[01:44] Why is my age not going up on 94b? [01:45] There have been a bunch of battles against that hill but I get no emails about it, and my warrior's age doesnt go up... [01:48] unless you make it onto the hill.. the age doesn't cheange [01:48] only successful challenges that push something off [02:00] whats up datagram? you around? [02:25] Yeah [02:25] I'm ON the hill [02:25] And have been for about a month, I think [02:27] yes, but your age only goes up when someone pushes something on the hill off with a new submission. [02:27] attempts that don't actually push someone off, don't increase your age. [02:28] ok.. my house is now wired for gige with 9k jumbo frames. [02:28] very nice :) [02:38] Oh, I did not know that [02:38] I assumed it went up every time a challenge was made [02:42] naw, that'd be silly.. it doesn't count a challenge as aging unless there was a chance you could have been pushed off. [02:42] *shrug* [02:42] Seems illogical either way you look at it [02:52] why either way? [02:54] Because it makes sense to have it only when someone gets dropped off so people cant send crap warriors to increase it [02:55] But having it that way is not representative of how many battles that warrior actually survived through [03:00] yeah, but battles where there was nobody pushed off aren't worth considering.. [03:01] you don't consider fights where your a 50lb kid tried to attack you battles :) [03:01] amusing anecdotes maybe. [06:37] Join: fiveop joined #corewars [06:38] MSG: Client Quit [06:38] Join: fiveop joined #corewars [09:15] Join: pak21 joined #corewars [10:35] Join: caesar joined #corewars [10:48] Join: Mizcu joined #corewars [11:11] hi mizcu [11:15] hi caesar [11:18] MSG: Read error: Connection reset by peer [11:25] Join: caesar joined #corewars [11:34] MSG: Ping timeout: 252 seconds [11:40] Join: caesar joined #corewars [12:00] MSG: Ping timeout: 252 seconds [12:06] MSG: Ping timeout: 252 seconds [12:09] Join: fiveop joined #corewars [12:10] Join: Roy joined #corewars [12:11] Hrm... anybody here? [12:12] ja [12:15] I want to brag about my highest optimax score :-) [12:15] 154.66, has any warrior higher..? [12:16] MSG: Ping timeout: 252 seconds [12:17] Join: fiveop joined #corewars [12:17] Actually might have [12:17] But i dont recall immediately [12:18] You might? What kind of warrior? [12:18] Heh, not me [12:20] I know the old Hullabaloo (and two) scored about 152, but Nenad said yesterday his warrior scored 153, so I had to improve it a little [13:19] yay 155.33 (using optimax now!) [13:54] hrm [14:15] MSG: [15:07] Join: Neogryzor joined #corewars [15:07] hello [15:23] greetings neo [15:27] greetings bvowk, how are you? [15:28] pretty good.. [15:28] except my upstream freebsd feed site mangled my freebsd mirror :/ [15:29] I have to download via *ugh* 100Mbit [15:29] :) [15:32] doing anything interesting neo? [15:34] some coding for DAQ system [15:35] and looking at CSEC Round2 results [15:35] cool.. [15:35] is your daq work or play? :) [15:36] they are distributed I/O modules already. They work well; my code not at all... [15:37] I had to recode my postprocessing widget last night :( [15:37] (again) [15:37] (for the fourth time) [15:38] one day it'll be able to process the whole hill without needing 20GB of memory [15:38] woah! [15:40] you know what thats for right? my 152,000 warrior nano hill :) [15:40] works out to about 11552000000 battles or so [15:41] I got all the scoring done, but generating the score matrix and writing the report turns out to be a big pain.. [15:41] A really brute hill!. Haven't you thought to remove a*few* of them? [15:41] mysql says it'll do it, so I got that fired up yesterday :) [15:41] heh [15:42] yeah, but whats the fun in that? :) [15:42] speed? [15:42] 152k = pretty much every nano warrior I ever evolved [15:42] it took a couple weeks to run all the scores [15:45] * Neogryzor suspects bvowk is frenzy about the nano-hill :) [15:45] heh [15:48] And the main question: What are you going to do with the hugh results list? [15:48] post it to my webpage.. [15:48] and wait for someone to beat my record :) [15:49] I wonder if I can get in the world record books for largest corewar hill :) [15:49] (and maybe use the score matrix to conquer nano!) [15:52] Surely you would do, unless an insane redcoder kidnaps one of those supercomputers... [15:53] heh.. [15:54] why kidnap when you can buy big machines off ebay? :) [15:54] Hasn't appeared a nano-scanner among your evolved warriors? [15:54] yeah.. [15:54] well.. it wasn't evolved.. [15:54] it came from the not evolver [15:55] er.. kinda.. [15:55] the old evolver made an almost working one-shotty thing.. [15:56] and the not