[00:01] Join: eva`5|fur joined #corewars [00:21] eh? [01:12] hrm. [01:18] Join: Waknuk joined #corewars [01:20] Part: Waknuk left #corewars [03:09] Well, I'm about to kill a legend. [03:17] Hrm. [03:17] I don't really want to kill Petro. [03:24] Join: ares_ joined #corewars [03:25] hey ares. [03:26] * bvowk prods ares [03:26] MSG: Ping timeout: 255 seconds [03:39] Join: ares joined #corewars [03:42] MSG: Ping timeout: 255 seconds [04:00] MSG: Quit: 15,1[ 4,1HIX-Script v2.2 15,1]2,1:::15,1[ 4,1Download from 15,1]2,1:::15,1[ 4,1www.rupertonline.ca/hix/ [10:29] Join: Fluffy joined #corewars [10:29] :) [10:34] aye [10:35] Mizcu: You should send another warrior to nano. Bvowky must be stopped :) [10:36] darnit, where will i get the processor -power to run both evo and my newest game at the same time [10:36] why don't you use an old computer for evolving? [10:36] i aint got no oldie one to use [10:37] :( [10:37] Ask bvowk to use his computers ;) [10:38] great, my ramdisk has decided not to work properly [10:38] Why do you need a ramdisk? [10:38] less use of HD [10:39] hmm [10:39] i have Yace -settings that use HD more than the Metcalf-settings [10:39] But HDs were created to be killed ;) [10:39] maybe yours [10:39] Join: sf_ghoul joined #corewars [10:39] Hello [10:40] I have flu :( [10:40] So far I've only killed 3 HDs in 15 years! [10:40] Hi sf_ghoul! [10:40] Hi Fluff [10:40] Just stopped by to catch up on the latest nano stuff [10:41] Zul won't like that I've killed hi Petro "I'm old" Warrior :( [10:42] *his [10:42] :( [10:42] Right here goes, clicking on the link to nano [10:43] brb, boot+chkdsk [10:43] John won't like this :( [10:43] sf_ghoul: But 1787 is a nice age for a nano warrior and I'm quite sure that we'll soon have even older warriors. [10:43] another nanohof required [10:44] yes, I really hope that he's automated the creation of the HoF [10:45] no [10:45] Oh, have you read r.g.cw! [10:46] which post do you mean? [10:46] inversed [10:47] I'll post a link but I'm not commenting [10:47] http://groups.google.com/group/rec.games.corewar/msg/5f8dab53a5a51d9e [10:48] hmm [10:49] I don't think that inversed is (entirely) right. [10:49] Yes, spamming the hill with thousands of warriors takes a little bit of the fun, but so far bvowk hasn't succeed in wiping the hill! [10:50] So inversed can still create a warrior, which easily kills all the evolved stuff [10:51] ... and I don't see that bvowk will do the same on tiny-*/94* [10:51] Even he doesn't have that much computrons :) [10:57] back [10:57] Mizcu: Does everything work now? [10:57] yup [10:57] :) [10:57] some voodoo-error with the ramdisk-driver [10:58] thar, yace running' [10:59] ah, dinner set to marinate [10:59] Mmmm... dinner :) [10:59] Hmmm ... yace ;) [10:59] marinade, what ever the sp? [11:05] 51 42.4 48.5 9.1 Bombus Nasutus S.Fernandes 136.3 490 [11:05] * sf_ghoul makes a voodoo figure of bvowk [11:06] * Fluffy gives sf some needles [11:06] *gives [11:08] Thanks fluffy [11:08] Don't hesitate to ask if you need more ;) [11:09] Topic Change: Mizcu sets topic: http://corewar.co.uk - http://www.corewar.info - official "hate Bvowk" day today! [11:10] aaaand another bunch of settings tweaked to prevent mov-towers in yace.. [11:10] Mizcu: You should op all bvowk-haters. Then we can really kick him ;) [11:11] maybe i should send the piesoup to John-hill [11:11] yes, do that :) [11:12] Yes :) [11:12] At least it has a proper name ;) [11:14] Fluff/Sf, either of you run nano-evolvers at the moment? [11:14] no, I run an optimizer [11:17] no, I write stuff by hand and sometime run a few hundred benchmarks with different constants [11:19] bvowk seems to be having a problem killing those 1000+ warriors [11:20] yes :) [11:20] Which is strange because I can easily write something with 70%+ wins against any of them [11:21] Maybe it isn't that easy [11:22] No, I think there are some things which are impossible to evolve [11:22] Because there's no route to them [11:22] Like little islands in the solution space [11:23] Join: fiveop joined #corewars [11:23] Hi fiveop [11:23] Hi fiveop [11:23] well buck