[00:08] Join: brx joined #corewars [00:11] Join: Mizcu joined #corewars [00:19] hey miz [00:21] MSG: Quit: leaving [00:26] Join: brx joined #corewars [00:57] MSG: Quit: leaving [02:05] MSG: [03:32] MSG: [03:44] Join: Philb joined #corewars [03:45] Part: Philb left #corewars [05:54] Join: |caesar joined #corewars [05:54] <|caesar> hi [05:55] <|caesar> is anybody here? [06:47] Join: Mizcu joined #corewars [06:47] hi caesar [07:41] Join: willvarfa joined #corewars [07:42] hi Will [07:42] good morning miz [08:51] Join: Jens joined #corewars [08:51] :) [08:53] Engfeh ad infinitum [08:53] Cant find my passnumbers to my netbank.. [08:53] Just gotta keep searching [08:59] "Engfeh"? [09:00] also spelled *ngh* [09:01] But what is that? [09:02] sound you can often hear out of stressed out or blanked geek [09:02] ok, now I get it [09:02] :) [09:29] my heads about to blow soon [09:32] much happen after I left last night? [09:32] don't think so [09:33] ah, so we gave up on brute force? [09:35] Good question, but I would say no, but first I have to think about it a while [09:36] So pissed off right now i feel like scooping my prains out with a melon-baller [09:36] ignore the typo [09:40] well we might have, but you drew attention to it, miz' [09:41] is netbank giving you a pad of one-time-use numbers for login, or something? [09:41] well, the note wasnt under my 'puter screen.. [09:41] login/password and onetime-pad [09:41] excellent [09:42] two factor authentication, classic [09:42] how many digits are they? [09:43] 8/5 and 4 i think [09:44] do you tear off used numbers and burn them, or some such thing? [09:44] pen them over [09:44] wha, you got my papers? [09:45] my bank uses little computers about the size of a small calculator [09:45] much the same effect, prehaps [09:46] good luck in the search! [09:48] It seems, that Nenad has woken up. Already two submissions to the 94nop-hill [10:11] Join: cy joined #corewars [10:11] Part: cy left #corewars [10:17] Quite a difficult task to talk to somebody, if he leaves after a few seconds [10:19] sorry [10:19] ? [10:19] @work [10:19] ah thought you meant I seemed to be distracted [10:19] which I usually am [10:20] Oh, I forget, that everybody outside Germany has to work today. [10:21] holiday in Germany? [10:22] meeting /me waves [10:23] Yes, the reunification between East- and West-Germany is "celebrated" [10:26] For most people it is just another reason to get drunk [10:26] OK, "most" is the wrong word, but there a some of them [10:40] Join: brx joined #corewars [10:52] MSG: Ping timeout: 252 seconds [10:55] Join: brx joined #corewars [10:57] <|caesar> hi [10:57] Hi |caesar :-) [10:57] <|caesar> i come back [10:57] and we'll be here [10:57] <|caesar> why not mush people here [10:58] <|caesar> it....just some quiet,i think [10:59] <|caesar> jens,can u tell me why that? [10:59] Yes, but we cannot talk all day. We couldn't make new warriors then :) [10:59] <|caesar> heh....thx [10:59] <|caesar> yeay,u're right [10:59] <|caesar> and how about u? [10:59] It is a little bit more crowded/budy here in a couple of hours. [10:59] *busy [11:00] <|caesar> u say u busy now? [11:00] <|caesar> i'm sorry,i frm Asia [11:01] As long as I can understand you it is no problem :) [11:01] I'm not that busy at the moment, but working on some ideas for corewars [11:01] So, if you like to talk about corewar, go ahead :) [11:02] <|caesar> i leave some times for supper... [11:02] :) [11:02] <|caesar> i don't think so,my english not good... [11:02] I can understand you and that is enough :) [11:02] <|caesar> err...leave some times [11:10] MSG: Ping timeout: 252 seconds [11:11] Join: datagram joined #corewars [11:12] Hi datagram! [11:12] Hello [11:12] * willvarfa stretches [11:12] How is everyone doing today [11:12] MSG: <|caesar> Quit: GoodBye , My Friends... [11:12] :) [11:14] I was looking to get into Corewars a bit [11:14] Do you have any specific questions? [11:15] I already understand the redcode and basic strategies used by most of the recent warriors I looked at [11:15] Jens, Not at the moment, I don't thik [11:15] Think* [11:16] Then I think, it is best, if you try to write your own warriors and send it to the beginner's hill :) [11:17] But we'd be happy to help, if you have any problems [11:17] I was wondering [11:17] Join: caesar joined #corewars [11:18] How hard would it be to try and enter the tournaments such as this one: http://www.corewar.info/tournament/cwt.htm [11:18] I realize those are mostly professionals [11:19] It is not that hard. As the rules for the turnament change with each round, even a beginner can win. [11:19] You really should give it a try. [11:19] Indeed I will for the next tournament [11:20] Do they announce them early in the month? [11:21] As far as I know the tournaments aren't held in regular intervals. [11:22] If I understand the last porsting about round 24 right, you can send in your warriors until October 10 :) [11:22] Oh I thought it was once a month [11:22] *posting [11:22] Round 24: The Shrinking Core Knockout Round deadline: September, 30 2005 [11:22] ? [11:22] You should look into the cw-newsgroup: news://rec.games.corewar [11:23] or http://groups.google.com/group/rec.games.corewar [11:23] There are often more current information available and new rounds are announced there [11:25] Ah I see his post from today [11:25] Thank you [11:26] :) [11:28] I suppose, you have already head about the beginner's hill at http://sal.math.ualberta.ca/hill.php?key=94b ... don't you? [11:29] *heard [11:31] Yes I researched all of the links on the main page [11:31] :) [11:31] I saw the post about needing more beginners [11:31] I am part of a few hacker groups, and I will probably be giving a presentation about the game to them in the near future [11:31] Nice to hear :) [11:32] A few already want to start a mini tournament within the groups [11:32] So I will probably drag a few people here too eventually [11:32] good good :-) [11:33] Join: fiveop joined #corewars [11:34] What is the most popular strategy? [11:34] For a new warrior? [11:34] In general [11:34] I don't think there is one, but people have their preferences [11:35] I see quite a few Paper warriors on all the ladders I find [11:35] Also Oneshot [11:35] Even "outdated" strategies can become to life again. on the current 94nop we have a vampire and it was thought, that they were too weak [11:35] I think vampire is the most interesting strategy [11:36] on tiny you find mostly clears and oneshots currently [11:36] Please explain oneshot [11:36] I lost my bookmark to the glossary [11:36] And cannot seem to find it again : / [11:37] http://www.corewar.info/lexicon/ ? [11:38] No, it was a part of another document I think [11:38] White background [11:38] http://freebiology.org/wiki/REC.GAMES.COREWAR.FAQ ? [11:38] the FAQ perhaps? [11:39] oneshot is a scanner that attacks the first thing it finds and becomes a bomber. It is a good combination which compromises two strategies for a good overall effect [11:39] Ah, that is it but the one I saw earlier was not the same document. It was all text based. This one has code examples along with definitions. [11:39] Thanks. Jens [11:40] You might find http://home.pipeline.com/~corewar/corewarrior/ interesting, too. [11:41] It is an archive for all issues of the CoreWarrior [11:42] Again thanks [11:42] Jens: what do you think of the idea some have mention in the newsgroup after nenads post about one 'central' corewar page? [11:43] fiveop, I think it is a good idea. I was going to suggest a wiki, but then I was linked to the wiki : ) [11:43] hosting issues come to mind; corewars has been going for over 20 years now. but not always the same people [11:43] fiveop: It might be a good idea, but some links on koth.org to the most important other pages would be enough for a start [11:44] we might make a portal now, but what about in two years time? five years time? [11:44] Personally I think the hardest part for a beginner is not necessarily the redcode. That is well documented, and