[00:00] MSG: Quit: humhum [00:33] MSG: Quit: leaving [06:44] Join: brx joined #corewars [07:59] Join: fiveop joined #corewars [08:32] Join: Mike joined #corewars [08:32] Hello [08:44] MSG: Remote host closed the connection [08:56] Join: Nenad joined #corewars [08:56] Hello Nenad [08:56] Hi [09:00] I just turned the heating on and it smells strange and I've got a headache. [09:01] Switching off again I'd rather sit here and shiver than get Carbon Monoxide poisoning [09:06] :) [09:15] it is cold here, too. [09:15] Did you come up with a name for Jen's next warrior? [09:18] Join: Core29 joined #corewars [09:18] Hi Core29 [09:19] Hello Core29 :) [09:19] No name yet. [09:20] but I think it should have a strategy line like [09:21] ;strategy visit www.mikes-page.com to purchase fantastic warrior names [09:23] on-speed is no more the master of the nano hill :( [09:25] ? it fell off? [09:31] Yes by a warrior with a rubbishy name [09:38] well, names don't matter much, unfortunately [09:38] besides, it lived long enough [09:39] and nano is a hill where most action takes place recently [09:41] You should be there too [09:41] :) [09:41] well, coding for nano is a little boring in my opinion [09:42] I've made some nano warriors, but they died quickly. [09:42] optimize! [09:42] well, I could do that, of course :). But I'm optimizing some 94nop warriors at the moment. [09:43] Do you think most of the evolved warriors are core-clears? [09:44] yes. [09:44] probably. [09:44] maybe also some papers in larger cores. [09:44] and it also depends on the method [09:45] if you evolve from scratch, you'll get mostly clears [09:45] if you combine existing warriors randomly, you'll get many things... [09:45] I don't know what beats clears [09:45] WhiteNoise is a oneshot, for example... [09:45] and was KOTH on the tiny hill for a long time [09:45] it is still alive... [09:45] and with a good score [09:46] evolved by MicroGP corewars collective [09:48] Evolving might suit me :) [09:49] tha make an evolver! [09:49] *than [09:49] I meant downloading the program and using it! [09:50] well, you can do that, too :) [09:54] What strategies can be made in 5 lines? [09:54] And which can't? [09:55] coreclear, paper, scanner, stone, oneshot, qscanner, ... you name it... [09:55] I don't know about any vamps, though... [09:55] and hybrid strategies can't fit into 5 lines [09:56] s/p, s/i, and such [09:56] Neogryzor even made a nano zoom trick scanner! [10:08] so nothing has an advantage because everything is possible [10:10] I didn't say that. some things are advantegous. but many things are possible [10:11] core-clears seem to be good. [10:13] yes. they are. but try to code one by hand. the chances are that it won't enter the hill. the warriors on the hill are very optimized... [10:13] qscans did ok some time ago. and they are still on the hill. [10:18] Are any of the qscans published? [10:18] hmm... I wouldn't know. [10:21] how did you learn about corewar? [10:22] Someone told me about it [10:22] Same person who told me to shut down the chatbot! [10:22] :) [10:22] so, when did you first come here? [10:23] August 16th [10:23] I've got a CD full of Corewar stuff and I haven't even looked at most of it [10:24] * Nenad thinks that you should look at that CD [10:27] I've read most of the Corewarrior but skipped the bits I don't get yet (most articles) [10:28] There should be a newbie CW mag [10:28] Just like to get newbie computer mags and techie computer mags [10:28] a good idea! [10:28] (although there are newbie tutorials..) [10:29] the only problem is that there aren't enough beginners to interest someone into making a small beginner magazine [10:30] If there was a magazine there might be more beginners [10:32] that is true [10:32] but to have more neginners, we first need to make more people find out that there is such a thing called corewar [10:32] *beginners [10:33] I was thinking of maybe organizing a presentation at my university.. [10:33] Talk about it in newsgroups blogs and forums [10:34] and a series of introductory lectures, perhaps.. [10:34] Someone arrange to start a thread and others reply the new people will join in or look at the websites mentioned [10:34] Presentation is a good idea. [10:34] Put a video online! [10:45] Even just a beginners website would be good. [10:50] Join: Roy joined #corewars [10:51] hi Roy! [10:51] Howdy Nenad [10:52] how are you? any new warriors in preparation? [10:54] Well, not really.. [10:54] Out of ideas at the moment [10:54] lol [10:54] hI rOY [10:54] I'm never out of ideas. only out of time... [10:54] You? [10:55] mE? [10:56] Well, I was asking Nenad, but do you Mike? [10:57] No ideas No time No Hope [11:26] Part: Mike left #corewars [11:28] MSG: Ping timeout: 252 seconds [11:47] MSG: [11:51] Join: Mizcu joined #corewars [12:30] Join: Jens joined #corewars [12:30] :) [12:38] hi jens [12:39] Have just read Nenad's posting. [12:39] He's right. [12:41] Yeah.. he always it [12:41] *is [12:42] always? [12:46] Most of the time :) [12:48] I'm now thinking about the idea to make sth. like a CoreWarrior for beginners. [12:49] But we already have the CoreWarrior. Maybe it is enough to add a beginner section [12:49] We could do that too.. but the problem is finding the information, and making it public [12:49] The CoreWarrior is only published 2/3 times per year now [12:49] There is much to write about. [12:50] Just think about the explaining the basic strategies [12:50] It would take quite a lot of time. [12:51] Then sth. about improving warriors from beginners [12:52] Thats true, but we have already done that.. [12:52] The old CoreWarriors [12:53] Sure we could write it again, but in a year that is also old information; like the first couple of CoreWarriors are now [12:53] Yes, but it should be done several times per year [12:53] But that'll take a lot of time, rewriting the same information [12:54] If 3 or 4 people would contribute regularly, it would be enough. And we would get enough of material for beginners [12:57] I don't know.. [12:57] A short overview about the main events on the hill and especially on the beginner's hill (maby 94m, too) and two or three short articles would be