[00:12] MSG: Read error: Operation timed out [00:13] MSG: Read error: Operation timed out [00:14] MSG: Read error: Operation timed out [00:14] Join: johnkw joined #corewars [00:14] MSG: Read error: Operation timed out [00:14] Join: michal joined #corewars [00:16] Join: asw joined #corewars [00:21] Join: Mizcu joined #corewars [00:22] My puter is so screaming for mercy that it cant keep up irc-connection [00:23] But thats just a sign of well sliced processor time (Optimax > Irc) [00:38] Mizcu: http://solace.is-a-geek.org:8080/projects/screens/rfc.mp3 [00:39] comment that please :) [00:43] heh, it crashes my winamp at 3 seconds [00:44] Too sharp for my ears [00:44] too sharp? [00:44] explain [00:45] high-pitched voices [00:46] ah [00:46] eh a-7 g-7 e-7 / g-7 d-7 e-7 [00:46] does the string sound okay? [00:47] Part: michal left #corewars [00:47] you cant really do anything to the samples with a tracker [00:47] well I can alter the pitch [00:47] i can transpose it to an octave lower [00:47] and the drums(?) dont fit [00:48] which one [00:48] the "bassline" ? [00:49] actually it does sound better if I make it snare only [00:49] i hear really only 4 layers: humming in the background, strings, "chump chump" (what i called drums(?)) and the hihat-thin drum behind strings [00:50] there are 5, the drumpattern is made up of two different drums [00:50] the higher sounding is a different drum realy [00:50] but yes [00:50] it's only 4 layers [00:50] (4 tracks) [00:52] I have actually a cd with a nice bunch of 4-channel mods, could try diving into it again some day [00:52] from 1996.. [00:55] http://mirror.gbar.dtu.dk/AmiNet/mods/xceed/Synthmania06.lha , this one was sorta favourite of mine when i got the cd [00:56] (Yes, i still could remember the composers handle and last name to google with, i have said i have quite selective memory) [00:57] lha? [00:57] lharc, archiver for amiga [00:57] lemme send unpacked then [00:59] old modules sound half-a dozen times better on Mono rather than in Stereo.. [01:00] Mizcu: same url with rfc2.mp3 [01:01] Mizcu: that is nice [01:01] kind of [01:02] it's really fast.. [01:03] well mine is actually running faster :) [01:03] the "chump chump" needs lower volume, it sounds like another lead, though it should be support [01:03] * brx_ nods [01:04] I don't know if I am allowed to change volume :) [01:05] hmm [01:07] 119 Megs of stuff of mod's on the cd [01:08] (guess i need to go get some sleep, writing getting wierd) [01:10] ? [01:19] (You stuck into listening that, or copying ideas into yours?) [01:19] copying now [01:19] no [01:19] but listening [01:50] this. [01:50] I don't know how I made this sound like tetris [01:50] but I did [01:52] heh, i say [01:52] wait [01:52] heh [01:52] you gotta listen to this... [01:57] link/dcc! [01:58] yes [01:59] ! [02:00] I am repositioning the noise a little [02:06] okay [02:08] listening :) [02:09] Mizcu: whack.mp3 [02:09] http://solace.is-a-geek.org:8080/projects/screens/whack.mp3 [02:11] that second lead at 3rd second sounds exactly like out of NES [02:11] which one? :) [02:12] well what new lead is added at 3.5 secs? [02:12] I check [02:12] ah that [02:12] how do you like the noise? :) [02:13] that noise has to go [02:13] Doesnt fit [02:13] yes :) [02:13] you want to help doing this? I will send you the xm with the samples [02:13] :) [02:13] dunno [02:14] would like to concentrate on the drums, because atm they suck [02:14] and I do think you could make something interesting :) [02:15] http://solace.is-a-geek.org:8080/projects/screens/whack.xm [02:18] Wonder if i should reinstall buzz.. [02:22] can buzz edit xm? [02:22] dont think so.. [02:23] come on miz :) [08:29] MSG: Read error: Operation timed out [08:30] Join: asw joined #corewars [09:13] Join: elkauka joined #corewars [09:13] hi kauka [09:13] hi miz [09:13] still working on my vampire. yet no boot/ q scan - 123 against fsh3.0 [09:14] if i can manage to get 130 i send it to the hills [09:20] Join: Roy joined #corewars [09:21] hi roy [09:26] MSG: Quit: Quit now, you'll never make it. If you disregard this advice, you'll be halfway there. [09:33] hi [09:34] topscore of 115.9 .. [09:35] but i dont like the look of pattern in that step [09:39] heh, that is without attack? [09:39] yes [09:40] should i choose pattern with least losses, or just plain scoring.. [09:41] I think the plain scoring.. but I'm not a big fan of optimizing without attack either :) [09:42] In Hullabaloo (1+2) the qscan is even optimized with the paper+stone. When I changed the qscan with a more effective one (proved by Jens) it didn't score better, but much worse [09:50] boot [09:50] MSG: [09:53] Join: Mizcu joined #corewars [10:24] Join: Jens joined #corewars [10:24] Hi everybody! [10:24] hi jens [10:25] :) [10:25] Roy: About your evolver ... altough I promised to look into it, I totally forgot about it. Sorry ... [10:26] Hehe thought that that happened :) [10:26] Did you change anything? [10:27] Nah, didn't do anything with it yet, can't find the stupid bug [10:28] Then I give it another try and will look into it after I have uploaded all new score surfaces. [10:28] Sascha has sent a lot of then and I'm way behind with publishing them. [10:28] Wow nice [10:29] I have now all score surfaces for Tiny Stone and YAP. [10:29] But creating all pages and images will take some time [10:30] As several people have suggested I should take a simpler scanner as a prototype than Hired Sword. [10:30] All the surfaces? Nice, now you can do the math and find the best stepsize for the warrior against the benchmark :) and make a picture of that [10:31] For YAP I've already done that [10:31] ... and sent to the hill (score 130) [10:31] but with a qscan (from JM's Digital Swarm) [10:32] The average score surface is at http://corewars.jgutzeit.de/score_surfaces/yap/average.en.html [10:32] hmm, a big red line parallel to the big blue diagonal line [10:33] That's only an artefact of interlocking [10:33] Without interlocking it looks like the "normal" score surface, i.e. like the one against Muskrat, but with lower scores [10:34] Sascha was so kind to explain it to me. Interlocking is a dodger and copies itself behind a paper-instance. When the paper-steps are almost identical, the copy of interlocking will be overwritten quite soon. [10:34] That's why the red lines parallel to pStep2 = k * pStep1 [10:35] ah.. [10:36] That the fun-thing about score surfaces, you can learn a lot about how warriors work [10:37] Can you think of interesting things to do with the data for the score surfaces? I'm quite desperate to do sth. with it [10:37] Hmmm, not really sure what to do with them now.. [10:39] To tell good and bad paper-steps apart, I let YAP fight a jmp $ 0, $ 0. Where YAP scores > 290 is a good step-pair. It might be useful to have a list of good paper steps for optimax [10:41] hmm.. [10:42] I would really like to do a score surfaces for a standard core, but as we would need about 1025 times the time to complete one score surface, it is out of reach now. At the moment the fastest score suface needs about 12 hours (Tiny Stone vs. ...), it would take about 16 months on one computer :( [10:43] and against a jmp? is it slow then too? [10:43] TS vs jmp $ 0 might be a little bit faster, but not that much [10:43] maybe you could fight against a djn.b 0,#200 (or something) to see how soon a paper can kill a warrior [10:45] I don't want to fight "random" warrior. I already know, that if I can make one score surface, I want to do many [10:45] I know one thing you could do.. [10:46] Get a huge loan and buy a supercomputer :) and let us use it too! [10:46] Yes, I've already thought about that idea, but there is an easier solution. [10:47] Wait 10 years, then the computers are way more faster [10:47] Heh and then we are old.. [10:48] Nah, until then the secret of rejuvination is found ;-) [10:49] And if I had a faster computer, I'd probably use it for optimaxing my warriors [10:49] Hell yeah :) nice entry btw, paper? [10:49] :)))) [10:49] I mean: What kind of paper? [10:50] fsilk/fsilk/fsilk/mov.i # iStep, { 1/mov.i x, < y [10:50] heh almost the same as I used in Thingamabob [10:51] Did you tell me that? [10:52] I mean, when I wrote it, I knew I've already seen sth. similar, but couldn't remember the warrior [10:52] Don't know... but it uses mov.i #step,{1 - mov.i >x,{y [10:53] OK, then I can still claim, that is was my idea ;-) [10:53] Lol no, you just admitted you had seen it before :-O and we have logs! [10:53] damn [10:53] Its not my idea either btw... I stole it too ;-) [10:54] nPaper has mov.i # 1, {1/mov x, < y/jmz.a [10:54] err .. nPaperII [10:54] yeah, but it doesn't have a open loop [10:55] I've only written it down and added a qscan. I was very surprised, that it ended 7th (or so) [10:55] I've made one run with pmars to make sure, that it does suicide. That's all. [10:55] Well, it IS optimized isn't it? [10:55] After that I optimaxed it a little bit further and it became 2nd [10:56] Ah k :) [10:56] And the Fizmo was so kind to send Zoom in on a Stone and it became koth [10:56] It is quite interesting, that it really likes a lot of scanners [10:57] Never managed that before [10:57] Now the only problem is to make sure, that nobody touch the '94nop-hill again [10:57] *touches [10:57] I've already made the suggestion to open another '94nop-hill ;-) [10:58] No, you want to touch that because then you could age it :) you can always save the html file and put it online somewhere [10:59] Hmm ... that the other possibility, but it works only under the assumption, that Harmless Fluff* Bunny gets old enough [11:02] I'm going to make a very oldscool paper now :) should be fun to watch that score [11:04] '88-style? [11:05] Tiny Stone vs. bestwar4 is now