evolver made it work.. [15:56] but its not terrifically strong [15:57] "not evolver"? is it just a optimizer, or it makes your warriors more primitive? [15:59] well, its not an evolver, its kinda makes little code snippets and tests them, and stores the good ones in a database.. [15:59] and then it looks for commonality between the chunks, and tries to extract useful information it can use to make better blocks [16:00] it *kinda* works.. but the problem space turns out to be *WAY* bigger than I estimated.. [16:00] so its kinda sitting waiting for me to figure out what to do with it.. [16:00] I'm contemplating rewriting it in scheme so I can do some funky self rewriting things... [16:01] and do less testing and more thinking ;) [16:06] Join: fiveop_ joined #corewars [16:06] MSG: Read error: Connection reset by peer [16:12] Did you try it for 94nop hill warriors? [16:12] or tiny? [16:13] just nano so far.. [16:13] but even nano is very huge.. [16:14] I don't think I'd get anything usable for any of the larger cores. [16:14] although, there is alot of existing code that I could import instead of trying to build it from first principals.. [16:14] which could very well bootstrap.. [16:15] almost everything done for nano is either something I did, or something based on one of my warriors, or something that my evolver can't seem to work with (scanners) [16:15] (or something completely different, but I've got 150k warriors vs a few (thousand?) human coded [16:18] What a disadvantage! :) [16:19] heh [16:19] I'm just saying that 94nop might actually be better if I imported other warrior bits to work on :) [16:20] but the space to test would be insanely huge. [16:22] Does it test the bits of code against warriors or against other bits? [16:23] warriors for the most part.. [16:23] although.. thats kind of up in the air.. [16:24] it could do it any which way.. [16:48] MSG: Read error: Operation timed out [16:54] Join: pak21 joined #corewars [17:06] * Neogryzor has to go now [17:06] cya neo [17:06] bye [17:06] MSG: Quit: Physical Wreck IRC webscript: http://www.pswclan.com/cgi-bin/irc/irc.cgi [EOF] [17:15] MSG: Quit: Lost terminal [17:15] Join: elkauka joined #corewars [17:20] Part: elkauka left #corewars [17:45] Join: Fluffy joined #corewars [17:45] :) [17:52] Join: Metcalf joined #corewars [17:52] Hi [17:52] Hi John! [17:52] Anything new? [17:53] Don't know ... [17:53] Join: fiveop joined #corewars [17:54] Hi Fiveop [17:54] Anything new on your site over the last few days? [17:54] re [17:55] nope [17:55] I miscalulated the number of different 94nop warriors possible [17:55] it's not 4587520000000000 [17:55] but [17:56] (18*8*8*8000*8*8000)^100? [17:56] MSG: Ping timeout: 252 seconds [17:56] 1438126416563134739158006617624379486167197185926170979429666255378680450058419567567313865808954197764313331793539142218484475003675406228336919490172190617446272369770760697057845691615546438505022332282615344444651442857582955091738481751529612721109821508665559043584363819875677544018789362661349925311416234985144957126344826198464074708291144316069493690325126504157947958827926644427019882817130784071555910596577648375568286475996587006664 [17:57] ((16*7*8*8000*8*8000)^100)*100 [17:57] Are you sure? [17:57] yes [17:57] Why *100 at the end? [17:57] 16 not 17 cause cmp == seq [17:57] end start? :0 [17:57] Of couse. [17:58] fiveop: You result isn't right! It has to end with at least 2 zero, but doesn't! [17:58] Then why not a double zero at the end of your enumeration [17:58] Fluffy: 10^....? [17:58] if the *100 is correct [17:58] ups [17:58] someone's gonna have nervous breakdown ;) [17:58] - * 10^ [17:58] + many many 0 [17:58] so stupid :/ [17:58] Maybe you should write it down somewhere ;-) [17:59] http://rafb.net/paste/results/M8USjn17.html [17:59] ... [17:59] Join: fiveop_ joined #corewars [17:59] MSG: Client Quit [17:59] hehe [18:00] imagin (?) to run all these warriars against each other [18:00] with -P! [18:00] You'd need a Quantum 'puter [18:01] or wait some years ;-) [18:01] how many fights for -P on nop settings? [18:02] 15602 [18:02] it's 1522 for tiny [18:02] thanks [18:02] 15602 [18:02] hmm ... too slow ... as always [18:02] You was quicker with the ending in 00 comment! [18:02] http://rafb.net/paste/results/mbUXDT51.html [18:02] Join: Roy joined #corewars [18:03] fight's with -P in a round robin ;) [18:03] Join: Nenad joined #corewars [18:03] Hi [18:03] what a incredible huge number ;) [18:03] hi [18:03] Hi Nenad [18:03] Hi all [18:03] training now [18:03] by [18:03] e [18:03] (hi Nenad) [18:03] bbl [18:03] but still onilne ... as always ;0 [18:03] Hi Roy! [18:04] That is a nice big number ;-) [18:04] Congrats :-) [18:04] Roy: so, you've decided to make a better optimax score :) [18:04] Thanks.. [18:04] Nice warrior and a perfect name :) [18:04] on what? [18:04] Heh yeah :-) what is your highest score nenad? [18:04] well, still 153 - haven't finished yet, though [18:04] -> Congrats on beating him this round [18:05] oh :) [18:05] I'm mostly doing some 94t warrior making today [18:05] Nice scores indeed, too bad s/p's don't do well on the current hill, see Halcyon (good score for example koenigstuhl, 2nd, but bad on hill) [18:05] yes, true... [18:05] Mandragora is evil [18:05] Yeah its good [18:06] Bvowk: you can't get into the record books for anything to do with Corewar. [18:06] I read the submission guidelines on the Guinness Book of Records site. [18:06] Hehe, and..? [18:06] There is the Corewar Book of Records if you'd like to submit something for consideration however! [18:07] Why isn't it allowed? on what grounds? [18:07] it's not "interesting" [18:07] probably [18:07] Hrm well I can live with that, but for me its more intresting then what the highest number of pingpong balls you can put in your mouth :-) [18:08] ... does anyone know what is considered a good score against 88 fsh? [18:08] No idea... [18:08] Try some koenigstuhl 88 warriors maybe..? [18:08] I thought that 145 was ok for my first 88 sbi, but it scores bad on the current hill... [18:09] Join: Core29 joined #corewars [18:09] so I was wondering if the score limits are different, or maybe the bench isn't balanced? [18:09] Never really made 88 programs (beside the occasional tournament) [18:09] it is not so easy... at least for me. [18:10] I'm used to exploiting every possible aspect of redcode, so. [18:10] ... I don't feel comfortable with using only default modifiers. [18:10] Yeah indeed, its hard to go back one step (at least thats how it feels) [18:10] I could live with only <, but that modifiers thing is just annoying [18:11] I made a really good stone and then realized that I used add.x by accident ... and couldn't fix the problem later [18:12] run pmars with -8 so it tells you there's a problem [18:12] Hehe yup [18:13] yeah... I realized the mentioned problem AFTER I already optimized the stone, so... I wasted a few hours.. [18:13] never going to make that mistake again, though [18:14] MSG: Quit: Trillian (http://www.ceruleanstudios.com [18:15] well, atm I want to make a large and good 94t fsh, so that I could use optimax to aid me in creating future 94t warriors. everything so far is not optimized, well a little optimized, in the extent that is possible by handpicking the values. [18:16] Roy: which warrior did you optimize to 155? [18:16] HullabaThree :-) [18:16] :) [18:16] are you going to submit it? [18:16] Or something like that :-) [18:16] or did you do it just for fun? [18:16] Maybe.. I guess not, maybe on the 94 hill [18:17] why not? [18:17] if it's better... [18:17] Because I made the Hulla's to age, not to be killed by me ;-) [18:17] heh [18:17] 209 is a good age, and the first one was also something like that... I could have had a pretty old warrior by now! [18:17] yes. you could've [18:18] but Hullatwo is much better [18:18] you remember the difference when you submitted the new version? [18:18] According to the hill, but not according to optimax I think [18:18] optimax warriors are not alive on any hill [18:18] Yeah, it had a big lead on the hill once I submitted it again, this one will probably have it too [18:19] besides, Fizmo will probably release the new benchmark soon [18:19] Heh, then we can optimize it further ;-) [18:19] :) [18:19] And THEN I might resubmit it [18:20] did you make any plans for this tournament round? [18:20] No not yet.. [18:20] Still lot of time left though [18:21] I have a plan ;-) [18:22] you always do :) [18:25] you just have to sign your name and your warrior will win ... [18:25] Submit Type-1 nobody will win that! [18:25] Why not just tweak the fsh? [18:25] ? [18:26] swap some current hill warriors for some equivalent old warriors. [18:26] Then your optimax results will be closer to the hill results [18:26] Lots of warriors beat Type-1 [18:27] well, I don't really have hill warriors... except for my own [18:27] Less warriors beat Prototype-1 however :-) [18:27] and it's not my goal to optimize against them [18:28] There's a new name almost on the nano hill. Richard Halstead [18:30] OOPS! my parkinson hands pushed send twice, now I send my warrior 2 times and will kill a innocent warrior :-( [18:30] Yeah I was just going to ask you about the new name [18:30] Roy: you murderer [18:30] I think I might turn up at ICC2006, just to get some revenge! [18:31] how are you going to kill it? will both warriors die if you