enough. Every 2 or 3 months [12:57] Its always fun to hear that, and write articles about new findings etc, I also plead for more CoreWarriors :) [12:58] I would also love to give lectures, at schools etc, to teach them the basics about CW [12:58] At college I did it this year for all my teachers, they really liked it/the ideas but they don't have time :( [13:03] The number of dat-warriors, that were sent to the hill has increased quite a lot during the last days [13:03] yeah, but they don't really count the age etc :) [13:04] The hill ages a lot at the moment, no artificial aging required :) [13:04] I meant: dats don't increase the age.. [13:06] And so I've understood it [13:06] Ah ok, so there isn't any artificial aging by sending datwarriors ;-) [13:12] Roy, you should have learned it by now. Sending a warrior, which name is "test", won't score good, because of its name ;-) [13:12] How do you think, a warrior feels, when his name is "test". [13:16] Oops, forgot about that.. [13:16] lol [13:16] I'll send a fluffy name next [13:16] Hey, that is not allowed [13:16] I've trademarked all "Fluffy" warriors ;-) [13:17] "Hairy Furry Blue Eyed Bunny" is allowed! [13:17] OK, you're right. [13:17] (A warrior with such a name won't even enter the hill ;-) [13:17] But think of the name-length-limit [13:18] You suggest something good then, something untrademarked! :) [13:18] Let me think [13:18] (it is hard) [13:18] ... [13:18] I'm also in a dilemma.. Solo 4, to qscan or not to qscan [13:19] Basic Equation [13:19] How about that? [13:19] Fluff + Fluff = KOTH [13:20] "Ruined in a day" [13:20] "Touched by the hand of Roy" ;-) [13:20] Cozy Fuzzy Furry Squashy [13:21] That'll do ! [13:22] Actually to keep it simple it has to be called Solo 4 :-( [13:22] pfff [13:22] Join: sf joined #corewars [13:23] Hi sf! [13:23] Hi Jens [13:23] sf: What are you thinking about sth. like CoreWarrior, but especially for beginners? [13:24] That is a good idea [13:24] If I remember correctly, you have just begun to do cw. [13:24] Like Steve's Guide [13:25] No, I've been around a while [13:25] oh, sorry [13:25] As usual my memory isn't that good [13:25] sf = S.Fernandes = FatalC = http://corewar.atspace.com/ :-P [13:26] I know your name ... it is on my homepage :) [13:27] Your history page is nice [13:28] Thanks ... I'm already working on similar pages for the hills at SAL [13:34] Nenad's post has some interesting ideas [13:35] Yes and I'm still thinking about a CoreWarrior for Beginners [13:38] Reading logs. [13:38] :) [13:38] Corewarrior for Beginners ought to contain results for 94b and multiwarrior [13:39] Beginners are often lucky on multiwarrior. I was very lucky! [13:39] Don't you have the oldest warrior for 94m? [13:39] I mean, you have written it, didn't you? [13:40] or am I wrong again? [13:40] I had one age 260. [13:41] That has been beaten at least twice. [13:41] Metcalf should make a 94m hall of fame [13:41] Hmm ... once I have made the image for 94m, I'll now that exactly [13:42] For 94nop I have only included all warrior, that have reached an +1000-age. Can you think of anything else to add? [13:44] Something about who is the current oldest [13:44] Like http://corewar.co.uk/logs.htm says [13:44] Good idea. I just check the logs and change the page [13:45] And you could mark suggested causes against drops in age [13:46] ? [13:46] There might have been a big drop in average age after Optimax was published. [13:47] When was optimax published? [13:47] If there was it should be marked on the graph [13:47] Or when I new kind of warrior was popularised/invented/published [13:47] Exactly [13:47] But I couldn't think of an example apart from Moore paper and bvowk's snippet [13:47] Moore style papers are maybe the reason for the new long down [13:48] Stone/paper age, after Reepicheep [13:48] According to corewar.info optimax was published July 11, 2004 [13:48] It is version 0.63. Was it the first? [13:49] Not long before Son of Vain dropped off [13:49] No idea, could be [13:49] And what do you think about mentioned the most recent events first [13:49] instead of the current Iist I have [13:50] When Moore published his warrior or when Fizmo made spiritual black... something like that [13:50] Hmm ... ok, I have some ideas now. Let me work a little bit :) [13:51] Jens: we are going to make lots of work for you! [13:51] But is interesting [13:51] 11 jun 2003 <- Birk posts on r.g.c that Sunset is posted and king on koenigstuhl 94 [13:53] Let's find sth. like "Stone age", "Paper age", "Optimax age" ;-) [13:54] :-) [13:55] lol that would be nice [13:55] Hrm, I think I'll stop making warriors for today, I have 7 on the hill :-O [13:55] can you make a graph of the balance of the hill [13:56] What should that be? [13:56] How many paper, how many scanners and so on [13:56] Thats hard to categorise... some warriors are unknown, some are hard to place [13:56] Or a simple one, the balance of tie prone warriors to warriors with few ties [13:57] Average score is also intresting, or the difference between the top warrior and the bottom one [13:57] I've already made an average-score-graph. I is almost a flat line at about 135 [13:58] Heh.. [13:58] But then you can zoom in and see the small fluctiations [13:58] Do you know when Blacken was sent to the hill [13:58] ? [13:58] No.. [13:59] it must be somewhere between Sep 4, 1998 -Nov 29, 1998 [13:59] Sun Nov 29 10:59:43 EST 1998 [13:59] Blacken is 11 then :) [13:59] Yes, I have JM's logs [14:00] It is quite a pity, that so many data about the first challenges is missing [14:00] It is age 63 in corewarrior 70 [14:01] Maybe Ian knows [14:01] I'll soon post r.g.c. with a set of questions [14:01] Probebly not.. he's planning on maybe making a new warrior based on Blacken [14:01] It was age 2 on Sun Oct 18 03:31:52 EDT 1998 [14:02] Where did you find that..? [14:02] Yes, where? [14:02] r.g.cw Koth results. [14:02] url? [14:02] And it wasn't there previous week [14:03] http://groups.google.com/group/rec.games.corewar/msg/1277042a20fc7aec?dmode=source [14:03] Wow, cool [14:04] Hehe Vladimir Bychkovskiy...? WHO?? :) [14:04] By the way, does anybody know the most successful completely evolved warrior