online: http://corewars.jgutzeit.de/score_surfaces/tiny_stone/bestwar4.en.html [11:06] Now I have an idea. We have programs like corestep, which calculate good bombing/scan-steps. I should test, whether good score are at such steps for Tiny Stone [11:06] *good scores [11:11] ah yes, good idea [11:11] benchmark for stepcalculaters :P [11:13] I find the idea behind the step-calculator quite good, but as far as I know, nobody has tested it beyond checking, that own scanners/bombers get better scores with it [11:33] Hmm ... I forgot to do Tiny Stone vs. Tiny Blowrag :-( [11:45] MSG: [11:49] Join: elkauka joined #corewars [11:54] yeah! 94nop place #9. my highest entrie yet on 94nop! [11:57] Perfect!! [11:57] err .. Congrats! [11:57] Hi elkauka! [11:58] MSG: Ping timeout: 252 seconds [12:19] Have to go, too! [12:19] bye! [12:19] Part: Jens left #corewars [12:47] MSG: Ping timeout: 252 seconds [13:08] Mizcu: ! [13:08] hi [13:17] Mizcu: 2nd gig is finished [13:57] Join: michal joined #corewars [14:15] Join: elkauka joined #corewars [14:39] MSG: Quit: Quit now, you'll never make it. If you disregard this advice, you'll be halfway there. [15:14] Join: Jens joined #corewars [15:28] Join: brx joined #corewars [15:32] MSG: Ping timeout: 252 seconds [15:38] Join: Mike joined #corewars [15:39] Hi [15:45] Part: Mike left #corewars [16:47] Join: sf joined #corewars [16:47] Hi sf! [16:48] Hi Jens [16:50] What did Fizmo man here? [16:50] Still got a request from a php magazine. They would like to have a 6-month trial version of Corewar [16:50] *mean [16:51] Has Bvowk gone away? [16:52] Join: willvarfa joined #corewars [16:52] evening all [16:52] Hi Will [16:52] http://corewar.atspace.com/ [16:53] sf: Don't know, what Fizmo meant, but I'll ask him ;-) [16:53] Hi willvarfa! [16:54] sf: Nice page. Did you send your successful warriors to Koenigstuhl? [16:55] I think so. If they are not I will do. [16:55] Thanks Jens. No real content yet though [16:55] But at least a link to my page ;-) [16:56] Your page is the most frequently updated I think [16:56] No, koth.org and SAL is most often updated [16:57] Automated updates though [16:58] updates are updates [17:09] Just working on an update for both now [17:09] both? [17:09] two different updates [17:11] Did you change Dark Energy a lot? I mean the version, that you've just sent to the hill. [17:11] No, just testing the original. [17:11] blah, optimax-scores sucked, back to all-at-one optimizing -_- [17:12] Mizcu: What is your current best score? [17:12] 123, 5 hours [17:12] hmm [17:12] but i did first without attack, then opted attack [17:13] Did you test a qscan-version? [17:13] I should get higher even without qscan [17:14] * willvarfa waves [17:14] MSG: Quit: Chatzilla 0.9.68.5 [Firefox 1.0.4/20050512] [17:15] Mizcu: I've made the experience, that often a little bit tweaking by hand until you have an acceptable score is often better than optimaxing hundreds of hours. [17:15] Join: sascha joined #corewars [17:15] The score surface for Tiny Stone vs. Tiny Black Knight is now online. [17:15] Hi Sascha! [17:15] Hi sascha [17:16] Hi, email for you on the way, i'll be back in 5 minutes... [17:16] Hi sf [17:16] ok [17:16] sf: Nice. Now it has 10 points more. A little tweaking and it might enter the hill [17:18] I am working on that now [17:21] Argh! [17:21] ? [17:21] I had the code on the clipboard then I copied something else :( [17:22] Does WinXP have a clipboard viewer that lets you go back to previously copied stuff? [17:23] You ask the wrong person, I haven't use Win for years [17:24] no to previous [17:25] Strange. Win 3.11 used to [17:29] sf: That was close [17:29] Join: elkauka joined #corewars [17:29] Hi elkauka [17:30] hi [17:30] I see vampires are back! :) [17:30] yeah! [17:30] Hi elkauka! Nice entry, altough it dropped after the hill was dat-ed [17:31] place #12 now [17:31] Hi elkauke, finally your scanner/vamp hybrid fizmo told me some time ago seems to work...Congrat [17:31] MSG: Read error: Connection reset by peer [17:31] Join: michal joined #corewars [17:32] it has to get older than 3 to be my longest living entrie on 94nop...lets hope it'll do well [17:33] It will, I think [17:33] My current maxage is 19 [17:33] * sf once had 24 [17:34] sf: You are my hero ;-) [17:34] The code is mine, but the constants are Fizmos, so I don't know if it counts [17:35] my overall best was 69 on 94m [17:35] * sf had a good run on 94m too! [17:36] age 261! [17:46] * elkauka waves [17:46] MSG: Quit: Quit now, you'll never make it. If you disregard this advice, you'll be halfway there. [17:49] The 94 hill is mostly warriors which tie [18:36] MSG: Quit: Bersirc 2.2: All the original sexiness of Bersirc, open to the world. [ http://www.bersirc.org/ - Open Source I [19:51] MSG: [20:08] * Jens waves [20:08] Part: Jens left #corewars [20:08] time to go... [20:08] Part: sascha left #corewars