do ;kill warriorname [18:31] yeah, both will die.. (I think)( [18:32] ohw hmm, no... the first one I send had a bug and will be last so I'll push it off..! yay [18:32] no innocent victims! [18:32] Nice score Roy [18:32] Metcalf, you'd better turn up, I expect you :-) [18:33] Hmm, maybe I misread the optimax score, its not 155, it was 15.5 ;-) [18:33] err ... it seems, I've missed this conversation ... it is an improved HullabaTwo? [18:33] yeah [18:33] again? [18:33] well, should be :-P [18:33] Roy seems to think it's improved, but looking at the score I wouldn't agree [18:33] :) [18:34] Do you intend to kill HullabaTwo on 94nop? [18:34] Roy was probably drunk while optimizing it [18:34] Just had this conversation with Nenad :-) [18:34] No, I want it to age.. [18:34] But its always nice to have a predecessor waiting, and 94d is a testhill ;-) [18:35] Now I know, why no warrior has come near SoV yet. Everybody seems to be quite eager to kill good warriors to replace them with supposedly better ones. [18:35] and everyone else are also eager to kill those warriors [18:35] Roy: But the way ... 53.92 is a nice improvement to you killed HullabaTwo ;-) [18:36] That one was buggy ;-) [18:36] Now you say that ;-) [18:36] The new one is 6th while the old one was 13th, that seems a good improvement.. [18:37] Let's wait until the hill is dated [18:37] Then its 5 or 6 or 7th... depends, but not higher or lower [18:38] A general question: Is there a good guide out there, which describes how to "breed" good warriors (evolve), i.e. how to create a new warrior from two others? [18:38] And this was only by optimizing the qscan a bit, and one new constant, now I'm actually changing all the others [18:39] Just cut them up and paste them, and the cutpoint is randomly chosen (general GA) [18:39] hmm ... I thought, that there is more "intelligent" way [18:40] Maybe there is, but this works.. :-) [18:40] You should check some GA sites, most work like this, but maybe there are other solutions [18:40] k [18:41] Some people tried to work with CW-genes [18:41] Join: Core29 joined #corewars [18:41] (parts of warriors that work together etc) [18:41] But no idea is that worked [18:44] There is an article in an early issue of TCWN about Corewar Genes [18:47] JM: you can do better than that. incendiaries can do well now at that hill [18:48] according to my tests, at least [18:48] It's doesn't have an incendiary bomb. [18:48] oh, so Tornado-style? [18:48] spl/jmp so it would work under lp too [18:49] like WW3? [18:50] Like this http://www.corewar.info/tournament/cwset2005/round2/toffee.red [18:51] Looks like I broke it anyway :-( [18:51] yes. it should've scored better, I think [18:55] JM: Where can we find the old TCWN's? the ftp seems to be down [18:55] Planar's page. [18:55] But it isn't in one of those online [18:55] Hrm.. so you have it? [18:55] Upload them fool :-) [18:55] Yes. [18:56] However I have them as paper copies. [18:56] Fizmo should include them on his website too, in the newsletter section [18:56] And they are copyright W.Buckley [18:56] Ohw hrm, thats a bit harder to upload [18:56] scan them and convert to pdf [18:56] And W.Buckley needs to give permission before I can do that. [18:56] No, scan them with a pic-text thingy [18:57] I don't mind retyping the whole lot, if I'm allowed to put them online. [18:57] Heh if he is still alive (acient stuff) [18:57] I did that with the Darwin articles which are in http://corewar.co.uk/temp/ [18:57] There is a copy of TCWN there too, which I converted to html [18:57] He's still alive [18:58] <- still works..? [18:58] Wrong person! [18:59] Mark Durham's issues of TCWN are all online. [19:00] Ohw hehe, nice article about how to 'train a warrior' hehe [19:00] MSG: Read error: Operation timed out [19:27] * Nenad waves [19:27] * Metcalf waves [19:27] MSG: [19:31] Time for me to go too. [19:31] * Metcalf waves [19:31] MSG: Quit: mov.i #1,1 [19:46] MSG: [20:57] Metcalf could bring his TCWN copys to ICC2006 [20:58] So far only 146 in stone-bench, need better.. [20:58] on second thought, maybe not.. [20:59] stone-bench? [21:00] mostly scanners, meant for optimizing stones [21:02] 5 numbers to optimize.. need to cut that down [21:02] hmhm [21:03] Jens should get "Biocorewarlogist" talking about warrior genes and such [21:04] Part: Core29 left #corewars [22:05] 139 Wilfiz with no optimization so far <3 [22:08] however; creativity: 0 [22:26] cpy paste? :P [22:27] no, it just has nothing original [22:27] Just a version of Hulla.. [22:29] It had interesting boot-idea, but it wasnt good enough [22:33] MSG: Quit: humhum [22:52] MSG: Read error: Connection reset by peer [22:55] Join: Mizcu joined #corewars