Re: ON Topic: Ada for CP/M

["Randy Brukardt" <randy@rrsoftware.com>]

Adrian Hoe wrote in message
<9ff447f2.0205252157.78b290ff@posting.google.com>...
>dewar@gnat.com (Robert Dewar) wrote in message
news:<5ee5b646.0205251457.788ad9b6@posting.google.com>...
>> "Frank J. Lhota" <NOSPAM.lhota.adarose@verizon.net> wrote in message
news:<WdPH8.6060$Np5.1189@nwrddc01.gnilink.net>...
>> > "Robert Dewar" <dewar@gnat.com> wrote in message
>> > > I think I still have my copy of CPM :-)
>> >
>> > That reminds me of a former co-worker who was a big CP/M fan. He
told me
>> > once that there was a CP/M - based Janus compiler for a small
subset of Ada.
>> > Has anyone here heard of it? I'm curious about how much of the
language they
>> > could do under CP/M.
>>
>>
>> You could certainly do a full language compiler for a
>> CP/M machine if there was a demand ...
>
>
>I was once a CP/M fan. I am still having the Apple CP/M Ada compiler
>in 5.25" diskette. I dont' know if the diskette is still readable or
>not (it has been almost 20 years which I started to have the diskette
>in mid 1980s). I don't know what Ada compiler it is.
>
>I can still remember faintly the compiler messages:
>
>lexical analysis++++++++
>semantic analysis++++++++++++
>
>something like that. Any idea which Ada it was?

There were two that I know of, the Stupidsoft (sorry, old habits:
Supersoft) Ada, and our Janus/Ada. Looking at the messages, you probably
had the bad guys. :-)

It was quite a trick to compile a useful subset of Ada on a 56K machine.
OTOH, targetting that machine is not particularly difficult. For Ada 83,
the Janus/Ada runtime never exceeded 12K on the 16-bit machines, Z-80
code was quite a bit more compact and also much more aminable to
compression.

So, I don't think that there would be any particular difficulty
targetting Janus/Ada 95 to such a machine (with compiler running on
Windows or Linux). You probably couldn't use every possible Ada feature
in programs targetting such a machine, but the sorts of programs
developed for such memory-limited systems don't do that anyway.

If anyone is interested in such a target, I'd be happy to try to figure
out how to read the 8" CP/M floppies that contain the source code to
those versions. (Our last CP/M machine doesn't boot anymore, couldn't
figure out why.)

           Randy Brukardt
           R.R. Software, Inc.