(defun moment_of_silence (print "*"))
October 24, 2011 1:21 PM Subscribe
John McCarthy, AI pioneer, ACM Turning Award winner, National Medal of Science winner, Professor Emeritus of Computer Science at Stanford University and inventor of the Lisp programming language passed away suddenly last night at the age of 84.
(print ".")
posted by Foci for Analysis at 1:23 PM on October 24, 2011 [2 favorites]
posted by Foci for Analysis at 1:23 PM on October 24, 2011 [2 favorites]
Shit, they're all going. Has anyone checked on Knuth?
posted by Leon at 1:23 PM on October 24, 2011 [7 favorites]
posted by Leon at 1:23 PM on October 24, 2011 [7 favorites]
If you people keep unbalancing the parentheses like that, his wraith is going to come back and destroy the place.
posted by Wolfdog at 1:28 PM on October 24, 2011 [18 favorites]
posted by Wolfdog at 1:28 PM on October 24, 2011 [18 favorites]
Coding in Lisp was a heck of a lot of fun in school.
posted by ZeusHumms at 1:32 PM on October 24, 2011
posted by ZeusHumms at 1:32 PM on October 24, 2011
(list '1 '2 'boo)
posted by Bunny Ultramod at 1:33 PM on October 24, 2011
posted by Bunny Ultramod at 1:33 PM on October 24, 2011
(write-line ".")
posted by kmz at 1:33 PM on October 24, 2011 [1 favorite]
posted by kmz at 1:33 PM on October 24, 2011 [1 favorite]
Where's the confirmation that he's passed away? The wikipedia page lists nothing.
posted by ZeusHumms at 1:35 PM on October 24, 2011
posted by ZeusHumms at 1:35 PM on October 24, 2011
Ritchie and McCarthy in the span of a couple weeks. Damn.
posted by chunking express at 1:35 PM on October 24, 2011
posted by chunking express at 1:35 PM on October 24, 2011
Also, are there any links indicating he is dead besides that wikipedia page?
posted by chunking express at 1:38 PM on October 24, 2011
posted by chunking express at 1:38 PM on October 24, 2011
Turing Award
posted by pracowity at 1:40 PM on October 24, 2011 [1 favorite]
posted by pracowity at 1:40 PM on October 24, 2011 [1 favorite]
ZeusHumms, this post on Laughing Squid has a couple of references from moderately reliable sources. It appears, like rms, the family so far has just told some friends/maybe some early journos who haven't published articles yet.
posted by skynxnex at 1:40 PM on October 24, 2011
> (CDR '(MCCARTHY))
NILposted by skynxnex at 1:40 PM on October 24, 2011
This isnt much of a source, it has been pulled from wikipedia. Strange month in tech if he has passed, three great men who have contributed in their own ways have passed.
posted by adventureloop at 1:40 PM on October 24, 2011
posted by adventureloop at 1:40 PM on October 24, 2011
A twitter from a journalist who has done articles about him in the past.
posted by zabuni at 1:41 PM on October 24, 2011
posted by zabuni at 1:41 PM on October 24, 2011
just read his story The Robot and the Baby. Pretty good.
posted by leibniz at 1:49 PM on October 24, 2011
posted by leibniz at 1:49 PM on October 24, 2011
There were a couple posts about this on the programming reddit, and the sources seem to be very sketchy and circular. Comes down to one tweet. It'd be great if it turned out not to be true.... cause wow. Ritchie and McCarthy? :(
posted by edheil at 1:51 PM on October 24, 2011
posted by edheil at 1:51 PM on October 24, 2011
So, who invented BASIC and ... how are they doing? Because that would be the trifecta.
posted by seanmpuckett at 1:55 PM on October 24, 2011
posted by seanmpuckett at 1:55 PM on October 24, 2011
The book Hackers by Steven Levy is an amazing history of the early computer movement/industry that covers McCarthy and others. It's a great and entertaining read.