for 94nop? [14:05] I already have the first completely evolved warrior [14:05] Ask bvowk, he'll know :) [14:05] Where is bvowk? [14:07] Gone..? [14:07] He posted some days ago, that he'll ge away for some months, but gave no reason. [14:08] Ah.. he said something about a roommate giving him a ride home or something [14:08] fiveop: The words in names can be found at http://corewar.co.uk/records.htm [14:10] While googling found this website, is it new or do you guys know it? [14:10] http://www.srcf.ucam.org/~pak21/corewar/warriors.html [14:11] No, that has been around for a long time [14:11] It used to be on geocities [14:12] Lol, my Solo 4 was 8th three challenges ago, and now last...ready to be pushed off... [14:13] Hsa scanners just don't seem to survive as long as they used to :) [14:27] OK, I've changed http://corewars.jgutzeit.de/history/index.en.html a little bit [14:27] MSG: Read error: Connection reset by peer [14:28] Heh, reordered, I like this better (first thing I notice!) [14:29] I'm now writing two or three lines about "current events" and then some useless things like lowest score ever to enter the hill, ... [14:30] Yay, thats nice [14:30] And JM is busy on his who-is-who, curious what he is going to make from that :) [14:30] But I need some more ideas for "useless information" [14:32] most ties/biggest lead/most consecutive king/etc? [14:32] why not [14:32] Hard to find that sort of stuff out.. [14:32] Apart from Arrow being the oldest warrior, any other "news"? [14:33] No it isn't hard to find out. For the graph I've written a script, that parsed all logs for 94nop, that I have. I only have to change two or three lines of get that sort of information [14:34] I have 7 on the hill, Fizmo 5, Nenad 3... [14:35] Most warriors with the same score (3/4 is possible, I have seen that!) [14:39] Highest Score ever (according to incomplete 94nop-logs): Inky by Ian Oversby, score: 172, Februar 24, 2002 [14:39] Wow, that is a lot :-O [14:39] On the same hill the lowest score was 70 [14:40] 18 26/ 43/ 32 Kernel Trojan Joshua 108 9 [14:40] 19 23/ 75/ 2 jBBB James Huggett 71 1 [14:40] 20 22/ 75/ 3 jBBs James Huggett 70 3 [14:41] Heh, not fair, I can break that record too if I kill two warriors and let loose some warrior that HullabaTwo trashes ;-) [14:42] Do that and I'll mention it on my page :) [14:43] Nah, that wouldn't be fait [14:44] But you will be mentioned anyway for lowest score ever on hill [14:45] Hmm, but there are so many warrior with an age of 5 [14:45] *score [14:46] papertest 5, SpTest, paper test, Uninvited (hehe!!!), The Three-Handed Knight, test, SPtest, Detonation, Permanent Delusion [14:47] Lowest score ever on hill? :) [14:47] 5 [14:48] Maybe lowest score to enter the hill [14:52] Lowest score to enter hill: 21 - Yippie-ya-yeah by Uncle Pain, January 23, 2003 [14:52] I have left all warrior, that were killed/replaced, but stayed on the hill, out [14:53] And Yippie-ya-yeah died at the age of 3 (!) [14:56] Hehe it actualy survived challenges :) [15:00] Apr 22, 2004: Last time the highest number of different authors (=16) was reached [15:01] Unfortunately the number isn't correct as some author use different names :( [15:01] Roy = Roy van Rijn [15:01] Fizmo = Christian Schmidt [15:02] Yeah.. sorry about that, HullabaTwo's Roy was a mistake :) [15:04] Most silly name: "t.A.T.u Girls Rule" by John Metcalf ;-) [15:04] Let's whether I can find a more silly name [15:04] lol, I remember that [15:05] The first warrior to have the a "test" in its name: SB-test by Paul-V Khuong [15:05] That was the 8th warrior every to be sent to the hill!! [15:05] Heh, and its till being used [15:05] wrong [15:06] the hill had an age of 17, when the warrior was submitted [15:06] not 8th [15:08] Let's see, whether I can find the hill with the highest number of warriors, which have "test" in their names [15:08] I think not more then 2/3 [15:09] one moment [15:15] September 10, 2003: Hazy Test 63, devilish 2.02 test, Sequential Test 18, psptest, aiptest, sikill-test [15:15] 6 test warriors [15:16] Wow 6.. [15:16] aiptest and psptest could been mine :P [15:16] one mo ... I have to find the corresponding hill [15:18] psptest by Roy van Rijn [15:18] Looked familiar :) [15:18] hehe ... and on the same hill, "test1" was pushed off by "sikill-test" [15:18] Hehe [15:19] Hrm, I have to go, dog is waiting and oven is pre-heating [15:19] * Roy waves [15:19] MSG: [15:40] April 25, 2002: The oldest 94nop ever (total age of all warriors on hill = 8319), 8 +500-warriors, only 6 warriors with age below 100 [16:25] Part: Jens left #corewars [16:35] Join: Jens joined #corewars [17:12] Join: Roy joined #corewars [17:12] Hi [17:13] hi Roy [17:13] Working on anything atm Miz? [17:15] finishing my dinner [17:16] Needs to be done too :) What was for dinner? [17:16] Pile of Spaghetti, Can of Tuna, and 2 Cans of Pineapple-rings [17:17] Simple but tasty ;-) [17:23] tasty? [17:23] omg [17:24] Better then Can of Beans, Pack of Bacon... [17:25] hmmmm .... no ;) [17:44] Join: willvarfa joined #corewars [17:44] hey will [17:44] Hi willvarfa! [17:45] hello folks [17:47] willvarfa: Can it be, that you know a little bit about evolving warriors? [17:48] I'm looking for the best (completely) evolved warrior, that has was on the 94nop-hill [17:48] Join: Nenad joined #corewars [17:48] Hi [17:49] Hi Nenad! Back again for yet another hill-aging-party? [17:49] no. unfortunately not. [17:49] well I've never got an evolved warrior on any hill [17:49] :( [17:50] well maybe a few competitions, but then you only have to enter to get on those types of hill ;-) [17:50] although I need to make more papers soon to save KL - it'll probably fall off soon.. :( [17:51] the rest of the time I was working on my future page, more specifically on oneshot tutorial. [17:51] :) [17:51] yay, then we'll know the secret behind KL! [17:52] how goes the project, Roy? [17:53] Pretty bad :) I got my points, but it isn't really working [17:53] It has a bug that I can't find/explain [17:54] When you start it works well, creates new and better warriors but after a while it stops making better warriors and suddenly creates random shit :-S [17:55] what kind of points