Knuth Plans 'Earthshaking Announcement' Wednesday

I Don't Believe in Imaginary Property writes "Donald Knuth is planning to make an 'earthshaking announcement' on Wednesday, at TeX's 32nd Anniversary Celebration, on the final day of the TUG 2010 Conference. Unfortunately, nobody seems to know what it is. So far speculation ranges from proving P!=NP, to a new volume of The Art of Computer Programming, to his retirement. Maybe Duke Nukem Forever has been ported to MMIX?" Let the speculation begin.

Comments

Re:That he is...

[jordan_robot]

That he is a computer simulation fooling all of us for over 50 years...

I think you mean that we're all a computer simulation he has been running for over 50 years...

Re:I don't think proving P!=NP is earthshaking

[mark-t]

A proof that P=NP would have much more potential to genuine change things simply because it would disprove a ubiquitous assumption: that P NP. Historically, when universally popular assumptions have been proven wrong, the resulting paradigm shift in the way people think about the matter produces some fascinating changes. P!=NP would give closure to an open problem but would not be so earth-changing because we already operate under the assumption that the premise is true.

or just

[Kohath]

drink a beer, relax, and wait until tomorrow for the announcement. Which is sure to be disappointing now.

I predict he announces that computer programming is best practiced as a semi-automated assembly-line-style set of interchangeable tasks rather than an "art". He'll say that programming as an "art" is anachronistic. inefficient, and impractical, and that the conventional approach and the people who promote it have been holding back progress in software creation because a faster, cheaper, more modern, dumbed-down approach doesn't appeal to them professionally or aesthetically.

And then he'll announce his new software construction method that can be done by ordinary people with a short period of training for 1/5th what computer programmers make. It works great, but it's boring and repetitive and never creative. It delivers software in a predictable amount of time with a predictable budget and reasonable (also predictable) quality. And the development costs less than half of conventional approaches.

That's my prediction.

<--- Flamewar starts here

[SpazmodeusG]

Knuth doesn't stand out amongst his peers in his field as much as those examples you've mentioned. Peers such as Turing, Shannon, Dijkstra, Boole, Babbage, von Neumann, Hopper... (etc.) are all more important
http://en.wikipedia.org/wiki/List_of_computer_scientists

(disclaimer: i knew who Knuth was but i'm just not bothered by those that don't when there are so many prominent computer scientists)

Re:John Carmack

[Martin Blank]

Works on space projects and still designs game engines.

He also married one of the world's most awesome women, Anna Kang. On their honeymoon, she let a pair of computers be set up in the hotel room so that he could program when the mood struck him. No woman I know would allow such a thing to happen.

Re:P!=NP

[hitmark]

one potential worry today tho is the lack of "downtime". That is, there are so many ways for us to not be bored that we basically have no real time to sit down and form grand mental models.