[00:21] bvowk, Supposed to be ;) [01:13] MSG: Quit: Trillian (http://www.ceruleanstudios.com [06:56] Join: fiveop joined #corewars [08:02] MSG: Quit: Lost terminal [08:46] Join: Mizcu joined #corewars [09:01] Join: pak21 joined #corewars [11:12] MSG: Quit: leaving [11:43] Join: Core29 joined #corewars [11:50] Join: FARAMIR joined #corewars [11:59] MSG: Quit: Trillian (http://www.ceruleanstudios.com [12:01] MSG: [13:24] MSG: [13:35] Join: elkauka joined #corewars [15:05] Join: pak21 joined #corewars [15:28] Join: Mizcu joined #corewars [15:37] Join: Core29 joined #corewars [15:54] hrm. [15:56] Join: aFlag joined #corewars [15:58] I tried to make a scanner, but it sucks a lot hehe [15:59] hi [15:59] well, scanners are hard to make [15:59] aFlag's new too? [16:00] what i do is jump 4 to 4 instructions and, if some instruction is not dat #0, #0 it writes spl 0 then jmp -1 jmp -1 jmp -1 [16:00] or to be more exact, not hard to make, but hard to make good [16:00] fiveop, yes [16:00] is there something conceptually wrong in what i do? [16:00] yeah, it took me several years of evolving to get a scanner :) [16:00] (and it doesn't even stick on the hill!) [16:00] do you dat-wipe the core after you have overwritten enough opponent's with spl's? [16:00] :P [16:01] yes [16:01] i do a core clean [16:01] bvowk: ever written a warrior by hand? ;) [16:01] and you might want to use 4 spl 0's instead of spl/j/j/j [16:01] core_clean spl 0, -1 [16:01] mov 1, <-1 [16:01] where the next position is supossed to be a dat [16:01] fiveop: a couple, but I'm not very good at it :) Its more fun for me to write the code to write the warriors :) [16:01] aFlag: self killing! ;) [16:01] ok, that should work against basic papers [16:02] never did like doing anything easy, when I could do it ass backwards :) [16:02] fiveop, it's not self killing [16:02] when it overides the mov 1, <-1 [16:02] it hangs [16:02] self stunning! :) [16:02] :P [16:02] well, it has to end sometime heh [16:03] doesn't have to [16:03] like keep doing core clean? [16:03] yes [16:04] do you use one mov per each bomb written, or do you use a loop? [16:05] what do you mean? [16:05] well, after you have found your opponent, do you use something like: [16:05] Hey [16:05] mov.i splbomb, >target [16:05] mov.i jmpbomb, >target [16:05] mov.i jmpbomb, >target [16:05] mov.i jmpbomb, >target [16:06] or something else? [16:07] i do this [16:07] bombardeio mov bomba, mov bomba, mov bomba, mov bomba, jmp achooo [16:07] where bomba is spl 0 [16:07] ok, you could use [16:07] and achooo is my finding code [16:07] mov bomba, djn -1, #4 [16:07] add.ab #4, -1 [16:08] it is a bit slower, but you save one line [16:08] i did like that because i was bombing with other stuff other than spl 0 only previously [16:09] yes, i thinked so [16:10] anyhow, what's a fast scanner? [16:10] i thought mine should be fast [16:10] but it isn't apparently [16:10] hehe [16:10] to say shortly, the bigger your scanner is, the faster it is [16:11] Heh, but that also makes it a bigger target [16:11] problem is that the bigger you are, the easier you are to hit [16:11] like the more code it has? [16:11] yes [16:11] aFlag, I wrote a scanner/paper that did OK in the tournament two rounds ago [16:11] But it might be too complex [16:11] hum, i thought a fast scanner was the smallest possible hehe [16:12] add/jmz -scanner checks one position per two instructions run [16:12] add/sne (or add seq) -scanner checks two per three instructions [16:12] bombers are usually better than scanner, then? [16:12] add/sne/add/jmz checks three per four, add/sne/add/sne checks four per five [16:13] hum [16:13] Not really better, and bombers have lots of problems beating replicators [16:13] should the scanner check every postion then? [16:13] checking every position takes too much time [16:13] that's what i thought [16:13] i'm jumping 4 to 4 [16:14] so you are using add #4, #4 jmz -1, @-1 ? [16:14] or add #4, 1 jmz -1, somewhere? [16:14] the problem is that adding 4 at a time is not good [16:15] why? [16:15] i thought there shouldn't be code much smaller than 4 [16:15] well, in theory, best scan searches for big thing first, then small, and then smaller things [16:16] i'm doing sub instead of add, i'm going backwards [16:16] I found a step of 11 worked well [16:16] Depends on the core [16:16] Depends on how you decode, I guess [16:17] a good scanner named hazy lazy uses stepsize 5690 [16:17] but i dont think it will work with your scanner [16:17] wow [16:17] that big? [16:17] It doesnt matter how big it is [16:18] recon 2 (another very good scanner) uses 6557 [16:18] but as datagram said, it is not the size, it is the pattern it makes [16:21] crazyshot2 checks every 10 instructions [16:22] but it doesnt use normal scan-bomb-scan-bomb-...-endgame -tactic [16:27] greets datagram [16:27] Hey bvowk [16:27] I need to get to bed [16:27] 08:30 [16:30] pattern? [16:31] this corewars thing is more complex than I thought hehe [16:31] ;) [16:31] What happens is that they scan 6 apart [16:31] and then move ahead however much [16:31] and scan those 6 apart [16:31] They dont scan 6, then the next 6, and so on [16:31] You have to move it around alot to maximize finding the enemy [16:32] It's why you dont bomb 1 locationa fter the other, too slow, and low probability of hitting the other person's code [16:32] i see [16:34] Where are you from [16:35] ? [16:38] brazil [16:38] where are you from? [16:38] USA [16:39] Finland [16:40] cool, people all around the world :P [16:40] a place with much to low average temperatur [16:40] Mostly europeans :) [16:40] I think I'm one of three USA people [16:40] Me chip and John? [16:41] Hurkyl, Philb [16:43] i think i'm one of one brazil people [16:43] hehe [16:43] :P [16:43] We're all pals anyways [16:43] Except for Mizcu [16:43] But shhhh don't tell him that [16:44] i have seen once a Brazilian person send a warrior to corewars.sourceforge.net when it was still up [16:44] I am not sure if it was you or someone else [16:44] hehe i'm sure there are other brazillians playing hehe [16:44] it was probably not me [16:45] br brazil ... thought of belgium ... don't know why :) [16:46] hehe we're a little far away and slight bigger than belgium, otherwise you'd be right :P [16:51] Join: cube joined #corewars [16:51] hi @ all [16:54] * cube is away: have a shower after training [16:58] training corewars? hehe [17:03] MSG: Quit: Hometime [17:11] * cube is back (gone 00:16:19) [17:11] no...cycling [17:14] do you go to big tournaments and stuff? [17:18] local races and national [17:20] does anyone knows a good source that explains a lot about scanners and replicators? [17:20] Fizmo's site (corewar.info) has a good paper tutorial [17:20] Check the lexicon under P [17:21] Jens has a quickscanner tutorial, but I don't think theres a normal scanner one unless it's in one of the core warrior magazines [17:23] cool, thanks [17:52] you can find some stuff about scanners at Metcalf's website [17:52] Join: sf joined #corewars [17:52] Hi [17:53] heya sf [17:53] I pushed one of your warriors off nano last night [17:54] * sf is astonished! [17:54] Just going to SAL to check you're not joking with me [17:57] How dare you. I thought you knew better [17:57] heh [17:57] uh oh.. [17:57] its a good thing I'm across a continent and the atlantic. [17:57] although, if you want to come kick my ass.. I'll buy the beer and pinball just to make it up to you. [17:58] I don't drink beer! A nice glass (or three) of wine would be better [17:58] heh [17:58] but the pinballs ok right? [17:58] otherwise the deals off! [17:58] Never tried it. Pool maybe [17:59] really? [17:59] when I was in enjunear skool, we joked