On 5/19/2021 9:40 PM, Bob Smith wrote:
Larry died peacefully with his family nearby on Sunday, May 16, 2021.
I met Larry in 1971 upon joining STSC where he was Vice President for Systems. We hit it off right away as we shared a love for many things including word games and puzzles, and of course APL.
Before I joined STSC, Larry had achieved a significant milestone in 1972
by jointly writing one of the world's first worldwide email systems,
named Mailbox, or 666 BOX to its users where he was known as LMB.
The annual Dictionary Rally was started by three of my friends to which
Larry caught on immediately, eventually winning the competition along
with his first wife Donna, an unheard of twice, undoubtedly because of
his keen sense of detail and love of words. He started the meme of Lost Positives where we used his 13 volume Oxford English Dictionary to look
up “gruntled” where we found that “dis” is an intensifier, not a negater.
He shared with me many stories of the early days of APL design
discussions. Once, the group was deciding whether to keep the symbols
for the “and” and “or” functions since their result was duplicated by “min” and “max”. Larry settled the argument by noting that those two
sets of functions had different identity elements, and so separate
symbols for those Boolean functions remained, which, later on, pleased
me greatly.
I remember that when he learned that he shared the 1973 Grace Murray
Hopper award for his work implementing APL\360, we bundled into my car
and brought back bottles of cold champagne so we could celebrate
properly in the middle of the day.
After he went back to IBM, we kept in touch, often exchanging small
gifts, such as a beautiful nautilus which still sits on my book shelf,
along with a very early plot of a prime spiral.
At some time, I made it a point to visit Larry and Beverly annually,
just to keep in touch. In time, his health deteriorated, but I was able
to convince him to attend the Minnowbrook APL Implementors Workshop in 2017. He flew to DC and we drove to the workshop in upstate NY where he showed off more prime spiral plots as well as a collection of Dictionary Rally dictionaries and instructions over the years. The 450 miles of the
trip up and back with Larry melted away as we reminisced about the many
good times we shared.
Due to ill health, he, Jim Brown, and Roger Hui couldn’t make it to the 2019 Minnowbrook gathering, so we asked the attendees to say a few words
of encouragement to each of them which Jon McGrew adeptly videotaped and
sent off.
Throughout the time I knew Larry, he was my mentor. I looked up to him
so much that I followed him into becoming an APL implementor and
language designer, all of which has given me such great pleasure over
many years. I can fully appreciate how he so much enjoyed those roles.
To say I miss him greatly is an enormous understatement.
Please share your own thoughts of your experiences working
I met Larry in 1972 at STSC in Bethesda, MD after graduating high
school. He was one of a number of brilliant people there who inspired
me and helped me broaden myself both professionally and personally.
I remember there was a storage room containing office supplies and STSC informational material, including bios on the senior staff there. One
was of Larry, containing a picture of a totally beardless and
short-haired Larry, looking (as I remember it) very, very young. I wish
I had saved that picture.
The following year when I was at college I wrote Larry a letter, which I don’t have a copy of. Among other things, apparently I made a number of suggestions about compiling APL, which although I wouldn’t have had the
first idea of how to implement, I would certainly have appreciated.
Larry mailed me back a hand-typed letter, which I do still have the
original of. I think his response is interesting.
It begins “This is a genuine offline letter -- the system’s been down
for several hours because of power failure, and I'm using the powerful, flexible IBM 2741 in Local Concept Mode (or whatever IBM is calling it).
Note the low number of typos -- my specialty is ability to switch finger-reflexes from APL to Courier to keypunch on short notice.” “Your four ideas make sense (although deleting comment lines would probably
mess up programs written by people like [name deleted] who believe in
absolute line numbers) and could probably be incorporated in APL/360. I
suspect you haven't considered some of the problems with 'compilation',
though, whatever that means. (Back in the good old days I used to know,
but with APL I'm not so sure.) It sounds so straightforward to say ‘type
and rank must be known at compile time‘ -- but just what do we do with,
say, the argument and result, respectively, of FAPPEND and FREAD? Those functions must be able to handle any APL quantity! Phil Abrams and I are starting to consider seriously the problems of (ahem) efficient APL
execution; so far, it looks like whatever clever tricks we come up with
we'll need a full APL interpreter at run time, to fall back on when
compiled code fails us.”
Commenting on my major and extra-curricular activities, he approved and
said that they would “probably involve you in a far greater breadth of activities than would computer-bumming. I am not enthusiastic about
computer science undergraduate majors; it just seems there must be more significant things to be learning at that age. I know computers can be
lots of fun, but I also know that they're an unsurpassable temptation
for flunking German!”
Also apparently in my letter I reported that my first Fortran program
had compiled without errors and I was finding it all very interesting
(yes, I learned APL first, in high school, THEN learned Fortran and
assembly language. Boy did THAT make me appreciate APL even more.). I
love his response, particularly considering APLs longevity (today):
“Your comment on Fortran's acceptability is depressing … it merits
Twain's comparison to a dog walking on its hind legs. I earnestly hope
that your praise is motivated by, say, the understandable exhilaration
of getting your first program to actually run. No, this isn't my APL
chauvinism coming out -- it's more a belief that in a field developing
as rapidly as computer languages, a seventeen-year-old language just
can't be good. We've learned too much in the meantime.”
Goodbye, Larry. And thank you. Paul
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)