[16:04] Join: John joined #corewars [16:04] Hi :-) [16:05] Deserted here today? [16:05] quite [16:07] mail.google.com [16:07] Hmmm... wrong application ;-) [16:23] Join: Fluffy joined #corewars [16:23] :) [16:34] Hi Fluffy [16:34] :) [16:35] Seems as if I've been kicked out from #corewars. Was the server down? [16:38] your shell bonged [16:38] * Fluffy kicks bvowk [16:38] * Fluffy feels better now [16:39] Join: fiveop joined #corewars [16:40] HI Fiveop [16:40] Does anyone have a copy of fmars? [16:40] not me [16:41] Oh, not to worry, the page is working today :-) [16:43] John: What is the highest speed you can bomb with? [16:43] with a loop [16:43] Depends how big your willing for the loop to be [16:43] max 5 or 6 insns [16:43] (coresize-1/coresize) * 1 [16:44] Mizcu: huh? [16:45] Well with real bombs you can do .6c in 5 lines [16:45] 7999 mov -instruction and a jmp [16:45] thats pretty close to 1 [16:45] Mizcu: ha ... ha [16:45] John: With the Tornado engine? [16:45] Yes [16:45] nothing faster? [16:46] Well you could add a decrement [16:46] add.f inc, ptr [16:46] Hmmm... that won't work :-) [16:46] :( [16:47] Anything else involves scanning of some kind [16:48] Maybe a tornado look with a SNE *ptr, @ptr / ADD.F inc, ptr at the beginning? [16:49] About .71c [16:50] You could also use the look instruction to check somewhere [16:51] hmm .. thanks :) [16:53] look EQU loop ;-) [16:55] Everyone seems to have jumped on the website update bandwagon :-) [17:11] Join: Fizmo joined #corewars [17:11] hi hi [17:12] Hi Fizmo [17:13] Hi John [17:16] Is anyone using reddit? [17:19] Join: Fizmo_ joined #corewars [17:20] MSG: Ping timeout: 240 seconds [17:20] Nick Change: Fizmo_ changed nick to Fizmo [17:20] My girlfriend nicked my photos for her blog - http://clearmystudentdebt.blogspot.com/2007/04/holidays-are-over.html [17:20] Are you here now? :-) [17:43] Join: Fizmo_ joined #corewars [17:45] MSG: Ping timeout: 240 seconds [17:45] Nick Change: Fizmo_ changed nick to Fizmo [17:50] Join: Fizmo_ joined #corewars [17:52] MSG: Ping timeout: 240 seconds [17:52] Nick Change: Fizmo_ changed nick to Fizmo [18:00] Join: Fizmo_ joined #corewars [18:03] MSG: Ping timeout: 240 seconds [18:03] Nick Change: Fizmo_ changed nick to Fizmo [18:08] Who wrote the probability mars and implementation tricks pages? [18:10] pvk [18:11] Hmmm [18:12] hmm what? [18:14] ??? [18:15] Hi Fluffy ;-) [18:15] :) [18:15] Hmmm = pvk doesn't have a backup of the wiki [18:15] :( [18:15] and it's not in the web archive [18:15] But I have my Solyanka :) [18:15] Solyanka? [18:15] hmm??? [18:17] http://en.wikipedia.org/wiki/Solyanka [18:17] http://de.wikipedia.org/wiki/Soljanka even has a picture, but that's not very similar to mine :) [18:17] (apart from the red color) [18:19] hehehehehe [18:19] looks somehow [18:19] tasty ;-) [18:20] It *IS* tasty :) [18:20] Other opinions will be met with dealy force from my side! ;) [18:20] *deadly [18:23] Join: Fizmo_ joined #corewars [18:24] * John waves [18:24] MSG: Quit: mov.i #1,1 [18:24] * Fluffy waves [18:26] MSG: Ping timeout: 240 seconds [18:26] Nick Change: Fizmo_ changed nick to Fizmo [18:26] * Fizmo waves [18:27] too late [18:29] MSG: Quit: ChatZilla 0.9.78.1 [Firefox 2.0.0.3/2007030919] [18:39] Join: Pavel joined #corewars [18:40] hi there [18:40] Hi! [18:41] Hi Pavel! [18:41] how may i/we be of assistance? [18:42] Few things .. [18:42] :) [18:42] 1) Fizmo invited me :-) [18:42] 2) I just released nMars 0.9.2, your feedback is welcome [18:43] 3) maybe I would like to talk about my changes to pMars [18:43] so thats why the name was so familiar [18:44] what kind of changes to pmars? [18:44] http://nmars.sourceforge.net/pmarsdll.shtml [18:45] I would like to merge it to trunk in source control [18:46] Did you test your changes under *nix? [18:46] Yes [18:46] which *nix? [18:46] suse [18:47] that's hardly *nix :) [18:47] bsd ? better ? [18:48] :-) [18:48] No, I was talking about several different *nix-variants and not just one [18:48] and what about macos? [18:49] I have no Mac hardware [18:49] any volunteer ? [18:49] I just want to know :) [18:49] aint got no smelly Mac [18:49] try rgc [18:50] as you know, I tried some time ago, no response ... [18:50] darnd [18:52] why don't you maintain you own "fork" for a while [18:53] that way you can implement whatevery you feel fit and maybe users will switch eventually [18:55] Joonas forked long time ago and i feel that is time to merge [18:55] but if everybode happy with that, i'm ok with branch [18:56] that would probably be the easiest way [18:56] i think that comunity is too small to create lot of branches .. [18:57] and for newbie it is very complicated too [18:57] it will probably be that way forever :) [19:03] ok another topic, my borning warrior strategy [19:04] first prototype is here [19:04] https://nmars.svn.sourceforge.net/svnroot/nmars/warriors/my/train.red [19:05] It is something which is moving slowly forward, so that instruction is sliding together with task ip [19:07] You main problem is that you are too slow [19:08] motivation is save some cycles by eliminating jmp of djn instruction [19:08] MSG: Read error: Operation timed out [19:08] yes, you are right it is currently [19:08] slow [19:09] but worker threads could be multiplied, so rate will be better [19:09] apart from that you only seems to bomb your opponents with DATs [19:09] which doesn't work very against multi-process-warriors [19:09] exactly [19:09] any advices ? [19:09] a common strategy is to stun them with "SPL # 0" [19:10] and kill them later with dat-clears [19:10] i know, but that code do not slide well [19:11] also multithreaded bootstrap is slow :-( [19:11] apart from that you can try to use other bombs than "dat 0, 0" [19:11] everything else (except dat $ 0, $ 0) can act as a decory against scanners [19:11] *decoy [19:12] currenty i'm just moving blocks... [19:12] which seemed to me fast [19:12] Join: fiveop joined #corewars [19:13] but multi-copy-process warriors could survive anyway [19:13] papers? [19:14] almost anything [19:15] the warrior i've send is bit complicated ... [19:15] simplest form is: [19:17] https://nmars.svn.sourceforge.net/svnroot/nmars/warriors/my/train_base.red [19:17] this one is pretty fast [19:24] which seems to be some kind of stone [19:24] why don't you check some more common designs? [19:34] ok [19:50] MSG: Read error: Operation timed out [20:01] Join: fiveop joined #corewars [20:23] Part: Pavel left #corewars [20:26] MSG: Read error: Operation timed out [20:38] Join: fiveop joined #corewars [21:29] MSG: Quit: Lost terminal