[00:00] yes, but the pipes will carry you bits far enough away that I won't have to listen to them [00:00] Do you already have pipes in Canada? [00:00] ;-) [00:01] yes, I've got a full service igloo. [00:01] And aren't they used for da Internet? [00:01] I've even got an automatic waterer for my dog sled dogs [00:01] (because I can snowshoe everywhere) [00:02] er.. s/can/can't [00:02] Hey, you have dogs? Why not eat them? [00:03] I like my dogs :) [00:03] i wouldnt eat animals, that i dislike [00:04] you prefer eating people? [00:04] yes, but they have to be properly fet [00:04] so you like eating americans?! [00:05] I said PROPERLY! [00:05] i would not recommend that [00:05] yes, that would be another reason to invade Germany [00:06] I don't think they'd invade.. [00:06] ok, then they'd nuke us [00:06] no, they will buy us [00:06] with the current problems in iraq, I think they're on a strict "bomb you back to the stoneage" policy [00:06] or send Britney Spears clones [00:06] aaaah [00:07] i suddently feel, that i have need for a space ship [00:07] no! wait! That's would be too cruel [00:07] Virgin Galactic is offering short trips for a couple of $ [00:08] fluffy: you still have to come back [00:08] yes, but I think, that no one has seen an A-bomb go off from space [00:10] * Fluffy hugs bvowk and replaces the "shoot me" with "I use LITHP" on his back [00:16] bvowk: Why don't you evolve something for tiny or even 94nop? Isn't it time to give the other hill a new try? [00:17] by fluffy [00:17] bye [00:18] no, I'm not going [00:18] so, youre gone, and are staying -- [00:19] I've only killed Jabber, ICQ, ... [00:21] I'm waiting for my generator to finish and then I'm done for today [00:22] you can't sleep, while your computer is running, did'nt you say so? [00:22] But unfortunately it decided to improve my warrior in tiny steps [00:22] at the moment I have score changes of about 0.1 every minute or so :( [00:27] Yes, I can't slep with my computer running [00:47] hmm [00:54] * Fluffy waves [00:54] MSG: Quit: fluffy.i < 1, # 42 [05:37] MSG: Ping timeout: 255 seconds [08:24] Join: fiveop joined #corewars [09:08] evenin' [09:30] MSG: Quit: Lost terminal [09:54] Join: Fluffy joined #corewars [09:54] :) [10:26] ^^ [10:26] And again we have MNS [10:54] bbl, getting a flapjack [10:54] :) [11:09] Join: John joined #corewars [11:09] Hi :-) [11:12] Hi John [11:12] Hi Jens [11:12] sf has been busy on nano [11:13] And you and Roy :-) [11:13] and bvowk has promised to start his new evolver soon [11:13] :-) [11:13] but unfortunately tlotd is 40th [11:14] But bvowk's warriors always give plenty of points to mine, so it'll be okay :-P [11:14] yes, but the question is whether you'll survive that long [11:15] 37th :) [11:16] but my warrior will soon return [11:16] the one that you've pushed off was only a little test [11:17] someone keeps pushing off my Christmassy warriors [11:17] (how do you spell Christmassy anyway?) [11:18] (You are asking the wrong person.) [11:21] that wasn't much of an improvement, 47th now :-( [11:22] but tlotd is 34th now [11:23] Not sure why, the score is within 10 points of the first version [11:23] And tlotd doesn't score very well against hey mister snowman [11:24] Now you only need to survive 13 successfull submissions :) [11:25] On the other hand there might be the problem, that Zul's warrior might soon become older [11:25] than Roy's and your warrior [11:26] That's not a problem [11:26] why? [11:26] I mean, don't you want to have the oldest warrior? [11:27] It would be nice, but I'm sure whoever gets it deserves it [11:29] I'm off on the train to go Christmas shopping soon [11:29] rightey [11:29] Hi Mizcu [11:30] strange, an imp seems to get 63-67 points against every hill [11:30] :-) [11:30] Wow, look at the scores of that imp [11:31] It nicely kills some warriors :) [11:31] never expected that [11:31] and it is good against tlotd, too [11:31] Last try for the snowman, don't think this new trick is going to work [11:31] fluffy: it's all those warriors containing imps which are hurting tlotd [11:32] Then you won't like my next warrior [11:32] :-( [11:33] But maybe it won't enter