Sounds accurate to me

@Kerm: *cough, cough, source plz?, cough, cough*

Simms wasn't working right when I talked to him. He wouldn't answer me for a long time and then suddenly said "Morning Simms" and ignored me for another 5 minutes. Then he suddenly repeated himself, took a short break and decided to greet me 5 more times, even though I was no longer speaking to him.

Time: 1:00 - 1:30 AM GMT (The usage bars on your sig Kerm said first two machines were full usage and others nil.)

[Edit: 1:48 AM GMT]

I'm posting on the forums and suddenly I get a message from Simms through Kerm's AIM. It sent me a message consisting of a period. I got three of those. It seems I've made a friend.

Yeah, since I fixed part of the wiki module its been crashing itself ever more frequently. I'll try to take care of that today, which will hopefully repair its response time.

Also, welcome back to Cemetech!

Slightly off topic, but not to much...

http://www.wired.com/techbiz/people/magazine/16-02/ff_aimystery

Interesting/depressing article on 2 AI researchers...

Fascinating article, and as you say, quite depressing. That seems to be an excellent overview of the progress of natural-learning AIs over the past few decades, though.

*bump* One million connections in less than a month!

1,000,185 connections currently link 246,735 nodes in LTM.

I saw that, interesting article, sad story.




Kerm: does this mean its working better now?

Yup, definitely. It's going seven or eight hours between MySQL crashes, and the distribution seems to be working substantially better. Part of the improvement is from more efficient wiki parsing.

*bump* Over the past two weeks, Simms has started driving itself to load averages of 110+, then recovering down to nominal, then quickly driving itself to a high load again. I'm looking into it.

I haven't seen Simms on AIM in a while. what's he been up to besides reading wikipedia?