coming from an assembly background I had no problem picking it all up in a day or two. A site with a good discussion of general strategy would be good, though. [11:44] The old issues of CoreWarrior/PushedOff contain some descriptions of basic strategies [11:45] (I think all corewarriors ought to make their pages "GPL" and explicitly encourage each other to mirror their content. So that it doesn't get lost) [11:45] The My First Redcode Book texts were a good start, but too basic [11:46] CoreWarrir #89 (http://home.pipeline.com/~corewar/corewarrior/cw89.txt) contains interesting information about scanners [11:46] Well if I ever get good I will write a primer on strategy [11:46] Jens: for a start ... but as you're currently gathering and edit much information about corewars, I thought you'd be interessted in building such a site [11:46] And call it something esoteric like "A Treatise on Corewars" [11:47] datagram: For papers you can find a lot at http://www.corewar.info/lexicon/paper.htm [11:47] datagram: I don't like the idea of a "community" page with a forum or something ... we've got irc and rgc [11:48] fiveop: I will continue to write about the history of the various hills and of course about score surfaces and quickscanners [11:49] but I don't think it will become a good page for beginners. Altough I'm playing with the idea of publishing sth. like CoreWarrior for beginners [11:49] about 6 to 8 issues/year [11:49] otoh with corewar.info fizmo's got a (in my opinion) quite 'complete' infopage already [11:57] How do I get the debug area to show what each program is doing, like how the pmars page on koth.org has it shown? [11:57] -e [11:57] pmars -e warrior1.red warrior2.red [11:59] Does it work? [11:59] I am using pshell to load pmarsv, I believe [12:00] Do you use Windows? [12:00] Not primarily, but the box running pMars is yes [12:00] Then http://www.geocities.com/corewin2/ might be of interest [12:00] I have that bookmarked but I haven't downloaded it yet [12:01] As I'm only using Linux it doesn't work for me, but it seems to be good for debugging [12:02] But the screenshots look promising [12:04] I see [12:04] What is the library I am missing for Linux? [12:04] Which library? [12:04] It craps out at make, I believe it was called itermlib? [12:05] what does it say exactly? [12:05] MSG: Quit: Download Gaim: http://gaim.sourceforge.net/ [12:05] Making pmars with compiler flags -O -DEXT94 [12:05] Linking pmars [12:05] /usr/bin/ld: cannot find -ltermlib [12:05] collect2: ld returned 1 exit status [12:05] make: *** [pmars] Error 1 [12:05] : ( [12:05] Join: Jens joined #corewars [12:05] Making pmars with compiler flags -O -DEXT94 [12:05] Linking pmars [12:05] /usr/bin/ld: cannot find -ltermlib [12:05] collect2: ld returned 1 exit status [12:05] make: *** [pmars] Error 1 [12:05] Sorry to spam [12:06] Do you have pmars-0.9.2? [12:07] I mean the version from http://sourceforge.net/projects/corewar ? [12:07] No 0.8.6 [12:07] I did not know a newer version existed [12:08] Please download 0.9.2, I think I can help then :) [12:08] MSG: Ping timeout: 252 seconds [12:09] datagram: You only have to make a little adjustment to your Makefile, then it should work [12:13] Do you have the new version now? [12:13] I got it installed and everything [12:13] And running [12:13] :) [12:14] I know my way around the pewter, don't worry none bout that : ) [12:14] :) [12:16] And I like the idea of founding local groups at universitys or something :P [12:19] Yay, my generator script works now. After a few rounds it has already generated a warrior with a score of 68 (for tiny) [12:20] wow [12:20] Join: Roy joined #corewars [12:21] Hi Roy [12:21] hi Jens [12:37] Hi Roy [12:38] Hi data [12:40] I have never evolved warriors. Does anybody know how long it usually takes before sth. emerges with a decent score, say >100? [12:41] That would depend on how you are evolving it, wouldn't it? [12:42] I was talking about the time you need with sth. like redrace/yace/... [12:43] I have