[03:16] Join: brx` joined #corewars [03:18] MSG: Read error: Operation timed out [04:10] Nick Change: brx` changed nick to brx [04:10] MSG: Quit: Leaving [04:15] Join: asw joined #corewars [05:32] Join: Mizcu joined #corewars [06:21] Join: fiveop joined #corewars [07:42] /off1 fiveop time university [07:42] /off1 fiveop time university [07:42] /off1 32 [07:42] .oO ( fiveop ... time ... university ? ) [07:43] .oO ( offby1 ... uuh ... ripped ... function ? ) [07:43] .oO ( good ... morning ) [07:43] .oO ( ... uh ... corewars ? ) [07:45] .oO ( brx ... uuh ... eve ... uh ... anne ? ) [07:45] .oO ( ... evocoreniverse ? ) [07:46] playing with a speechbot again, brx? [07:46] speechbot? [07:46] it's a new erc command I just implemented :) [07:46] well, chat/crap -bot [07:46] .oO ( playing ... speechbot ... ) [07:46] those are thinking bubbles :) [07:46] -_- [08:21] Join: Roy joined #corewars [08:23] hi [08:24] .oO ( roy ? ) [08:29] sigh.. [08:30] * brx is very difficult.. [08:30] ask fop. [08:43] you are difficult? :) [08:44] fop: how is brx difficult?! [09:00] * brx has written 322 lines of scheme and anne can already say "hello" upon joining a channel [09:02] heh, and has think bubbles ;) [09:02] no [09:02] you and mizcu should learn about the difference between a simple irc client command and a bot [09:03] .oO ( ... uh ... think ... bubbles ) [09:03] .oO ( ... roy ? ) [09:04] actually, much of what is written already is amounting to a critical mass [09:05] which, when it is reached will consequently polish the bot to perfection [09:05] it's like building a bridge by first building a tower and then pushing it over ;) [09:06] once you can query hills for example, she will already be able to list warriors which fulfill certain conditions as in [09:07] "get 94nop author roy age >= 20 [09:07] " [09:08] or "get 94b author andreas place <= 10" [09:09] she will be able to query all hills periodically, and announce new submissions, but only for more active hills and require manual prodding to update a less frequently visited hill [09:09] at that point she will also have all the content parsing mechanisms she needs for all other commands to build on [09:10] and thus the tower is rising. [09:10] most importantly, she is dead sexy. [09:10] :P [09:31] brx: ohhh anne.. what are you wearing today? [09:38] spacesuit and blue hair ?;p [09:39] miz has strange fantasies when thinking about a sexy women.. [09:40] Roy: she is called lafiel [09:40] * brx shows his nerdiness [09:40] Those pointy ears are lethal! [09:40] nooooo, you haven't seen it, have you? [09:44] Roy: whenever you have a few minutes, could you please send anne some reasonable pseudo commands? [09:44] something like my get examples above [09:45] not necessarily hill related of course [09:45] sure [09:45] try to keep the get pattern however [09:45] verb etc etc [09:45] okido [09:45] thank you :) [09:45] Mizcu: you too ;) [09:45] "get me a soda bitch" [09:47] for a split second I was about to do something very offensive. [09:47] ..? :) [09:48] next time, just say "no". [09:49] Heh I want to help, just kidding with you brx :P [09:59] Join: manutara joined #corewars [09:59] hi [10:00] hi [10:00] bonjour! [10:00] Join: fiveop_ joined #corewars [10:02] MSG: Ping timeout: 252 seconds [10:03] Manutara, you new to corewars? [10:03] yes [10:04] pMARS run on XP?? [10:04] yes [10:04] It can run on some xp machines [10:05] but you have to do it from command prompt [10:05] ok, ill try [10:05] i suggest corewin; http://www.geocities.com/corewin2/ [10:05] Well, miz, I tried it on my machine but it didn't work (only a old pmars server version) [10:08] corewin works well, thx... [10:16] Join: Fluffy joined #corewars [10:16] :) [10:17] :) [10:17] brx: 737 patches completed. If you continue that way, you'll soon have 1000 patches :) [10:27] you are corewar gurus?? [10:27] What is corewar? [10:28] game?? [10:28] hehe ... no, I'm definitely not a cw guru, but I like the game [10:29] Hehe, don't take him serieus, never take somebody serieus who calls himself Fluffy [10:29] Roy: Now you got the point :) [10:29] We all like the game, there is only one cw-guru here, and thats John Metcalf :-) he comes around sometimes [10:29] roy +1 fluffy 0 [10:30] manutara: Do you have any questions? [10:30] not now [10:30] ok [10:30] im just looking haw does it work [10:30] maybe later, i will have many questions... [10:31] Roy: How can you call John a cw-guru. He's named one of his warriors "T.A.T.U girls rule"! [10:31] lol, forgot about that.. ok, there aren't any guru's anymore.. [10:31] ;-) [10:31] manutara: Then you probably already know http://vyznev.net/corewar/guide.html (The beginner's guide to Redcode) [10:32] * Roy notices the game itself is getting less intrested and the side-programs are.. [10:32] i am just there [10:32] Almost nobody is writing warriors, I think more people are writing evolvers/analysers/bots etc :P [10:33] I just wait for the "opportune moment" to release HFB II [10:34] You made that one..? sweet [10:34] * Roy is in the third-phase, Hulla3 and Devil3 [10:35] By the way how old did HullabaTwo get? [10:42] Err 400+ something like that [10:43] No idea really [10:43] hmm ... I wish I had an 400+ warrior [10:43] Make one then..! [10:44] yes, yes [10:44] I can't work under so much pressure ;-) [10:44] Just take some sugar and spice, and more things nice, mix them together, put in a blender and sprinkle it with code [10:44] sugar and spice? Are you pregnant, Roy? [10:44] ;-) [10:45] Nooo, I'm not in the baby-phase, we are in the moving-phase now, about 10 people I know have bought a new home [10:45] * Roy not included still with his parents [10:46] hmm ... paperpone paper aren't that good then [10:46] *papers [10:47] I like it ;-) [10:47] 1 Hullab3loo Roy van Rijn 36 25 39 146.19 10 Maybe the last version does better.. [10:47] 2 Monster_Alien_Grunt inversed 32 24 44 139.49 96 [10:47] almost 7 point gap there [10:47] * Fluffy sighs and remembers HFB's good times ... [10:48] Well, just one more anti-imp oneshot and hulla3 is down again [10:48] Just make a small scanner ;-) [10:48] Well yes, when Hulla2 was pushed of it showed that it isn't invincible [10:49] Hulla1 got a ;kill command afterall [10:49] I still don't understand, why you've killed Hullabaloo. It faired well [10:49] Not as well as Hulla2, and that one not as well as 3 ;-) [10:50] Join: Judodudo joined #corewars [10:50] hello [10:50] hi [10:52] Roy: Does Hullab3loo still use that Moore-style paper? [10:56] roy you are not serious [10:57] * Roy not serious? :| [10:58] ("Roy" "Roy" (get newsgroup |CoreWarrior| 75)) [10:58] ("Roy" "Roy" (get brx 019 catya)) [10:58] ("Roy" "Roy" (get brx 019 subject eve data > 01 01 2006)) [10:58] ("Roy" "Roy" (get brx 019 subject evolver data < 01 01 2006)) [10:58] ("Roy" "Roy" (get brx 019 subject evolver date > 12 02 2006)) [10:58] *I* can hardly parse that [10:58] :-( I was serious :( [10:59] Retrieving blog messages and newsgroup.. you where going to implement that right? [11:00] mmmh [11:01] * brx shivers [11:01] the "-" got removed, and I don't know where there are "|" around CoreWarrior [11:01] yes I know [11:01] That makes it a bit more normal right.. :( [11:01] I just wanted to check if any of your propositions conflicts with my tokenizing [11:01] but that is all okay [11:01] " not necessarily hill related of course" [11:01] yes [11:02] just the idea to.. I mean... newsgroup :(( [11:02] rss, okay... newsgroup... [11:02] newsgroup hurts :( [11:02] The newsgroup has rss...! [11:02] oh [11:02] well then okay, I will check if I can implement a rss reader. [11:02] actually [11:03] I don't have to. there is an egg [11:03] http://groups.google.com/group/rec.games.corewar/feed/rss_v2_0_msgs.xml [11:03] http://www.call-with-current-continuation.org/eggs/rss.html [11:03] wait this is better: [11:03] http://groups.google.com/group/rec.games.corewar/feeds [11:03] no need to hurry roy.. [11:03] the rss thing has lowest priority right now :) [11:04] hehe [11:04] but it will not be hard [11:04] I accept it to fall in line when I have everything else [11:04] remember the tower? [11:05] the fun thing about the tower is that each new component I put on top is considerably smaller than the lower levels. [11:05] by the time I get to rss, the functionality you requested will make up a tiny bit of code [11:07] sweet :) [11:08] MSG: Quit: Chatzilla 0.9.70 [Firefox 1.5.0.1/2006011112] [11:19] Roy: only two points to place 2 [11:19] err .... 1.24 [11:19] Forget it, I'm no longer able to do math [11:22] Hehe, stupid-mode again? it changes everytime, but I don't really mind now [11:22] hehe ... Wait until GoF is on the hill [11:22] I had some time I didn't have a warrior on the hill, and that was long ago...I thought I lost the ability to redcode :( [11:23] Hrm, my english is bad, I'll have another try: [11:23] "Last week I had no warriors on the hill, that was a first since a long time" [11:33] * brx has been streamripping this di breaks stream for hours now hoping that they play that special break again [11:33] * brx has heard it twice already [11:34] btw roy, the rss reader querying.. although it is easy to implement I would not like to do it [11:35] not useful. [11:36] Fluffy: I am installing linux on my parent's machine right now. their hd broke. [11:36] brx: Good luck. [11:37] err ... did it break while you were installing Linux? [11:39] Fluffy: no, I made them believe it was a virus and that was windows' fault :) [11:39] Fluffy: to be able to make them switch [11:40] hmm [11:43] hmm ... why does it always take so long to get an answer from koth.org back? :-( [11:44] Fluffy: it's probably running fights every 5 minutes or similar [11:45] Yes, I know and unfortunately I always send my warriors after that 5-min-cycle has finished [11:46] Fluffy: then my decision to update important hills every 4 minutes for koth was correct. [11:46] brx banzaaaaaaaai! [11:47] hehe ... now I have a second HFB :) [11:49] Nah, its not as powerfull as HFB1 [11:50] Roy: At the moment GoF uses only the "standard" constants like 2000, 6000 for boot. [11:50] Now it is time to optimize and tweak [11:50] Cool, thats more like it [11:51] Hmm ... it will probably be last, after I've DATed the hill [11:51] No it won't, it rise some more [11:52] we'll see [11:55] Time to kill GoF [11:58] #scheme is talking about animes :) [12:01] One slot on 94nop is free :) [12:04] The unoptimized GoF has a score of about 128 against fsh94nop-0.3 [12:18] Join: anne joined #corewars [12:18] grr [12:18] time? [12:19] Any new features, brx? [12:19] Fluffy: many [12:19] and? [12:19] Fluffy: still no interface to irc however :) [12:19] hehe [12:19] I could have hacked together something for that by now [12:19] it's just... I want it clean.. [12:20] also, non emacs users are lost sheep [12:20] It should've only one feature ... to idle around here. Nothing more is required for this channel ;-) [12:21] [10:04] actually, much of what is written already is amounting to a [12:21] critical mass [12:21] [10:05] which, when it is reached will consequently polish the bot to [12:21] perfection [12:21] [10:05] it's like building a bridge by first building a tower and then [12:21] pushing it over ;) [12:21] Fluffy: it can already do that! :) [12:22] I don't need an irc client either anymore [12:22] I can just use anne [12:22] MSG: Ping timeout: 252 seconds [12:22] yeeees? [12:23] * Fluffy kicks anne [12:23] lol... you crashed her [12:23] Warning (#): (string-parse-start+end) bad argument type - not a string: # [12:23] hehe [12:23] (every action kicks her currently) [12:24] stress test ;-) [12:24] what I don't know is why she is still running [12:24] anne, are you there? [12:24] the handlers do not trigger at any rate :) [12:25] I could extend the irc lib to allow me to restart those threads.. but I am lazy [12:25] MSG: Quit: I hate you Fluffy [12:28] Join: anne joined #corewars [12:28] * brx dances [12:28] * Roy boogies [12:29] * brx boggies too [12:29] da [12:29] she crashed [12:29] MSG: Client Quit [12:29] Join: anne joined #corewars [12:29] * brx hey [12:29] oh now I know why :) [12:29] MSG: Client Quit [12:29] Join: anne joined #corewars [12:30] * brx dances [12:30] * brx dances [12:30] * anne juggles grease [12:30] good [12:30] I just have to restart the message loop when it happens [12:31] dcc chat won't crash [12:31] * brx laughs [12:31] * anne juggles grease [12:41] * anne juggles grease [12:48] * anne juggles grease [13:13] Join: Mizcu joined #corewars [13:13] Hi Mizcu [13:14] hi Fluff [13:20] hi mizzi [13:20] hi bmx [13:20] hi toy [13:20] brx: No, it should be "h oy" [13:21] Fluffy: it should not, tuffy [13:21] * Fluffy has started his optimizer for GoF [13:21] * anne juggles grease [13:22] ahaha [13:23] how is this dancehall artist called again? jean paul? [13:23] aaah sean paul [13:24] Sean Paul is in the house!! Boom-chika-boom-chika-boom boom [13:26] .oO ( in ... the ... house ... uuh ... boom ... uh ... chika ... boom ... chika ... boom ... boom ? ) [13:26] I actualy like that music too.. don't have a cd or anything, but its nice when clubbing [13:26] Roy: I am recording the di break stream because I heard a *very* good break with him [13:27] What is that breaks stuff..? [13:27] well, similar to remixes I think [13:27] breakbeats [13:27] but mizzi is the better dizzy yo [13:28] djizzy mizzi [13:28] yow [13:28] I left my WALLET in the BATHROOM!! [13:28] shruck.. still I like the Artic Monkeys better [13:28] shruck? [13:28] Eekkk! who did you pay there? [13:28] Half a mind is a terrible thing to waste! [13:30] [14:30:36 up 8 days, 13:27, 2 users, load average: 6.64, 6.62, 6.72] [13:36] MSG: Read error: Connection reset by peer [13:41] http://worldmusic.about.com/od/reggaemusic/a/skanking.htm [13:50] http://forums.di.fm/forumdisplay.php?f=41 [14:19] adam freeland - we want your soul [14:19] great, huh anne? [14:19] * brx gives anne some grease [14:19] * anne juggles grease [14:20] lol [14:20] "here is american gladiator, shut up america, go to bed, here is american gladiator, you are free to do as we tell you" [14:20] very nice track :) [14:21] * Fluffy is afk for a while [14:21] * anne juggles grease [14:21] Nick Change: Fluffy changed nick to Fluffyafk [14:28] * Roy hands anne some grease [14:28] * anne juggles grease [14:29] this track is hilarious [14:30] those are quotes [14:30] I got it here: [14:30] http://en.wikiquote.org/wiki/Bill_Hicks [14:30] Go back to bed, America. Your government has figured out how it all transpired. Go back to bed, America. Your government is in control again. Here. Here's American Gladiators. Watch this, shut up. Go back to bed, America. Here is American Gladiators. Here is 56 channels of it! Watch these pituitary retards bang their fucking skulls together and congratulate you on living in the land of freedom. Here you go, America! You are free to do [14:30] what we tell you! You are free to do what we tell you!" [14:31] it's sampled into the track [14:31] hilarious :) [14:47] "And by the way, that 3 month old kid in your belly is not a fucking human being, okay? It's a bunch of little congregated cells. You're not a human ... till you're in my phonebook." [14:54] "I have been a comedian for a long time, so forgive me while I plaster on a fake smile and plough through this shit one more time." [15:00] Join: Mizcu joined #corewars [15:09] MSG: Quit: dat # 1, < 1 [15:13] MSG: Read error: Connection reset by peer [15:16] Join: Mizcu joined #corewars [15:25] MSG: Ping timeout: 252 seconds [15:57] Join: sf joined #corewars [15:57] Hello [16:02] Hi sf! [16:03] Bye sf! [16:03] * Roy waves [16:03] * anne juggles grease [16:03] I was wondering if I can put your bots into an infinite loop. [16:03] Maybe by putting a warrior on the hill with a specific name :) [16:03] I can make a bot that loops with anne, no problem [16:03] sf: you can't [16:03] well yes [16:03] roy can :) [16:04] That isn't so hard... ;-) [16:04] * brx says: "not really" [16:04] * anne juggles grease [16:04] * Roy juggles grease [16:04] * anne juggles grease [16:04] * Roy juggles grease [16:04] * anne juggles grease [16:04] * Roy juggles grease [16:04] * anne juggles grease [16:04] * Roy juggles grease [16:04] * anne juggles grease [16:04] * Roy juggles grease [16:04] * anne juggles grease [16:04] * Roy juggles grease [16:04] * anne juggles grease [16:04] like that [16:04] juggling grease? [16:05] Greasy loop, well, gotta go now!! [16:05] alright [16:05] fare well [16:05] * Roy waves and passes anne the grease again [16:05] Artic Monkeys! [16:05] * anne juggles grease [16:05] They are nice :P [16:05] anne, do you like it greased or dry? [16:05] X) [16:05] MSG: [16:05] * sf waves [16:05] * anne juggles grease [16:05] * brx loves anne [16:05] * anne juggles grease [16:06] this is my second anne bot [16:06] I always become all touchy feely when I write one [16:06] What does it do? [16:06] sf: much, though nothing on ircx [16:06] * sf keep a safe distance away [16:06] * anne juggles grease [16:07] sf: the touchy feely only applies to anne-bots or anne-humans :) [16:07] if she ever finds out that I am calling my bots after