December 17th, 2000


awake, damnit.

I just woke up, thinking it was morning after only 4 hours of sleep. Damnit.

Upon getting to my computers, both my Linux and Windows machines were reset... sitting at their login screens. Cartman and Kenny have uptimes of exactly 2:55 each... looks like the power went out about 3 hours ago, or about the time I went to bed.

That, or somebody's in the house and reset them.

My cough is back. Stupid sickness.

Why can't I sleep?? I fell right asleep earlier.... why couldn't I stay sleeping? Grrrrr.

a record

only 10 messages in my inbox!

I know once I fall asleep I'll wake up to 70 again, but oh well. It was fun getting it this low. Got a lot of work done.

I'm starting to get tired-hungry ... that feeling of hunger you get when you stay awake so long that you're hungry, but you're so tired that you don't want to eat and food sounds gross. The only solution is to fall asleep and wake up normal-hungry later.

I should go to bed.

(no subject)

working on a quick* non-LJ programming project.
i won't announce it before I'm done. maybe in an hour.
if all goes well I'll have pretty graphs to show.

* it's turning out to be more difficult than I thought.

bowling statistics

I knew this program I'm writing would take a ton of CPU, but I guess I didn't think realize how much it. What I'm doing is calculating bowling statistics over all possible games. Last time I went bowling I got it in my head to do this and I can't stop thinking about what the three-dimensional graph of [x=balls thrown, y=score, z=occurences] must look like.

I wrote it in Perl initially but it was way too slow so I rewrote it in C. In C it takes 43 seconds to compute the incredible small subset of the larger problem... in Perl that same subset takes 4 minutes and 10 seconds. Stupid Perl. :-)

If I disable the prints to stdout, the time of the C program drops to 10 seconds. On Kenny (the P3-933) it runs in 4 seconds.

Solving the entire problem is going to take approximately (50**6) times longer because right now I'm running it assuming only 4 frame game.

I need to think of way of a new way to solve the problem by breaking the game up into separate parts which don't interfere with the rest of the game (i.e. sets of open frames following an open frame).

Or, I need to make the program start at arbitrary points and run it on all four computers in here while I sleep. That's kinda hard to do right now because of how it's written (using recursion) ... An iterative solution might be faster, and more easily parallelizable, but it's harder to think about.

I guess I'll graph what I have so far, and then think about this later.

Also, I decided to use glib for the C port, but I don't like it much ... g_list_append appears to be O(n). But then again, so is strcat .... :-/

(no subject)

played some pool with my dad then watched the simpsons... pretty funny. :-)

need to eat, then maybe head over to garrett's for a bit.


my goal for this vacation is to learn a bunch of new computer things and to work on my spinning. garrett and I spun yesterday ... my little brother watched me today and tried for awhile.

as for computer stuff: I'm going to undertake a major project now in JavaScript ... I've always hate JavaScript, but I think it's time I give it another look. I need it now so I might as well accept it and keep an open mind.