[Simh] RT-11 source

[Ethan Dicks]

On Wed, Oct 26, 2016 at 4:13 PM, David Holland
<david.w.holland at gmail.com> wrote:
> (Not that I can add much to the thread.)   Other than the Minix, and split
> I/D discussion tickled an old memory..
>
> I'd say it depends on the the version of Minix.
>
> My Google Foo Finds these two threads (quickly) WRT Minix & split I/D.
>
> http://www.verycomputer.com/79_f57ba779880523a6_1.htm

Ah... well it seems to matter on the version, but MINIX 1.0 through
1.2 might be small enough to fit.

> https://groups.google.com/forum/#!topic/vmsnet.pdp-11/Azzk6fQHmLE
>
> The latter is funny in this context, that thread is only 12 years old.

Indeed.

The mention in one of those threads of a "free 'for-real' C compiler"
is well said.  ISTR the thing that made MINIX possible in the first
place was the Amsterdam C compiler which, for reasons of cost, was not
available to most of the readers of Tanenbaum's book so it wasn't
possible for most of us to recompile from source anyway.  Back in the
1980s, we had a project at work that was based on our other
68000-based products, and the development environment chosen was a
Perkin-Elmer 4-user UNIX workstation in large part because it _came
with a C compiler that emitted M68K object code_ that we then
transmuted into .o files that worked with our in-house assembler to
build the kernel for our product.  The cost of a standalone C
cross-compiler for a "real machine" in 1984 was hideous compared to
the cost of a UNIX workstation that already had a compiler as part of
its necessary tool set.  The world is different today, but I'd still
approach the project of a new UNIX-like OS port to the PDP-11 by
answering the question of what C compiler is going to be used for this
task, and add to that the question of how well does it integrate with
a few percent of PDP-11 assembler source.  Define the toolchain first
and the source kinda falls into place.

-ethan