did you get? [17:56] studypoints at school, nothing really intresting [17:56] top marks? goes towards your degree? [17:57] Going for my bachelor in informatics/programming [17:58] Have to go on a final internship and write my final thesis soon [18:00] and what kind of marks did you get? a lot? [18:00] Pfft hard to tell.. [18:00] Its just a 'did he work hard enough? y/n' [18:01] * asw waves [18:01] * Roy waves [18:01] who is leaving? [18:02] And because I held a presentation and noted I had read "On the Origin of Species" they said I did more then enough work :-S [18:02] I guess awm is leaving..? [18:02] replied to Nenad's post about updating the REC.GAMES.COREWAR.FAQ -- Anton had agreed to put it into my name a long time ago (so weekly postings would be up-to-date but he has not been able to to do it as of yet.) [18:03] good work asw [18:03] I guess you're leaving, which is a shame; I was wondering about gnu-pmars, and a bit about quantum whatsits [18:04] Please update the wiki page at http://freebiology.org/wiki/REC.GAMES.COREWAR.FAQ [18:04] we'll all contribute to updating the FAQ :) [18:04] willvarfa? I'm not going anywhere [18:05] Nenad - yah - it's just a pain that we can't stop the OLD FAQS FROM POSTING! [18:05] Why did you wave then..? :) [18:05] asw: how is the quantum corewar project going? [18:05] Roy - =^P [18:05] good good, thought you were waving. So how goes thoughts about gnu-pmars? [18:05] Or was it a welcome wave? [18:06] a mexican wave, maybe; only Roy joined in though, making all us others party poopers [18:06] I'm in the middle of migrating from our cluster in the lab (35 dual processor nodes, 4gig ram) to a larger cluster at HMS. [18:06] migrating the app, or migrating the data too? [18:06] * asw just waving hello. [18:07] when are we going to be able to make interventions in the ecosystem (or can we do it now?) [18:07] migrating all the evolvers and optimisers he has secretly been running there [18:07] btw. I'm now the maintainer of both pmars and pQmars at sourceforge. [18:07] http://corewar.sourceforge.net [18:08] pqmars? /me goes looks as he is unsure what it is exactly [18:08] aha!!! obvious!!! [18:09] pQmars is the "Quantum" version of pMARS. I'm so new to php and mysql that I didn't want to announce the project until it was further along. But I'm happy to give you guys an update. [18:09] http://coreworld.med.harvard.edu/qcw/ [18:09] (um. yah it's an empty page.) [18:10] It [18:10] :( [18:10] ah, and I was thinking that qmars (Martin Ankerl) might become portable... lol [18:10] I'm trying to develop a standalone version that can run a single machine (not just a version for our cluster.) So I'm spending a lot of time trying to keep the code semi managable ... keeping track of mysql accounts.... security... blech. [18:12] pop quiz: anyone know what 'package' pthreads normally comes in on linux? [18:13] I have a cluster version, a standalone version (and a development/production version of each.) I also have the standalone version running on a AMD64 box and a more standard X86 laptop. [18:13] Today I'm testing connections to mysql from php and perl (it works differently for various idiosyncratic security concerns.) [18:14] Anyway. I'll probably have regular corewar hills running when the quantum coreworld is done. And the quantum coreworld itself will permit 1994 style programs to run unchanged... [18:15] (although the 1994 under coreworld is still vaporware - like too much of this.) [18:15] That's my update. [18:18] ah do you recall explaining a bit about quantum cryptography to me a while back? [18:19] maybe. [18:19] I remember a conversation with Joonas in Berlin... [18:21] I didn't make it to berlin :-( [18:22] but basically quantum cryptography doesn't seem to solve the man-in-the-middle type of attacks [18:22] it can spot someone observing a channel, but if that channel is completely routed through the enemy they can play both sides flawlessly, it seems; about right? [18:23] no. [18:23] qc (at least the bb84 protocol and friends) should be called "key extension". [18:24] Alice and Bob start with some "shared randomness". [18:24] ok that makes sense; if you have an agreed key, you can use a channel you control entirely and be sure noone has tapped it..? [18:24] They can use this randomness (via bb84) to get even more shared randomness then they started with. [18:25] the point is not that they have total control of a chanel. (eve could just cut the fiber optic cable between alice and bob ) [18:25] but that they can use the randomness BOTH for authenetication and for secrecy... [18:26] sorry for the typos. [18:26] making sense? [18:26] yeah [18:27] but 'channel' is a rather good term, because qc using photons is a basically line-of-sight thing? [18:27] you can't broadcast nor jam photos? [18:27] photons [18:27] but if alice and bob DONT share randomness (say they haven't even met) then there is a problem... later research in quantum crypto has tried to address that (something like public key quantum cryptography) but I'm less knowledgable about that. Since I've been studying biophysics these last five years. [18:27] :-) fair enough [18:28] you can jam photons. [18:29] I was thinking that with the promise of quantum computing and the first 'program' being to factor a very large number, then without some kind of pkc-version of quantum cryptography, governments would win? [18:30] people are trying to do bb84 by bouncing off a satelite (so alice and bob could be using satelite phones) but besides blowing up the satelite you could "jam" it by flooding its general direction with EM radiation and degrade the quality of the signal to the point that alice and bob can no longer execute the protocol. [18:30] you'd have to bombard the receiver, blinding it, I would guess; radio receivers with ariels that are not uni-directional would be far easier to blind as you wouldn't have to know where they were..? [18:30] the physics is out of my league. [18:30] mine too [18:31] :) [18:31] asw: do you think "we" are really that close to commercial Quantum-solutions as atomchip.com presents? [18:32] so my next problem was understanding how quantum computing is supposed to work at all. It seems that things that are unobserved are in all states at once, or roughly. And by submitting a question and then observing it, you seem to get the right answer. How come you don't get a purely random answer instead? [18:32] (though atomchip is horribly slow for a real quantum-processor, but thats what you get with electron-Motherboards) [18:33] Mizcu - don't know about "atomchip.com" but Harvard is part of major consortium to build "quantum repeaters" so that a worldwide fiber network for quantum key exhange could be practical for banking (and persumably? military applications) in the next decade or so. [18:34] You could cut the fiber (this stuff would be running under oceans etc.) but you couldn't tap it. [18:34] (but you can bend the fiber) [18:35] it's hard to "explain" how quantum computing works. We have lots of experimental evidence that it does but no real understanding of why or if the weirdness "scales" or changes behavior as the quantum states get larger and larger. [18:36] The coolest part of the new quantum corewar release is that I'm simulating a subset of quantum mechanics that only requires "quadratic resources" in the number of qubits to simulate. So I get classiclly impossible behavior but I can simulate it for even thousands of quantum bits. [18:37] It should be very fun. [18:37] * Roy lost it... about 20 minutes ago :-) [18:37] errr... [18:37] Roy: you're not the only one [18:38] And people say corewars is hard to understand and play ;-) [18:38] yeah - sorry guys - I didn't mean to interrupt so abruptly. Just saw Nenad's message on RGCW and thought I'd "wave" [18:39] :) [18:39] Now we can play with experimental quadratic qubits which have classically impossible behaviour :-) [18:39] (??) [18:39] well, this was closer to a "wave equation" than to a "wave" :) [18:40] asw: I am really interested in quantum computing, actually, but never had the chance to learn about it. [18:40] can you recommend some introductory texts? [18:43] Would be fun if they teach this sort of stuff at school.. I'm really fasinated at the moment, they are now teaching us AI (neural nets/knowledge systems/vision+speech recognition etc) [18:44] its the kind of thing that the more you study, the less you seem to know ;-) [18:44] I'll listen to AI lectures next semester [18:45] will: isn't everything like that? [18:45] the more math I know, the more I realize that I know very little.. [18:46] well, I have to go now. :( [18:46] * Nenad waves [18:46] * willvarfa waves to Nenad [18:46] The idea I get from the lectures I get is that they just explain how it could/should work, but still you have no idea how it works :-( only the simple things are explained in detail (A*, Minimax/AlphaBeta) [18:46] * asw waves - nice chatting... [18:46] * Roy waves to nenad [18:46] MSG: [18:46] so asw, is quantum electro dynamics still 'current' - and widely accepted, or superceded? [18:49] hmm getting more beginners into corewar [18:50] anyone think how we can get better awareness within the programming community? [18:50] The idea of having a single 'portal' is good, we have a lot of seperate pages now, but nothing really major [18:51] All pages have links to 20 other pages.. but that doesn't really help a lot :) [18:51] so we might have two problems: 1) finding new people, 2) helping those people and getting them addicted before they give up in boredom/fustration [18:51] Because koth.org is quite important, we should add some pages there. [18:52] qed is still current. so is plain qm. it just depends on the domain of interst. [18:52] there is no theory that unifies qm and gravity yet. Lots of attempts... [18:53] i'm biased but I think the unification of qm and gravity will (explicitly) involve information processing and some notion of quantum information. [18:54] Maybe we should make some kind of news/blog page, where all new warriors on the hill are noted, for example: "Today was a busy day, we have seen a lot of new warriors from Nenad and Roy on the hill. Hot discussions on r.g.c about publicity and asw dropped by to puzzle us with quantummagic" [18:54] been reading feynman's lectures on it [18:55] maybe we ought concoct a reason to be slashdotted? [18:55] There's a class of theories called "topological quantum field theories" that can unify various qm & gravity phenomena. For a while people thought these theories might lead to a "stronger" notion of quantum computing but it was discovered that the exisiting quantum computing model can simulate those theories efficiently. [18:56] willvarfa: We push Arrow a little bit further and have a new SoV. Reason enough? [18:56] Roy - willvarfa - all - I'm definitely interested in helping to make corewar more popular. But I should get back to work. I'll read periodically. [18:56] I was thinking that if we could get an evolved warrior to be a true koth, we could announce on slashdot that "machine beats man" and thus get some of the (choicer) slashdotters to come prove us wrong - a challenge, as it were [18:57] thx asw [18:57] Thats a good idea will [18:57] well I never made a koth :-( [18:58] Or just shout: "Corewars helped prove unified qm & gravity theory" [18:58] I should think that 'computers write better programs than humans' would be something that would get widespread coverage in the programming online world of blogs, portals etc [18:58] It easy as soon as you know the secret code to add to your warrior. Ask JM. He knows. [18:58] Sssst, thats a secret Jens! :o [18:59] willvarfa: Unfortunately bvowk isn't here. You could probably help at lot. [19:00] But evolving a warrior isn't necessarily needed. It could be enough to generate one. [19:01] i.e. take a benchmark, create a random warrior, calculate benchmark, randomly change an instruction, see if better and repeat [19:01] We can just say HullabaTwo was generated from Hullabaloo, and basicly...it is :-) just with a slight touch of human hand [19:01] Err Jens, that IS evolver... [19:02] OK, kind of evolver, but I would only have one warrior to deal with instead of thousands [19:02] hmm Jens is making a proper distinction which, generally, evolving folks don't [19:02] the difference between optimisation and evolution