posted by formless at 2:03 PM on October 24, 2011 [3 favorites]
posted by formless at 2:03 PM on October 24, 2011 [3 favorites]
skynxnex: "It appears, like rms,"
dmr unless there's been a new development!
posted by mkb at 2:05 PM on October 24, 2011 [4 favorites]
dmr unless there's been a new development!
posted by mkb at 2:05 PM on October 24, 2011 [4 favorites]
.
posted by shakespeherian at 2:05 PM on October 24, 2011
posted by shakespeherian at 2:05 PM on October 24, 2011
.
posted by qxntpqbbbqxl at 2:07 PM on October 24, 2011
posted by qxntpqbbbqxl at 2:07 PM on October 24, 2011
dmr unless there's been a new development!
Darn. Yes, dmr, not rms. I saw someone mention rms related to this, so I conflated when I came over here.
posted by skynxnex at 2:10 PM on October 24, 2011
Darn. Yes, dmr, not rms. I saw someone mention rms related to this, so I conflated when I came over here.
posted by skynxnex at 2:10 PM on October 24, 2011
Who is Killing the Great Cooks Programmers of Europe the US?
posted by IAmBroom at 2:12 PM on October 24, 2011
posted by IAmBroom at 2:12 PM on October 24, 2011
One tweet and a Portuguese news story? I'm skeptical.
posted by CheeseDigestsAll at 2:15 PM on October 24, 2011
posted by CheeseDigestsAll at 2:15 PM on October 24, 2011
.
Steven Levy a hero in his own right. Seems to have confirmed his death.
posted by Ad hominem at 2:27 PM on October 24, 2011 [1 favorite]
Steven Levy a hero in his own right. Seems to have confirmed his death.
posted by Ad hominem at 2:27 PM on October 24, 2011 [1 favorite]
(quit)
posted by wachhundfisch at 2:33 PM on October 24, 2011
posted by wachhundfisch at 2:33 PM on October 24, 2011
My introduction to coding was in the derivative of LISP that was AutoCAD's original scripting language called AutoLISP. It was a very beautiful language in its simplicity, elegance, consistency and symmetry. Moving on to Basic, and even C never gave me the sense of fulfillment that could be achieved crafting an efficient LISP routine.
Saying "cudder" never sounded very elegant however.
posted by Mei's lost sandal at 2:44 PM on October 24, 2011
Saying "cudder" never sounded very elegant however.
posted by Mei's lost sandal at 2:44 PM on October 24, 2011
So, this story isn't exactly LISP related, but my freshman year data structures and algorithms course was taught in Scheme, which obviously owes a lot to LISP and is a close sibling. We were encouraged to work in pairs on some of the program sets, and a friend of mine that lived down the hall was my go-to partner in these cases.
The final program set of the year was a pair set, and my friend and I spent a whole night working on the solution. Throughout the night -- driven by caffeine, lack of sleep, and the brain bending that recursive programming demands -- my friend took small post-it notes that were sitting on my desk and wrote "I hate Scheme" on every one of them and stuck them up around the room.
Then my roommate came back from wherever he was that night and was... confused. We're all still friends now, though, so I guess it turned out ok. Also, I came to appreciate Scheme and LISP for their minimalist recursive beauty.
So, thanks McCarthy for the beauty of a recursive function reaching its termination and gracefully falling/collapsing back to the calling function, correct answer cradled in its little perens.
posted by Inkoate at 2:53 PM on October 24, 2011
The final program set of the year was a pair set, and my friend and I spent a whole night working on the solution. Throughout the night -- driven by caffeine, lack of sleep, and the brain bending that recursive programming demands -- my friend took small post-it notes that were sitting on my desk and wrote "I hate Scheme" on every one of them and stuck them up around the room.
Then my roommate came back from wherever he was that night and was... confused. We're all still friends now, though, so I guess it turned out ok. Also, I came to appreciate Scheme and LISP for their minimalist recursive beauty.