the hill [11:33] Oh, the new trick did work after all :-) [11:33] :) [11:33] There are some warriors low on the hill which hurt tlotd [11:34] So I'm hoping they'll get pushed off before me :-) [11:34] Plasmodium circularis will soon die [11:34] :-( [11:34] the question is, whether my replacement will hurt tlotd evenmore [11:35] John: My other warriors are doing fine and Vera City I will soon become +100 :) [11:35] I'll update the HoF that day ;-) [11:36] Then probably later today [11:36] Then I'll do it tonight when I get home [11:36] :) [11:37] I'd like to see impossible [11:37] Yes, let's hope, that gnik eventually decides to come to #corewars [11:39] Okay, have to go, train to catch [11:39] * John waves [11:39] * Fluffy waves [11:40] MSG: Quit: WLP.I #0, {0 [11:40] oh no! [11:41] what does this WLP do? [11:43] wallop? [11:43] huh? [11:44] could be shortened of that word [11:44] Didn't John want some special opcodes implemented some time ago? [11:53] Sorry, John, but Plasmodium giganteum isn't any good for your tlotd ... [12:16] Join: Zamboch joined #corewars [12:17] Hi Zamboch [12:18] Are you new to corewar? [12:18] MSG: Client Quit [12:19] right [12:19] :) [12:19] It must have been a very frightening experiece to be talked to by me. [12:20] by a noseless man, yes [12:24] hrmm [13:28] Join: Core_old joined #corewars [13:40] Join: fiveop joined #corewars [13:40] Hi old Core, fiveop :) [13:45] Join: ares joined #corewars [13:46] 'morning [13:46] ares: A little bit late for that, isn't it? [13:47] no, late would be 22h [13:47] who is Dinchen? [13:50] read yesterday's logs :) [13:51] oh fresh blood [13:52] i was online, when dinchen came... i really did not see it... well i was busy, i suppose [14:23] MSG: Quit: Trillian (http://www.ceruleanstudios.com [14:37] Join: Roy joined #corewars [14:37] Hi Roy! [14:37] Howdy [14:38] I might have a nice nano warrior on my home pc, all warriors scored about 180 against the benchmark but one scored 189 :o [14:38] Then send it :) [14:38] Was in a hurry this morning so I shut it down and went to work, which is where I currently am [14:41] So how long do I have to wait? [14:50] Don't know, couple of hours? [14:50] * Roy is sad about Rezoomscanner :( it sucks :( [14:51] One of the few pieces of new code I made and its fundamentelly broken :-D [14:51] :( [14:52] But I had a new idea, working on it now ;-) [15:10] You making anything for the normal hill currently? [15:10] no [15:14] oh :( [15:15] at least not this year [15:23] Ah thats better! Just fixed another flaw in my latest scanner, now its scoring like I hoped it would! :D [15:34] Now I need to over-optimise it just to annoy Inversed! [15:35] :) [16:12] Roy: you've sparked my interest in 94nop again, but I won't send new warriors this year [16:13] Hehe cool ;-) [16:13] I'll post my new warrior soon too, it didn't make the hill :-( [16:13] Its a slight variation on He Scans Alone (mixed with Zoom+Blur)! [16:16] why is your warrior 93 instructions long? [16:21] Because it is... its a scanner with decoy, nothing intresting about the size [16:21] hmm [16:22] Want to see it? :) [16:22] yes :) [16:22] wPtr spl #0 , <0 [16:22] tPtr spl #7673 , <2986 [16:22] for 10 [16:22] dat 0,0 [16:22] rof [16:22] nop }sEnd , {pScan-1 [16:22] sTop mov.b tPtr , wPtr [16:22] mov.i tSpl , >wPtr [16:22] tScan add.x #-10 , tPtr [16:22] sne.x *tPtr , >tPtr [16:22] add.x *pScan , tPtr [16:22] jmn.f tSelf , jmz.f tScan , *tPtr [16:23] pScan mov.x @tScan , @tScan [16:23] tSelf slt.b @tScan , #tEnd+4-tPtr [16:23] sEnd djn sTop-1 , @tScan [16:23] djn *pScan , #10 [16:23] tSpl spl #0 , }0 [16:23] tClr mov tEnd , >tSelf [16:23] djn.f tClr , {tSelf [16:23] tEnd dat >1 , 2-tSelf [16:23] Here you go :-D [16:23] 93 million miles from earth to sun [16:23] Silly nop line kicks the wipe in action the first time it finds something [16:23] ares: Has Austria been invaded by the US or why do you use "miles"? ;-) [16:24] thelemic warrior? [16:24] because he googles it, and what is thelemic?? [16:24] thelema, a philosophy about true will [16:25] greek word "thelema" has numeric value of 93 [16:25] as "agape" also has [16:25] (love) [16:25] huh? [16:25] mystical