is .. blurred, in corewars at least [19:03] We should have something shocking.. [19:03] "computer programs better than .. you!" [19:03] great headline on slashdot, osnews, etc [19:03] "Yesterday a 22 year old student from Holland commited suicide when his 'corewarrior' was pushed of the 'hill'" [19:03] and we should have a server, that can withstand the slashdot-effect [19:04] i.e. killing servers [19:04] crikey, we could go further [19:04] "This morning a 22 year old student paid $100.000 for the code of a warrior named Arrow" [19:04] ever heard of that idea that to send spam you had to compute something? [19:05] When people think they can get rich... we're populair ;-) [19:05] an aside: I think that hills ought to have an optional ;license line, and if the license specified was open source, the hill would publish said source [19:05] Yeah, that would be cool [19:06] willvarfa: Yes, it but it works the other way around [19:06] Heh look at my warrior that didn't get on the hill 10 4 86 116.10 [19:07] willvarfa: Spammers need to send a lot of emails fast. Normal user don't need. For every email they solve a little "riddle" and then can continue to send the email [19:07] the spam thing? my joke. I was thinking that to withstand a /., you could make the visitor answer some corewars trivia before proceeding [19:08] willvarfa: hehe, but it was really proposed [19:08] it didn't address authentication, it addressed horsepower; spammers have more horsepower than authentication in my estimation.. [19:09] its like catchyas (sp?). They are ineffective against automated attack, but effective only because people make their own [19:10] Cool, http://corewars.sf.net has now a link to SAL :) Just got the email from Walter Hofmann [19:10] err ... it redirects to SAL [19:10] cool [19:11] was there ever sourcecode anywhere to their java applet? [19:11] which applet? [19:12] corewars.sf.net had an applet that let you watch battles [19:12] I assumed it was a mars in java [19:12] thought it would be worth getting the sourcecode to [19:15] Try to send an email. Maybe he'll release the sourcecode [19:15] Roy: What is your TieMaster? A paper? [19:16] Its only available on TellSell [19:17] hehe [19:18] But yes, it is a paper [19:18] (multible papers actualy) [19:18] I want to make the next Die Hard ;-) [19:18] * Roy is chasing records [19:18] How about "Die Hardly" ;-) [19:19] Die Less.. [19:19] Or Tie Hard [19:19] Undieing [19:19] oops, Undying [19:19] I'm still looking for the silliest name ever used on '94nop. [19:19] But its pretty hard to make something that nothing can beat :) [19:20] t.A.T.u Girls Rule is still my favourite, but I haven't published it my 94nop-page yet [19:20] You should collect a few and make a poll ;-) [19:21] 614235-9746-xt430-3-eai2 should do as well [19:21] Thats evolved :-) [19:22] Yes, bvowk always uses such nice names ;-) [19:22] Speaking of spam, we could also start a media offensive by spamming [19:23] "Get a your C.O.R.E.W.A.R. stronger?" [19:23] "Don't be an I.M.P.!" [19:23] "Be an IMP-SPIRAL" [19:24] "Want girls? Become corewar-master!" [19:24] "See it yourself, Herbal A.V.E.N.G.E.R" [19:24] lol [19:25] :-) [19:25] willvarfa: You said sth. about creating an computer-generated warrior to be koth. We only need some computers to do that [19:26] and some time [19:26] If we start now, we might have a good warrior at the end of the year. A perfect time to become slashdotted ;-) [19:27] IT: Evolved corewar program turns into a x86 virus [19:27] That is easy. Take a paper, that spreads to new cores, too. [19:28] But it will do well on slashdot ;-) [19:30] MSG: [19:35] hmm a concerted effort to make a 94nop warrior ... [19:35] brute force will simply never make one [19:35] its not within bounds of human endeavour, sorry [19:36] you need an algorithm to determine which warriors to evaluate [19:37] (it would be feasible to brute-force a hill that was -l 2 -d 4 -s 8 -p 8 or so, but not beyond that, in my rough finger-counting [19:37] evolution is just an effective way to search a very large keyspace [19:37] Wait a second, we have BIG cw news comming up [19:37] hence preferring evolving warriors to random twiddling [19:37] ... we do? [19:37] JM says he has a unbeatable warrior, and he has prove [19:38] He just wants to work it out a little [19:38] I have written some python scripts, that use fmars to create my nano-warriors, not evolving, but optimizing. I could adjust them for 94nop. [19:38] to work like optimax [19:38] But we have sientific proof there is a unbeatable warrior (well, not yet) [19:38] proof that such a warrior exists? or that he actually has it? [19:39] To prove the existance would be quite a feat [19:39] No made one, and he has the proof (thats what he said anyway) [19:39] "Corewar Holy Grail found solved after 50 years" [19:39] did he only tell you Roy, or is this something I have completely missed since I never look at logs?? [19:40] willvarfa: Roy only make future headlines [19:40] *makes [19:40] He mentioned his unbeatable warrior a couple of times, but its still a bit of a secret [19:40] ahh ... you are talking about JM's warrior, that no one has seen [19:40] yet [19:40] (I could imagine JM solving -l 1 -d 1 -s 3 or so.. [19:40] Well I think its something like that yes [19:41] Maybe slightly bigger code, but he is working on that [19:41] well then you have to consider the shape of the hill.. [19:41] an infinite hill, where he submits all other perms of warrior, that he might be koth of [19:41] and you have to think about psychology, because you have to take into account, what your opponents will do [19:42] but for the 'true koth' there is a subset of warriors that, overall don't do so well, but within a subset of the warriors reordered into a partial hill, would beat it.. [19:42] * willvarfa has been thinking about this too much, sorry [19:42] How about actually calculating the score of all warriors of all warriors for sth. smaller than nano. [19:43] If you have a core and you have a warrior, and you test every possible combination against eachother, then you'll know when you have the unbeatable warrior [19:43] Just to see, what the result will be. [19:43] Yeah, what