It is rare enough to be called 'Earthshaking' but the probability is really thin.
An "idea implemented." Unfortunately is rarely turns out to be something concrete - it's usually an idea with a significant buy in... But an idea that's a step in the right direction... A call to action...
Examples:
1) Cisco announced a really fast router that was going to "change the Internet forever." (http://tech.fortune.cnn.com/2010/03/09/what-cisco-lacks-in-s...)
2) Bloombox (http://bloomenergy.com)
Of course, as others have pointed out maybe he's just playing the fool.
Did not mean to garner offense.
Valve at E3. Bastards.
Not trying to be a dick, just wondering if you have a point if you're not kidding.
I just can't think of anything besides P=NP (or proving P!=NP) that Knuth is likely to announce that I'd consider earth-shattering.
In the end, it's likely to be a decent but not earth-shattering discovery. It better not be a TeX announcement.
If he walks out, says "42" and leaves, I'm selling my copies of "The Art of Computer Programming".
OTOH, maybe Twitter will continue their overengineering by porting everything to MMIX.
That's what I hope we'll find out tomorrow. Too hopeful?
I'm curious to see other's guesses here. My personal hunch is that he is going to declare TeX complete, and move the version number to Pi (as he had planned to do upon his death.)
Besides, this is an announcement at a TeX conference; it's going to be TeX-related.
Incidentally, I went to his last Stanford lecture, the lecture hall with maybe ~400 seats was totally full, and there were exactly 3 women in the audience.
If it was something as significant as P{=,!=}NP, I really doubt he would have been able to keep this hidden for any length of time.