So, thanks McCarthy for the beauty of a recursive function reaching its termination and gracefully falling/collapsing back to the calling function, correct answer cradled in its little perens.
posted by Inkoate at 2:53 PM on October 24, 2011
.
posted by paulus andronicus at 2:53 PM on October 24, 2011
posted by paulus andronicus at 2:53 PM on October 24, 2011
.
I spent many years coding in LISP for a living and still miss the beauty and elegance it affords when I'm forced to work in other languages. Sigh.
posted by Runes at 3:07 PM on October 24, 2011
I spent many years coding in LISP for a living and still miss the beauty and elegance it affords when I'm forced to work in other languages. Sigh.
posted by Runes at 3:07 PM on October 24, 2011
.
posted by vogon_poet at 3:15 PM on October 24, 2011
posted by vogon_poet at 3:15 PM on October 24, 2011
This Hacker News link confirms via Paul Graham.
posted by Bourbonesque at 3:15 PM on October 24, 2011
posted by Bourbonesque at 3:15 PM on October 24, 2011
(mccarthy (mccarthy))
posted by Matt Oneiros at 3:24 PM on October 24, 2011
posted by Matt Oneiros at 3:24 PM on October 24, 2011
)
posted by GuyOnABuffalo at 4:30 PM on October 24, 2011
posted by GuyOnABuffalo at 4:30 PM on October 24, 2011
Oh jeepers. I've been getting more and more into Lisp and Scheme in the last year and it keeps blowing my mind. So sad to see him go.
posted by DU at 5:46 PM on October 24, 2011
posted by DU at 5:46 PM on October 24, 2011
...taught in Scheme, which obviously owes a lot to LISP and is a close sibling. We were encouraged to work in pairs...
(see-it (what (did (you there))))
posted by DU at 5:50 PM on October 24, 2011 [2 favorites]
(see-it (what (did (you there))))
posted by DU at 5:50 PM on October 24, 2011 [2 favorites]
.
posted by LobsterMitten at 6:28 PM on October 24, 2011
posted by LobsterMitten at 6:28 PM on October 24, 2011
λ
or perhaps
#<DOT>
I always liked his grumpy rants about how with enough technology there could be enough of everything for everyone forever. We are losing the CS greats one by one. I wonder when we will start looking back keeping track of computer history in a formal and good way, as opposed to the ad-hoc way we do now.
posted by pmb at 6:29 PM on October 24, 2011
or perhaps
#<DOT>
I always liked his grumpy rants about how with enough technology there could be enough of everything for everyone forever. We are losing the CS greats one by one. I wonder when we will start looking back keeping track of computer history in a formal and good way, as opposed to the ad-hoc way we do now.
posted by pmb at 6:29 PM on October 24, 2011
]
posted by Obscure Reference at 6:43 PM on October 24, 2011
posted by Obscure Reference at 6:43 PM on October 24, 2011
I had been meaning to post some of his crazy ass essays at some point; now is as good a time as any. There is some really interesting and maybe controversial stuff on his website--it is worth exploring.
I've never coded in Lisp, unless you count the time in college when I implemented a matrix multiplication program in Scheme for a class. But I've been coding in Ruby for the last year, and because some claim that Ruby is an acceptable Lisp, I read On Lisp last winter, and it helped me to understand why Lisp macros are so cool and useful. I believe it was a worthwhile read. I am lucky enough to count Guy Steele as an acquaintance. (He is alive and well the last I knew.) A while back I mentioned to him that I was coding in Ruby and how much I liked it, and his response was something like "Ah yes, as computing power increases, languages increasingly come to resemble Lisp."
.
posted by A dead Quaker at 6:50 PM on October 24, 2011
I've never coded in Lisp, unless you count the time in college when I implemented a matrix multiplication program in Scheme for a class. But I've been coding in Ruby for the last year, and because some claim that Ruby is an acceptable Lisp, I read On Lisp last winter, and it helped me to understand why Lisp macros are so cool and useful. I believe it was a worthwhile read. I am lucky enough to count Guy Steele as an acquaintance. (He is alive and well the last I knew.) A while back I mentioned to him that I was coding in Ruby and how much I liked it, and his response was something like "Ah yes, as computing power increases, languages increasingly come to resemble Lisp."