Jens says [19:43] * willvarfa was going to submit 250K of warriors to a -l 2 -d 2 -s 8 -p 8 hill and be koth forever!! [19:44] yeah I only got it down to 250K by making assumptions about warriors that would suicide [19:44] but those warriors themselves would affect the scores of all others [19:45] Even -l 2 -d 2 -s 20 -p 20 would be easily computable [19:45] http://www.geocities.com/varfar/redcoder/multi-manics.html [19:45] Jens, that is a bold claim [19:45] That is what JM is worried about, another warrior that beats the others just a bit better, thus still better [19:45] 2 instruction -> 17*17*20*20*20*20*7*7 different warriors [19:45] (without pspace) [19:46] how many lines? [19:46] 2 lines [19:46] -l 2 [19:46] how many starting positions for each combination of two instructions? [19:46] 16 [19:46] ?? [19:46] -l 2 -d 2 -s 20 -p 20 [19:47] two might suffice ;-) [19:47] ah ... as usual I'm talking about different things [19:47] sorry [19:47] 20 instr, 4 lines are warriors and 2 distance -> 14 [19:47] err.. [19:47] never mind :) 16 positions it is [19:47] nope [19:47] 17 [19:47] forget it [19:48] 17 instructions [19:48] how many modifiers? [19:48] 2 * (CORESIZE-MINDISTANCE+1) [19:48] all [19:48] perms of line: 17*7*8*8*coresize*coresize [19:49] yes [19:49] perms of line, including empty line: +1 [19:49] 11..22222222222222.. (warrior 1 = '1', and warrior2 = '2')... is 14? [19:49] (a completely empty warrior is completely legal) [19:50] for -l 2 -d 2 -s 20 -p 20 we'd have 3046400 different "instructions" per line [19:50] 3046400**2 different warriort [19:50] *2, because of possible starting positions [19:51] not that much ... most warrior should be dead quite soon anyway [19:51] 12 million? [19:51] why 12? [19:51] sorry, I'm used to python, 3046400^2 [19:51] * 2 [19:52] yeah oops [19:52] but actually we would have less [19:52] spl.b = spl.f = ... [19:52] 18561105920000 [19:52] 18,561,105,920,000 [19:53] 18.5 whats? petrasomethings? [19:53] 18.5 * 10^12 [19:53] hehe [19:54] (Jens, I worked out spl.b, spl.f etc; it is roughly half; yet they look different to sne etc [19:54] we can't truely bruteforce -l 2 -s 20 can we? [19:54] why not? [19:54] we might start by making some assumptions about what warriors we can cull [19:55] Jens, imagine that ten of us all worked on this, and each of us could do 100 million battles per second.. [19:55] .. how long would it take? [19:56] But we wouldn't need to do it all. First think about removing unneded fight [19:56] *unneeded fights [19:56] that is a bit against 'brute force', no? [19:56] that would take 31min? [19:57] 2*10^13/ 1*10^8 seconds = 2*10^5 seconds [19:57] yeah, I just realised that Roy :-) so 100 million seems a workable number ... [19:57] div 10 people [19:57] 2*10^4 seconds [19:57] but [19:58] remove 0,000,000 per second per person, you'll have 1856 seconds = 31 minutes [19:58] we have to make a fight each warrior each other warrior [19:58] arrgh ... bad English [19:58] We have to let each warrior fight against all other. [19:58] that would square the number of fights [19:58] crikey, we don't want to _list_ all perms, lol [19:59] in a 20 core, what is -P ? [19:59] 34 [19:59] yes [19:59] 34 fights? [19:59] yeah [19:59] surely only 18? [19:59] nope.. [19:59] it says 34 (for some reason) [19:59] warrior 0 never needs to move [19:59] I thought, we were looking for the unbeatable warrior for -l 2 -d 2 -p 20 -s 20? [20:00] the legal positions of warrior 1 are governed by -d 2 ... [20:00] correct [20:00] == 16.. [20:00] but borth warriors have to start the fight once [20:00] == 18 surely? [20:00] no, -d 2 on both sides [20:00] aha! of course [20:01] yeah, 16, * 2 [20:01] * willvarfa is getting rusty [20:01] 17*2 [20:01] 17*2 = 34 :) [20:01] but I don't understand the 17, why not 16!? [20:01] 2*(CORESIZE-MINDISTANCE+1) [20:01] what about length 1 warriors? [20:01] 2 because of each warrior has to start once [20:01] No need to think of length 1, because of mindistance [20:02] it is like 1 instruction + enough dat's to fill the space [20:02] of mindistance [20:02] but with warrior 1 static, you only have 16 legal positions for the other warrior if you keep 2 distance between the warriors [20:02] no 17 [20:02] please count them all [20:02] I made the same mistake [20:03] I did.. 11..22222222222222.. [20:03] 14.. :-S [20:03] no [20:03] ok, let's count together [20:03] warrior 1 is fixed at position 0 [20:03] the first possible position for warrior 2 (w2) is at position 2 [20:04] The last possible position is at position 18 [20:04] 2,3,4,5,6,7,8,9,10,11,13,14,15,16,17,18 [20:04] hrm ohw, I always thought the first position was 0 + -l 2 = 2 + -d 2 = 4 :) [20:04] Jens wins [20:04] that makes 17 positions [20:04] thats how I remember it ... now you've jogged my memory :-) [20:05] :) [20:05] so basically full circle - you worked out it would take us just 55 hours..? [20:05] For writing Quickscanners - Part III/IV I had to deal with such things and couldn't work it out quite a while [20:05] why 55 hours? [20:06] So with -d 2 -l 2 you can have this: DAT DAT DAT|MOV bla,bla/MOV bla,bla|MOV bla,bla/MOV bla,bla|DAT DAT DAT [20:06] Roy: yes [20:06] Lol, never knew that :| [20:06] hmm 582 hours maybe? [20:06] I always thought you would have mov/mov/dat/dat/mov/mov (-d...2!) [20:07] Try to throw a bomb at position length+1 and then fight a jmp 0 with pmars -P, you will get exactly 2 wins [20:07] willvarfa: Which fights do you want to make? [20:07] So on 94nop a warrior can start at position 100 :-) argh, I've always set my clears etc to start at pos 200 :-| [20:08] Position 200 is good, because it is the first position to get the best probability of finding sth. [20:08] willvarfa: all warriors against all warriors? [20:10] right, there are (nwarriors*nwarriors)/2 fights (incl. selffights) right? [20:10] all the 18,561,105,920,000 warriors, need 34 round against 18,561,105,920,000 warriors...? seems like a big number comming up :-) [20:11] 2*10^12*2*10^12 = 4*10^24 fights [20:11] (done this kind of thing before, but rusty: http://redcoder.sourceforge.net/opexfreq/94-nop-100.html ) [20:12] (warrior-n vs warrior-x == warrior-x vs warrior-n) [20:12] but that is only