[TUHS] Re: Trying to date "A Supplemental Document For Awk"

[segaloco via TUHS <tuhs@tuhs.org>]

> But what did people use before grog and why was the compilation line
> never placed in a comment in the file?

The primary macro packages I see come up between Bell and UCB are man, ms, mm, and me.  Man of course finds use in the manual pages (although there are different representations of manpages in nroff over time.)  From what I've seen (someone who was there can surely correct me) it seems that ms macros were more commonly used on the research side of things while the mm macros proliferated more in the supported side.  Finally the me macros were a BSD component.  Given these separations, the origin of or relative vicinity from which a paper originates provides much context as to which macros may be present.

To a finer point, the papers published with V7 are ms macros papers while the new additions in PWB lineages are mm macros, while some papers that crop up in BSD likely use me (although I haven't gotten too far into BSD with doc research yet.)  Papers from UNIX consumers such as universities are likely in ms or me most of the time.  On the flip side, mm was the macro package touted with Documenter's Workbench, so many commercial operations using System V for documentation would've produced documents in mm.  I'd be curious whether the earlier "Phototypesetter" package included ms or mm (or both.)  I don't think I've seen a "papers" set with both the Lesk ms document and the Smith and Mashey mm one, so couldn't say how common both in the same Bell offering were.  Additionally, my research hasn't touched on any officially sanctioned use of mm in BSD, so that's an area ripe for some more study.

As for other breadcrumbs, Bell mm macros papers do often include a comment at the top indicating to print with nroff -mm or mm(1).  I don't recall seeing similar in research papers, but haven't necessarily gone looking.  In any case, the paper sets with UNIX itself typically had scripts included with the necessary command-lines, as many papers additionally needed some eqn and/or tbl processing.  I imagine any other such formally distributed document sources would likewise include scripts in lieu of commentary, but it depends.

- Matt G.