numerology [16:25] But what do you think of it Fluffy, its simple/funny ey? :) [16:26] Roy: If you mean the warrior, then yes [16:27] Yeah the warrior, it uses the same amount of lines as the usual Hescans be now has zoom+continues attack [16:28] Because it goes the opposite way its a real HSA killer btw [16:31] Time to go home, bye! [16:31] MSG: Quit: dat [17:15] aha!! the second warrior of koenigstuhl allready causes problems... if pmars only had an option to add the source line number of everey instruction to the output... [17:23] It does in the pmars -V output :) [17:23] oh - thats sounds VERY interesting... no more compatibility problems? checking... [17:25] why didnt i know that one week ago? [17:25] hrm [17:25] * bvowk wonders about fluffy [17:26] ares: Because there've been sunspots on the sun? [17:26] bvowk: I didn't do it! [17:26] http://www.kitco.com/images/live/gold.gif [17:26] hrm. [17:27] oops. [17:27] wrong window ;) [17:27] although, maybe the rest of you care about gold and the US dollar [17:27] Yes, but only if I posses large amounts of both [17:29] bullionvault.com fluffy ;) [17:29] you can acquire as much as you like [17:29] and they've got a really slickass xml interface [17:30] and their tech people rock [17:31] bvowk: and why did you wonder? [17:31] about you? [17:31] yes [17:32] just to make you wonder. [17:32] hrm. [17:32] the US dollar is taking some flak today [17:34] and the Canadian dollar? [17:34] http://finance.yahoo.com/currency/convert?from=USD&to=CAD&amt=1&t=5d [17:34] meh. [17:34] nothing to report.. [17:35] which likely means we're going down with the american [17:35] s [17:37] hehe [17:44] bvowk: Ready to start your evolver? [17:45] yeah, I'm test running it right now [17:45] I hope that you've changed the naming scheme [17:45] why does everyone hate the naming scheme so much? [17:45] does it really matter? [17:46] hate the warrior thats kicking your ass, not his name ;) [17:46] I don't really hate it, but it is fun to tease you with it and so far I couldn't see what those numbers and letters mean [17:47] Maybe you can enlighten me [17:48] it was a pool number, randomly chosen at start time for a run, then a warrior number in the pool, incremented every time one was created.. and then it was a machine name where it came from usually. [17:48] the hex string ones were the crc32's of the instuctions in each level of the form.. [17:48] 9988-449-xt642-8-nano-eve78 [17:48] so it was instructions-imods-amods-bmods-a-b [17:49] yeah, that was has the identifier of the evolver version on the end [17:49] so it was nano-eve78 [17:49] machine name xt642-8 [17:49] Who has chosen that name? [17:49] (xterm #8 in rm 642) [17:50] hmm [17:51] you know that was a distributed system right? [17:52] That settles it then. Please use a better naming scheme ;-) [17:52] yes, I know [17:52] so there was a main collection server... and then a bunch of clients doing the actual churning [17:52] k [17:52] suggestions then? [17:52] I've currently got it doing sha-1 for names ;) [17:52] 128bit hex string [17:53] I'm betting that won't be popular. [17:53] How about the following: [17:53] Use only strings with max. length 20, which have an MD5 or SHA1, which ends with lots of zeros [17:54] You could use one or two of your computrons, which do nothing but search for such values ;-) [17:54] both md5 and sha1 are longer than 20 [17:55] You don't use the MD5/SHA1 as the name, but the string [17:55] your method is flawed! [17:55] which has that MD5 [17:55] no, it isn't! [17:55] flawed! [17:55] * Fluffy kicks bvowk [17:55] lets hear your next idea [17:56] Name your evolver properly (like Bvowk Mega Supa Dupa Evolver) and then append the number of the generated warrior [17:56] BMSDE 42 [17:56] *Bvowk's [17:57] currently I've got 42-BMSDE and yo udn't like it! [17:59] Third scheme: Idicate the position in the number pi, where you warrior is encoded as (whatever you like) [17:59] no, won't work [17:59] the number if probably too large [17:59] *is [18:00] heh [18:00] do you have any good suggestions? [18:00] Choose a random poison and use that as a name [18:00] that works for the first