[00:03] Oh, nevermind, I was mistaken. [00:04] Do you still want me to send them to you? [00:04] not if there is nothing wrong with them : ) [00:04] heh. [00:05] I don't know what I was doing wrong, But I started over from scratch to test them, so as not to embarrass myself, and found that the results were very similar. [00:05] Maybe you just set the core setting incorrectly the first itme [00:05] time* [00:08] I don't know what I could have been doing differently, but obviously something was different. [00:08] It could possibly be that I was using a different version of a warriors code than I thought I was. [00:08] Well, time to go home. [00:11] MSG: Quit: leaving [00:40] Join: CL joined #corewars [00:40] MSG: Client Quit [04:43] MSG: Ping timeout: 252 seconds [05:29] Join: Mizcu joined #corewars [05:51] Join: fiveop joined #corewars [14:15] MSG: [14:40] Join: Fluffy joined #corewars [14:40] :) [15:11] hrm [15:24] Join: heath joined #corewars [15:24] hi heath [15:24] Hello. [15:24] Hi bvowk, heath :) [15:24] where you from? [15:25] Redlands, California. You? [15:25] Edmonton, Alberta [15:25] not many north americans in here usually [15:26] Interesting. [15:27] bvowk: In your evolver ... do you know how many fights are usually executed to benchmark one warrior? [15:27] (on one node) [15:27] Or does that change from warrior to warrior? [15:27] depends.. [15:28] I usually run a quick validation round [15:28] to weed out the useless ones.. [15:28] it'll be really small.. like 5-25 battles [15:28] and then I'll run the scoring runs, that'll be anywhere from 50->500 battles [15:28] depending on how accurate the scoring needs to be.. [15:28] the better the pool is, the less noisy the scoring can be [15:29] Do you reuse the old results (from 5-25 battles) in the scoring runs? [15:29] Maybe someone should write a croewar@home. [15:29] err... corewar@home [15:30] corewar@bvowk will probably have more computers than corewar@home ;-) [15:30] I already did :) [15:30] no, the validation is usually against something different [15:30] bvowk: It's distributed and runs as people's screen saver? [15:30] :) [15:30] heath: its distributed and runs on piles and piles of cpus [15:30] hehe ... I haven't used a screen saver in years [15:30] Wow. Cool. [15:30] but since I'm unix, its not a screen saver [15:31] I was just thinking about those folding at home and golem projects. [15:31] http://demo.cs.brandeis.edu/golem/ [16:20] Join: sf joined #corewars [16:20] Hello [16:20] Hello. [16:21] Hi heath are you new? [16:21] Yes. [16:21] Nice to see you sf [16:21] Hi Bvowky [16:21] Hi sf! [16:21] Hi Fluff [16:21] Is corewar.co.uk working for you? [16:21] I'll never talk to you again, sf. bvowk gets an 'y', but not me ;-) [16:21] :/ [16:23] Waltz, bad nymph, for quick jigs vex. [16:23] ? [16:23] It's a pangram. [16:23] Oh :) [16:23] Only 26 letters long, mind you. [16:23] You've read some of dewdney's other stuff too :) [16:25] I think there is a tyranosures rex trying to crush into my building. [16:25] I Count 28 letters [16:26] sf: Oops, sorry. For a second I thought there were only 24 letters in the alphabet. I stored the information as (2+[letters in alphabet]) and got the second number wrong. [16:26] bvowk: Do you like to hear some benchmark results for PyCorewar? [16:27] bvowk: (And no, I won't release it yet ;-) [16:27] sf: In my brain that is. [16:28] I only know of two pabgrams that make sense that only repeat 2 letters. [16:28] That bein one, anf the other: [16:28] Sphinx of black quartz, judge my vow. [16:29] Hmmm... Quartz might be a good word in Scrabble [16:30] fluffy: Lau Ludoviko Zamenhof bongustas fresa ceha mangajo kun spicoj [16:30] Shure would. Can you imagine getting a triple word on that? [16:30] :) [16:30] sf: :) [16:31] Dang, there is scarry bashing and shaking going on above me on the roof. It is scaring me. [16:32] heath: wild animals? [16:33] pMARS: 7.8405 MIPS, exhaust: 9.0484, exhaust-ma: 11.7426, exmars: 12.4855, PyCorewar 0.2.0: 13.3659, PyCorewar-devel: 16.5030 MIPS :) [16:33] sf: I hope not. It must be people. Some kind of construction? [16:34] The other week a crane and it's load fell through the roof of a local shop, destroying the shop [16:34] sf: Scary, was everyone ok? [16:37] Yes but the shop had to close [16:40] The beginner hill finally has 25 warriors :) [16:41] sf: I keep trying to get number one on the beginner hill, but my warriors never seem to be as good as I think they are. [16:42] heath: keep trying :) [16:42] heath: do you know about optima numbers? [16:42] heath: Have you already started to benchmark/optimize your warriors before submitting them to the hill? [16:42] sf: No, what are they? [16:43] Fluffy: Well, I run them agains other warriors to see how