.
posted by A dead Quaker at 6:50 PM on October 24, 2011
I was forwarded this by another old Lisper:
Here lies a Lisper
Uninterned from this mortal package
Yet not gc'd
While we retain pointers to his memory
posted by Runes at 6:50 PM on October 24, 2011 [6 favorites]
Here lies a Lisper
Uninterned from this mortal package
Yet not gc'd
While we retain pointers to his memory
posted by Runes at 6:50 PM on October 24, 2011 [6 favorites]
A story I heard from my programming languages teacher, which I'm sure I'm garbling: McCarthy was traveling in the Eastern Bloc, and had with him a reasonably-sized collection of his notes on CS-related matters. He was stopped at a border checkpoint, and one of the guards came across his notes and worried that he was smuggling some kind of confidential information. To prove that they were his, McCarthy started writing out the LISP code for a basic meta-circular interpreter, which of course he could do cold because he was John McCarthy.
Learning to code in LISP was one of the most mind-expanding things I've ever done. I still get a little misty-eyed and wobbly-kneed thinking about just how awesome a language it is.
)
posted by grimmelm at 7:15 PM on October 24, 2011 [1 favorite]
Learning to code in LISP was one of the most mind-expanding things I've ever done. I still get a little misty-eyed and wobbly-kneed thinking about just how awesome a language it is.
)
posted by grimmelm at 7:15 PM on October 24, 2011 [1 favorite]
"Ah yes, as computing power increases, languages increasingly come to resemble Lisp."
I can't find it now, but a person of my acquaintance liked a quote similar to this: "All programming languages grow until they contain a half-assed implementation of Lisp".
posted by DU at 7:16 PM on October 24, 2011 [2 favorites]
I can't find it now, but a person of my acquaintance liked a quote similar to this: "All programming languages grow until they contain a half-assed implementation of Lisp".
posted by DU at 7:16 PM on October 24, 2011 [2 favorites]
I can't find it now, but a person of my acquaintance liked a quote similar to this: "All programming languages grow until they contain a half-assed implementation of Lisp".
Ah, right. Greenspun's tenth rule:
Any sufficiently complicated C or Fortran program contains an ad hoc, informally-specified, bug-ridden, slow implementation of half of Common Lisp.
This kind of boils down to “any sufficiently complicated computer program will need to implement a domain-specific language”, with the assumption that Lisp is the best way to do so because it has such a flexible grammar.
posted by spitefulcrow at 7:30 PM on October 24, 2011
Ah, right. Greenspun's tenth rule:
Any sufficiently complicated C or Fortran program contains an ad hoc, informally-specified, bug-ridden, slow implementation of half of Common Lisp.
This kind of boils down to “any sufficiently complicated computer program will need to implement a domain-specific language”, with the assumption that Lisp is the best way to do so because it has such a flexible grammar.
posted by spitefulcrow at 7:30 PM on October 24, 2011
The impressive part about LISP isn't that it exists. It's so simple and clean that something very similar would have been invented eventually.
The impressive part is that it was one of the very first programming languages, along with COBOL and FORTRAN. There was nothing at all to learn from and he got almost everything right in one go. It's kind of like if some guy took a careful look at the first Model T, worked for a couple of months, and came out with a 2011 Honda Civic.
Utterly unbelievable. And yet that's how it happened.
posted by miyabo at 7:42 PM on October 24, 2011 [6 favorites]
The impressive part is that it was one of the very first programming languages, along with COBOL and FORTRAN. There was nothing at all to learn from and he got almost everything right in one go. It's kind of like if some guy took a careful look at the first Model T, worked for a couple of months, and came out with a 2011 Honda Civic.