not checked out the evolution source for those programs yet [12:43] Roy, you should know as you already have created some evolved programs [12:44] I'm just curious, because my current generator creates acceptable (score about 100) warriors after about 300 rounds (i.e. 300 runs against a benchmark) [12:46] No idea.. I don't benchmark the warriors against a normal benchmark so often, I just wait a while (unsupervised) and when I return I take the best warriors :-S [12:46] What hill are you evolving? [12:46] fshtiny [12:46] forget about the fsh-part :) [12:49] heh, can't really help you there, you can do tests yourself, installing Yace isn't very time consuming :) [12:49] ok [13:03] http://www.google.com/search?complete=1&hl=en&q=lernen+mit+karteikarten&btnG=Google+Search hm ... :P [13:03] :) [13:10] Join: Nenad joined #corewars [13:10] Hi [13:11] :) [13:11] Join: SarG3n70 joined #corewars [13:11] * Nenad goes to have a look at todays logs [13:12] Hi SarG3n70! [13:12] Hi Jens! :D [13:13] SarG3n70: I've never seen you here before. New to corewars? [13:14] Hello [13:16] MSG: Ping timeout: 252 seconds [13:17] Join: SarG3n70 joined #corewars [13:17] Join: Fizmo joined #corewars [13:17] hi [13:17] hi! [13:17] Hi Christian! [13:17] wow, lot of new names here ;-) [13:17] Im new, Im looking for a good tutorial to learn how to play Core Wars [13:18] Hi Fiz [13:18] SarG3n70: http://vyznev.net/corewar/guide.html [13:18] Sarg73nsoef: Also take a look at Fizmo's website (www.corewar.info) [13:19] let me see... [13:21] Nenad, Hi [13:21] Hi, datagram [13:22] Hi Fizmo, also [13:22] Nenad, I saw your post about the need for beginners [13:22] * Nenad greets Fizmo [13:22] I told Jens that I will probably give a presentation to my hacker groups about the game [13:22] that would be great! [13:23] :) [13:23] I am a Neophyte but I already understand all the code aspects [13:23] * datagram comes from a _heavy_ programming background [13:24] It's just strategy that seems tough to grasp for beginners, methinks [13:24] MSG: Ping timeout: 252 seconds [13:24] Join: SarG3n70 joined #corewars [13:25] Well, maybe.. but the basic strategies explained with examples aren't so hard to understand (IMO, but my mind is colored) [13:25] The best way to learn is just to analyse existing (new) warriors [13:25] That's what I've been doing [13:25] well, to have a complete understanding of strategic issues, you have to run many battles in Corewin and look at what happens. then analyze the scores, etc. it is rather easy to do ti that way [13:26] And then just test/change things in programs and see if you can improve it [13:26] Is there tournament scheduling code for Linux? [13:26] Or is a shell script in order [13:26] but it's best (in my opinion) to first try and make something by yourself, without trying to copy existing warriors [13:27] Just two ways to start out, we both master both using a different path [13:28] :) well, if there is one thing I like about corewar, it's the possibility to achieve the same result in many ways [13:29] btw, since so many of us are here, maybe we could organize ourselves and start making that beginner magazine... [13:29] no reason to wait [13:29] I think that most of the people at r.g.c. agreed about that [13:29] yup [13:30] should we just let someone make it one month, someone else next month... or do some teamwork? [13:30] MSG: Read error: Connection reset by peer [13:31] ... suggestions? [13:31] Teamwork would be more efficient and organized [13:31] Unless you are all ronins or something [13:32] : P [13:32] :). how about it? Roy? Fizmo? Jens? [13:32] Nenad, I will offer my newb help if you need it [13:32] Join: brx joined #corewars [13:32] any help is always welcome. [13:32] hi brx [13:32] heyhey :) [13:32] many people here... that's nice to see. [13:33] * Roy would like to help a little, I could write something about a basic strategy [13:33] datagram: "Inside view from a newbie" [13:33] Roy: how about high-tech imps ;-) [13:34] Nah, more about basic process