[00:09] hrm.. [00:09] fizmo eh? [00:55] MSG: Read error: Connection reset by peer [01:49] MSG: Quit: humhum [03:09] MSG: Ping timeout: 252 seconds [03:11] Join: datagram joined #corewars [03:15] MSG: Ping timeout: 252 seconds [04:43] Join: pkhuong joined #corewars [04:44] Roy (when you read logs): Speedierscanner is now 7 lines long + an invisible line. [04:44] (foo 0, 0) [04:44] :p [04:46] MSG: Read error: Connection reset by peer [04:46] Join: pkhuong joined #corewars [04:56] Join: AI_coder joined #corewars [04:57] hello [04:57] I'm new to corewars... [04:59] hi [05:00] no exactly the best timing [05:00] *not [05:00] but heh [05:02] So how does execution work? [05:03] It just goes follows standard eip sequence. [05:03] lol [05:03] I'm thinking as I'm writing and that made no sense. [05:05] Forget what I'm saying I'm reading the beginners guide [05:06] k. [05:07] Basically, yes, it just falls through to the next instruction [05:07] by default [05:07] jumps/seq/seq[cmp]/dat change that. [05:07] the most important thing to remember is the everything is always relative to the currently executing instruction. Sort of like PIC. [05:08] (not the ucontroller, position-independent code) [05:08] So how does the game start? Where are the programs loaded? [05:08] random? [05:08] yes. [05:08] with a given minimal distance. [05:08] Execution starts in a random place too? [05:08] no. [05:09] first instruction of each warrior by default [05:09] can be changed with the end or org pseudoinstructions [05:09] So there are two eips? [05:09] yes. [05:09] -e [05:09] or more [05:09] 1 for each process. [05:09] clock you mean? [05:09] no. [05:09] you can have more than 1 process/warrior. [05:10] They timeshare like unix multitasking? [05:10] but they're scheduled fairly between each warrior. [05:10] yes. [05:10] How do you have more than 1 process/warrior? fork? [05:11] In the debugger, the first warrior's first instruction is given the address 0, by convention. It doesn't mean anything really. [05:11] spl. [05:11] It queues spl's target at the end of the process queue. [05:11] read the guide. [05:11] Am right now :D [05:13] Oh ok spl=split [05:14] So as an example... spl is loaded, enemy moves, new process one move, enemy moves, address following spl moves enemy moves... [05:14] no. [05:14] the new process is queued last [05:14] so process 0 executes [05:15] other warrior dispatches [05:15] wait. [05:15] *thinks* [05:15] test it in the debugger ;) [05:15] I haven't got that far... [05:15] I started 20 min ago. [05:15] k. [05:15] the specifics usually don't affect the way your program will work, so you can wait until the guide explains it. [05:19] Does mov mean move or copy? [05:23] copy. [05:23] the same thing mov does in asm. [05:23] Actually I just played w/ asm, I never learned it well. [05:31] Is it possible to watch a corewar? [05:31] Or spectate in some form or another? [05:32] the debugger [05:32] and/or the display at a slow enough speed [05:32] Is there any form of multiplayer? [05:32] Networking perhaps [05:32] uh? [05:33] you just load n warriors (files) in the core. [05:36] So as an example I would do pmars -e aeka.red flashpaper.red [05:36] yes [05:36] After that what do I do? [05:37] i believe h brings up the help [05:37] Ok step (enter)n times [05:39] How do I get it to step more than once say 1000 times? [05:39] again, read. [05:39] Oh cool it worked. [05:40] sk 1000 [05:41] In the debugger purple is a dat? [05:41] no. [05:41] What is purple? [05:42] pmarsv? [05:42] Just pmars I think... [05:43] read the manual [06:32] MSG: Quit: Lost terminal [08:52] Join: Mizcu joined #corewars [10:31] Join: Roy joined #corewars [10:31] pk: did you solve the problem of the double-scanning-no-hitting? [10:33] (as inversed described on r.g.c) [10:34] Nenad solved that but didn't post his solution yet [10:43] * Roy is afk again [11:06] Join: fiveop joined #corewars [11:36] Join: Core29 joined #corewars [12:29] Part: Core29 left #corewars [13:00] Roy: double-scanning no hitting? [13:00] back to r.g.cw [13:03] ah k. [13:03] needs one more line, though :/ [13:03] could be done... Likely to mess up the stepping, though. [13:39] Join: Neutrino joined #corewars [13:39] hi [14:33] MSG: Ping timeout: 252 seconds [15:19] Join: Metcalf joined #corewars [15:19] Hi [15:33] Join: pkhuong- joined #corewars [15:33] MSG: Read error: Connection reset by peer [15:36] Hi [15:36] MSG: Read error: Connection reset by peer [15:36] Hi Roy [15:36] Join: pkhuong joined #corewars [15:36] pk: Maybe you can ask Nenad how he fixed it [15:36] Heh, just in time :) [15:36] Hi JM, no luck yet with the primes :) [15:37] I found some of 731 size (by accident) [15:37] And one sized: 30526 [15:38] :-) [15:39] But those aren't so big I guess [15:39] Join: pkhuong- joined #corewars [15:40] 3 years ago 30526 digits would have made it into the top 5000 [15:40] MSG: Ping timeout: 252 seconds [15:40] Heh too bad we live in 2005 [15:41] Now you need 63142 [15:42] Only twice as big... could do that probably (if this only took half a day) [15:42] Size 63659 now :( [15:43] Don't bother starting a search for anything with less than 64500 digits though. By the time you find one it will be too small for the list [15:43] Heh true.. [15:43] I found two actualy 23523*2^101388 and 23523*2^102882 [15:44] MSG: Ping timeout: 252 seconds [15:45] You will notice the higher N gets, the longer it take to test each candidate [15:46] Yeah, I've restarted it, now starting from 200000+ [15:46] Those numbers will be larger [15:47] I have one machine left, shall I also search for a 2667? What range do you have left yet? ;-) [15:47] I would start from N=215000 otherwise you might mind something just under the list [15:47] Roy, I have searched up to N=300000 [15:48] How do you calculate the size of those primes, no program I have here can do that easily :P [15:48] Erm... [15:50] hey met [15:51] Multiply N by 0.30103 should be a rough estimate [15:51] Hi Bvowk [15:51] Within about 10 digits [15:52] Ok, thanks! [15:55] Bvowk: do you have access to the original Core War articles from the U.S.A. editions of Scientific American? [15:55] erm.. [15:55] maybe [15:55] I thought someone had already put scans online [15:56] Yes, they have. I am not sure if they are from the American editions of the magazines [15:58] * Roy is away again [16:00] Does everyone know what Mars stands for? [16:04] well.. I can take a peek [16:04] :-) [16:40] Roy: found a new prime already? :P [16:44] Roy did any of the smaller primes you found have a comment? [16:44] If they have a comment, they might make another list. [16:44] Time to go [16:44] * Metcalf waves [16:44] MSG: Quit: mov.i #1,1 [17:00] Memory Array Redcode Simulator? [17:09] Join: pkhuong joined #corewars [17:10] Roy: well. fixing the scan but no bomb would require an additional insn :/ [17:11] otoh, it's relatively easy to beef the scanner up w/ some bombing. [17:18] No idea, I don't know what your scanner looks like :) [17:24] :p [17:24] 7 lines + 1 invisible one :p [17:25] 7+1 is with or without clear? [17:25] I guess without, but with you I never know :) [17:29] without clear [17:29] :p [17:29] heh. [17:29] I might be able to put a self bomb in it and have a weak clear. [17:30] but yeah, if we have to go bigger, it'd probably be worth it to make it a bombing scanner [17:32] Well, just wait until (maybe) Nenad published his findings (sometime) then you can make a real version, have you tried this version yet in some sort of bench? [17:32] Does it have any potential? [17:33] I don't even have any mars on my computer [17:33] and no bench suite..and no constant ;) [17:33] doubt it's got any potential, though [17:34] I'm trying to trim away another insn [17:34] we'll see [17:35] want me to pm you what I have so far, or do you want to try and find it? [17:42] there's always the hack of starting to bomb between our 2 scanned locs [17:42] (if we scan x and x+small_int [17:45] Join: fiveop_ joined #corewars [17:47] MSG: Ping timeout: 252 seconds [19:07] Join: pkhuong- joined #corewars [19:07] MSG: Read error: Connection reset by peer [21:58] MSG: [22:38] MSG: Read error: Operation timed out [23:11] Join: soarkalm joined #corewars [23:11] welcome, soar [23:11] thanks [23:12] I just remebered the term corewars from some reading I did some time ago, and now i'm looking into it [23:12] and ended up here from the koth web page [23:13] this irc-channel tends to be a bit quiet, especially during europian nighttime [23:14] ok, that