[17:10] Join: Altair joined #corewars [17:10] hi there [17:32] can a ask you guys for some help. I sure the solution ist very simple, but i just cannot find the mistake in my little warrior. its just a warrior which copies itself, so thatthere is an exponential growth of tasks ( 2^x). but somehow its stops. the 16 task doesnt start the way it is supposed to start. [17:33] the spl points to a wrong place -.- [17:33] but i dont know why. [17:33] start mov 9,313 [17:33] sub #1,start [17:33] djn.a start,start [17:33] copy spl 307 [17:33] sub #50,backup [17:33] sub.a #50,copy [17:33] mov backup,start [17:33] jmp start [17:33] backup mov 9,313 [17:33] dat #0,#0 [17:48] the first problem is that your paper confines itself to a very small area in the core - when all copies are so close, there will be some overwriting [17:49] yeah, okay, i know. but i tried very different values and simply watched what happens. everytime there is some task which tries to etablish a new task at a place where isnt any code [17:50] you consider that is the problem? [17:51] or is there some ...mm....did i simply made in error in troughtinking this? [17:51] with papers, there will always be collateral damage [17:52] but at a later time......the problem appears after 16 tasks. [17:52] well, 1,2,4,8 are working fine. [17:54] so this is not a problem of the structure of the warrior but that papers damage themselves a lot? [17:56] the paper 16 is malformed, yes, it doesnt have the djn.a copied [17:57] the source paper has two processes instead of one [17:58] o.o? [17:58] something that is worth to fix [17:58] ? [17:58] the problem is that the paper itself is not really worth "fixing", it needs an overhaul [17:59] ^_^. well, im a beginner, so dont mind pls. okay. [18:00] i just focused on a paper that should recreate itself very fast [18:01] is there any kind of advise you could give me. probably there is a solution which is faster? well, there should be. :D [18:01] you dont need tha sub #1, start [18:02] you dont need to keep the backup for first line with you [18:04] mm.........but how to decrease/increase the b value of the mov? [18:05] @_@ [18:05] of course [18:05] put the targetvalue somewhere else, and use an < [18:05] i forget the simplest things >.> [18:06] and you change the the djn to an jmn with an #-1 [18:06] no.....mm........ [18:06] with an -1 [18:07] okay. thank you very much. [18:12] ugh, i never was good with "old-style" papers [18:13] ? [18:15] i started developing that paper of yours forward to give a hint or two, but i noticed that mov x, then i changed into a jmn based-loop, and gave up [18:15] well, its not that bad because modern papers are lot easier to write anyway [18:16] ^^. that's my problem too. you prefer what warriors? [18:16] modern papers @_@? [18:17] yes, in '94 a "silk engine" was created, which doesnt use a loop for the copying (or rather, the copyloop is infinite in size..) [18:18] sounds suspicious :D [18:18] its makes copies like pair of epileptic rabbits [18:19] get a warrior named "timescape" and youll see what it looks like [18:21] this one is supposed to be one of the very first new style papers :D : http://www.ociw.edu/~birk/COREWAR/94/HILL/silkwarrior13.red [18:21] true [18:26] kind of scary :D [18:29] well, thx again^^ [19:17] sry, one more question. the next thing I will focused on should be the new papers, more time on old papers, or scanner? [19:18] a little more time on old papers, then scanner [19:20] k, thanks [20:09] Join: Metcalf joined #corewars [20:09] Hi :-) [20:09] h [20:09] Oh, missed Lukasz [20:11] Hi Mizcu [20:12] you got an explanation why } scored better in my weirdopaper? [20:14] Join: Fizmo joined #corewars [20:14] hi hi [20:15] hi John [20:16] I've had a tough weekend, walking across the fells and crawling down little holes in the ground [20:16] ohh ,really? [20:18] hi? [20:20] MSG: [20:23] MSG: Quit: ChatZilla 0.9.82.1 [Firefox 2.0.0.14/2008040413] [21:15] Hmmm... [21:29] ? [21:56] MSG: Ping timeout: 245 seconds [21:58] Join: Metcalf joined #corewars