in a fucket, it was supposed to weaken mov-towers, and now i have spl mov mov djn -warrior brewing O_o [11:24] I used to run what I thought was an evolver but Fluff and bvowky said it wasn't [11:29] hmm ... I can't remember that [11:53] Zul posted to r.g.cw too [11:58] MSG: Ping timeout: 255 seconds [12:04] *sigh* [12:04] 163 points, but another movtower [12:06] why don't you make restrictions about the type of warrior to be "evolved"? [12:08] needs more coding than i am capable of [12:09] they become too easily competitive [12:09] hrm [12:37] * Fluffy waves [12:37] MSG: Quit: fluffy.i < 1, # 42 [15:10] fluffy! [15:10] why did you kill petro! [15:10] I was gunna let him live :( [16:04] hrm. [16:54] Join: sf_ghoul joined #corewars [17:13] hey sf [17:13] I'm about to commute.. stick around? [17:14] stick? [17:14] * bvowk glues sf to the floor [17:14] * sf_ghoul skewers bvowk on a metal stick! [17:14] ow! [17:14] thanks for the idea [17:14] girls who drink beer?! [17:14] and I get called out all the time for dumb names?! [17:14] ;) [17:14] Have you finished making the nano hill look ugly now? [17:15] yeah, I feel bad for zul [17:15] I shouldn't have left petro at the bottom while I went to sleep on it. [17:15] at least this dumb name is understood by 99% of people [17:15] well, mine is a unique identifier for a single warrior :P [17:15] Unique? [17:16] So you've multiple submitted several warriors? [17:17] mine are unique identifiers too [17:18] aw damn. [17:18] I really wasn't paying attention.. [17:18] yeah, there's a few dupes.. [17:20] * bvowk commutes. [17:28] Join: Roy joined #corewars [17:29] Hi :) [17:33] Hi Roy [17:35] * sf_ghoul votes max-amount = 33% [17:37] Maybe when you send a warrior with a bvowk-like name you should put ;author bvowk [17:37] Then he'll be confused about which ones to target [17:40] hehe, I knew it was rubbish :P [17:41] (the warrior would never make the hill) [17:41] But I'm testing some new stuff, to beat bvowk ;-) [17:41] :) [17:43] But I'm not so good at nano, except starting yace and waiting ;-) But the problem is we don't have a testset with bvowks warriors [17:44] Against my current benchmark (koenigstuhl) I get a lot of points, but well... that doesn't do well on the hill [17:44] So I think its time for myself to create a benchmark with koenigstuhl-beaters [17:45] One of those B1F3 warriors is on Koenigstuhl [17:46] And there's stuff on infinano which hasn't made it to Koen yet [17:47] Join: Fluffy joined #corewars [17:47] Hmm well, we'll see :) [17:47] Hi Fluf [17:47] * Fluffy kicks Roy [17:47] Hi Fluff [17:47] Hi everybody [17:47] * sf_ghoul kicks Fluff [17:47] * Fluffy kicks sf_ghoul [17:47] not quick enough ;) [17:48] * Roy kicks sf_ghoul and fluffy and wins! [17:48] :( [17:48] Aww.. wins is hurt bad.. :( [17:48] Why are we kicking each other? [17:48] * Fluffy increases global warming by a factor of 100 and waits for the Netherlands to be flooded within seconds. [17:49] We should be kicking bvowky! [17:49] * sf_ghoul kicks bvowk [17:49] Because it seems as if nobody can spell my name correctly [17:49] greetings humans. [17:49] ow. [17:49] how's it going floppy.. [17:50] why'd you kill petro :( [17:50] It was you, bvowk, who pushed it that far down! [17:50] I've only pushed it off, but regretted it nonetheless [17:50] its true, I shouldn't have left him there.. [17:50] :( [17:51] roy: you want some warriors? [17:51] Yes pushed it down so far the next submission was bound to push it off. [17:51] I think I'm done with my crushing of nano.. [17:51] Then left it to someone else [17:51] well, I pushed it down so far and then I felt too guilty to kill it.. [17:51] so I decided to sleep on it.. [17:52] Didn't you feel guilty about any of the others? [17:52] bvowk, just publish some of the best ones then ;-) [17:52] Especially when you was ;killing your warriors which gave them too many points to bring them down :P [17:52] not that guilty [17:53] Bvowk, you should publish 34cf7eac-40fea321-52d3 [17:53] I'm going to publish everything including the evolver here shortly.. [17:53] And 8c09fc1a-852cb1ba-575a [17:53] and I'm planning on restoring the hill to a balance again.. [17:53] (its pretty wonky right now) [17:53] 10:51AM up 1000 days, 5:44, 0 users, load averages: 0.10, 0.09, 0.08 [17:53] (w00t!) [17:53] And 34cf7eac-b6b64199-b366 [17:54] What do you mean restoring the hill to a balance? :/ [17:54] I don't like the sound of that [17:54] its pretty much one strategy right now [17:55] and kill my dupes [17:55] You should pick out the best warriors (i.e. longest lived or with highest benchmark scores) and send to Koen [17:55] hey, does zul = inversed? [17:55] Why do you ask that? [17:55] We'll restore it for you if you just publish ;-) [17:55] just curious.. inversed snarled at me on r.gc [17:56] While we're asking about alt names, does fizmo = metcalf = kline [17:56] heh [17:56] sf = metcalf! [17:56] aliens from the planet corewar. [17:56] You're just going by code quality there [17:57] Because we can write stuff which you're evolver can't find [17:57] Like little islands in the search space [17:57] well, it can find it eventually. [17:57] You can include inversed in that too ;) [17:57] sf = metcalf = inversed [17:57] I found some interesting things I didn't know existed when I was wandering around looking for code. [17:58] like that whole troop of impy warriors [17:58] You won't find something when it so delicate, the slightest changes completely breaks it [17:58] There's no path to it through the warrior space, the evolver will always take a different route. [17:59] * sf_ghoul has impy warriors bvowk could never dream of :P [17:59] gnik has impy warriors no-one else could ever dream of! [17:59] * sf_ghoul wonders about rumplestiltskin [18:00] sf: I thought the same thing about scanners.. [18:00] but I've got hsa's coming out of the woodwork [18:00] Ah, scanners are no prob [18:01] Because they're not so delicate [18:01] You can change all sorts of stuff and they're still effective [18:11] bvowk: What exactly do you mean by balancing the hill? Have as many different strategies on the hill as possible? [18:11] We can do that bit [18:12] I was going to take all my warriors, and then submit the one that beat them the best... and then take the new hill and repeat [18:12] hmm [18:12] hey fluffy, you mind if there are a few of your warriors in my nanopit and I publish it? [18:12] Or you could leave it to naturally balance [18:12] sf: it might take a long time [18:13] as long as it doesn't contain White Moon, I'm ok with that, bvowk [18:13] gack. [18:13] it might.. [18:13] damn white moon. [18:13] Then you have to remove it :) [18:13] bvowk: you should pick out the best warriors and submit to Koen [18:13] sf: will do. [18:13] Christoph probably won't take the lot! [18:13] hehe ... John's updated the HoF ... and it is already out of date ;) [18:13] you don't think he wants 522,000 nano warriors? [18:14] No [18:14] bvowk: How about the best 10 warriors? [18:14] I don't submit the 1000 or so warriors I generate when I optimize something [18:14] fluff: I was going to pick the best ones out and publish them.. [18:14] but I thought people might like the nanopit for evovling [18:14] I would suggest sending the top 5 HoF warriors [18:15] sf: but I've *REALLY* hit on some good stuff, and it just keeps getting churned out by the thousand [18:16] I wouldn'y suggest submitting all HoF warriors since there are already 70+ bvowky warriors in the HoF [18:17] And then maybe submit another 5 or so warriors which you feel are adequately different from each other and the warriors currently on Koen [18:17] sorry.. when you're good you're good ;) [18:17] * bvowk ducks. [18:17] Or have something interesting to show. [18:17] Christoph wouldn't take many submissions anyway, they'd just go in his unpublished directory [18:18] hof is the closest to being up-to-date it's been for a week or so [18:26] Ok, gotta go! [18:26] MSG: Quit: dat [18:30] MSG: Ping timeout: 255 seconds [18:37] lol [18:37] + N 1308 Feb 5 KOTH (9826) SAL "SorrySF!" challenges Nano hill [18:37] it is a nice way to communicate [18:38] ... and if I were you, sf, I'd send a "* sf kicks Roy"-type of warrior ;) [18:41] Roy didn't do it on purpose :) [18:41] Anyway, sending a better 'dumb' [18:42] * bvowk goes for some lunch.. [18:42] Yes, I like your warriors! White Moon isn't far from koth now :) [18:42] and then I should be ready to publish the evolver and "stuff" [18:42] "Won't do it again!" by "Roy van Rijn" [18:43] although, maybe I should push off the damn white moon before I quit being a nano abuser. [18:43] :) [18:43] or I'll never hear the end of it. [18:43] * Fluffy yawns [18:43] what do you think sf? [18:46] bvowk: No one will support your evil doings ;-) [18:47] I think you should leave White Moon alone [18:48] and let it drop off naturally [18:48] naturally? [18:49] Not bad though, a warrior submitted 3 weeks ago is about to enter the list of the oldest warriors ever! [18:49] :) [18:50] naturally = not manipulated off the hill with specifically chosen submissions and ;kills [18:50] that's hard to distinguish from un-natural causes ;) [18:51] It is when you can see the ;kills and check ;) [18:52] It's a bit suspect when you're warriors in the lower half of the hill, then suddenly the 5 or so warriors which give you most points are ;killed [18:52] sf_ghoul: Don't give bvowk ideas! [18:53] Fluff: I'm not, it's what he's been doing [18:53] Join: Core_old joined #corewars [18:53] Hi old Core! [18:53] hi [18:55] Hi core_old [18:57] The evidence! http://sal.math.ualberta.ca/hill.php?key=nano&t=6283&how=late&sortby=score&dir=desc [18:58] Here bvowky ;killed the three warriors which gave most points to White Moon [18:59] Guilty! :P [19:00] *grmpf* [19:00] If you look through it happens when most of the old warriors are pushed off [19:00] bvowk: You really shouldn't leave that much evidence around! [19:01] Luckily this was a failed attempt ;) [19:01] What happened to nenad's tournament? [19:01] I have no idea [19:01] And while I'm asking, what happened to datagram's tournament? [19:02] no idea, too :(/ [19:02] * :( [19:03] And while I'm asking, what happened to the next chapter of the qscan tutorial? :P [19:03] I'll be working on it until it is ready [19:03] :) [19:09] MSG: Quit: Lost terminal [19:09] * Fluffy waves [19:09] MSG: Quit: fluffy.i < 1, # 42 [19:11] * sf_ghoul waves [19:19] MSG: Quit: Trillian (http://www.ceruleanstudios.com [19:31] MSG: Quit: I was using TinyIRC! Visit http://www.tinyirc.net/ for more information. [19:39] hrm. [19:43] http://sal.math.ualberta.ca/hill.php?key=nano&t=6757&how=late&sortby=score&dir=desc [19:43] 8 point lead ;) [19:55] Join: Core_old joined #corewars [19:57] hi core [20:57] Join: fiveop joined #corewars [21:24] Join: Fluffy joined #corewars [21:24] Hi fluffy [21:24] STRIKE! [21:24] bvowk: Look at the nano-Koenigstuhl! [21:25] oh dear [21:25] I'm never going to hear the end of this am I? [21:25] Nope :) [21:25] 6 of the top 10 .. [21:25] Do you prefer to be teased or taunted? [21:26] not shabby [21:27] Why did you kill those 3 or 4 warriors, bvowk? [21:27] they were dupes [21:27] figured they were fairly young (in the nano abuse sense of young), so why not. [21:27] But you still have dupes on the hill ( 90ae3a01-4d7b2a5d-ea135101.rc) [21:28] hey fluffy.. wanna go halfsies on ebay #: 300076435547 [21:28] I missed one? [21:29] alright, I'll kill it next sub. [21:30] That SunFire is a little bit too expensive :) [21:30] ... and probably too noisy :) [21:30] thats why I asked if you wanted to split it with me ;) [21:30] ... and I don't need those 72 GiB RAM (at the moment) [21:30] 72 processors :) [21:30] I think they're dual core processors too [21:31] so 144 processors.. thats only a half gig of ram each.. [21:31] ... and who will pay the power bill? [21:31] (we'd need an upgrade!) [21:31] see any other dupes fluffy? [21:32] no [21:32] k [21:33] So ... who would pay the power bill? [21:34] you? [21:34] HWMNBM? [21:34] you worry too much [21:34] once you've got giant cool toys.. people go out of their way to help you play with them. [21:34] no, I don't even have the money to buy it [21:34] you're not made of money!? [21:35] no, mostly out of water [21:35] how disappointing. [21:36] 270083790996 [21:36] how