[madnak]

[ QUOTE ]
Things Im interested in tend to grow factorially. So even in a log scale we will still see some serious growth for those problems. This could be a function of my interests though, and the whole set of programs may not tend to show that growth pattern.

[/ QUOTE ]

I think you're right here. It's hard to categorize these things. I mean, solving checkers, building a rainbow table, rendering a video, these all fall within the middle range. I'm sure there are many more examples. And if computers continue to get more powerful the whole "window" will change. But maybe there is a gap. I don't know what that would imply... Coincidence? Or are we "missing something?"

[ QUOTE ]
Also, at least for right now, the difference between something like 10k and 10 million years to solve is, for intents and purposes, trivial. In as much as neither can produce the data we want, and thus fails at its sole purpose. (obviously, to a certain extent, Im taking a selfish viewpoint w/ that statement, since its reasonable to believe that, at some point, those programs will be runable in normal time, and, of course, at some point the difference will no longer be trivial).

[/ QUOTE ]

Right, but the difference between 1 microsecond and 1 millisecond is also trivial to us, because it produces the data effectively instantaneously. Time scales of hours and days are more significant to us because those are the time scales we deal with in our lives. Exceptions abound on the small end of things, and would abound on the large end if we were immortal.