they fare before submitting them. Is there a more organized precedure? [16:43] heath: basically if you are bombing core, bombing every fifth location would be bad [16:43] Yes, there are some well-designed benchmarks out there [16:43] because most opponents are longer than five instructions [16:44] At the moment probably the wilfiz-benchmark is good enough, but later you might want to try the fsh-benchmarks [16:44] Both can be found at corewar.info [16:44] Fluffy: And I just run my warriror against those warrriors and see how they do, right? [16:44] yes, usually you average the results [16:44] but you could bomb every 35th instruction, and when the end of core is reached, your bombing pattern would wrap around. eventually you would bomb every fifth location [16:44] sf: Ah, that makes sense. [16:44] and scale them to 100 rounds [16:45] sf: I generally pick a large prime, like 2153. [16:45] but by bombing every 35th location you are more likely to bomb an opponent longer than 5 instructions sooner [16:45] heath: that sounds ok [17:10] Join: Mizcu joined #corewars [17:15] Hi Mizcu [17:16] . [17:16] Hi Mizcu [17:17] hi sf [17:22] MSG: Ping timeout: 252 seconds [17:27] * sf waves [17:27] * Fluffy waves to sf :) [17:27] Join: Mizcu joined #corewars [17:28] * sf pokes bvowk with her soldering iron [17:28] MSG: Quit: "Help! I've been g:lined from my mIRC!!" Bersirc 2.2: less n00bs [ http://www.bersirc.org/ - Open Source IRC ] [17:34] MSG: Ping timeout: 252 seconds [17:38] ow! [17:43] bvowk: 10 minutes to react? What took you so long? ;-) [17:44] It wasn't plugged in when she first poked it into him. [17:44] Join: Mizcu joined #corewars [17:44] It took a while to heat up. [17:44] Koth doesnt like me -_- [17:45] Mizcu: What's the problem? [17:45] Keeps disconnecting from Irc [17:45] other servers work fine [17:46] Try another nick, to see whether it really doesn't like you ;-) [17:50] im not laughing [17:51] Yes, I can hear that ;-) [18:06] Ok, let's see how Gogo does. [18:07] Oh well... [18:08] Could be better [18:08] Yesh. [18:11] some sort of scanner or oneshot? [18:12] It was an attempt at bombing and blindly splitting into nowhere, hoping to steal enemy code. [18:13] What is a one shot? [18:13] its a scan into a multipass coreclear [18:14] usually with one or two runs using spl's and then multiple using dat's [18:15] it usually works quite well against papers [18:15] Though i know of a version that has one pass with a special mov [18:29] Join: Core29 joined #corewars [18:31] Join: sascha joined #corewars [18:32] Hi sascha! [18:32] Good evening gentlemen ! [18:35] Hello. [18:35] It's before noon here, however. [18:35] heath: then you live in the wrong part of the earth ;-) [18:36] Heh :). [18:38] I was that I could touch a book and instantly know and understand everything in it. [18:38] s/was/wish/ [18:39] That would eliminate a lot of fun. [18:40] Fun? Reading? I am such a slow reader. It is sad. [18:40] Then learn how to read [18:40] ;-) [18:40] :) [18:59] Time for lunch. [19:01] :) [19:21] MSG: [19:53] MSG: Quit: using sirc version 2.211+KSIRC/1.3.12 [19:55] MSG: Quit: Leaving [19:55] Join: asw joined #corewars [19:57] MSG: Quit: leaving [20:04] MSG: Quit: humhum [21:13] MSG: Quit: Trillian (http://www.ceruleanstudios.com [21:22] Join: brx joined #corewars [21:52] You are all extremely quiet lately [22:09] * brx has had stuff to do [22:17] brx: Haven't I seen you somewhere? In #fluxbox I beleive. [22:52] heath: ah in #ratpoison [22:52] you have seen me there :) [22:52] Ah, that's it. [22:52] What a small world. [22:53] indeed. [22:53] I suppose the kind of people who are into corewars would be the same kind of people who are into ratpoison, perhaps. [22:53] hrm [22:54] well, two kind of weirdos at the very least. [22:54] :) [22:55] I used to lurk here til a few weeks ago. recently I had no real motivation for anything creative or thought provoking beyond math [22:55] +since [22:55] Are you a math major? [22:56] I have recently started studying math [22:56] On your own? [22:57] no [22:57] in university with a hoard of people [22:59] Ah. [23:00] it's been great fun so far. [23:01] How many quarter's have you put in so far? [23:04] I have started recently, as in a few weeks ago. [23:04] Ah, cool. My major was computer science, but I had to take wuite a bit of math. [23:04] s/wuite/quite/ [23:06] well, I had to decide for one or the other [23:06] I considered math to be more challenging and interesting, so I went with that. [23:06] What are your professional intentions? [23:07] I have none. well, besides becoming a pirate. [23:07] Ah, well then ok. :) [23:07] I just want to have a good time doing something fun :) [23:07] It's just that computer science is more marketable, I