[00:13] Join: NW joined #corewars [00:13] NW [00:14] ehm... well, NW to you, too [00:19] MSG: Quit: NW [00:19] Join: NW joined #corewars [00:20] HELP [00:20] SUPPORT. [00:21] i'll be back with more coffee [00:22] whats up? [00:24] those, who never ask, are keeping silent, i suppose. [00:25] MSG: Quit: NW [00:26] Join: BIG joined #corewars [00:27] hi big, can i help you? [00:28] HI [00:30] HI ALL [00:31] MSG: Client Quit [00:31] Join: alternate joined #corewars [00:32] HI [00:32] are you using the Java-IRC-Client? [00:32] YES [00:38] MSG: Quit: alternate [00:38] argl [00:40] i want pizza [00:46] and intelligent users [01:16] orly [01:18] yarly [01:35] ill go nyarly myself to bed [01:35] *g* [12:06] back [12:49] Join: Core_old joined #corewars [15:05] Join: John joined #corewars [15:05] Hi :-) [15:07] hi John [15:07] Hi Mizcu [15:07] How's your Christmas been? [15:07] not that special [15:08] http://corewar.co.uk/nanohof.txt updated (again) - Loren Jamison enter the top 50 [15:08] please read the last two days logs and give a comment about my little project [15:08] Okay [15:20] Okay, it's a bit difficult to follow the logs because everyone else keeps saying stuff [15:20] But I take it you're planning to make some benchmark with a grading system, like easy, medium, difficult [15:21] i will most likely suggest wilfiz as "difficult" [15:21] http://www.ociw.edu/~birk/COREWAR/94/HILL/fatallure.red = post HSA [15:22] yes, looked too like to be coincidence [15:30] I always used to use these warriors for a benchmark - Quicksilver, Son of Vain, Olivia, Behemot, Vanquisher II, [15:30] Hazy Lazy, Willow, MyBlur2, Win!, CrazyShot 2, G2, Firestorm, [15:31] Stylized Euphoria, Disincentive, Reepicheep, Return of the [15:31] Fugitive, Digitalis 2002, Origami Harquebus, Self-Modifying [15:31] Code v0.11 and Shapeshifter [15:31] yes, that could be for "difficult" [15:32] Wilfiz contains no p-spacers [15:32] i doubt about using pspacers in a non-pspacing benchmark [15:32] Ah, okay [15:33] It's important to keep the same balance as the hill [15:33] if youve read the bench.txt , youll notice that i have thought about the medium-bench [15:34] but id rather take ideas on the easy-bench [15:36] I know it isn't the idea of a benchmark, but you should pick warriors where the code is easy to follow where possible [15:37] i dont mind cleaning and commenting the code if required [15:37] (for the newbies) [15:39] Have you looked at Wilkies, Wilbez, Wilmoo and Wilfiz? [15:39] yes [15:40] i have them all available from commandline if there is need to use [15:41] How easy do you want the easy benchmark [15:41] I would suggest including d-clear [15:41] easier than wilkies [15:42] preferable pre-94 warriors on the easy [15:42] Have you looked at Moore's 88 benchmark [15:43] havent heard of it [15:44] http://corewar.co.uk/moore88.zip [15:45] I like this stone http://www.ociw.edu/~birk/COREWAR/88/HILL/armadillo88.red [15:46] And this scanner http://www.ociw.edu/~birk/COREWAR/88/HILL/beholder17.red [15:47] those look awfully competitive in the moore's bench [15:49] In my opinion as easy benchmark required a few medium warriors. (maybe 3 out of 20) [15:50] possible, yes [15:50] To add a little bit of challenge for the newbies ;-) [15:54] but the easy-bench would still need some extention, i am not familiar of warriors that would fit [15:57] and i am a bit at loss about the plain stone for the benchmark [15:57] though Juliet Strom without imp should be good enough [16:01] http://www.ociw.edu/~birk/COREWAR/94/HILL/poke01.red [16:02] thatll do [16:05] Simple '88 paper http://www.ociw.edu/~birk/COREWAR/88/HILL/xeroxII.red [16:13] second problem would be the paper/stone [16:13] most PS's seem a bit too strong for medium-bench [16:15] but one should be included [16:18] flashpaper is too weak [16:37] MSG: Remote host closed the connection [16:55] Join: Fluffy joined #corewars [16:55] :) [16:56] hi Fluffyer [16:57] *grmpf* [17:00] Join: John joined #corewars [17:00] * John kicks fluffy [17:00] Hi John [17:00] Hi Jens [17:01] Have you had a nice Christmas? [17:01] I've seen that you've been busy on nano [17:01] Hmmm... I've been submitting but I can't make the new scanner work [17:01] It was nice enough :) [17:01] And the scanner/imp still only just enters the hill [17:02] Maybe that's because scanner/imp doesn't work? [17:05] hrrmmnn [17:09] technically this warrior should work, but it doesnt.. [17:10] Or the hill is just stupid [17:10] John: You could send me your warrior for optimizing :) [17:13] this paper is just absurd.. [17:13] huh? [17:14] my eXperimental paper still acts weird [17:14] scanner/imp at 40th on nano hill now [17:14] Weird in what way? [17:15] John: you remember my two-in-one paper? I made it heavier and it still scores relatively well against scanners, but crap against everything else [17:15] Yes [17:15] * Fluffy congrats John again for tlotd :) [17:15] Thanks :-) [17:16] Did you update the HoF? [17:16] Yes, today [17:17] :) [17:19] AAARGH [17:20] a little bit mor