On Tue, 18 Nov 2003, Gert Doering wrote:

> sfortier@wishbone-tech.com (Stephane Fortier) writes:
>
> >MAKE_DIRS_PERMS=	/var/spool/fax ${ROOT_USER} ${ROOT_GROUP} 755
> >MAKE_DIRS_PERMS=	/var/spool/fax/incoming ${ROOT_USER} ${ROOT_GROUP} 755
> >MAKE_DIRS_PERMS=	/var/spool/fax/outgoing ${ROOT_USER} ${ROOT_GROUP} 1777
> >MAKE_DIRS_PERMS=	/var/spool/fax/outgoing/locks ${ROOT_USER} ${ROOT_GROUP} 777
>
> For 1.1.29 and up (which includes 1.1.30) the fax/outgoing/ directory
> should not be mode 1777 any longer.

I'll let seb respond to that. It should be easy to fix. At one time, I
had trouble getting our build framework to create the "fax" user before
build-time, but that may be fixed by now.

> >do-configure:
> >	${MV} ${WRKSRC}/doc/fax.1in ${WRKSRC}/doc/sendfax.1in
>
> Hmmm, what's that good for?  sendfax has its own man page, which is very
> much different from "man fax" (which is the overview page)?  I don't think
> it's a good idea to create a sendfax.1 and a sendfax.8 page.

That's this:

    revision 1.20
    date: 2002/03/06 19:19:02;  author: fredb;  state: Exp;  lines: +4 -1
    Rename the fax overview manpage from "fax" to "sendfax", to avoid any
    conflict with the "efax" package. "efax" already has an "efax.1", the
    actual man page, but "mgetty"'s "sendfax" manpage is "sendfax.8", so
    "sendfax.1" is available. Closes PR pkg 14254.

With several thousand packages in NetBSD pkgsrc, this sort of thing comes
up from time to time. The general solution is tag the packages as conflicting,
(so only one can be installed at a time), but the submitter of that PR wanted
to install them both at the same time. I don't see why anyone would want to
use anything but "mgetty", myself. ;-)

> >--- mg_utmp.h.orig	Fri Sep 19 15:48:15 2003=0A=
> >+++ mg_utmp.h
> >@@ -4,6 +4,22 @@=0A=
> >  * highly SysV / BSD dependent=0A=
> >  */=0A=
> > =0A=
> >+#if __NetBSD_Version__ >=3D 106020000 /* NetBSD >=3D 1.6B has =
> >SVR4-style utmpx */=0A=
>
> I'd appreciate if portability things like this would be sent back to
> me...

Yes, of course. I just resent the email I sent you in February of this year.

> (Policy things, like the listed Makefile changes, are local to a specific
> environment anyway.  But the C code changes are useful for people installing
> from tarballs, so they should be included upstream)

I couldn't agree more.

> (Besides this I'm not really happy with the style those changes are done)

Some of the contortions were driven by the desire to keep the diffs near
the top of the file, to make subsequent merges easier. Any other oddness
stems from my own peculiar way of looking at the world. It's a given that
you'll redo them your own way, anyhow.

Frederick