think. But that doesn't mean that you get a fun and intereting job. Trust me :(. [23:08] apparently math is _very_ marketable [23:08] But if you are going to go all the way (aka PHD), then either will do, I think. [23:08] brx: Really? [23:08] indeed. [23:08] With a BS? [23:09] bachelor? [23:09] Yeah. [23:09] My brother has his BS in math. [23:10] The only job he has found is a high school teacher. [23:10] well, probably not if that's all you get. but I suppose a bs in math might help a great deal if you have a higher degree in something else [23:10] But he is not very ambitious, I guess. [23:11] brx: Yeah, that is what I meant. You can get a job fairly easily with just a BS in computer science. It seems. [23:11] neither am I, really. only the horrors of ending up as a teacher might save me from a similar fate :) [23:11] really? [23:11] I don't know about that over here in germany. [23:11] Might be different. [23:11] If you go further though, you could be a professor. [23:11] :) [23:11] no thank you. [23:12] time to deal with interesting stuff, good. time to deal with stupendously writing stuff on a blackboard for dozens of bored students, no. [23:14] I suppose, but the best professors guide the learning of math in a way in which you do not become bored. [23:14] I enjoy sleeping on sheets of paper with weird symbols on them, spending time on hogwash and getting to treat it seriously in front of others.. [23:14] fun [23:15] :) [23:16] though many of those who started with me have already given up [23:17] I have heard of that happening. I didn't really know many people in school, so I didn't really see it happen. But it is said that most people drop out as freshmen pretty early. [23:17] Not willing to put in the work when no one is forcing them, I suppose. [23:18] Or possibly they are no longer the biggest fish in the sea, so they would rather leave the sea. [23:18] we all suffer together. no big fish in sight. [23:19] School is hard. Math is hard. But it is all worth it. Sometimes I wish I had worked harder to learn more while there. It's embarrasing the things that I don't remember that I know were covered. [23:21] Oh well, now I have to put myself on a strict learning regimen. To make up for the lacking CS program that my university provided. [23:22] I am attending some compsci lectures in univ too [23:22] actually, one. [23:22] Which one? [23:23] oh, introduction to cs [23:23] formal languages and automata. very laid back. [23:24] Ah, ours was called "Discreet Structures" basically logic and methods of proofs (like mathematical inducton). [23:24] from what I can tell, I won't hear any challenging things in compsci til my bs in math [23:24] Formal languages was an advances course for us. My favorite of the whole program, I think. [23:25] Converting regular expressions to NFA's then to DFA's is not challenging? [23:26] not really. the proofs we were confronted with or had to formulate in exercises were rather simple too so far. they all had a "constructy" feel to them. [23:27] Interesting. I have "fond" memories of useing the pumpink lemma to prove whether or not somthing is a regular language. [23:27] s/pumpink/pumping/ [23:27] ah, we had that one too, I can't recall it however. [23:28] Yeah, I would have to go back to my notes. [23:28] Compiler design was the next class after that. Man that class was bad ass, very tough though. [23:29] I am not sure if that is among the 5 I am supposed to attend [23:30] Hmm. I but foundations of algorithms will be one of them. [23:30] s/but/bet/ [23:35] anyway [23:35] I gotta crash for now [23:35] talk to you later [23:35] ok. [23:35] Hold up a sec. [23:35] ya? [23:35] I think I found some of my hoework from that class. [23:35] Yeah: [23:35] http://www.csupomona.edu/~hncaldwell/hw3.pdf [23:36] ah, what about it :) [23:37] Just like to brag, It looks sweet, no? [23:37] it looks annoying, but not overly complex [23:38] Yeah, the process of going from a DFA to an NFA is tedious, but fairly straight forward. [23:38] what did you use to make the diagrams? [23:39] LaTeX. [23:39] the drawings? [23:39] With pstricks, [23:39] Yeah. [23:39] pstricks.. [23:39] The pstricks package is beautiful. [23:39] Lets you do all kinds of cool stuff. [23:40] sounds incredibly sweet [23:40] It is. [23:40] what can it do? [23:41] It's probably best to see examples: [23:41] http://tug.org/PSTricks/main.cgi?file=examples#arc0 [23:41] I can't remember the resource I used mostly when studying it, but it had some cool examples. [23:43] looks nice [23:43] wow [23:43] thank you :) [23:43] and now I gotta crash [23:44] Sure, you're welcome. LaTeX and pstricks were invaluable for me in school. [23:44] Ok, have fun. [23:44] Nice talking again. [23:44] yes. talk to you later :)