6 ms·

The Art of Computer Programming (interview with Don Knuth)

140 points by tjr 15y ago

Jun8 15y ago
Even in a casual interview Knuth feeds you enough nuggets to make you think for some time, e.g. the Jevons Paradox he mentions at the end (http://en.wikipedia.org/wiki/Jevons_paradox http://en.wikipedia.org/wiki/Jevons_paradox). His faculties and rate of work does not seem to be affected by age (he's 73!), he's the Euler of CS (although I hope he lives longer than him).
andos 15y ago
Now, to be put into that category of one of the best books of the century, that’s a little bit embarrassing as they rank me with Einstein and Feynman - I’m not in that league really, I just didn’t have as much competition - they had to have a token person in computer science! There might be a little truth here but, nonetheless, his devotion to those books made him a legend of our field. And to think people—well, I!—brag so much about so much less. His humility is disconcerting.
- wglb 15y ago
  I think this is simply extreme modesty. Many researchers in the field often think they have discovered something new. Digging a little further, they discover that Knuth has been there already.
- horacio 15y ago
  Andos, I don't feel his humility (expressed here and elsewhere) is at all disconcerting. I feel it's both endearing and appropriate. Each of us who have accomplished much less, but have boasted of much more, need only keep Knuth's example in mind as a guide for our future lives.
pingswept 15y ago
The American Scientist list mentioned in the opening: http://www.americanscientist.org/bookshelf/pub/100-or-so-books-that-shaped-a-century-of-science http://www.americanscientist.org/bookshelf/pub/100-or-so-boo...
bdhe 15y ago
The CS department at Stanford has a weekly Theory Lunch during the academic year and we're fortunate enough to get Don Knuth to speak at least once every quarter[1]. His talks, although they start off from basically very little quickly become intensely technical and mathematical. It is quite astounding to see a man of his age have the kind of passion he demonstrates. Even more so the ability to translate interesting and quirky insights into deep mathematical and algorithmic statements (which is very apparent to those who go to his talks). [1] http://theory.stanford.edu/theory_lunch/ http://theory.stanford.edu/theory_lunch/
udzinari 15y ago
There is also a very long and interesting interview with him at http://www.webofstories.com/play/17060 http://www.webofstories.com/play/17060
huherto 15y ago
One of my favorite parts... Looking forwards, what do you perceive are the biggest issues / challenges computer science faces, now and in the near future? The challenge of how can we go to sleep at night when there are so many things left to be done! (....) but I think so much more could be done by having better health care records and better ways of describing and visualising combinations of symptoms and combining statistical methods as well as visual methods, this would help doctors to understand things more clearly and quickly. Not to mention the ways in which computers can help biologists to design better drugs. Everywhere you look there’s something to be done which needs a good programmer to help achieve it. There’s no shortage of challenges and little chance of ever running out of them. It gets me a little personal validation that we still have many exiting challenges ahead.
- cheez 15y ago
  > It gets me a little personal validation that we still have many exiting challenges ahead. I read that one of the US Patent Office people quit at the turn of the 20th century because he was convinced there was nothing left to do. If I have learned anything, it's that for every problem humans solve, a few more pop up that we never thought would exist. It makes me happy.
  - jleader 15y ago
    It appears that that story about the director of the Patent Office saying there's nothing more to invent is mostly made up, see http://en.wikipedia.org/wiki/Charles_H._Duell http://en.wikipedia.org/wiki/Charles_H._Duell or http://www.quotationspage.com/quote/22779.html http://www.quotationspage.com/quote/22779.html or the most detailed debunking: http://www.myoutbox.net/posass.htm http://www.myoutbox.net/posass.htm
yarapavan 15y ago
I liked his answer for the question - why don’t you do email? Someone has to not be tweeting all the time, someone has to be thinking about things which need a long attention span and trying to organise material and build up strong foundations instead of rushing off across the frontier. It takes a long time to put out something that has the right style; I have to really think about it and if I’m going to do it right I have to spend a lot of time focussed on it. And I was being treated like an oracle, lots of people from around the world were asking my opinion about whatever, so after 15 years of email I decided that was a life time’s worth.
- a3_nm 15y ago