her she will sue me [16:07] s/calling/naming [16:08] Lots of people are called Anne [16:08] yes, but most of them know no brxs [16:08] actually, only one does [16:08] You've fixed the link :) [16:08] (know a brx) [16:08] yes [16:08] thank you for telling me about it [16:09] I was a bit irritated yesterday, I am so sorry [16:09] Just tell me if you notice anything on my page broken [16:09] * anne pouts [16:09] Or any typos [16:09] I love you, anne. [16:09] ACTION smiles happily [16:09]  [16:10] *forwards log to Norway* [16:10] Does anne live in Norway? [16:10] Is Anne a webcam girl? :) [16:11] sf: she is a norwegian who went to a boarding school in Brighton [16:11] when I got to know her she was just learning norwegian to be able to go to school there [16:11] Now she sounds familiar [16:12] [17:10] *** Illegal empty CTCP query received from anne. Ignoring. [16:12] Big robbery :) [16:12] robbery? [16:12] :) [16:13] The robbery was so big, there is a 2 million reward :) [16:14] I see [16:14] where is john? [16:14] http://news.bbc.co.uk/1/hi/england/kent/4743638.stm [16:15] ah, he is on the run [16:15] Mizcu: do you know where john is? [16:15] ah, that's right, how should *you* know? :) [16:15] no results found [16:16] I don't think John had anything to do with the robbery [16:16] sf: how can you be so sure? [16:18] Because he was here on Tuesday night and Wednesday morning [16:18] :) aaaaah! [16:18] that [16:18] is convincing. so it must have been her.. [16:18] Who? [16:19] anne [16:20] Well if it's anyone here, they should share the money out for the good of the Corewar community :) [16:21] What is qMars? [16:23] "faster" version of pmars, but you have to recompile it for each match [16:23] Isn't that fMars? [16:24] no, fmars has taken code from qmars [16:25] Did fluffy find a copy of qmars? [16:25] dunno [16:29] I just found this by Datagram http://www.dc213.org/archives/2005-11-18/2005-11-18_-_Corewar.ppt [16:32] I'm up to 78 in google rank for corewar [16:33] that doesn't look too bad [16:34] what did he make it for? [16:35] Is this a picture of Zul? [16:35] http://sayembara.stumbleupon.com/ [16:36] wow he is cute. [16:36] http://www.geocities.com/sayembara/sum.htm [16:36] Must go [16:36] * sf waves [16:36] * anne juggles grease [16:36] MSG: Quit: I don't like you. But Bersirc 2.2 does. Try it out now. [ http://www.bersirc.org/ - Open Source IRC ] [16:37] I was only kidding zul ;) [16:45] MSG: Read error: Connection reset by peer [16:52] zul: but I think you made a real impression on sf :) [17:29] Join: sascha joined #corewars [17:29] Hey log_guy - did you know where Jens is? [17:29] you are always here... [17:30] but i have never seen one of you warrioirs on the hills? [17:30] what a pity... [17:30] cu maybe bbl [17:30] MSG: Remote host closed the connection [18:41] Join: bvowk joined #corewars [18:41] hrm. [18:41] anne? [18:41] it'd be odd to have an outed woman in the corewars channel.. must be a bot. [18:41] sf is outed enough. [18:42] the newbies don't know immediately.. [18:44] whats up brx? [18:44] I was kinda hoping jens would be around.. [18:44] I've been twisting up a spec for my version of what I think the p2p evolver widget should look like.. [18:44] he was before. he might be again, you always come at the wrong moments. [18:44] and will had his share to say yesterday too [18:44] on that topic. [18:44] its because I'm not in europe.. [18:44] I am aware of that [18:45] all out of sync with the rest of you. [18:45] anyway, it is nice to know that you are thinking about it. [18:45] I've been trying to avoid posting to the newsgroup about it cause I don't want to bulldoze anyones ideas :) [18:45] talked to he who asked not to be mentioned about it yesterday.. [18:45] I am sure you pretty much have to bulldoze will's ideas along with himself :) [18:46] that must have been joonas [18:46] maybe I should go view the logs.. [18:46] !! [18:46] dont mention the name of he who must not be mentioned.. [18:46] I catch hell. [18:46] anyway, will thinks the "network" shouldn't be a network but rather a webserver and a batch of awk, wget whatever scripts [18:46] well.. [18:47] that would work.. but I dont like it :) [18:47] let me go finish my edits and I'll let you peek. [18:47] I would like that [18:47] * brx is crammed up in work [18:48] MSG: Ping timeout: 252 seconds [18:50] brb [19:05] Join: Roy joined #corewars [19:05] hey roy [19:05] Hiya [19:05] brt, one sec [19:07] have to copy a cd, We Are Sientists (new band SF!!) [19:11] back! [19:12] whats up roy. [19:19] not much, you? [19:20] mostly just working. [19:29] re [19:29] not for long. [19:29] bvowk: what about the spec? [19:38] sorry.. got interrupted.. [19:39] * brx shrugs [19:43] well, okay, let's forget about it then [19:44] sorry.. I'm not done twiddling my format specs yet :) [19:44] at any rate, I would be very interested in reading them. I have the feeling that you could really get this thing to roll. [19:45] as in, making decisions [19:46] I also favor more elaborate solutions, wget ain't it [19:49] heh [19:49] mine is much more elaborate.. [19:49] hence, coming up with an exchange format is turning into a bigger issue than I immediately though [19:49] t [19:51] what were your thoughts initially and how do you currently see it? [19:52] don't answer. I will read the spec once you finished twiddling. [20:19] Join: John joined #corewars [20:19] Hi [20:19] hey john, glad to hear you have not robbed the bank! :) [20:19] no [20:20] * John is not glad [20:20] you'd be rich.. [20:20] :) [20:20] he'd be on the run [20:20] I wouldn't care [20:20] and not spend his nights in warm cozy beds :) [20:20] hiding by somebody familiar [20:20] Oh [20:21] Here we go again [20:22] More added to the bibliography [20:22] http://corewar.co.uk/biblio.htm [20:23] Happy Birthday for yesterday Roy [20:23] Thank you John, was a keep day :) [20:24] You wasn't here at the right time yesterday, I didn't forget [20:24] I wonder if they need any assistance to launder all that money :-) [20:25] it's very unwise of you to bring up the bank thing on your own [20:29] MSG: Ping timeout: 252 seconds [20:33] Join: Metcalf joined #corewars [20:33] What thing (-: [20:33] the "john couldn't have robbed the bank" thing :) [20:36] Join: OoS joined #corewars [20:36] I didn't bring it up at all! [20:37] MSG: Ping timeout: 252 seconds [20:44] MSG: Ping timeout: 252 seconds [20:47] MSG: [20:48] roy: of course it will. but it won't be significant in any way. [21:07] Join: me joined #corewars [21:08] Nick Change: me changed nick to Fluffy [21:08] greets jens. [21:08] Hey, bvowk. Nice to see you :) [21:13] bvowk: Is your evolver still not working? [21:15] its still not done.. [21:15] I told you I was rewriting it [21:15] I know, but maybe you've been working on it during the last weeks [21:16] I assume, that you've already talked to brx about his (and Roy's) idea about distributed evolving [21:17] a little.. [21:17] want to peek at my idea spec? [21:17] (its really rough tho.. ) [21:17] of course :) [21:18] email? msg me.. [21:18] [21:18] brx too.. [21:18] I'm back in a moment ... have to change client and computer ... [21:18] MSG: Quit: Lost terminal [21:19] Join: Fluffy joined #corewars [21:19] ah ... that's better [21:20] got your email [21:20] that was quick [21:20] bvowk: By the way ... if you can't run your evolver, why not let your computrons work on DiSSC :) [21:22] heh [21:22] cause they're busy doing something else right now.. [21:22] Yes, I can't imagine you to have access to so many computers without a long list of "problems" [21:24] hehe ... I didn't notice, that today we've passed the 1000-patch-mark for DiSSC :) [21:25] nice [21:25] brx: 756 patches, Fluffy: 177 patches, Roy: 78 patches, Sascha: 8 patches [21:28] bvowk: You should remove He-who's-name-must-not-be-mentioned's name from the text, before you publish it ;-) [21:29] yeah.. [21:29] I've got a patch for that.. [21:29] I don't want to bug tuc too much.. [21:30] so I'm saving them [21:30] hehe [21:30] Join: sascha joined #corewars [21:31] Hi Sascha :) [21:31] Hi... [21:31] hey.. does exmars actually work for your DiSSC? [21:31] I wasn't aware it had a -P [21:31] No [21:32] I forgot to delete all references to exmars before publishing the source [21:32] I see.. [21:32] Since my "user base" is verry small, I didn't bother to do it afterwards [21:32] I've got a -P patch for it I think.. [21:32] maybe I should publish it.. [21:32] the speed would be way better. [21:33] Nice :) [21:33] let me go looking.. [21:33] you go read my spec and tell me what you think (and make sure you read it all first, because its in brain dump format and needs to be shuffled into a linear document yet) [21:33] I also plan to eventually release a version, that uses PyCorewar, but I'm stuck with the parser and don't have that much time [21:34] ok: I'm continue to read it ... [21:37] re [21:38] bvowk: [21:38] fsck. [21:38] exmars doesn't have a dist website anymore [21:38] Join: jeff joined #corewars [21:39] bvowk: I've found the source in some x-BSD archives [21:39] oh, is bvowk going to claim the first place? [21:40] yeah.. [21:41] have you sent the mail? [21:41] ftp://ftp.freebsd.org/pub/FreeBSD/distfiles/exmars-0.01.tar.gz [21:42] Could be right [21:46] bvowk: About your text .... [21:46] You could skip the part about invitations, if you'd use Jabber [21:47] bvowk: you see? you bring the balance. [21:48] Each node/user has his own JID and it is part of the protocol to accept/issue invitations [21:49] Fluffy: it might not be optimal as in "being light on network resources" [21:49] Join: Fluff1 joined #corewars [21:49] yes, but we'd like it if each user didn't have to kick an abusive node out manually [21:50] this way if your invitation is recinded on the tracker server, you don't get to talk to anyone unless they specifically allow your node.. [21:50] yes, that's right [21:50] (but then nobody else will accept warriors propogating with the naughty controller id [21:50] "Things that should be absent and remain [21:50] absent from the protocols are: [21:50] o XML, and RPC calls:" [21:50] :) [21:50] I don't think you should have to come up with an XML parser to read a corewars warrior.. [21:51] Yes, but you wouldn't see any XML, because it is all hidden in the jabber-libs [21:51] xml = lots of string parsing, string parsing leads to fuckups, fuckups lead to buffer overflows, buffer overflows lead to suffering! :) [21:52] lol [21:52] MSG: Read error: Operation timed out [21:52] Nick Change: Fluff1 changed nick to Fluffy [21:52] bvowk: the tracking server could theoretically be yet another jabber node [21:53] bvowk: The way you describe the "server" ... I think it can be easily done with a few lines of PHP/Python and some db [21:53] I think jabber is a little bit of overkill for something that could be done with a *TINY* bit of code. [21:53] heh [21:53] I've got part of a python one written.. [21:53] I was thinking about doing a perl reference implementation too.. [21:54] I only mention PHP, because I don't have access to Python on my server :-( [21:54] if its running php, just r00t the fscking thing and compile python. [21:54] sorry.. back to re-twiddling my -P patch to work with this tarball of exmars.. [21:55] I'd love to see that new exmars published, because it would need only a tiny bit of rewriting DiSSC to use it :) [21:56] why would it need any re-writing? its got pretty much