a little factor [20:12] (18,561,105,920,000 * 18,561,105,920,000) / 2 fights...? [20:12] let's say 10^24 fights [20:13] 10^26 ... sorry [20:13] it is 1.8*10^13, not ^12 [20:14] so, how many fights are there, in long-hand numbers? [20:15] 172,257,326,486,729,523,200,000,000 ? [20:15] yes [20:17] hmm, anyone know what kind of prefix such a number has? petra tetra giga whatever? [20:17] much much ;-) [20:18] I always like to count the following way: 1, 2, many 1, many 2, many many 1, many many 2, ... [20:18] troll [20:18] lol [20:19] It is far mor easy to remember than 1,2,4,3,5,6,8, [20:19] 172 yottabattles! [20:19] Not that much [20:19] ;-) [20:19] 10^24 is yotta, I think...? [20:20] http://en.wikipedia.org/wiki/SI_prefix [20:20] ah, my mistake; I went to wikipedia with 'orders of magnitude' [20:21] Quadrillion! now that is a number we can feel the size of! [20:22] http://en.wikipedia.org/wiki/Binary_prefix [20:23] Y? [20:24] so it is 172 yottabattles after all? [20:24] yes [20:24] can we brute force it? [20:24] nope [20:25] well I did some thinking, and decided you could cull all but 250K-ish for -l 2 -s 8 ... [20:25] 172*10^24/100, 000, 000/10 = 172*10^15 seconds :( [20:25] you could calculate that the others all suicide [20:25] yes, might be easier [20:26] (but with -l n -d n, what if the warrior simply runs into the code of warrior 2? could this really impact the overall hill significantly, and could you _prove_ it didn't?) [20:26] I think we'd stuggle to prove -l 1 -s 3 -d 2 ..? [20:27] -s 3? [20:28] hmm ... which warriors wouldn't suicide after the first instruction ... [20:28] spl,mov, jmp 0, jmz 0, jmn 0 [20:28] its an interesting thing to think about [20:29] hmm, what the opponent put a 'continuation' into the cell behind them so they didn't die? [20:29] When you have the complete data, you can post to slashdot "Bet warrior ever found" [20:29] *Best [20:30] then it would live a little bit longer [20:31] Probably djn.f # 0, < x would be quite a good warrior [20:31] "DJN.F-Stream to be the best warrior ever" [20:34] Another problem! [20:35] 94nop-hill - The average age of a warrior, when it is pushed off, is about 21 [20:35] It know, that there has to be a good reason for that, but I cannot think of any [20:35] for a hill with 50 slots, it would 51 [20:36] but why? [20:40] interesting pattern [20:40] is there a pattern to the average position to enter a hill? or a correlation between the two? [20:41] I would mostly expect correlation with author ;-d [20:42] Haven't checked so far. [20:43] For 94nop the average age when falling of the hill was 21.71. I am quite sure, that the theoretical value is 21, but why ... [20:43] I have always hated stochastics [20:45] It is a pity, that I haven't started with cw earlier. [20:46] While scanning through the 94nop-logs I have noticed, that I have missed all the interesting events :( [20:52] all those ... so far! [20:52] don't study history, make it!! [20:53] OK ... hmm ... HFB will make an +2600-age-warrior :) [20:53] How about that? [20:54] *make -> will be [20:55] HFB = Harmless Fluffy Bunny [20:56] completely possible [20:56] if you submit the next 2600 challenges..! [20:57] I often wondered about that in my evolving days [20:57] making 'no go' areas on the spatial corewars globe whatever, so that your new challenges kicked butt but never adversely affected things already on the hill that you cared about [20:57] a bit like the attack of the first spirals, if I've read that right [20:59] Hmm, look at the current 94nop-hill. The number of different authors soon will reach 0. Maybe I really should submit the next 2600 challenges [20:59] It would almost make no difference [20:59] * Roy won't submit more warriors then now, maybe I'll even kill some :( [21:00] why? [21:00] And which ones? [21:00] me and fizmo are the reason of the low nr. different authors :) [21:01] You could always clone yourself and call the clones differently ;-) [21:01] Myr because its old and was pushed off already, but I killed one and filled it with this one to not keep a open space [21:02] You know, that I would prefer more a-imps on the hill [21:02] don't you? [21:03] You do? I'll send a couple of imps instead of Myr, try to kill them [21:03] hehe [21:05] I just can't get more ties :-( [21:05] YottaImpz ;-) [21:07] fizmo is infact four different authors; well known fact [21:07] yeah.. [21:07] I think he's Nenad too :-( [21:07] Maybe even Jens :-\ [21:07] willvarfa: We should add this fact to the FAQ in order to inform all beginners [21:08] Jens: link for the wiki: http://groups.google.co.uk/group/rec.games.corewar/msg/566fe2d5b302caa4 [21:08] Is there a special reason for myself being four different authors? [21:09] Yes, bvowk (I think) has already introduced me to the fine art of cw-humor [21:09] art? never thought of it that way before [21:10] time for me to go get some sleep; work tomorrow!! arggh [21:10] * willvarfa *waves* [21:10] MSG: Quit: Leaving [21:10] * Jens waves [21:10] and is too late as usual [21:11] 86% ties, woo, thats pretty good I guess [21:11] But it won't survive long, altough (Fluffy Bunny)'s food won't kill it [21:12] Nothing can easily kill it, but its incapible of killing too :-) [21:12] Did you add a qscan? [21:15] No..! then the losses will go up [21:16] Have sent another test warrior to the hill [21:16] with a better name [21:17] hehe 37.10 [21:17] but not the lowest score ever [21:19] 9 4 87 9% wins, is that the lowest ever on a hill? [21:19] one moment [21:21] no [21:21] And you should know the answer!! [21:21] April 14, 2004 - DieHardest by Roy van Rijn [21:21] :-| [21:22] I already made that? :-S [21:22] wins 5 %, 6 % losses, 89 % ties [21:22] * Roy faints [21:22] score 105 [21:22] Heh well, I have less losses atm, so it dies harder! [21:22] pushed off at age 2 [21:23] But no, not the least amount of losses, thats Die Hard [21:23] (is on JM's site) [21:24] Oct 30, 2002 - 13 point paper/imp by John Metcalf, w 7, l 5, t 88 [21:24] sorry, wrong values [21:24] w 8, l 4, t 88 [21:48] Part: Jens left #corewars [22:15] MSG: [22:28] nenad has started an interesting discussion on rgc [22:41] Join: X-Scale joined #corewars [23:45] MSG: Quit: humhum