hundred or so [18:01] The names of Bacteria and Viruses are already taken ;-) [18:01] perhaps I need to come up with latin names for the various warrior classes [18:01] and then we'll speciate them based on ancestry [18:01] yes, why not. That would be at least better than the current scheme [18:01] there'll be a committee [18:01] I like that idea [18:02] * Fluffy is still waiting for your first warrior for nano ... [18:03] you could name the first warrior "Bob" [18:03] HSA + coreclear not work, mmkay [18:04] oh, roy quitted alrdy [18:04] If I remember correct "mmkay" means "fuck" [18:05] mmkay [18:12] Benchmarking old warriors is frustrating. They are always better than the current warriors :( [18:12] huh [18:12] I'm talking about my own warriors [18:19] (everything is better than fluffys warriors.. I'm holding off with my evolver to let his ego have a couple breaths before I crush it) [18:19] * Fluffy yawns [18:20] You haven't even been koth on nano [18:20] no, I just held most of it ;P [18:20] (programatically!) [18:20] yes, rank 40-50 [18:20] I'm so good I could hold the hill using my brain at one level removed! [18:20] ;) [18:21] of course, of course [18:21] Then go on and start to send your warriors, bvowk. [18:21] heh [18:22] I wonder if there's enough current stuff on the nano infinate to bootstrap my hills [18:23] Give it a try [18:24] fluffy: i am still shocked because of -V [18:24] ares: ok, continue to be shocked ;) [18:25] hey fluffy: why don't you send met a selection of your finest ;) [18:25] I can do that [18:25] Do you really want them? [18:25] *g* heheh [18:26] yeah [18:26] bvowk: I've published about 30% of my best warrior to Koenigstuhl lately [18:26] there's a nanok? [18:27] of course [18:27] http://www.ociw.edu/~birk/COREWAR/NANO/hill_rec.html [18:27] cool. [18:27] My best one is 3rd and your best one 22nd. Thas says everything [18:27] heh [18:27] *That [18:27] I haven't ever submitted any! [18:28] that must have been met! [18:28] yes, he sends all new infiniano warriors regularly to Koenigstuhl [18:31] bvowk: I've sent all of my nano warriors (except those currently on the hill) [18:31] the worst ones missed the hill by max. 3 points [18:33] you've got 2.9k compressed? [18:33] * bvowk rolls his eyes. [18:33] I usually go for quality instead of quantity [18:34] remember: 22nd of Koenigstuhl ;-) [18:34] *on [18:34] heh [18:37] And please notice the inspiration that is needed to choose such wonderful names :) [18:41] I really start to think that you don't have an evolver [18:43] heh [18:43] would you just keep your pants on? [18:43] its not like its ever been a quick process :P [18:43] I'm still debugging too. [18:45] Let me guess ... your program consists of 3 lines (including the 2 lines with comments) and it still kills itself immediately, right? [18:45] Join: Roy joined #corewars [18:45] heh [18:45] yes, thats exactly it. [18:46] how did you know? [18:46] bvowk, stop asking me for nano warriors... write your own evolver will ya! [18:46] Because you your crappy naming scheme, bvowk ;-) [18:47] *Because of [18:47] heh [18:52] Roy: and your nano warrior? [18:53] Ah! yes, lets send it and laugh :-) [18:53] LOL [18:53] nice conversation [18:53] you two like each other, do you? [18:54] * Fluffy hugs bvowk [18:54] * Roy joins and hugs Fluffy [18:55] * bvowk eats fluffy's brains [18:55] * Fluffy now has a warm feeling in his pants [18:55] (putting a "kick me" sign on his back) [18:55] * Roy keeps pouring more coffee down Fluffys pants [18:55] (puts kickin' shoes on) [18:56] 43 45.1/ 48.3/ 6.6 Serious Request Roy van Rijn 141.8 1 [18:56] Blegh.. [18:57] what is an abstract error? *g* [19:01] maybe it is an error, which is abstract [19:01] Whoops, I'm optimising against my own warriors :-D [19:02] Stupid as I am / Evolving against myself / Should stop that right now [19:02] yes, remove yourself from the genepool [19:03] Clearly its messing with my brains [19:03] you probably know this already: http://www.darwinawards.com/ [19:03] yes [19:05] I need 3 more successful challenges! [19:12] Do something about that then ;-) [19:15] Actually I'm waiting for bvowk to finish his "debugging", but I seriously doubt, that he'll finish it this century. [19:15] yeah yeah yeah. [19:16] Don't talk, bvowk. Finish debugging! [19:30] yay, i am debugging a VPN through remote desktop, and now confing the adsl-box [19:30] :) [19:34] note to self: vpn disconnects everything for a moment [19:35] note to mizcu: and won't reconnect [19:36] yes [19:37] did it really? [19:38] hey, i irc through a shell [19:38] ??? [19:38] so my comp might explode and you wont notice [19:39] only my connection to the shell borked [19:40] wouldnt closing the connection close all progs running in your shell, and cause the irc client to logout? [19:40] ever heard of program screen? [19:40] ares: Not necessarily. For Unix-systems there a program called screen [19:40] ah [19:40] *is [19:40] hehe [19:42] Just start it as a job (programname &) should keep running in the back once you exit the shell [19:43] Roy: that might be tricky for interactive programs [19:43] Join: John joined #corewars [19:43] Hi :-) [19:43] i will remember that, if i ever start any chat program in a ssh session [19:43] Heh yeah.. I only use it for servers etc [19:43] Hi John [19:43] Hi Johnny [19:43] Hmmm... 13 people on IRC, only 12 in #corewars [19:43] Who's lurking? [19:43] sf? [19:44] John: tlotd has been pushed off by bvowk's warriors [19:44] No it hasn't :-P [19:55] John: bvowk won't send any warriors anyway [19:56] That's a shame, I was hoping to show however much computing power he throws at them, my hand-written 15 minutes on the back on an envelope warriors are better ;-) [19:57] But it is obvious, that hand-written warriors are better than those evolved on his machines. [19:58] His warriors don't even have proper names ;) [19:58] Yes, that's a big problem [19:59] and I can't imagine, that they can spread chrismassy feelings like yours do [20:00] I imagine bvowk would be the first person who'd submit one called "Humbug" :-P [20:00] lol [20:00] also lol [20:02] or rather 1409-12880-humbug-eve78 [20:03] Don't give him ideas [20:04] how about encrypted warrior names? [20:04] * Fluffy kicks ares [20:04] or is this all about encryption? [20:05] Join: gareththe joined #corewars [20:05] hi [20:05] hi [20:05] Hi [20:06] ho, ho, ho? [20:06] not sure how to answer that [20:06] ho, ho, yes! [20:06] gareththe: Always assume, that I've disengaged my brain for a while [20:06] who says I haven't? [20:07] I figured out why I couldn't submit that warrior last night [20:07] tell us [20:07] I had a label called 'endgm' and sal parses this as the end command and stops reading [20:07] that's why it was ok in pmars [20:07] lol [20:08] does anyone know, what sal is using, or why this happened? koth-preprocessor? [20:08] ask bvowk [20:19] Join: Pavel joined #corewars [20:19] Hi Pavel [20:19] hi pavel [20:19] Hi [20:19] ho [20:20] Hi Pavel :-) [20:20] I'm going to read thru your emails and reply, so give me few minutes :-) [20:20] huh? [20:20] Hi ho it's off to work we go :-) [20:24] John: It is haiku time [20:24] hrm [20:24] hey met [20:24] bvowk: ready? [20:25] fluffy: no, I've got a 64bittedness issue somewhere. [20:25] Haiku? [20:25] bvowk: In 3 lines of code? [20:25] John: yes, read yesterday's logs [20:25] lol not again [20:29] Expected Haikus / Just like we had yesterday / None were forthcoming [20:31] something stirs in core / looks around - crushes paper / scanner rules the core! [20:34] but then stone attacks / bombs indiscriminately / scanner is no more! [20:34] fluffy torments me / revenge will be sweet he will / fall beneath the horde [20:36] the paper returns / many copies fall on stone / we begin again! [20:36] has two comments here / got three lines of bad lisp there / evolver still bad [20:38] metcalf nice to see / you again, we should catch up / in a private msg [20:39] lol [20:42] bvowk do not chat / still has tons of bugs ahead / got nothing on hill [20:43] new assembler done / hard work after all it was / proud i was again [20:43] then i here complained / why dont pmars show source lines / would have saved much work [20:43] i've been told a switch / fluffy said use minus V / pmars would have done [20:43] i cant really say / my assembler