is fine...i'm often on a irc channed for my LUG...and only 4 people are ever there [23:14] but people read the logs at www.koth.org/irc-logs/ quite often, so you can just leave here a question and someone has probably answered it by next day [23:14] www.corewar.info is _the_ corewar portal these days [23:15] thanks :) I'm still reading the web page at koth [23:15] ahh [23:15] i'll follow that link [23:18] any thoughts on the corewars ebuild on gentoo? [23:18] not currently [23:18] ok [23:19] i'm watching the java applet from the website [23:19] interesting [23:19] hmm? [23:20] i though corewars.sourceforge.net was dead [23:20] and its only place i remember a java applet with view [23:20] hmm [23:20] http://www.corewar.info/emulator.htm [23:21] very last link [23:21] ah [23:23] http://packages.gentoo.org/ebuilds/?corewars-0.9.13-r1 <- the corewars -emulator there is no longer supported [23:23] or not.. [23:23] hmm [23:23] yes, its the dead one [23:24] see http://sourceforge.net/projects/corewar [23:24] pmars is the emulator for general use [23:24] looking... [23:25] for windows-users there is corewin, that needs no command-line tinkering [23:27] ok.. I might load that up just to see what it is [23:28] to do a fight in pmars, run pmars warrior1.red warrior2.red [23:28] ok.. i'm going to see if the source code will compile and run on my amd64 linux box here [23:28] for pmars [23:29] i have no idea how/if it will work under 64 [23:29] me neither :) [23:29] but it won't take long to find out [23:32] eek.. this might take a little longer :) [23:32] I am not much help about pmars on a 'nix, since i use combination of corewin and Dos-pmars [23:33] that's fine.. i'll tinker a little bit.. [23:33] can't break anything to big :) [23:34] Join: Fluffy joined #corewars [23:34] hi Jens [23:34] :) [23:34] kinda late for you, isnt it? [23:35] yes, but I was kind of busy :) [23:35] I've been playing around with my xmas presents [23:36] Neo: I hope, that you will read this ... could you please tell, which of data files for the score surfaces you couldn't read [23:37] Anything new today? [23:38] the source forge code seems to work nicely on amd64 :) [23:38] thats good [23:38] Fluffy: not really [23:38] Hi soarkalm :) I guess, that you are talking about the pmars sources [23:39] Fluffy, yep...remembered reading about corewars somewhere and finally decided to look into it [23:39] Mizcu: At least now we've made it officially, that nothing has happend ;-) [23:41] soarkalm: I hope, you will have some fun with it :) [23:41] i'm working at it :) and thanks [23:42] interesting [23:42] the pspace.red warrior is giving me an asser failure [23:42] *assert [23:42] argh ... as usual I'm quite stupid. Why do I leave here a message for Neo, if I can reach him by email [23:43] soarkalm: Each warrior file can (but not has to) contain an ;assert line [23:43] Error in line 10: ';assert VERSION >= 80 && ROUNDS > 1 [23:43] with it you can make sure, that the warrior runs under the conditions, it was developed for [23:43] ahh ok [23:43] so if i set those conditions. it'll then work [23:44] Simply replacing it with ";assert 1" will always work [23:45] ahh and pspace warrior is more of a demo of features than a warrir [23:45] bah [23:45] *warrior [23:45] You should go to the Koengistuhl page [23:45] It is some kind of warrior archive [23:45] Koeningstuhl might be kinda overwhelming in its stock [23:46] *Koenigstuhl [23:46] Mizcu: Do you have the URL? I don't have my bookmarks at hand [23:46] http://www.ociw.edu/~birk/COREWAR/ [23:46] :) [23:47] thanks [23:47] Of course it might be overwhelming, but at least better than the example warriors :) [23:49] heh redcode is reminding me of the motorola hc11 assembly i did for a class [23:51] :) [23:51] :/ I could write a compiler for redcode [23:51] that'd be nasty ;) [23:51] Nick Change: pkhuong- changed nick to pkhuong [23:52] Hi pkhuong! [23:52] if you can get it both faster and atleast as trustly as pmars's, then why not [23:53] Hey, Mizcu, don't tell all about my secret ambitions for PyCorewar ;-) [23:53] everyone knows them already [23:54] Everyone? knows? [23:55] well i didn't [23:55] :) [23:56] MSG: Ping timeout: 252 seconds [23:56] Mizcu: Maybe it is just vaporware ;-)