A consolidated message.

[chris.brown1@juno.com (Chris Brown)]

In the intrest of consolodating messages (I get mail & see the message
"You have
59 new messages", Aaarrrgghhhh!)

---Rich Thomson wrote...
> I just remembered another program I wrote at DELTA.  The directory
> listings would come out unsorted (like DOS' dir, unlike unix's ls).  I
> wrote a program which exploited a side-effect of how the internal
> directory structure changed when you renamed files and made a
> 'directory sorter' program that you could run and it would rename all
> your files in such a way that when you did another DIR, it listed them
> all out alphabetically.  Does anyone else remember this?

Yes, Rich.  I liked the program but was afraid to run it on my own
directory
out of fear of trashing it.  But nice work just the same...

---Tom Conte wrote...
> As for my favorite hacks, Steve Mentzer rewrote COMMUN and WHO using
> TECO macros.  He had discovered that you could make SYSCALLs from
> inside TECO.  So when you were editing something, you could load in
> his macros and respond to someone's COMMUN message or see who was
> logged on.

I too used SYSCALLs in TECO to get directory listing while in TECO.  But
this
brings up the point of writing code in TECO in the first place.  Was it 
masochism, egotism, the thrill of the bizarre, or what?  I have NEVER
since 
written in, or seen, code as cryptic, awkward, un documentable, &
generally
illegible as TECO code.  By comparison FORTH is beautiful.  One had to
have
an excellent memory to keep track of what you are writing as you code. 
Going 
back to the same code fragment a month later it would seem as if someone 
set your LA36 to the wrong baud rate.

---Rich Thompson wrote...
> Does anyone remember the name of that program on the Burroughs machine
> that would parse natural language queries and attempt to answer them
> from a database of facts?  (Chris, is this the QUEST program you
> mentioned in an earlier message?)  It was kind of rudimentary AI-like
> at the time.  I remember being quite impressed with the program
> ( although the half-duplex nature of the Burroughs machine itself
always
> struck me as REALLY odd).  I remember you could ask it where to get
> something to eat in Newark and it would tell you about Gino's
> (recently moved to Main Street at the time) and other places.
> (Speaking of which, I still remember eating gigantic turkey subs from
> Leonardo's deli many an evening in Willard Hall!  The closest we get
> to subs out here in Salt Lake City is "subway" and no cheese steaks
> although we can finally get Tastykakes by mail order now.)

> At one point I got to be friends with the guy in the CC who wrote the
> program (or perhaps he was just maintaining it at the time I met him).
> It was written in Algol and he was extolling the virtues of Algol to me
> (I had barely been introduced to compiled languages at the time).  Does
> anyone happen to remember this guy's name?  He had a DELTA attitude
> (i.e. irreverance of authority :), but he was never involved in DELTA
> per se.  There seems to have been a much smaller crowd of people who
> got involved through the Center, although it seems that the Center
> didn't encourage such "extracurricular" computing.

Yes Rich, that is the program, Quoting from it...
" QUEST is a computer program which can answer over one
thousand questions.  These questions are divided up into
subject areas - the FORTRAN subject contains FORTRAN
questions, the CNADE subject contains CANDE questions, etc."
It used free-form user questions from which it tried to find matching
or similar questions it already knew the answers to.  I've got the ALGOL 
source code as well as the RSTS/E ALGOL compiler source code.
If you hadn't asked the question of who wrote it, I'd have been able to 
tell you.  Can you believe he did not place his name in the source code?
He DID like Algol, I did too, I wrote a few things in it on the B7700 and
the
DEC-10.  My understanding is that Wirth wrote Pascal as a subset of 
Algol to teach algorithmic programming without the computing resources
Algol required.
Ah, Leonardo's!  I remember going there late at night for a steak
sandwich &
the cook (owner?) knew exacly how I wanted my sandwich (fried onions &
bell peppers, melted swiss cheese, salt & pepper, heavy mayonaise.  You
are right, we just don't seem to have good steak sandwich places here in
Salt Lake City.  I've actually given serious thought to getting out of
the
technology industry & opening a diner myself.

---In  Alan D. Flippen's "Unofficial History of Project DELTA" he
writes...
> Members have been more well-behaved than the staff.  'Incidents' were 
> fairly common, ranging from disk corruptions (as Rich Thomson and Chris

> Brown discovered one night when they said PIP 
> [1,38]/MO:16384.=MESSAG.TXT and corrupted the UFD) to utility wars, in 

Actually, Rich had, while non-priv, written a program that would copy his
LOGMES
mail to his account on the backup hard drive.  He did this with code that
opened 
the backup drive something like OPEN "DK01:" AS FILE 1%.  A non-priv
person
doing this would default to a file like [20,20]NONAME.  Of course Ed's
coding
standards would have had rich specifying the filename properly.  Well,
when Rich
first ran his mail backup program when he became priv, it opened THE
WHOLE 
DRIVE as file 1%, it then blithely copied his mail to the MFD of the
backup disk.  
That disk was where all the 'for profit' source code was kept.  And, as
chance
always happens in such cases, there were no tape backups of that drive. 
But
fortunately Ed, Bob, & others with help from DEC documentation were able
to 
search the drive to find the Radix-50 representation of 'MFD' and 'UFD'
at the
end of every block that used to have directory information in them & hand
build
a new UFD gain access to the drive.  I mention all this for several
reasons:
1)  to clear my name in the incident (though I was present), 2) note the 
advantages to coding standards, 3) highlight Deltoid's (Ed, Bob, & etc) 
abilities.

---Rich Thompson write...
> Ralph Gonzalez' prolific output on the Tektronix 4014 graphics terminal
> inspired me to a career in computer graphics.  I would never have
> gotten where I am today without that inspiration, Ralph!

I agree, I was always impressed by Ralph's graphics work, but not
necessarily his spagetti code ;-)  I have often wondered in these
intervening
years what studio he'd end up at (Industrial Light & Magic and the
others) or
perhaps there at Evans & Southerland.  Bob, you should solicit personal
histories after DELTA to place on the web site  (perhaps requiring a
login
to ease privacy concerns).


Chris Brown

_____________________________________________________________________
You don't need to buy Internet access to use free Internet e-mail.
Get completely free e-mail from Juno at http://www.juno.com
Or call Juno at (800) 654-JUNO [654-5866]