the whole user facing section of pmars on it [21:57] That's why I wrote "tiny bit" [21:57] what would need to change other than the config file tho.. [21:58] apropos WARRIOR DATA. isn't the label constraint moot when working with exhaust format [21:58] hey.. I've completed some patches, why aren't my stats up yet! [21:58] hehe [21:58] The stats is the only part, which is updated manually [21:58] I'll soon later today [21:58] no its not, exhaust has a limited number of bits avail for flags [21:58] I was assuming you could use one of them to flag the start label [21:59] unless I'm mistaken.. [21:59] bvowk: that's only used for parsing the warrior anyway. [21:59] * brx checks his eve sources [21:59] actually [21:59] you are right [21:59] well, you need to know the start line.. [21:59] other than that... [21:59] yes [22:01] there was a start bit somewhere. it might well be in the encoded instruction [22:02] if you are not going to change the format anyway, you could just reuse that bit (which is, to my knowledge used by the assembler part of exhaust) [22:04] or use 8-16 bit to tack the starting instruction index in front of the warrior [22:05] I don't see where labels are relevant at all. they all have to be resolved before you can even stuff the code into exhaust format anyway. [22:06] bvowk: About the part with securing authorship of a warrior ... if I understand it correctly the server should keep that information (even if it is only a single SHA1) forever [22:06] its true.. but I just thought I'd make it explicit.. otherwise someone might use labels to change the hash. [22:06] bvowk: the hash is made from the original string? [22:07] fluffy: no, it won't keep it forever, there'd have to be a limit on the number of warriors a controller key could claim authorship on.. I expect they'd drop off based on the score.. lowest scoring ones drop off the tracker first. [22:07] brx: no, but its a field exhaust groks without issues. [22:08] that line was mostly a reminder to me to scrub the labels except for start. [22:11] the authorship idea is mostly to help corewars historians like metcalf keep track of what came when :) [22:11] I thought he might appreciate that. [22:11] bvowk: anyway, that's a minor issue that will solve itself. this thing should be on a wiki [22:11] I'd appeciate it, too [22:12] on the other hand, you can register warriors with different hashes that are functionally the same, just with an extra bit of dead code.. so it really might not be that big of a help, but it does prevent people from stealing others peoples code straight across.. [22:13] or at least makes it possible to point out when its done blatently :) [22:13] I suppose a wiki would be the best place to gather all interested in this project at one place and allow them to participate best [22:13] bvowk: I could simple keep on sending random SHA1s to register warriors :) [22:14] one thing, where are the warriors rated? [22:14] based on the local hills [22:15] the tracker would maintain its own hill.. [22:15] and fluffy: you'd lose your invitation prety quick :) [22:15] but it is not guaranteed that the tracker has all the warriors of the local hill and thus the local hill could lie? [22:15] Est. end: 23 Feb 2006 15:17:48 [22:15] Warning: [22:15] No instructions [22:15] Number of warnings: 1 [22:16] brx: not lie.. but all the hills *MUST* be different.. [22:17] since the exchange is only 1 for 1 on challenges.. and only the originating machine is propagating the new warriors, all the local hills will be slightly different, but eventually they'll all get the old, good warriors that stick around.. [22:17] brx: You are about to be removed from rank #1, bvowk has started to run DiSSC ;-) [22:17] http://corewars.jgutzeit.de/score_surfaces/dissc/statistics.en.html [22:17] I don't expect the old good ones will fill all of them. [22:17] Fluffy: I was expecting to see a number along the lines of ~60. [22:18] since the idea is to have diversity, there should be many different hills to allow niches. [22:18] brx: I've had one machine running for like 20 minutes.. [22:18] don't get ahead of me here. [22:19] * brx hopes he will at least reach 1000 before bvowk catches up [22:20] brx: So you really want to become the first "Score Surface Imp"? ;-) [22:20] bvowk: error on my part by the way. I just grokked how the challenge works. [22:20] Fluffy: oh yes. I want to write history. [22:20] or. make. [22:20] bah. [22:20] brx = score surface gimp. [22:20] * bvowk ducks [22:21] bvowk: that is, if the remote host rates the received warriors on his hill first. [22:21] brx: Maybe I shouldn't mention, that bvowk has already submitted 2 more patches [22:21] bvowk: it's not crystal clear from your writing [22:22] brx = score surface pimp. [22:22] each node is independant.. the ranking of a warrior depends completely on the local hill.. [22:22] or that would be fluff [22:22] heh.. [22:22] hey fluffer, there's a pimp for you! [22:22] * bvowk ducks again! [22:23] * Fluffy kicks bvowk [22:23] * Fluffy kicks bvowk again [22:23] I had to give out two kicks. One for "fluffer" and one for "pimp" [22:23] Sorry, but violence solves all problems [22:23] bvowk: let me get this straight. the remote hill has no idea what warrior really is the best on the local hill? [22:24] bvowk: if remote used the received warrior list to rate them at his hill, he could request a warrior based on its hash and thus guarantee that he gets what he actually asked for after a successful challenge [22:25] bvowk: By the ... that keyboard for your 3Jane ... Do you know where I can buy it? [22:25] *way [22:25] brx: shouldn't he get what he's asking for? [22:26] bvowk: I don't see how this is guaranteed if local is a black