about that one ^ [21:36] Let's assume for a moment, that I've told you to buy that SunFire. What would you with it? [21:36] 96GB of ram, 24 Processors [21:36] whatever came to mind.. [21:37] recmath contests, corewars, computational chemistry when the mood struck me.. [21:37] looking for a 16 hint sudoku? [21:37] would any answer satisfy you? [21:37] yes, that would be a nice idea [21:38] But wouldn't it be cheaper to rent some 50 servers or so from $your_favourite_hosting_company? [21:39] likely not.. [21:39] and there are some problems you can't really work on without lots of tightly coupled processors and memory [21:39] (we'd need alot of disk too...) [21:40] but that 16-hint-sudoku can be implemented without connections between the computers :) [21:40] thats true.. [21:40] I've already generated many many GB of sudoku. [21:40] (hence, www.sudokutron.com) [21:40] how many 17-hint-sudokus do you have? [21:41] just a few [21:41] few = ? [21:41] I haven't checked the last 60GB or so.. but I had <50 17 hint sudoku [21:43] sorry, there's only 40ish GB waiting for me to process [21:44] Maybe you can check, if you've found a new 17-hint-sudoku [21:44] (compare to http://people.csse.uwa.edu.au/gordon/sudokumin.php ) [21:44] no, I hadn't found any new ones last I checked [21:45] Did you check for <16-hint-sudokus? [21:46] uh.. yeah! [21:46] since that was the point! [21:46] well, I haven't checked the last 40ish GB waiting for checkin.. [21:46] and do you still use that slow lithp-sudoku-finder or a fast C-version? [21:48] I use a fast C version [21:48] and then python to twiddle the bits [21:51] I wonder how fast it is [21:51] hey fluffy.. how about ebay #: [21:51] 110085882955 [21:52] its not my code, its Glen Fowlers code, its awfully quick. [21:53] HWMNBM and pk had designs to write a parallel implementation that was insanely fast and worked on multiple puzzles with SSE instructions.. but that didn't make it far. [22:04] where'd you get off to fluffy? [22:04] about that ebay thing: I'd need a rack [22:04] (I had an important telephone call) [22:05] g'night all [22:05] I've got plenty of racks [22:05] about sudoku: Why did they stop developing that sse code? The world needs it ;) [22:05] Nighty, Core_old! [22:05] MSG: Quit: Trillian (http://www.ceruleanstudios.com [22:06] because the topic in the channel we hang out in is "Procrastinators'R'Us" [22:07] huh? [22:08] ok, found a dictionary :) [22:08] now I get it [22:09] Could you please check how many 18-hint-sudokus you have (which cannot be reduced to 17-hint-sudokus)? [22:09] Forget that [22:13] 180082741628 how about that one fluffy? [22:14] I need more than 16 GiB of RAM ;) [22:15] heh [22:15] for what? :) [22:15] Don't know! Downloading The Intertubes? [22:16] you don't need that much memory for that.. [22:16] even running a dozen threads for each processor. [22:17] we'd need bigger disk tho.. [22:17] But I want to keep it all in memory [22:17] we'd need a fibrechannel router and a good couple racks of disk [22:17] ... and someone, who pay the bills [22:17] like I said.. don't worry about that ;) [22:19] Let's make the following. We simply wait until one US-$ is worth about 0.001 Euro. They I'll buy you that Origin 3000, ok? [22:19] *They->Then [22:19] should only take a couple weeks then.. [22:20] Maybe we should start some spam attack to make it drop faster :) [22:22] Maybe I could afford then the ASCI Thor's Hammer [22:24] Join: Roy joined #corewars [22:24] hi roy [22:25] wanna chip in for a giant computer? [22:25] MSG: Quit: Trillian (http://www.ceruleanstudios.com [22:25] fluffys too cheap [22:25] Hi Roy [22:25] Join: ares joined #corewars [22:25] Roy: We only have to wait until the US$ crashes :) [22:30] It is interesting that some warriors don't like imps [22:30] bvowk: Two questions! [22:31] one: Why is SAL named SAL? [22:31] 2010 [22:31] two: Have you/HWMNBN ever thought about redesigning the SAL-pages? [22:31] (with some niftly css-code) [22:31] *nifty [22:32] I doubt HWMNBN would be terribly interested.. [22:32] although, I think an auto-updating HoF would be cool. [22:33]