[TUHS] Brian Kernighan and very early *roff history

Noel Chiappa jnc at mercury.lcs.mit.edu
Fri Jan 14 23:07:52 AEST 2022


    > From: Angelo Papenhoff

    > to my knowledge no troff version before the C rewrite in v7

Apologies if I missed something, but between this list and COFF there's so
much low S/N traffic I skip a lot of it. Having said that, was there ever a
troff in assembler? I'd always had the impression that the first one was in C.

    > The v6 distribution has deleted directory entries for troff source but
    > not the files themselves.  I hope it is not lost. Maybe someone here has
    > an idea where it could be found?

The MIT 'V6+' (I think it's probably basically PWB1) system had troff -
i guess it 'fell off the back of a truck', like a lot of other stuff MIT had,
such as 'typesetter C', the Portable C Compiler, etc.

Theirs was modified to produce output for a Varian (I forget which model,
maybe the docs or driver say).

nroff on that system seems to have been generated from the troff sources; the
assembler nroff sources aren't present.

I looked at its n1.c, and compared it to the V7 one:

  https://minnie.tuhs.org/cgi-bin/utree.pl?file=V7/usr/src/cmd/troff/n1.c

and this one appears to be slightly earlier; e.g. it starts:

  #include "tdef.h"
  #include "t.h"
  #include "tw.h"
  /*
  troff1.c

  consume options, initialization, main loop,
  input routines, escape function calling
  */

  extern int stdi;

and in the argument processing, it has quite a lot fewer.

So that one is a "troff version before the C rewrite in .. v7", but it is in
C. Is that of any interest?

	Noel


More information about the TUHS mailing list