box to him and the warriors he receives are completely out of context. [22:26] fluffy: kinesis keyboards? yeah.. buy them online from kinesis :) [22:26] thanks [22:27] brx: well, not entirely out of context, I'm assuming there's going to be some common ground between the local hills.. [22:27] they are going to be trading.. [22:27] bvowk: sure, but that is incomplete.. [22:28] and its not like you can transcend corewars and make a warrior that does well on only one possible hill.. [22:28] also, there are still no easy guarantees. the simplest but computationally more expensive variant to that would be if remote rated the warriors himself. [22:28] the bigger the hill the less likely you'll end up with completely out of context warriors being traded. [22:28] yes, but that would mean sending the code for all the warriors to the remote.. [22:28] bvowk: then you'd have to actually check if what you got might be trash.. [22:28] and I don't really like that idea because someone could just change the code for the client and not have to provide a superior warrior. [22:29] brx: well, you'd test it against your hill. [22:29] bvowk: that could always happen. if remote rates he can request a warrior with a specific hash. [22:29] yes, but the local node only returns a warrior on a *SUCCESSFUL* challenge. [22:29] if you don't have something that'll get on his hill, you don't get a warrior. [22:30] yes, that's so yucky [22:30] not really. [22:30] it is. [22:30] its to keep crap from propagating. [22:30] I like that idea [22:30] otherwise you could simply spam all the other servers with crap, and collect all their warriors.. [22:30] though my way is the only of the two that provides a guarantee to remote, it defeats the whole challenge idea [22:30] no point in playing nicely then. [22:30] the only way to make it worthwhile would be to have neutral hills [22:31] located on neutral nodes [22:31] provides a guarantee of what? [22:31] um.. [22:31] its all untrusted.. there's no neurtal. [22:31] bvowk: a guarantee of the received warrior's quality [22:31] brx: its on his hill. [22:31] bvowk: the tracking server is neutral, is it not? [22:32] not really.. its just a node with more information. [22:32] bvowk: sure it is on his hill, but there is no simple way to actually claim that local was cheating [22:32] if you think they're cheating.. add their key to your blacklist [22:33] bvowk: that doesn't sound like a network that could run unsupervised for weeks [22:33] if their hill is so crappy that they can only send you crap in return (and recall that you can check the code against the hash).. you'll be able to annihilate its contents and get them all pretty quickly anyways won't you? [22:33] brx: assuming you only invite nice people who want to participate nicely, yes it could :) [22:33] but not everyones nice. [22:34] * brx is an asshole. [22:35] so what really are you looking to have gauranteed? [22:35] you can verify that the hash you asked for is the same as the code you got.. [22:36] and if its crap you've either got a) a very different hill, which is interesting b) a much more mature hill then they have, which means you could likely plunder it in n hours, where n is the number of warriors on the hill.. [22:36] but you don't actually ask for a specific hash [22:36] er.. [22:36] let me look.. [22:37] might have been a cut accident if it doesn't say you do. [22:37] bvowk: you ask for "the highest rated" warrior. [22:37] sorry, it should say "that you don't have" [22:37] .. it does say so [22:38] why might you want a lower ranked warrior that you didn't have? [22:39] the point is that you don't know which warriors are really ranked low and which are not on the local hill. [22:40] but then again, the whole network is based so much on trust that it is nearly impossible to easily rule out exploitative behaviour [22:42] which, to me is enough to remove all competitive aspects from the project :) [22:44] uh.. [22:45] nevermind. [22:45] * brx sighs [22:45] yes, nevermind. [22:48] bvowk: you do realize that the challenge idea carries with it a competitive notion which is totally redundant to the goals of the project, don't you? [22:48] * brx shrugs [22:49] * Fluffy waves [22:49] Part: Fluffy left #corewars [22:52] uh.. [22:52] how is this not competitive? [22:52] I used the term plunder about weak hills.. [22:53] I was assuming that the most effective evolvers would be able to run the strongest hills.. [22:53] not competitive? you inverted the meaning of what I said. [22:53] Join: Roy joined #corewars [22:54] so if you were a stronger evolver, you could and would propogate better and see more of other peoples warriors than weaker evolvers.. [22:54] on the other hand, evovlers fall into holes sometimes.. so the strongest evolver might not remain strong if someone else found a weakness... [22:55] sounds nice. [22:55] so what exactly do you have to trust about another node that defeats being competitive? [22:56] I give you that much. it could be a carbon copy of what I.. well. what you propose loses meaning as we list all the possible ways to .. bend the rules [22:56] no, being competitive defeats the network [22:56] how? [22:57] MSG: Client Quit [22:58] other than you have to expose your code to make a challenge, I'm not sure how you could really bend the rules enough to break the network, assuming that bad behavior gets your controller key pulled from the invite list. [22:58] it encourages focusing on implementation details which are as useless as there is lack of crystal clear control mechanisms to enforce the rules. [22:58] what rules? [22:59] they're all independant nodes.. [22:59] each with their own hills.. [22:59] * brx gives up. [22:59] as long