Utterly unbelievable. And yet that's how it happened.
posted by miyabo at 7:42 PM on October 24, 2011 [6 favorites]
The software I used for my master's thesis was written in Lisp.
.
posted by WalkingAround at 10:20 PM on October 24, 2011
.
posted by WalkingAround at 10:20 PM on October 24, 2011
A DSL?
Try garbage collection, catchable exceptions, high level control flow, a usable macro system (providing the ability to redefine syntax on the fly), reflection, hygeinic macros, the numeric tower, foriegn function interface, coroutines, grey streams, continuations, first class functions, on the fly compilation, stepwise debugging, editor/language integration, a built in documentation system, a REPL, a full featured object system, generic functions and the meta object protocol for starters. I am sure I missed a few things. Lisp had them before most of the other languages designers even knew what they were. And many (most?) of those features were invented in Lisp first. Not that every (any?) one Lisp version has all of these features. McCarthy did not invent or implement most of these ideas, but it is a testimony to his language that all of these things could be bootstrapped into his language from the small featureset he started with.
And the best versionof the rule I heard is: any sufficiently complex progam will incorporate an ad-hoc and poorly designed substitute for half of Common Lisp. Even your favorite Common Lisp implementation.
posted by idiopath at 3:59 AM on October 25, 2011
Try garbage collection, catchable exceptions, high level control flow, a usable macro system (providing the ability to redefine syntax on the fly), reflection, hygeinic macros, the numeric tower, foriegn function interface, coroutines, grey streams, continuations, first class functions, on the fly compilation, stepwise debugging, editor/language integration, a built in documentation system, a REPL, a full featured object system, generic functions and the meta object protocol for starters. I am sure I missed a few things. Lisp had them before most of the other languages designers even knew what they were. And many (most?) of those features were invented in Lisp first. Not that every (any?) one Lisp version has all of these features. McCarthy did not invent or implement most of these ideas, but it is a testimony to his language that all of these things could be bootstrapped into his language from the small featureset he started with.
And the best versionof the rule I heard is: any sufficiently complex progam will incorporate an ad-hoc and poorly designed substitute for half of Common Lisp. Even your favorite Common Lisp implementation.
posted by idiopath at 3:59 AM on October 25, 2011
It's kind of like if some guy took a careful look at the first Model T, worked for a couple of months, and came out with a 2011 Honda Civic.
And accidentally at that. I read somewhere recently that they originally intended to convert the s-expressions into something more traditional, but everyone kind of liked them and they never got around to it.
The powerful thing about Lisp is that it takes computers seriously. The whole power of a computer comes from the fact that instructions are just a kind of data. You can load a program. We're so familiar with that, but it's really remarkable. The instructions are separate from the machine. That's not true of mechanical machines (or even electronic ones without computer chips).
Lisp treats code as data. Lisp is basically another layer of computers on top of computers, while FORTRAN is a mechanical machine simulated on a computer.
posted by DU at 5:08 AM on October 25, 2011 [3 favorites]
And accidentally at that. I read somewhere recently that they originally intended to convert the s-expressions into something more traditional, but everyone kind of liked them and they never got around to it.
The powerful thing about Lisp is that it takes computers seriously. The whole power of a computer comes from the fact that instructions are just a kind of data. You can load a program. We're so familiar with that, but it's really remarkable. The instructions are separate from the machine. That's not true of mechanical machines (or even electronic ones without computer chips).
Lisp treats code as data. Lisp is basically another layer of computers on top of computers, while FORTRAN is a mechanical machine simulated on a computer.
posted by DU at 5:08 AM on October 25, 2011 [3 favorites]
Prof. McCarthy was a regular long-time participant in the usenet group rec.arts.books. (The conversation there ranged well beyond books and book reviews; the group's unofficial but universally acknowleged charter was "the place where bookish people talk about whatever they like.") His comments were always clever and insightful and often deep. Goodbye John, it was a pleasure to have known you, even if only online.
posted by jfuller at 7:09 AM on October 25, 2011 [1 favorite]
posted by jfuller at 7:09 AM on October 25, 2011 [1 favorite]
His list of accomplishments is awesome.
