFIX would then be too smart if a user runs
his home-compiled emacs in his home directory, and wants to install
system-wide. Hmmmmmmmm. On the other hand, the proposed configure test
explicitly states the prefix used.

/J=C5


--=20
`When you say "I wrote a program that crashed Windows", people just stare=
 at
you blankly and say "Hey, I got those with the system, *for free*".'
   Linus Torvalds


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click