generation and next month papers :-) [13:34] Heh [13:35] I suggested in my latest post that we could do something that has already been done before - construct a warrior together, and then explain every step in the creation process in the magazine... [13:35] therefore showing HOW to make good warriors [13:35] we could have several columns, though [13:35] MSG: Quit: humhum [13:36] someone should make a brief summary of happenings at 94b - that's for sure [13:36] Thats fun, I would like to do that! [13:36] which? a hill activity summary? [13:36] Nenad: I think JM might want to help with that, he already does that for the CoreWarrior [13:36] :) you're right. [13:37] so, you'd like us to make a warrior? [13:37] I'll write something about the basic strategies and let the readers mail me what warrior they want to see made (most votes count) and next issiue I'll start making it, and leave improvements up to the readers (they mail and I'll implement the best one) [13:38] one more thing ... how often should we do this? 1/month or 2/month? [13:38] So here is what I am going to work on for the next hour or so (until the sleepy meds kick in): I'm going to attempt to write a vampire program without having looked at vampire program source...dum dum dummm [13:38] 1 per 1/2 months is more then enough IMO.. [13:38] ok. :) [13:38] 1 month intervals gets quite hard, I've tried to write papers for the fractal art community at that rate but it is hard [13:38] and we'll publish it at r.g.c. [13:38] Vampire is hard and not so much used anymore... good luck ;-) [13:39] I would suggest something like 4-6 a year, so you have time to plan each one [13:39] Yes but I think it's the coolest strategy [13:39] Yeah, 2 month in between [13:40] MSG: Ping timeout: 252 seconds [13:41] Join: brx joined #corewars [13:42] Jens: you suggested some brochures. would you like to make some? [13:42] (in german and english) - we can all translate to our languages later.. [13:43] Name suggestion CoreExplorer (instead of the CoreWarrior) [13:44] :) a good name [13:45] MSG: Ping timeout: 252 seconds [13:46] Join: brx joined #corewars [13:49] Maybe one could also make some 'beginner challenges' like Neo's events [13:49] Can I give a SPL a B parameter? [13:49] To use as a counter [13:51] yes [13:51] no [13:51] ;-) [13:51] Fizmo: I could do that [13:51] spl is always executed, regardless of the value of the b-field [13:51] That's what I though [13:52] So it is okay, yes? [13:52] Jens: he meant: as a counter for something else [13:52] yesm you can [13:52] but you can use spl # test, < 42 to generate process and decrement the b-field of 42 [13:53] datagram: if you want to have spl 0, x , you can also have spl #data, x - it works the same [13:53] so, you can use spl to store two informations [13:53] I don't understand [13:53] I want the spl to catch them and get them stuck there [13:53] as a stun bomb? [13:54] As whatever it is called when you catch them and get them stuck, yes [13:54] I read the one paragraph on vampires [13:54] That's what I'm basing most of this on [13:55] you can use spl #x, y ,because immediate adressing mode is interpreted as "0" , so spl #x, y acts as spl 0, 0 [13:55] Ah okay [13:57] just a quick visit. Gotta go now. Maybe back later [13:57] Bye bye Fizmo [13:57] * Nenad waves [13:58] * Fizmo waves [13:58] bye [13:58] MSG: [13:58] bye [13:58] * Jens is too late... as usual [13:58] MSG: Quit: leaving [13:58] Jens: you haven't answered my question [13:58] which one? [13:58] or have you? [13:59] about the brochures? [13:59] I am curious, is it very bad to multitask? I was thinking of having it split early to become a few small processes that work together [13:59] I know how to do so, but would it impact the speed of my code ALOT to do so? [13:59] datagram: multitasking is ok. [13:59] Probably just 3-4 processes [13:59] Nenad: I can tell you now a preliminary yes, because I have to finish some other texts first. [14:00] Nenad: Please ask again in about 2 or 3 weeks [14:00] Nenad: ... as