Subject: Re: bin/7249
To: NetBSD Kernel Technical Discussion List <tech-kern@netbsd.org>
From: Robert Elz <kre@munnari.OZ.AU>
List: tech-kern
Date: 07/22/2000 04:02:32
    Date:        Thu, 20 Jul 2000 21:30:19 -0400 (EDT)
    From:        woods@weird.com (Greg A. Woods)
    Message-ID:  <20000721013019.25EAC90@proven.weird.com>

  | Indeed it's not portable.  I think groff was/is the only *roff to ever
  | implement .so internally.  (Mabye sqtroff had some support....)

Hmmm - .so is just about as ancient *roff as .br is ... I have been using
nroff for a long long time, and .so has always been there (I haven't been
using troff quite as long, as I used nroff back before I had any kind of
method of displaying troff output - but .so was in troff too).

There's nothing even slightly "not portable" about .so

However ...

  | Another issue is the fact that none of the troff preprocessors [...]

Yes, .so is preprocessor grief - though it need not be.  soelim helps
with that (unless the *roff code is broken (ie: the doc source) it
should always be possible to replace a .so by the contents of the named
file.

But soelim was really built to assist spell if I remember correctly - without
it, the included files never got spell checked.

kre