Dead Quaker is dead on about the crazy assed essays. I am not a professional computer programmer and I do not understand lisp yet. My first introduction to the dude was I was surfing the web and I stumbled on his debunking of the Chief Seattle speech.
The global warming page is interesting in that the last add-on there is him deciding he didn't know enough about the subject to have a reliable opinion after all. Full on agnostic, which is a sentiment I have no problem with. The most capable earth scientist I know has (or maybe had--he may have taken it down by now) a similar statement on his web site. It says he went to the library to try and find out more and he spent 12 hours digging through books and journals and he concluded he wasn't qualified to opine on the topic.
posted by bukvich at 9:48 AM on October 25, 2011
Dead Quaker is dead on about the crazy assed essays. I am not a professional computer programmer and I do not understand lisp yet. My first introduction to the dude was I was surfing the web and I stumbled on his debunking of the Chief Seattle speech.
The global warming page is interesting in that the last add-on there is him deciding he didn't know enough about the subject to have a reliable opinion after all. Full on agnostic, which is a sentiment I have no problem with. The most capable earth scientist I know has (or maybe had--he may have taken it down by now) a similar statement on his web site. It says he went to the library to try and find out more and he spent 12 hours digging through books and journals and he concluded he wasn't qualified to opine on the topic.
posted by bukvich at 9:48 AM on October 25, 2011
)
posted by supercoiled at 9:57 AM on October 25, 2011
posted by supercoiled at 9:57 AM on October 25, 2011
He will be mithed.
posted by ZenMasterThis at 10:26 AM on October 25, 2011 [1 favorite]
posted by ZenMasterThis at 10:26 AM on October 25, 2011 [1 favorite]
I have never been terribly fond of Lisp itself, but functional programming seriously rocks. Anyone who knows some programming should learn Haskell, Erlang, or OCaml, which all owe an incredible debt to Lisp.
There are vast swaths of stupidity found in object oriented languages like Java and C++ aimed towards emulating the most rudimentary functional programming tools. As Yaron Minsky notes 26 mins into his Caml Trading talk, object oriented languages impose an incredible amount of obfuscation upon simple abstractions through their layers of dynamic dispatch. All these problems simply evaporate in functional languages.
We would live in a better world if C++'s major features had been introduced in reverse order, meaning lambdas back in 1983, templates a little later, and classes only much later.
posted by jeffburdges at 6:00 PM on October 25, 2011 [2 favorites]
There are vast swaths of stupidity found in object oriented languages like Java and C++ aimed towards emulating the most rudimentary functional programming tools. As Yaron Minsky notes 26 mins into his Caml Trading talk, object oriented languages impose an incredible amount of obfuscation upon simple abstractions through their layers of dynamic dispatch. All these problems simply evaporate in functional languages.
We would live in a better world if C++'s major features had been introduced in reverse order, meaning lambdas back in 1983, templates a little later, and classes only much later.
posted by jeffburdges at 6:00 PM on October 25, 2011 [2 favorites]
JWZ tells how he once debugged McCarthy's dot-emacs file.
)
posted by Busy Old Fool at 3:52 AM on October 26, 2011
)
posted by Busy Old Fool at 3:52 AM on October 26, 2011
Just fyi, Alex McLean performs music live by live coding in Haskell, the world's sexiest functional language.
posted by jeffburdges at 4:44 PM on October 26, 2011
posted by jeffburdges at 4:44 PM on October 26, 2011
« Older Shari'a Law via democracy | Boom Bip: from abstract hip-hop to spooky sounds... Newer »
This thread has been archived and is closed to new comments
posted by outlaw of averages at 1:22 PM on October 24, 2011 [10 favorites]