[00:12] MSG: Quit: humhum [00:36] bleh. [00:59] MSG: Read error: Operation timed out [03:38] Join: flyduck joined #corewars [04:36] MSG: Read error: Operation timed out [04:37] Join: pak21 joined #corewars [05:31] Join: myndzi joined #corewars [05:43] Join: myndzi\ joined #corewars [05:46] MSG: Ping timeout: 240 seconds [06:16] Join: myndzi joined #corewars [06:18] MSG: Ping timeout: 240 seconds [06:19] Join: myndzi\ joined #corewars [06:20] MSG: Ping timeout: 240 seconds [06:22] Join: myndzi joined #corewars [06:25] MSG: Ping timeout: 240 seconds [06:25] Join: myndzi\ joined #corewars [06:28] MSG: Ping timeout: 240 seconds [06:30] Part: NorthStar left #corewars [06:40] Join: myndzi joined #corewars [06:43] MSG: Ping timeout: 240 seconds [06:50] Join: myndzi\ joined #corewars [06:52] MSG: Ping timeout: 240 seconds [06:52] Join: myndzi joined #corewars [06:54] MSG: Ping timeout: 240 seconds [06:55] Join: fiveop joined #corewars [07:36] Join: Metcalf joined #corewars [07:56] MSG: Ping timeout: 240 seconds [09:34] MSG: Ping timeout: 240 seconds [09:58] Join: nescience joined #corewars [10:14] MSG: Ping timeout: 240 seconds [10:16] Join: Mannerisk joined #corewars [14:38] Join: fish joined #corewars [15:57] MSG: Quit: z/ [16:53] Join: NorthStar joined #corewars [18:02] Join: myndzi\ joined #corewars [18:02] MSG: Ping timeout: 240 seconds [18:08] Join: myndzi joined #corewars [18:10] MSG: Ping timeout: 240 seconds [18:11] Join: myndzi\ joined #corewars [18:12] MSG: Ping timeout: 240 seconds [18:51] can anyone comment on nano warriors vs. full-length ones? [18:51] will a competitive full length warrior usually beat even the best nano warrior ? [19:09] depends on if you are running it under nano settings or not >:) [19:10] * flyduck gasps [19:10] why didnt i think of that, damn :o [19:10] there's nothing aside from complexity keeping you from writing short warrios for a normal hill [19:10] but the environment both are written for is very different [19:11] it's more feasible in a nano warrior, for example, to just hit random spots in core when possible [19:11] i mean you could write 5 lines that dec various spots and get a good section of the core without any loops etc [19:11] for the last day ive been evolving 5-line warriors against 3 classic full-length warriors in the full-length setting [19:11] and wondering why my fitness wasnt exceeding 0.3 [19:11] because you've only been evolving for a day? :) [19:11] hehe i dunno [19:12] in truth i need to work on my GA quite a bit [19:12] anyway, small warriors just have less options [19:12] right now it cant beat rave [19:12] i mean, cant evolve something to beat rave [19:12] or anything on-par with that [19:12] i have some improvements to make [19:12] which is why it makes some sense to work on evolving for nano, because it is potentially such a huge task that you want to sorta crop it back and solve a simpler problem [19:13] working on this GA has been a learning experience at every step [19:13] oh also [19:13] isn't scanning kinda a lesser useful strategy on nano? [19:13] so putting em up against a full length scanner is kinda unfair [19:23] but, keep in mind that nano programs are tiny [19:23] would a full-length scanner find it? [19:26] Join: yoR joined #corewars [19:26] MSG: Client Quit [19:28] sure, about as often as it'd find a small dwarf or something [19:28] so guess what you are pushing to evolve? :P [19:29] the question is, are you trying to evolve a warrior for nano or for the regular hill? [19:30] I'm working exclusively on the nano hill, myself [19:30] flyduck has been working on the beginner's hill [19:30] but now I'm working on ... going to work! be back later. [20:09] hrm. [20:31] Join: myndzi joined #corewars [20:33] MSG: Ping timeout: 240 seconds [20:45] Join: myndzi\ joined #corewars [20:47] MSG: Ping timeout: 240 seconds [20:50] MSG: Ping timeout: 240 seconds [20:50] Join: myndzi joined #corewars [20:54] Join: myndzi\ joined #corewars [20:56] MSG: Ping timeout: 240 seconds [20:59] Join: myndzi joined #corewars [20:59] MSG: Ping timeout: 240 seconds [21:02] Join: myndzi\ joined #corewars [21:03] hm [21:03] the way im doing things now, i may have an 'evolution toolkit' in a couple of days [21:04] MSG: Ping timeout: 240 seconds [21:05] MSG: Ping timeout: 240 seconds [21:06] Join: myndzi joined #corewars [21:08] MSG: Ping timeout: 240 seconds [21:11] Join: myndzi\ joined #corewars [21:14] MSG: Ping timeout: 240 seconds [21:15] Join: myndzi joined #corewars [21:17] MSG: Ping timeout: 240 seconds [21:20] Join: myndzi\ joined #corewars [21:23] MSG: Ping timeout: 240 seconds [21:28] Join: NorthStar joined #corewars [21:39] MSG: Quit: humhum [22:02] MSG: [22:02] Join: johnkw joined #corewars [22:13] Join: Metcalf joined #corewars [22:13] Hi :-) [22:14] 'lo Metcalf [22:17] man, i don't get this [22:17] nobody sells a multi line office phone with a wireless headset [22:17] plantronics has made a big business out of selling these headsets that are pretty much addon hacks [22:17] but i can't seem to find just a normal goodamn phone that plugs into the phone jack [22:18] MSG: Ping timeout: 240 seconds [22:18] Join: Metcalf joined #corewars [22:21] I still have a poor connection [23:00] Join: Neogryzor joined #corewars [23:01] hmm [23:04] Hmmm? [23:05] Hi Neo [23:05] heh, hi [23:05] Nice call / ret code :-) [23:05] thanks [23:06] i liked your fix method. The problem seems to be solved [23:06] Did you see my latest version? The first call is slightly slower and there are 6 extra instructions [23:07] But you can use "call subroutine" instead of "call subroutine fix" [23:07] a bit too triky to my taste... ;) [23:08] Join: OoS joined #corewars [23:09] i was thinking on a new problem but it is too difficult [23:09] What problem [23:10] Just don't say Heap Sort or Merge Sort :-P [23:10] A routine that controls created processes and destroy anyone when needed [23:11] MSG: Ping timeout: 240 seconds [23:11] Hmmm... definitely tricky [23:11] Yeah, mostly if you want to destroy one [23:11] The routine would have to trace the code the process is executing to keep track of where the process is. [23:12] too complex without parser's help [23:12] If it's tracing the code, it might as well execute each instruction individually [23:13] crazy [23:14] E.g. write a rountine that maintains a program counter and executes one instruction at a time. The easiest way would be to copy the instruction elsewhere and adjust the pointers [23:14] Then place a jump back to the control routine after it, then jump to it [23:16] jmz end,@ptr / mov.i >ptr,1 / XXX <- (inst. to trace) / jmp -3 [23:17] (just awondering...) ptr points to the code to trace [23:17] also djn.b -3,#0 as a counter [23:17] The only problem is if the instruction to trace is a jmp! [23:18] yeah, lol [23:18] or a non zero dat [23:19] unpredictable with SPLs [23:19] It could be done by splitting to it instead of jumping to it [23:19] And ensuring every location it points to is a dat [23:19] Then checking what changes the instruction made [23:19] Hmmm... [23:20] I think I'll skip trying to write that. [23:20] too complex maybe [23:20] Sounds trickier than Heap Sort [23:20] also i was thinking on a way to check how many processes are active in a certain moment [23:21] * Neogryzor doesn't like sort algoritms much.. [23:21] Should be easy if someone implements this opcode http://corewar.co.uk/opcodes.htm#EXC2 :-) [23:21] I like the simpler sort algorithms [23:22] Insertion, Selection, Bubble, Gnome, Quicksort, Comb [23:22] :-) [23:22] The first 4 can be implemented in 10 lines of redcode. [23:22] The last 2 in under 20. [23:23] * Neogryzor thinks it could be done easyly using p-space and an assistant :P [23:23] I haven't written a Shell Sort yet (should be about 20). [23:23] Also I haven't finished Heap Sort (should be about 75) [23:23] And haven't even started Merge Sort! [23:24] Does OoS stands for "Origin of Sorts"? :D [23:25] ok, silly joke.. [23:29] Almost :-P [23:29] Origin of Storms! ;-) [23:30] Roy's on twitter now :-) [23:47] time to bed [23:47] * Neogryzor waves [23:47] MSG: