Re: [TUHS] a book (was Re: PWB vs Unix/TS)

[Eric Allman <tuhs@eric.allman.name>]

On 2019-09-14 17:58, Adam Thornton wrote:
> I...have never been all that impressed with Salus's work.  It's not _bad_ but it's also not terribly insightful.
I think Peter's work was an amazing effort to collect and disseminate
facts, and despite a few gaps (inevitable) he did a great job.  But
Peter's works were more collections of facts than attempts to interpret,
contextualize, or otherwise put the facts into a larger narrative.

Honest historians can disagree on the role of written histories.  A pure
"just the facts ma'am" history avoids context and interpretation but
tends to be fairly dry.  This was Peter's approach.  But it's impossible
to completely avoid bias because you have to pick and choose the facts
you include.  Contextualizing history inevitably leads to interpretation
which leads to some amount of bias, but interesting or even gripping
histories read like a novel that unfolds before you.

I've believed for a long time that when the definitive history of Unix
is written, Peter's books will be a major (albeit not "primary", in the
technical sense) source material.  I salute him for all his hard (and
early) work.  I hope that someone will step up to do this larger history
(much of which happened after Peter's publication dates) before we all
die off.

And I have to say, It looks like Warner's research (with all the
abundant help from this group) the last week or two is amazing.  I've
learned tons of stuff I didn't know, some of which didn't match my
memory, e.g., about generations of *roff.  As the author of the -me
macro package, I'm probably one of the handful of people in the world
who have ever used every feature in [nt]roff, many of which looked crazy
until I needed them, when they suddenly seemed to be genius.  I deeply
regret that I never had an opportunity to meet Joe Ossanna.

eric