[TUHS] TUHS Digest, Vol 106, Issue 1

[TUHS] TUHS Digest, Vol 106, Issue 1

[Doug McIlroy]

>  CB/UNIX was developed to address deficiencies inherent in Research Unix,
>   notably the lack of interprocess communication and file locking

CB/UNIX was one of several versions in various divisions of Bell Labs
to implement IPC facilities beyond pipes and signals. Top management
in a division would declare that they wanted to use Unix, but needed
some particular IPC mechanism: semaphores, events, message passing, etc.--
and needed it right away. I always believed that these demands stiffer
as they percolated up through channels to the point that no alternative
mechanism would do. We in research would have preferred to seek a
general solution that would suffice to serve the various demands.
Besides, anything that we produced but didn't use ourselves would
automatically be suspect. We were very wary of featuritis.

Roughly speaking, each demand led to a different local flavor of
Unix, each (I like to think) reflecting the particular variant of
IPC with which one of its system designers worked in graduate
school. Somewhere between the wariness of research Unix, where
an ethos of generality ruled, and Linux, which offers a dozen ways
to do anything, there must lie a happy medium--a medium that I
believe would be much closer to Unix than Linux. That, alas, has
not proved to be the way of open source.

[TUHS] TUHS Digest, Vol 106, Issue 1

[A. P. Garcia]

On Tue, Sep 3, 2013 at 8:09 AM, Doug McIlroy <doug at cs.dartmouth.edu> wrote:

<snip>
> We in research would have preferred to seek a
> general solution that would suffice to serve the various demands.
> Besides, anything that we produced but didn't use ourselves would
> automatically be suspect. We were very wary of featuritis.
<snip>
> Somewhere between the wariness of research Unix, where
> an ethos of generality ruled, and Linux, which offers a dozen ways
> to do anything, there must lie a happy medium--a medium that I
> believe would be much closer to Unix than Linux. That, alas, has
> not proved to be the way of open source.

it happened to unix too, though. maybe not to research, due to the
ethos you describe, but it doesn't sound like Ken Thompson was
entirely happy with some of the things other people did:

"Probably the glaring error in Unix was that it underevaluated the
concept of remoteness. The open-close-read-write interface should have
been encapsulated together as something for remoteness; something that
brought a group of interfaces together as a single thing - a remote
file system as opposed to a local file system.

Unix lacked that concept; there was just one group of
open-close-read-write interfaces. It was a glaring omission and was
the reason that some of the awful things came into Unix like ptrace
and some of the system calls. Every time I looked at later versions of
Unix there were 15 new system calls, which tells you something's
wrong. I just didn't see it at the time. This was fixed in a fairly
nice way in Plan 9." --Ken Thompson

i also recall, well, a rant of sorts by tuhs's own Larry McVoy, where
he argued pretty vigorously to strip out all the cruft.