On Mon, Aug 25, 2014 at 01:00:45PM -0700, Lyndon Nerenberg wrote:
It was quite astounding to see the wide range of
performance impacts
this had on various systems. 3B* systems would tip over and die, except
for the (built by Convergent Tech) 3B1.
Sheesh, you people keep bringing up stuff from my past. My buddy Rob
Netzer (used to be a prof at Brown, now works on BitKeeper with me)
had one of those 3B1s. Neat machine. Sort of like a desktop VAX.
We were roommates so we both got to use it, my dim memory is I bought one
too but that might be wrong. What I do know is that we had a compiler
class together where the prof for the class had written a lex/yacc equiv,
we had to come up with a grammar for a subset of Ada and implement it.
Only problem was that the system was an IBM 360 or something miserable
like that.
So Rob went to the prof and said if we write our own lex/yacc equiv can
we do it on our own system? The prof said yes and that was the brief
moment in time where I vaguely understood the difference between recursive
decent parsers and LL(1) and LR(1) (just kidding, I never really got it
other than recursive decent is what I'd write not knowing any better.
I think at one point maybe I got the difference but Rob is light years
ahead of me).
Anyhoo, we wrote an Ada compiler on the 3B1. It didn't do everything,
I think we punted on late binding and some other stuff, but it did a
surprisingly large subset of Ada. It opened my mind to what could be
accomplished in a semester. Industry closed my mind to that because
I learned that all the shortcuts we took wouldn't work in industry.
But still. A couple of guys made a compiler in a semester. Kinda
Unix like.
--
---
Larry McVoy lm at
mcvoy.com http://www.mcvoy.com/lm