[Info-vax] The (now lost) future of Alpha.

[invalid]

On 2018-08-01, Dennis Boone <drb at ihatespam.msu.edu> wrote:
> > What? 200 passes? As someone who is likely more knowledgable than most in
> > this particular arena, just what are you talking about?  Please, site
> > examples and references, I'm very interested.
>
> The source for the PL/1-F compiler was, iirc, in the MTS distribution,
> and I extracted those files for a bit of light reading a while back.
> Here are the assembler TITLE statements for all the files in that set.
> It's not one phase for every file, but perhaps this will give some
> idea.  A key thing to remember is that this compiler was supposed
> to be "usable" in tens of kilobytes (perhaps not even plural) of
> main storage.

Yeah, the S/360 Model 20 had a 32K variant, IIRC, but I didn't ever see
one. I worked on monster machines with 6 and later 8 whopping megs of
core. We had PL/I X at some point and it was outstanding. Somebody at MIT
wrote a paper about how the IBM FORTRAN H compiler was innovative in the
area of optimization even decades after it was written. Unfortunately, there
is no known copy of H-Level Extended (128 byte float support) FORTRAN
available. Only the doc remains, to tease us...

I know almost nothing about MTS, since I did not use it. But the source for
the PL/I F level compiler (and I think probably G) source should also be
available somewhere in the MVS3.8J distribution.