[ih] Where's Multics now, was Internet-history Digest

[Clem Cole]

hmmmm ... fascinating history...

Remember, if you talk to anyone in a product group from DEC from those days
(particularly his managers), you will hear the same line that while he was
brilliant and fabulous at getting out version 1, "Dave never did version 2
of anything."

On Mon, Aug 18, 2025 at 1:35 PM Bill Nowicki via Internet-history <
internet-history at elists.isoc.org> wrote:

>  They also recruited Dave Cutler from DEC, who had just finished working
> the Vax PL/I code generator. Many said it was the best PL/I,

And before PL/1 shipped as a product, Leslie Kling (head of Technical
Languages and someone I eat lunch with fairly often these days) had to redo
what DC gave extensively because it was not yet there, including the minor
problems of breaking too many of the other parts of the toolkit, such as
the debugger, IIRC.   I don't remember everything.  But my memory is the
specifics (I'm a OS guy and >>was not << involved in any of this -- I just
was social and heard the stories/drank beers the engineers who were
directly effected) but I believe that at some point, PL/1 had a separate
incompatible runtime for PL/1, from which the rest of the compilers were
using or some such.  I just remember it was a horror show in Tech Languages
and Tools when it was tossed over the wall.

FWIW:   in the mid-90s, there was a huge celebration/party in ZK03 in the
VMS team when the last line of DC written code was removed from the VMS
kernel.   Those are some of the folks I can ask about the specifics of the
RSX to VMS heritage, BTW.

BTW:  About a month ago, it was noted by someone else who had joined us at
lunch for the first time in about 8 months, that Leslie remarked how often
this group drops into Culter stories.  We all tend to have them from those
days.

To be fair to DC, I tell similar stories about Joy and BSD. My line on Bill
is that he codes at 9600 baud.  While I respect him and think he is
brilliant, he types open curly brace, close curly brace, and he patches
faster than anyone I have ever known. Please open up the sources to the
original ex/vi or 4.1, and you can see the evidence of what I say.   Once
others, such as Sam, Kirk, and Keith, started to be more heavily involved,
the code started to get a lot cleaner.



> although the language never took off.
>
Hmm, it's not a fair statement. PL/1 was very popular in *some places*, but
not in the markets that DEC sold into, as it turned out. Remember, Culter
hated BLISS and C, and he looked at PL/1 as not being either.  DEC Sales
was saying that if they only offered PL/1, they could win sales in IBM
shops.   Tech languages had said it was a lot of work, and they lacked the
manpower, and it was not going to have any payback.

DC had proposed that he could write the PL/1 backend because he thought
Leslie's folks did not know anything, and he could do a better job at
writing a compiler backend anyway (since the front end was purchased from
Frieburghouse).  So Gordan said to DC, go ahead - this was what DEC called
an Advanced Development or AD project.

In the end, Leslie would be proved right in both cases (she's a pretty
smart lady, and I would listen to her advice on developing
production-quality compilers).