made of clay / but not works same way [20:43] so for my pitty / after all i must give up / include pmars now [20:43] fluffy, kiss my ass / I will fix my bugs at my / own pace, k, thx, bye [20:43] hehe [20:45] we do no more code / but writing silly poems / all join #haiku now [20:46] you infected me! [20:48] :-) [20:49] fuck haiku / boring shit / nobody cares [20:49] another sucessful meme [20:49] Mizcu: it is "mmkay" and not "fuck" ;) [20:49] go mmkary your mom [20:49] -r [20:51] Topic Change: John sets topic: I joined #haiku / but I'm all alone in here / where is everyone? [20:51] mizcu is bitter / studying surely is not / going very well. [20:52] Topic Change: John sets topic: http://corewar.co.uk - http://www.corewar.info - round 3 "Dual Core war" rules in rec.games.corewar [20:56] mm.. gareth seems to have missed the point [20:56] when i read source code / i hear an haiku rhytm / someone please kill me [20:57] my third fifth in a row on 94b everytime I improve I find the hill has got tougher :D wicked! [20:57] heheh *g* [20:58] Finally! I have the whole nano koenigstuhl in my benchmark! [20:58] Stupid evolver with the stupid parser rules.. [21:00] Evolver started / Now I have ages to waste / b.b.l. o.k. ? [21:00] roy: ha ha [21:15] Nano warriors away! [21:21] Are they gone? [21:21] Only sent one [21:22] Someone killed all but one of my Christmas warriors :-( [21:24] Roy: You got for one second [21:25] ? [21:26] how do i use pmars for checking, if a warrior does not die, if running allone? compete with a JMP 0 ?? [21:27] That is a way to do it yeah, there are others (better ones) I think, but can't remember [21:28] But if you fight a jmp 0 you have a warrior that doesn't die instantly (could still be suicidel though) [21:28] make it fight djn.b 0, #-1 [21:29] djn.f 0,{-1 <--- you could kill a jmp 0 with this warrior, but this warrior is still suicidal [21:30] MSG: Read error: Connection reset by peer [21:31] i am slightly confused [21:31] I make them fight nothing.. and then check how many cycles passed before the round ended [21:31] my first idea was to run only one warrior (double speed) and get to know, if it run MAXCYCLES [21:32] Thats the best idea I think, but some marses don't run single warrior battles if I recall right [21:32] but, how can i make pmars telling me, if there was a tie, when there is no opponent? [21:32] i am talking about pmars... [21:33] if run allone, a jmp 0 gets score 0 [21:34] so, if i want to find out, if a program is suicidal, i let it fight a jmp 0, for checking LIFETIME > MAXCYCLES, and let it fight DJN.b 0, #-1 for checking liftime = CORESIZE cycles, right? [21:35] but if it hits the single-instruction pseudo opponent? [21:35] eg coreclear vs "idiot" [21:35] something wrong with that? ;) [21:36] species (evolver) uses a similiar dummy in determining if the spawned warriors are worth anything [21:36] problem, if "idiot" is allways killed before suicide! [21:37] in which case you get win as a result from mars [21:37] jmp #0 is much harder to kill then a jmp 0 [21:37] so try that one ;-) [21:38] A single increment or decrement can kill a jmp $0, but a jmp #-1 or jmp #1 will still sit and do nothing [21:38] thx - good idea... i am thinking of an optimizer... in case, i accidently test a suicidal, its not worse than loosing time, is it? [21:38] thx for the tip, roy [21:39] MSG: Ping timeout: 255 seconds [21:39] Time is crutial in a evolver and checking suicidals before throwing them in a benchmark sure safes some time [21:40] thats, what this is about [21:49] funny - pmars warrior1 warrior2 -k --> 0 0 \n 1 0... [21:49] pmars w1 w2 -k -k --> Unknown by Anonymous scores 0 \n Unknown2 scores... [21:49] pmars w1 w2 -k -k -k results as -k allone [21:51] Seems like it flips a boolean ;-) [21:53] ok.. [21:53] evolver seems to be underway [21:54] what about a seti-screensaver like evolver? [21:54] its distributed like seti [21:54] i just had the idea [21:54] Thats cool, I imagined having that :) [21:55] I imagined evolvers being connected to interchange warriors [21:55] i can imagine, that some non redcoder-people would like it as a screensaver, if it visualizes