  Actually, you can still email Knuth, and his secretary will print the email out. See http://www-cs-faculty.stanford.edu/~uno/email.html http://www-cs-faculty.stanford.edu/~uno/email.html I can understand why he wants to be on the bottom of things, but I don't get why encouraging people to use postal mail rather than email is supposed to help. Email is just a medium, and it's a much more sensible way to move information around. The widespread habit of checking email every 5 seconds is by no means compulsory. I wonder why Knuth says that the batch mode communication he wants to do cannot be done properly with email.
  - jleader 15y ago
    I think it's just that he's trying to raise the barrier to entry a little higher. If you know you'll get a response in a few days or weeks on paper, instead of a day or two later in your email client, that makes you look for other, quicker ways to get the answer you're looking for, which reduces the load on Knuth (and his secretary).
dxbydt 15y ago
Coincidentally, I was thinking of what Knuth said just last week, when one of my comments was downvoted into obscurity for promoting a supposedly obscure/contrived math formula over a lame switch-case :) http://news.ycombinator.com/item?id=2817716 http://news.ycombinator.com/item?id=2817716 In reality, that formula is neither obscure nor contrived. I actually learnt it in high school ie. it was formally taught and we worked through how to come up with such functions given a set of inputs and outputs ie. essentially a bounded set mapping problem { 1->31, 3->31, 4->30,5->31, 6->30,7->31....} So anyways, Knuth gave an interview maybe 10 years back where he lamented the lack of dexterity in math in American undergrads. He said even Stanford has to offer "remedial courses in math" for CS undergrads, because there is an "interest vs ability" mismatch. At that time ( during the 1999-2000 tech boom ), a lot of Stanford undergrads were indicating CS as their major. So they clearly had the "interest". But upon testing them, it was clear they didn't have the "ability". Hence this interest vs ability mismatch, and a need for remedial math courses at Stanford. At the time, I was studying CS with a bunch of Russian, European, Indian & American classmates. I found that Russians & Eastern Europeans tend to favor clever 1-line math code. They themselves were trained in physics and math & looked at CS problems as essentially a laborious function-mapping exercise. Whereas the Americans invariably wanted "readable code" and ended up writing verbose switch-cases and if-elif-endif clauses ie. essentially unrolling for-loops. Knuth's Concrete Math ( http://www-cs-faculty.stanford.edu/~uno/gkp.html http://www-cs-faculty.stanford.edu/~uno/gkp.html ) was a sort of bible in my days. You had to know generating functions and recurrences and even stuff like complex powers. Like for instance i^i ( i raised to i ) leads to an infinite set of points with real and imaginary parts, scattered all over the 2D plane. You had to know cosets and Lagrange's theorem and set partitioning using cosets. Nowadays people simply enumerate everything as there is no penalty ( in performance terms ), so CS problems have a "select 1 answer from set of possible answers" flavor, rather than "contrive a function to compute the answer" flavor.
- kenjackson 15y ago
  I just upvoted your old post. :-) I did disagree with the solution (many who see the code would get confused), but it certainly was a positive contribution to the discussion, which is always my bar for upvote/downvote. And I love Concrete Mathematics. I wish people would say Knuth is the author of TAOCP and Concrete Mathematics. And Ronald Graham is another great computer scientist. With that said I do distinguish computer science from programming.
kragen 15y ago
There's a much longer interview with Knuth posted originally at http://www.webofstories.com/play/17060 http://www.webofstories.com/play/17060. The transcript is probably interesting reading for anyone interested in the man; I imported it into Git and started formatting it into paragraphs at https://github.com/kragen/knuth-interview-2006 https://github.com/kragen/knuth-interview-2006, although I haven't finished. I think it would be worthwhile for more people to (a) clone the git repository to reduce the risk that this primary historical source will be lost, and (b) reformat more of the transcript if they feel like it.
sebkomianos 15y ago
"I’m jealous of astronomers, for example, because people respect astronomers for doing astronomy because it’s interesting just for its own sake. I studied computer science because it’s interesting to study computer science." Probably my favorite part of the interview.
adharmad 15y ago
"I was born to be a computer scientist - I have a way of organising stuff in my head that just seems to make me a good programmer. I think everybody can learn to use computers, but only about one person in every 50 is a geek in the same way as I am. That means we can push the envelope and can resonate with the computer. The way we think helps to make it easier for us to know how to construct a machine." One of the few people who can state it as a fact, without sounding arrogant.

