Since I sent out party invitations, suppose I'm now committed to pulling it off. If even half the people show up it's going to be a blowout.

Suppose I deserve a reality check w/r/t my comments yesterday about the code red worm. The Morris worm only infected about 9000 computers or roughly 10% of the fledgling internet at the time. This one is reported to have hit 200,000 plus; although the number of connected systems is some staggeringly higher number now. So to put it in perspective, the Morris worm had more of a perceived effect, but code red probably is a bit more virulent in sheer spreading ability. Morris targetted Sun and DEC Unix boxes, so it affected me directly. I can laugh about code red because nyah, nyah; I'm not running win2k.

But we've started to enter the exponential curve. This is quite likely to choke traffic in the next day or two. An hour ago my hit count was 31, or roughly one per hour since the worm came back to life. We're now up to 41. Most of the infected sites are overseas, where they are less likely to know about or how to apply patches. My guess is that several of the backbone carriers are - even as I type this; making contingency plans to sever some problematic international routes.

Can't help but wonder about the news article of the football player dying of heatstroke while "working out in blistering temperatures south of Minneapolis". Yeah, I suppose it's possible, but are you sure somebody hasn't stretched either the definition of blistering temperatures or south of Minneapolis? I mean 75 degrees could be considering blistering if you've been through one of their winters and of course Texas is south...

There are three basic types of progress meters on computers and I hate them all. The first type you see during a software install - it just has a percent indicator. But you ever notice that it will cover 0-80% in like 3 minutes, and the remaining 20% takes about 25?. Why do they do that? The second type is the self-adjusting estimated download time indicator. It's the most accurate of the three; but its problem is that it tells the truth. It will read 14 minutes left, and then when the channel slows down (as it invariably does) it jumps to 47 minutes. Dontcha' just hate that? The third type is the estimated battery charging time indicator and Windows Me installer. The folks who programmed these were well aware that nobody likes to see estimated time jump backward, so they lie to you. Estimated 14 minutes until complete. Come back fifteen minutes later and it's still estimated 14 minutes until complete. The designers can't win. Maybe we're just better off not knowing. Before they had progress meters they generally just put up an info message "this could take between 16 minutes and three hours depending on how your computer is configured". I actually prefer that, as I have some basic idea of the range we're talking about. The best of the lot you don't see anymore. In this one the designers make very pessimistic assumptions when the operation begins. Estimated download time 6 hours 20 minutes. Then things go well and you come back 15 minutes later and it's an estimated 3 hours. Total physical time takes under an hour. Since it always takes shorter than the original pessimistic estimation, in the last minute or two the last estimated 15 minutes fly by at warp speed. I like that.