that we had a class in pinball.. [17:59] They only have pinball in amusement arcades here [18:00] because if you were really good, you could do all the course work in about 1/4 the required time, and then go to the lounge and play pinball for the next 45 minutes.. [18:00] omg, sf is the oposite of me [18:00] sf, do you walk with your hands? [18:00] tragic.. what a backwards place! :) [18:00] Hi aFlag [18:00] aFlag: oposite? [18:00] Do you mean opposite side of the planet? [18:01] during the last years of school, we played billard at any possible time [18:01] no, like you don't drink beer, i drink beer, you never played pinball i don't like pool [18:02] pinball is awesome, i wish i played it more, nowdays is hard to find a good old pinball machine :( [18:02] yeah.. its very hard to find good pinball machines.. [18:02] there is only one tiny company making them now.. [18:02] and its difficult to buy parts to repair them... [18:02] so you've got one in your flat/house/whatever? :P [18:02] yeah, now all you see is those dance machines [18:03] no, my current place is way too small :/ [18:03] I've tried those (once) [18:03] but I've got a friend who's working on building his own.. [18:03] i've tried them also [18:03] but it's too hard [18:03] he's built a CNC machine, aluminum caster.. [18:03] the lights keep on blinking all the time, you never know where to step [18:03] he's working on getting an injection moulding setup going.. [18:04] Part: Core29 left #corewars [18:04] I don't understand how people can use those dance machines outside of some place that serves large Q's of strong alcohol... [18:04] My email isn't working properly again [18:04] likely because of the latest windows worm.. [18:05] much of the intarweb is in email palsy mode.. [18:05] Incoming mail seems delayed, bvowk's apology hasn't arrived yet. [18:05] lol [18:05] I have issued no appology! [18:05] Then prepare for retaliation [18:05] bring it on! [18:05] :) [18:06] I'm quite impressed with how the evolvers working now.. [18:06] I'm getting a little more variation than I used to.. [18:06] grats :) [18:07] I think I will start evolving too when I've got time :) [18:07] I'm hoping the submission picker will give it even more of a boost by picking submissions that hurt my warriors less, and your warriors more :) [18:07] right now I just pick the top scoring one and send it off.. [18:08] but since I benchmark incestously, my highest scores score thusly because they're good at killing me. [18:08] ... ;) [18:15] fiveop: going write your own code or play with someone elses? :) [18:18] I've sent a couple of emails to SAL but no results yet [18:19] might take a while today.. [18:19] they've implemented some heavy filters upstream from me.. [18:20] Join: FARAMIR joined #corewars [18:21] Hi [18:21] +) [18:24] hi Fara [18:25] you new here? [18:25] hi [18:25] yes [18:25] +) [18:27] MSG: Quit: Kopete 0.9.0 : http://kopete.kde.org [18:27] bvowk: own of course! :) [18:35] MSG: [18:38] * sf waves [18:38] MSG: Quit: I don't like you. But Bersirc 2.2 does. Try it out now. [ http://www.bersirc.org/ - Open Source IRC ] [20:17] Join: Core29_ joined #corewars [20:30] MSG: Ping timeout: 252 seconds [20:55] in spl #5 the # is ignored and it's the same as spl 5? [21:10] MSG: Quit: ... cu [22:02] * bvowk wonders. [22:02] where did met and wil get to? [22:10] holy crap! [22:10] there's a freebsd port of fmars now.. [22:10] thats kinda cool. [22:15] is it on ports already? [22:16] yeah [22:16] cool [22:21] 21:55 < aFlag> in spl #5 the # is ignored and it's the same as spl 5? [22:21] nope [22:21] the 5 is ignored and it's like spl 0 [22:35] MSG: Quit: humhum [22:53] Part: Core29_ left #corewars [23:06] in slt a, b, being a and b labels, will the slt compare addresses? [23:23] hum... i don't get how this scanner code autochecks [23:24] s/auto/self/