140 points by tjr 15y ago

Jun8 15y ago

Even in a casual interview Knuth feeds you enough nuggets to make you think for some time, e.g. the Jevons Paradox he mentions at the end (http://en.wikipedia.org/wiki/Jevons_paradox http://en.wikipedia.org/wiki/Jevons_paradox). His faculties and rate of work does not seem to be affected by age (he's 73!), he's the Euler of CS (although I hope he lives longer than him).

andos 15y ago

Now, to be put into that category of one of the best books of the century, that’s a little bit embarrassing as they rank me with Einstein and Feynman - I’m not in that league really, I just didn’t have as much competition - they had to have a token person in computer science! There might be a little truth here but, nonetheless, his devotion to those books made him a legend of our field. And to think people—well, I!—brag so much about so much less. His humility is disconcerting.

wglb 15y ago
I think this is simply extreme modesty. Many researchers in the field often think they have discovered something new. Digging a little further, they discover that Knuth has been there already.
horacio 15y ago
Andos, I don't feel his humility (expressed here and elsewhere) is at all disconcerting. I feel it's both endearing and appropriate. Each of us who have accomplished much less, but have boasted of much more, need only keep Knuth's example in mind as a guide for our future lives.

pingswept 15y ago

The American Scientist list mentioned in the opening: http://www.americanscientist.org/bookshelf/pub/100-or-so-books-that-shaped-a-century-of-science http://www.americanscientist.org/bookshelf/pub/100-or-so-boo...

bdhe 15y ago

The CS department at Stanford has a weekly Theory Lunch during the academic year and we're fortunate enough to get Don Knuth to speak at least once every quarter[1]. His talks, although they start off from basically very little quickly become intensely technical and mathematical. It is quite astounding to see a man of his age have the kind of passion he demonstrates. Even more so the ability to translate interesting and quirky insights into deep mathematical and algorithmic statements (which is very apparent to those who go to his talks). [1] http://theory.stanford.edu/theory_lunch/ http://theory.stanford.edu/theory_lunch/

udzinari 15y ago

There is also a very long and interesting interview with him at http://www.webofstories.com/play/17060 http://www.webofstories.com/play/17060

huherto 15y ago

One of my favorite parts... Looking forwards, what do you perceive are the biggest issues / challenges computer science faces, now and in the near future? The challenge of how can we go to sleep at night when there are so many things left to be done! (....) but I think so much more could be done by having better health care records and better ways of describing and visualising combinations of symptoms and combining statistical methods as well as visual methods, this would help doctors to understand things more clearly and quickly. Not to mention the ways in which computers can help biologists to design better drugs. Everywhere you look there’s something to be done which needs a good programmer to help achieve it. There’s no shortage of challenges and little chance of ever running out of them. It gets me a little personal validation that we still have many exiting challenges ahead.

cheez 15y ago
> It gets me a little personal validation that we still have many exiting challenges ahead. I read that one of the US Patent Office people quit at the turn of the 20th century because he was convinced there was nothing left to do. If I have learned anything, it's that for every problem humans solve, a few more pop up that we never thought would exist. It makes me happy.
- jleader 15y ago
  It appears that that story about the director of the Patent Office saying there's nothing more to invent is mostly made up, see http://en.wikipedia.org/wiki/Charles_H._Duell http://en.wikipedia.org/wiki/Charles_H._Duell or http://www.quotationspage.com/quote/22779.html http://www.quotationspage.com/quote/22779.html or the most detailed debunking: http://www.myoutbox.net/posass.htm http://www.myoutbox.net/posass.htm

yarapavan 15y ago

I liked his answer for the question - why don’t you do email? Someone has to not be tweeting all the time, someone has to be thinking about things which need a long attention span and trying to organise material and build up strong foundations instead of rushing off across the frontier. It takes a long time to put out something that has the right style; I have to really think about it and if I’m going to do it right I have to spend a lot of time focussed on it. And I was being treated like an oracle, lots of people from around the world were asking my opinion about whatever, so after 15 years of email I decided that was a life time’s worth.