the battles [21:55] Like uhm webservices ;-) [21:55] enyone expirience with gridwars? [21:56] Yeah a little [21:56] is it nice? [21:56] I once tried it, but I was very bad :-) [21:56] They don't hold tournaments anymore, and it was just to complicated for me, you have to have some good grid-computing understanding [21:58] Join: Fluffy joined #corewars [22:00] Roy: You got me for one second, because you've send that bvowkish warrior to nano. [22:00] * Fluffy congrats bvowk for having his evolver up and running [22:01] bvowk: How long do you think it will take before it spits out its first warriors? [22:01] Hehe :) [22:02] My warriors will be the first target ;-) [22:02] Nicely grouped near the bottom [22:02] But probably Plasmodium circularis will die first [22:12] ares: I'm thinking about both screensave and distribution [22:14] sounds like an interesting combination - we could add a switch, that shows any "normal" screensaver, while battle is running - so you could apply it to desktops in a big company [22:16] would it be a new kind of troian, a "calculation time robber" ? maybe we write a worm, containing mars-clients... [22:16] lol [22:16] Heh, I'm a programmer for some of the largest companies in the Netherlands, we only work for the top 10 companies... I once had a idea to include a little java-corewar-library in all applications we deploy on the large servers.... but thats just evil :) [22:17] byee [22:17] not really, if the mars runs on low priority [22:17] ybee [22:17] :) [22:17] MSG: Quit: Ex-Chat [22:17] but the screensaver combo is a cool idea [22:17] They all have internet access so we could all access them and maybe even connect them [22:17] Roy: No, that's called practical thinking [22:17] +++fluffy [22:18] nMars components are safe .NET it could run even in browser, so one day ... [22:19] has anyone idea how evolver should work ? [22:19] it should be not much work to make such a screensaver... it only needs visualisation and the http-transfer [22:19] some design for nMars / [22:19] ? [22:19] Should/could... yeah [22:20] Pavel: its more about how much evolving and how much optimizing a program does [22:20] There was a wikipedia with very good ideas, and Will had very advanced plans [22:20] and the visualisation might be a simple moving text "MARS running. Thanks for supporting the core war community" [22:20] but only bvowk has a working evolver :) [22:21] F'y: really? [22:21] No... [22:21] ares: No, some porn would be better [22:21] redrace works [22:21] Yace works [22:21] I mean distributed [22:21] Species, sort of - almost - works [22:21] damn, i got a hunch to make an evolver again [22:22] :) [22:22] *g* [22:22] better go get some sleepy before i start making stupid things again [22:22] * Roy too.. [22:22] But all I can program now is Java :-) [22:22] * Fluffy wishs sweet dreams of an distributed evolver ;-) [22:22] *wishes [22:22] fluff and Roy know this already; I make my most horrible pieces of code between midnight and morning [22:22] Mizcu: java ? IKVM :-) [22:23] Shame only real programmin language i know i qbasic [22:23] Mizcu: It is 0:22, right? [22:23] Fluff: yeah [22:23] Making the evolver isn't the hard part I think, easy to do structured in Java, but integrating a fast mars is the problem [22:23] 23:21 here [22:24] If I only knew when the first of bvowks nano warriors enters the hill [22:24] nighty-night [22:24] roy: I have DLL and .NET wrapper around pMars [22:24] * Fluffy waves to Mizcu [22:25] ciao miz [22:25] Pavel: But pMars is quite slow [22:25] Yes, and I don't even know how to talk to a DLL lib in Java [22:25] shouldnt that be forbidden? [22:26] java sandbox aso...? [22:26] Roy: It's not you, who should do the talking, but your program [22:26] fluffy: it's just unit test validator for me, but i know how to cover fater engines if needed [22:26] * ares gives fluffy the kick back [22:26] faster [22:27] PyMars is good too, if Jens finishes that it would be easier to make a good evolver [22:27] who? [22:27] yeah, if jens ever finishes it. [22:28] he's just as bad as fluffy [22:28] lol [22:28] bvowk: Does it work now? [22:28] Its the fastest mars I know, and