[TUHS] Brian Kernighan and very early *roff history

[arnold at skeeve.com]

The name 'ditroff' for Device Independent Troff was how AT&T marketed /
sold / licensed it when they released it to the world.  The new name
was necessary to distinguish it from the original troff in V7 / System III /
System V.

It's understandable that inside Research no such distinction was made.
Nonetheless, the V8 and V10 archives show that the C/A/T variant was
still around (at least in /usr/src/cmd) under the name otroff.

HTH,

Arnold

Rob Pike <robpike at gmail.com> wrote:

> Dennis spent quite a bit of time cleaning up the troff code in the late
> 1980s, if I remember right, moving it to modern C. He got annoyed by it one
> day. It was the "ditroff" variant although honestly I don't remember us
> ever calling it that. It was just the current version of troff. Not sure
> where the name came from. Perhaps it was us but I think of it as a foreign
> name.
>
> -rob
>
>
> On Fri, Jan 14, 2022 at 11:05 AM Tom Lyon via TUHS <tuhs at minnie.tuhs.org>
> wrote:
>
> > Most of y'all are aware of Brian Kernighan's troff involvement. My
> > understanding is that he pretty much took over nroff/troff after Joe Ossana
> > died, and came out with ditroff.
> >
> > But Brian had much earlier involvement with non-UNIX *roff.  When he was
> > pursuing his PhD at Princeton, he spent a summer at MIT using CTSS and
> > RUNOFF.  When he came back to P'ton, he wrote a ROFF for the IBM 7094,
> > later translated to the IBM 360.  Many generations of students, myself
> > included, use the IBM ROFF (batch, not interactive) as a much friendlier
> > alternative to dumb typewriters.  I don't know if 360 ROFF spread beyond
> > Princeton, but I wouldn't be surprised.
> >
> > BTW, during my summer at Bell, nroff/troff was one of the few programs I
> > could not port to the Interdata 8/32 - it was just a mess of essentially
> > typeless code.  I don't think Joe Ossana got around to it either before he
> > died.
> >
> > --
> > - Tom
> >