[05:44] MSG: Ping timeout: 252 seconds [05:44] Join: Caelian|w joined #corewars [05:53] Join: Metcalf joined #corewars [06:16] MSG: Ping timeout: 252 seconds [06:16] Join: Metcalf joined #corewars [06:33] Join: bazziman joined #corewars [06:38] Join: OoS joined #corewars [06:39] MSG: Ping timeout: 252 seconds [06:43] Hi Bazziman [06:44] Are you new here? [06:45] oh, good catch Met, i was elsewhere [06:53] MSG: Ping timeout: 252 seconds [07:04] hello [07:04] Join: Metcalf joined #corewars [07:06] bazziman: hi, are you new to corewar? [07:07] yes i am [07:07] im just reading the beginners guide [07:07] and making myself familiar with ares [07:08] interesting i find some people here :) [07:10] It's quiet here due to Christmas [07:10] What kind of programming have you done before? [07:10] assembly, c++, ruby [07:11] in that order :D [07:12] What kind of assembly? :-) [07:12] oh what was the name of the c64 cpu ? 6502? [07:12] and 68030 [07:13] I never liked the 6502 :-( [07:13] i didnt do too much there [07:13] more 68000 [07:13] I found the z80 and 6809 much more friendly [07:14] i loved 68000 [07:14] z80 wasnt it the specrtum? [07:15] Yes, and quite a few other 8 bit computers [07:16] ah, never had one [07:16] buit a freind of mine [07:16] is there an active redcode scene? [07:17] i watched some pages this night [07:18] like that koh [07:18] Yes, we're active. [07:18] sort of [07:19] But there are only a few new submissions to the hill every week [07:19] afk, bbl [07:19] hm after watching some warriors [07:19] they looks all the same to me :) [07:20] and the strategies seem all the same too [07:20] just some tweaking with constants and optimizations or am i wrong? [07:20] Lots of tweaking and optimizing happens [07:21] But it's possible to come up with new stuff which is effective [07:21] really? [07:22] isnt everything ionvented there? [07:23] There are some general categories [07:23] like scanners, replicators, stones, vampires, oneshots, etc [07:23] yes [07:24] would u explain these for me in simple words? [07:24] Someone will come up with a new type of scanner, or a new hybrid scanner/stone or just a better way of implementing something. [07:25] some words i have heard already, like vampire, stone, paper, scanner [07:25] Have you seen this page? http://corewar.atspace.com/top10.html [07:25] i am here: Beginner's Guide to Redcode [07:26] After reading the beginner's guide, I recommend reading something which covers the strategies. [07:26] I recommend http://corewar.co.uk/mintard/index.htm [07:30] yes this is nice [07:30] did u send your own warriors to some king of the hill? [07:41] Yes, I'm currently on a few hills :-) [07:43] I have a few warriors on this one http://sal.math.ualberta.ca/hill.php?key=nano [07:46] seems like u did lost of versions on the warriors :) [07:46] lots* [07:50] I normally make several versions and keep the best one [07:50] And if something doesn't work, I'll put it to one side then try to improve it a few months later [07:52] since when do u code redcode? [07:53] about 1993! [07:53] wow 14 years now [07:54] u must be very familiar with all these addressing modes and such [07:54] i need much training i guess [07:55] The addressing modes aren't too bad [07:55] Basically you can refer to a location directly or indirectly [07:56] so there are $ (or nothing) to point directly to a location [07:56] yes [07:56] what irritated me was [07:56] or you can use a location as a pointer to a second location - < @ > or { * } [07:57] which partr of an instruction is selected a or b [07:57] but with for example move.ab its simple to see [07:58] You can always compile the code [08:01] i'll just read through these pages and try to write some simple warriors :) [08:01] i saw [08:01] there are begginer's kohs [08:02] i wonder how good the warriors there are [08:02] and really i dont knopw how good any warrior is :) [08:04] MSG: Ping timeout: 252 seconds [09:53] Part: bazziman left #corewars [15:46] Join: Metcalf joined #corewars [15:51] Join: pkhuong joined #corewars [15:52] john? [15:52] (metcalf) [15:55] Hi pvk :-) [15:55] Haven't seen you for a while :-( [15:56] hey. I have to bounce a random low level idea, and nobody's around [15:56] so I figured i'd come here ;) [15:56] :-) [15:56] have you worked on garbage collectors before? [15:56] What kind of low level idea? Not random numbers I hope :-P [15:56] No [15:56] did bv or he who must not be named talk about the PRNG stuff? [15:57] It's covered in a few books I have but I've never needed it [15:57] Everyone is rec.games.corewar was talking about PRNG [15:57] * pkhuong shall read [15:57] Problem was some of it was complete rubbish [15:58] As usual. PRNG's hard stuff :\ [15:58] Yes [15:59] I've been hacking on a common lisp compiler for a while, and am trying to add a concurrent garbage collector. One of the problems is that if the program is gonna be writing stuff while you're trying to figure out what's live and what's not reachable anymore, you have to somehow maintain a coherent view of memory. [15:59] Harder in corewar particularly if you don't know the coresize [15:59] impossible if you don't know the core size. [15:59] lagged fibonacci works regardless of coresize ;-) [16:00] just took a final year course on simulations, prngs and models. [16:02] ISTR smth re additive lagged fibonacci being surprisingly bad re period, even w/ large moduli. multiplicative lagged fib works fine, but there's no analytical way to appraise their goodness, so far. [16:04] additive should give a good period according to Knuth [16:05] Knuth's book is decades old ;) [16:07] So, re GCs, one of the ways is to basically work (conceptually, anyway) with a snapshot, from some time = T. I thought I could just fork and let the OS handle that part ;) [16:09] * Metcalf checks his bookshelf [16:10] Problem is, I've haven't read about anyone doing something like that, so there must be something lossy about it? [16:16] Not much in this book - just mentions concurrent garbage collection is possible [16:17] Argh. That moron Skybuck Flying has to be cut off the net. [16:17] :-) [16:17] He's also been polluting comp.arch for more than a year [16:17] :-( [16:20] Dijkstra, Lamport, Martin, Scholten and Steffens (1978) On-the-fly garbage collection: an exercise in cooperation. Comm ACM, 21, (11), pp 966-975 [16:21] Yeah, classic. Requires a lot of cooperation from the mutator. [16:22] That's the only reference listed here. [16:23] Does it sound even remotely sane? [16:24] I haven't read the paper and the book doesn't say much :-( [16:44] grtz pvk, met [16:45] Hi Mizcu [16:48] damn, found refs. [16:49] one's a patent, but it probably fails the non-obvious requirement [16:50] Join: OoS joined #corewars [16:51] MSG: Ping timeout: 252 seconds [16:58] Join: Metcalf joined #corewars [17:01] MSG: Ping timeout: 252 seconds [17:04] MSG: Ping timeout: 252 seconds [23:13] grumble, grumble [23:47] MSG: Quit: pkhuong