with python its pluggable :) [22:28] how much faster is fmars in comparison to pmars? [22:28] 2 - 3 times I think? [22:28] ares: Depends on how you compile fmars [22:28] fluffy: it seems to [22:28] Maybe even much more [22:28] * Fluffy congrats again [22:28] I'm having a couple problems with it dropping the odd packet on the floor.. [22:28] bvowk: How long will it take before the first warrior is send [22:28] t [22:29] I think neds got problems with its ethernet board tho [22:29] there's *WAY* too many tcp resends [22:29] Do you want me to log out? [22:30] No, kick bvowk until he sends something [22:30] I would NEVER do such a thing, Roy [22:30] * Roy kicks Fluffy [22:31] huh? [22:31] Didn't see that comming ey? :P [22:32] No, you now you got MNS again :) [22:32] *but [22:32] You too ;O) [22:32] I had it all the time [22:33] * Roy has MNS NNS and BNS [22:33] N, B? [22:33] Missing, normal and big :) :-) :OP [22:34] Hey, you can just invent new diseases. No sane mind would do that! [22:34] *can't [22:34] f'y: why? [22:35] a's: why not? [22:35] was?? [22:36] looks like you should go to bed, too [22:36] * Roy too! [22:36] * Roy waves [22:36] MSG: Quit: Ik ga weg [22:36] as me, also... i am just thinking of a distributed optimizer... what will also bring a distributed benchmark [22:37] when i add all the megaherzes in my flat,. i might be able to optimize as fast as you are with your modern pc [22:38] ares: bvowk should tell you how many computers he uses [22:38] i heard sth... but i suppose, these are not his private property? [22:39] I have access to somewhere around 1200 processors [22:39] ARRRR! [22:39] lol [22:39] wow [22:39] with various levels of access. [22:39] ares: look at http://www.math.ualberta.ca/~bvowk/ [22:39] would you let me access some of them? *BigCuteSmile* [22:40] there's a few I could.. [22:40] wow! [22:40] i'd have 2 ideas, for which i'd need really much processor power... [22:40] yeah? [22:40] bvowk: Before I forget again ... about that extinctionist game ... how can you be sure to have killed all monsters? [22:42] AI based on evolution (not evolving algorithms in the common meaning) and some sort of an atomic simulator for a cyberspace-physics engine [22:42] you have to understand ares that a) they're not tightly coupled or linked with high speed/low latency networking [22:42] fluffy: they stop showing up when you summon them [22:44] i allready made a renderer writes the ouput of a simulation of an "ideal gas" to an avi... but i want to get my simulator doing in real time some day... maybe in 40 years with quantum computers [22:44] b) there is wildly varying amounts of memory in them [22:44] c) and he needs them all to run his evolver ;) [22:44] better than my pcs, that most of you would put to a musem [22:45] some weeks ago someone gave me a commodore 128 [22:45] btw [22:45] does anyone have one or some double density 5,25" disks? [22:45] nice paper-weight, right? [22:46] or do you know, where i can get some? *g* [22:46] ebay? [22:46] thats an idea [22:46] a datasette would do, too [22:47] http://search.ebay.de/diskette-5-1-4_W0QQfkrZ1QQfromZR8QQsatitleZdisketteQ205Q201Q2f4 [22:47] i suppose, that will all be high density [22:48] yes, they are [22:48] c128 needs double [22:48] i will find some day or scrap that old thing [22:49] and what do you want to do with it? [22:50] find out, what i can get out of the hardware... i remember some programs i wrote in old days.... eg a multitasking like system... okay, i know, my cellular has 100 times more power [22:50] but arent sprites wonderful? [22:50] no? [22:50] banause [22:51] How dare you! [22:51] ;-) [22:54] * Fluffy wishes bvowk good luck with his evolver [22:54] time to get some sleep [22:54] MSG: Quit: fluffy.i < 1, # 42 [23:09] Jens: PyCorewars parser sucks [23:10] MSG: Quit: Lost terminal [23:11] 30-40 warriors per second.. [23:11] bah. [23:27] ares: some email in your mailbox, let me know please [23:28] oh [23:28] i thought you were bvok [23:28] sorry bvok [23:30] ares: and on r.g.c too [23:30] eh? [23:31] post for you on rec.games.corewar [23:31] i just saw... reading now [23:34] hrm [23:55] Part: Pavel left #corewars