usual texts about quickscanners :) [14:00] datagram: speed in general won't change. speed of separate pieces of code will. [14:00] Jens: :) [14:00] ? Explain [14:01] well, do you know how multitasking works in corewar? [14:01] Yes of course [14:01] Thats why I asked if it was considered bad [14:01] Join: fiveop joined #corewars [14:01] well, look at papers - they can have up to 5000-6000 processes [14:01] and they work fine [14:02] Heh [14:02] of course, any of paper copies is seldomly executed, but there are many of them [14:02] Papers are boring : ) [14:02] :) [14:02] scanners are probably most interesting [14:02] Suppose you have 2 processes, datagram. Each time slot one instruction is executed. So the overall speed of your program is constant, but the speed of the processes halves [14:02] Jens, I understand that [14:02] I think vampires are the most interesting [14:03] Nenad/datagram: It probably depends on your point of view [14:03] Because they are offensive and defensive at the same time [14:03] A scanner has to react; that takes more time [14:03] just like papers :) [14:03] Jens: I like playing with s/p's best. [14:03] but I like making scanners more... [14:04] well, you ARE making texts about quickSCANNERS, right? [14:04] But qscans are far more easier than normal scanners [14:04] btw, Jens: [14:05] have you considered adding an anti stone trick to qscanners? [14:05] the possible merit of that? [14:05] it would be just one line of code, once you know the exact place where you've found something [14:06] Havn't tried yet, but the the current q^4.5 needs 2*16 cycles for all scans. How many bomb can a stone throw during that time? [14:06] *bombs [14:06] a 0,6c stone? [14:07] compared to its size? [14:07] and if you take into account a possibility of a 2c decoy maker? [14:07] decoy can be made of several modified cells far away from each other, you know. [14:08] I can make junk as quickly as you can scan... :) [14:08] Join: brx joined #corewars [14:09] Jens: it would be nice to make a warrior with such antiqscan system, and then try both usual qscan and a stonetrickqscan attached to a warrior fighting it... [14:09] Let's see. The visible part of a warrior has about 50 instructions. Even a 2c decoy maker can only make 2*2*16 changes to the core. If the qscan finds sth., the possibility of it being the opponent, is still about 50 % [14:09] IF an opponent has a qscan [14:09] what if it is a 8 line incendiary? [14:09] with 20 line decoy maker? [14:09] yes, otherwise the qscan is useles [14:10] but, we're considering err... a kind of damage control here. [14:10] if someone is trying to trick the qscan [14:10] Join: sascha joined #corewars [14:10] Boing... [14:10] to have qscan capable of ignoring that decoy... [14:10] hi Sascha [14:10] Hi, [14:11] 140 against fsh but 119 against 94nop.... [14:11] that's very hard for me... [14:11] TheAntiImpBullet is a ccpaper+stone [14:11] and it is at the bottom... [14:11] but it didn't have such a high fsh score (as I remember) [14:12] * Nenad goes to look [14:12] hmm, i don't understand that stddev was 20.28..very good i think [14:12] 137,14 [14:12] 24,37 stdev here. [14:12] Hmm, same changes are needed.. [14:12] but 8 points more on the hill :) [14:13] I think i must increase the process pwoer of the stone, for more anti scanner behaviour [14:13] is this that you've sent a successor of pppc? [14:14] no i haven't any idea for pppc2 [14:14] sascha: I give very few processes to the stone in TheAntiImpBullet ... on purpose - that's why I call it paper/stone rather than stone/paper ... its main target are imps [14:14] but I keep the stone small [14:14] so it can't be stunned early [14:15] I can send you the code, if you like. I'll send it to C.Birk soon, anyway [14:15] Hmm that won't help..my paper is kind of unconventionell [14:16] based on Hydra..which is based on paperone.. [14:16] so i think the clear can be mor powerful than on standart silk-stylers [14:18] Join: Core29 joined #corewars [14:18] hi Core29 [14:19]