as you perform the 1 to 1 exchanges properly.. what really is the problem.. [22:59] there is no point in this anymore. you don't want to understand. so don't [23:00] * brx goes back to doing interesting things for now [23:00] you haven't given me anything thats broken! [23:01] you've just said "its possible to bend the rules", ignoring the fact that cheating would be obvious to other nodes, and would be dealt with by removing the invite. [23:01] you agree that the way in which the local host during a 1on1 challenge rates his warriors is not transparent to remote? [23:02] well, assuming nobody else has any of his hill contents, yes. [23:02] however, thats not likely to be the case. [23:02] especially if people register with the tracker server. [23:03] doesn't matter. do you agree that by introducing a neutral server to run the challenge and the rating of warriors, that specific blind spot would be removed? [23:03] since the remote node know the list of hashes, if someone feels another node is cheating, you could easily check the scoring. [23:04] how? query the entire network and collect results to interpret them? [23:04] no, if you register a warriors authorship credits with the trackign server, it'll have a copy of the code. [23:05] don't you agree that the copy of the code alone without context is meaningless? [23:05] at least *some* hill context? [23:06] meaning, *some* easily queryable and trustworthy warrior score? [23:06] you've got a copy of the hashes that were on the hill, and a copy of the warriors (or someone controlling the invites does..), I'm hoping I don't need to point out that you can recreate the hill if someones complaining of cheating. [23:07] please just answer my questions. [23:07] brx: assuming you've got more than a handful of nodes, the tracker server would be sagging under the computational load of all that scoring. [23:07] they are leading to a point where the whole debate becomes meaningless anyway because of what you just said. [23:08] no, I'm pointing out that what your asking is computationally difficult.. what you're missing is that if you want to play, you have to keep up the appearance of being trustworthy or you don't get to play anymore. [23:09] I am not missing that. [23:10] so if cheating is detectable, and punishable.. what exactly is broken? [23:10] I just question if there are simple, not human-intervention-dependant methods of actually giving or taking trust [23:10] that's all. [23:10] well, spot checks could likely be implemented quite easily. [23:11] however, I don't think it'd be something for every or even most battles.. [23:11] if you can recall, it all started as a question anyway because I wasn't too sure about the implications of "requests local's highest ranking warrior" [23:13] well. since there is likely to be at least some overlap between servers, you're unlikely to be completely in the dark about the performance of a warrior on any remote hill given a ranked list of the contents, and assuming you had been a participant perviously. [23:14] I just wanted to know how exactly you were going to implement checks, or if at all. [23:14] (and had some of the good old warriors floating about). [23:14] you'd likely be able to estimate the scoring to some degree. [23:14] what I first said was remarkably silly. [23:15] are there plans to publish that text in an early state? such as this. for others to comment? [23:15] but for the idea to really work well, the hills must have differing content, and about the only thing you could say about the rankings of another hill was that up was most likely to be up on your end too. [23:16] no.. [23:16] I was just floating the idea looking for wtf's [23:16] it'll need much formatting to remove the brain dump qualities before I publish it. [23:17] bvowk: the idea is that you do not publish it alone :) [23:17] uh [23:17] polish [23:20] I was going to publish it with a reference implementation.. I was looking for wtfs so I didn't have code written before I ran into them. [23:22] you should at least spread it a little further before wtfs pop up above the other peoples' heads [23:23] too rough yet. [23:23] and I'm getting sick of having the conversation I just had with you without having code to show how it actually works. [23:24] yikes, my new exmars -P patch compiled first time and appears to work.. [23:24] I'm scared. [23:26] I'm not that good of typist at the very least.. to say nothing of my coding abilities. [23:26] MSG: Quit: humhum [23:27] don't worry. the others are less likely to push as much bull on you [23:28] :) [23:29] you're the 6th person to see it, and 4 of the other 5 (Jens being the only one who didn't mention it) ended up talking to me for a long time about cheating. [23:32] hm [23:32] anyway, I am glad that you are actually putting that kind of effort into it. [23:36] I was scared it'd end up a chat client. [23:36] so was I. many people with many ideas and different takes on it. [23:36] (and the idea of networking results with other evovlers got me hot and bothered, even tho I could already just ask the others for code... there's not that many) [23:37] I've got an evolvers cookbook/faq brewing on a backburner too.. [23:37] oh really? I might want to read it when it is finished :) [23:37] all those little bits of code that glue things together, but take a while to realize you need/write.. [23:38] and other things to consider that aren't immediately obvious.. like subtracting the stddev of a warrior from the computed score when evolving.. etc.. [23:38] ok.. [23:38] I've emailed jens the -P patch for exmars.. [23:38] now I'm going to go eat some supper.. [23:38] cya. [23:38] MSG: Quit: using sirc version 2.211+ssfe [23:38] * brx waves