a3_nm 15y ago
Actually, you can still email Knuth, and his secretary will print the email out. See http://www-cs-faculty.stanford.edu/~uno/email.html http://www-cs-faculty.stanford.edu/~uno/email.html I can understand why he wants to be on the bottom of things, but I don't get why encouraging people to use postal mail rather than email is supposed to help. Email is just a medium, and it's a much more sensible way to move information around. The widespread habit of checking email every 5 seconds is by no means compulsory. I wonder why Knuth says that the batch mode communication he wants to do cannot be done properly with email.
- jleader 15y ago
  I think it's just that he's trying to raise the barrier to entry a little higher. If you know you'll get a response in a few days or weeks on paper, instead of a day or two later in your email client, that makes you look for other, quicker ways to get the answer you're looking for, which reduces the load on Knuth (and his secretary).

dxbydt 15y ago

Coincidentally, I was thinking of what Knuth said just last week, when one of my comments was downvoted into obscurity for promoting a supposedly obscure/contrived math formula over a lame switch-case :) http://news.ycombinator.com/item?id=2817716 http://news.ycombinator.com/item?id=2817716 In reality, that formula is neither obscure nor contrived. I actually learnt it in high school ie. it was formally taught and we worked through how to come up with such functions given a set of inputs and outputs ie. essentially a bounded set mapping problem { 1->31, 3->31, 4->30,5->31, 6->30,7->31....} So anyways, Knuth gave an interview maybe 10 years back where he lamented the lack of dexterity in math in American undergrads. He said even Stanford has to offer "remedial courses in math" for CS undergrads, because there is an "interest vs ability" mismatch. At that time ( during the 1999-2000 tech boom ), a lot of Stanford undergrads were indicating CS as their major. So they clearly had the "interest". But upon testing them, it was clear they didn't have the "ability". Hence this interest vs ability mismatch, and a need for remedial math courses at Stanford. At the time, I was studying CS with a bunch of Russian, European, Indian & American classmates. I found that Russians & Eastern Europeans tend to favor clever 1-line math code. They themselves were trained in physics and math & looked at CS problems as essentially a laborious function-mapping exercise. Whereas the Americans invariably wanted "readable code" and ended up writing verbose switch-cases and if-elif-endif clauses ie. essentially unrolling for-loops. Knuth's Concrete Math ( http://www-cs-faculty.stanford.edu/~uno/gkp.html http://www-cs-faculty.stanford.edu/~uno/gkp.html ) was a sort of bible in my days. You had to know generating functions and recurrences and even stuff like complex powers. Like for instance i^i ( i raised to i ) leads to an infinite set of points with real and imaginary parts, scattered all over the 2D plane. You had to know cosets and Lagrange's theorem and set partitioning using cosets. Nowadays people simply enumerate everything as there is no penalty ( in performance terms ), so CS problems have a "select 1 answer from set of possible answers" flavor, rather than "contrive a function to compute the answer" flavor.

kenjackson 15y ago
I just upvoted your old post. :-) I did disagree with the solution (many who see the code would get confused), but it certainly was a positive contribution to the discussion, which is always my bar for upvote/downvote. And I love Concrete Mathematics. I wish people would say Knuth is the author of TAOCP and Concrete Mathematics. And Ronald Graham is another great computer scientist. With that said I do distinguish computer science from programming.

kragen 15y ago

There's a much longer interview with Knuth posted originally at http://www.webofstories.com/play/17060 http://www.webofstories.com/play/17060. The transcript is probably interesting reading for anyone interested in the man; I imported it into Git and started formatting it into paragraphs at https://github.com/kragen/knuth-interview-2006 https://github.com/kragen/knuth-interview-2006, although I haven't finished. I think it would be worthwhile for more people to (a) clone the git repository to reduce the risk that this primary historical source will be lost, and (b) reformat more of the transcript if they feel like it.

sebkomianos 15y ago

"I’m jealous of astronomers, for example, because people respect astronomers for doing astronomy because it’s interesting just for its own sake. I studied computer science because it’s interesting to study computer science." Probably my favorite part of the interview.

adharmad 15y ago

"I was born to be a computer scientist - I have a way of organising stuff in my head that just seems to make me a good programmer. I think everybody can learn to use computers, but only about one person in every 50 is a geek in the same way as I am. That means we can push the envelope and can resonate with the computer. The way we think helps to make it easier for us to know how to construct a machine." One of the few people who can state it as a fact, without sounding arrogant.