verbose, please ;-) [17:20] *more [17:20] i have a nice paper-structure which i cannot make work, but it is too precious to give away [17:20] Then don't give away, but send to me [17:21] I'll reinvent it later and claim to be totally mine [17:21] ok? [17:21] ah, but ive already sent it to person smarter than me [17:21] to John [17:21] and i havent heard anything about it from him [17:21] :-/ [17:22] maybe, because he isn't John ;-) [17:22] I'm currently sending my warriors to Fluffy to improve ;-) [17:22] then who am I! [17:22] Not John? [17:22] John The Clone? [17:23] Darth Vader? [17:23] i optimized the hell out of the paper and it didnt reach the hill last august [17:24] though i guess Roy could do "something" to it, like he did to Exact Abstract [17:29] * John waves [17:29] MSG: Quit: WLP {-1, {-1 [17:29] Roy, come 'ere! [17:33] I think you need more than that. [17:33] Some dark ritual maybe? [17:35] such wont help [17:35] how about telepathically sending a migraine to hes girlfriend? [17:37] You really don't like Roy, right? [17:37] i dont hate him [17:37] i hate hullabaloo [17:38] I like it. [17:39] It must have been difficult to make a more-style paper work together with a stone [17:39] blah, just optimization [17:40] remember that sunset was Recon2 with Moorepaper/stone [17:40] hmm [17:52] * Fluffy waves [17:52] MSG: Quit: Leaving. [18:35] hrm [18:35] bvowkenstein [18:36] how goes? [18:36] average [18:36] muching away the last of a bag of corn chips [20:09] hrm [20:12] damn.. [20:12] I've got collisions in my hashing algorithm :( [20:16] yikes [20:17] yeah. Its not really *THAT* bad however.. [20:17] just a little bit more collidy than I had hoped [20:18] Jens: I think I've got pycorewar leakless.. you might want to note in the docs that you *REALLY* need to *COPY* the results returned instead of using the directly. [20:18] so results = py_corewar_results[:] [20:19] otherwise it causes oddness. [21:12] Join: Fluffy joined #corewars [21:12] hey fluff [21:12] * Fluffy is quickvisiting [21:12] I've just seen you nano warrior [21:12] 4666 bvowk 1 131 0 11824K 11092K RUN 21.3H 97.80% python [21:12] looks nice :) [21:12] looks like we're not leaking anymore [21:13] just showing one of the kids here what I was doing.. [21:13] yup :) [21:13] disregard that warrior ;) [21:13] you should drop the ".rc" at the end of the name [21:13] it would still be ugly [21:13] but it a little bit less ugly ;) [21:14] thats not the final naming scheme :P [21:14] look at nano :) [21:14] Nano Bob III is 36th :) [21:15] i just fired up this version on 3jane to see how it works [21:16] they've all just realized they can kill eachother.. [21:16] so we're a long ways off yet [21:16] MSG: Quit: Trillian (http://www.ceruleanstudios.com [21:18] How long does it usually take to evolve sth. useful, bvowk? [21:18] ... with your old evolver? [21:20] not very long [21:20] depends on how I boostrap it [21:21] I'm still playing with this one unbootstrapped too [21:22] I'm currently only given my generator a form and let it find sth. near-optimal against the given benchmark. [21:22] And even on ned that might take several hours [21:25] well, I've pulled the benchmark step out of this one.. [21:25] there's no real benchmark.. just a bunch of slightly randomized hills.. [21:25] the diversity seems to stay pretty high [21:25] Are you shure, that this will lead to acceptable warriors? [21:25] *sure [21:25] who knows ;) [21:26] but its interesting! [21:26] yes, that's true [21:26] they're slowly moving up the benchmark scale.. [21:26] I hope, that you will keep me informed with your progress [21:26] so they're not completely navel gazing [21:26] :) [21:26] And you only use 3jane at the moment, right? [21:27] yeah [21:27] its interesting to watch, because they really move by huge leaps [21:27] Do you have a GUI or just a print to the shell? [21:27] also had to peg the mutation rate *WAY* higher than I ever expected [21:28] its actually an interactive python process still ;) [21:28] it drops me to the >>> python prompt to look around every now and again [21:28] ok [21:28] would be nice to have page on your homepage, where we could see the progress [21:29] It shouldn't be that hard to implement ;-) [21:29] no.. not really [21:30] not really hard or not really interesting/nice? [21:32] not really hard.. [21:32] its pretty interesting :) [21:32] I hope that you make that page soon [21:36] so what are you working on? [21:36] I'm working on PyCorewar [21:37] since I haven't worked on the MARS itself for a while I had to retrace my steps for a while [21:38] ... and I've noticed that there are some problems with multithreaded Python-apps, which use PyCorewar [21:38] which I hope to solve soon [21:38] but it isn't sth. serious [21:39] apart from that I'm generating warriors for nano [21:42] MSG: [21:50] morning [21:51] ares: ooohhhkaaay ... [22:03] * Fluffy waves [22:04] MSG: Quit: fluffy.i < 1, # 42 [23:29] Join: nw joined #corewars [23:31] big [23:31] hi nw [23:37] Part: nw left #corewars