(no subject)

[timelets]

Today's breakthrough doesn't fit any model and it still looks too good to be true. I can't explain it even in hindsight. The young man who came up with the solution did it only after I pressed him hard to do the impossible. Actually, no. First, he said it was impossible and I asked him to solve the problem within a very specific, narrow set of scenarios. He did that. Then I asked whether it would be possible to generalize the solution. It turned out to be possible. Then another guy got up, walked up to the whiteboard and extended the idea a bit further. Then I asked whether it was possible to apply the solution in a different domain. Well, there seemed to be no fundamental barrier to do that. Then their program director got excited, got up and said, "I really like it, but what's the catch?" Currently, there seems to be no catch, but we have to verify the solution on Thursday.

I totally can't explain it. Maybe the key was to replace the current "# alphabet" with a similar, but a still different one. Let's hope and pray it works.