[00:17] MSG: [01:36] Part: Core29_ left #corewars [08:38] Join: brx_ joined #corewars [08:39] MSG: Ping timeout: 252 seconds [09:04] Nick Change: brx_ changed nick to brx [09:19] Join: Roy joined #corewars [09:19] woow :) me like this: [09:19] 1 38/ 27/ 35 Halcyon Roy van Rijn 150 20 [09:19] 2 39/ 29/ 31 HullabaTwo Roy 149 118 [09:19] 3 34/ 23/ 43 Made in Vain Roy van Rijn 144 41 [09:19] 4 37/ 31/ 32 Eccentric Roy van Rijn 144 64 [09:20] Hehe active day yesterday, but that was a suprise (ok, I dropped a couple of places again) [09:21] Nobody here to share my joy... too bad! [09:21] MSG: Client Quit [09:22] Join: Jens joined #corewars [09:22] :) [09:31] Join: Roy joined #corewars [09:31] Hi Jens [09:31] Just here for a second [09:31] MSG: Ping timeout: 252 seconds [09:32] Nenad> I'm not angry, the opposit! Aging all my warriors and not pushing any of, the 108 emails in my mailclient... I'll take that for granted [09:32] :) [09:32] They go into a seperate folder and the [Delete All] [09:33] Hehe ... but the 94nop-hill looks now far better [09:33] Naaah... ;) [09:34] Two challenges back, that was a nice hill (1/2/3/4!) [09:34] Depends on the point of view [09:35] Thats for sure.. Mandragora looks cool, but I think it'll always have some crutial weaknesses [09:35] If someone continues this challenge-drive some warriors might become very old [09:35] All other warriors from Fizmo dropped, thats too bad [09:36] Arrow 610 is getting very old atm [09:36] And Borgir (442) too... although its not going so strong anymore [09:36] I have to go, party!! [09:36] * Roy waves [09:36] MSG: Client Quit [09:36] bye, Roy ... too late as usual [09:55] Join: brx joined #corewars [09:56] Join: Mizcu joined #corewars [10:25] Mi brx, Mizcu! [10:25] hi Jens [10:25] It seems, that Nenad is back and is keep sending new warrios to 94nop [10:26] hi Jens , Mizcu [10:26] moving tomorrow [10:26] might have much time doing interesting stuff then, without internet. [10:26] Yes, I've heard, that there is life without the net ;-) [10:27] But I don't believe it ;-) [10:27] I was talking about getting more into corewars again [10:27] and ocaml, finishing my haskell raytracer [10:27] things like that [10:27] :) [10:28] hehe [11:30] Join: brx_ joined #corewars [11:31] MSG: Ping timeout: 252 seconds [11:32] Join: Nenad joined #corewars [11:32] Hi [11:37] MSG: Quit: leaving [11:38] Join: fiveop joined #corewars [11:53] hi fiveop [11:58] Hi Nenad, fiveop! [12:00] Hi Jens! [12:00] Nenad: You seem to continue your hill-againg-initiative ;-) [12:01] :) [12:01] I've just been testing some warriors [12:01] it seems that mandragora doesn't like fixed carpet scanners ,as well [12:02] but I was really happy to enter the hill with my new scanner (which is totally unoptimaxed at this point) [12:02] which one? There were so many recently ;-) [12:02] ... it has a really neat trick in it [12:02] AttackOfGreaterPanic [12:03] But I can't optimax it now - optimax is already busy with another s/p [12:04] btw, I've just posted the code of Uncertainty to r.g.c , if you want to take a look [12:05] I'm just reading your reply to Qscanners V [12:05] I'm already working on a new quickscanner [12:05] HFB will be in trouble when its food falls off - ElvenKing is at the bottom of the hill... But I think that it will rise higher once scanners start flowing in. [12:06] Join: brx joined #corewars [12:06] I hope, that HFB has enough power to survive the death of Elven King. [12:06] Hi brx! [12:06] Hi brx! [12:07] Nenad: It is quite difficult to write a new quickscanner. I have a lot of ideas, but there was always a tiny problem, which rendered it useless :( [12:08] hi [12:08] There are a lot of open questions for quickscanners and I really love to work on them, but it will take a lot of time to deal with them, especially, if I stay the only one, who works on them. [12:09] well, I could join, then :) [12:09] I'm already writing sth. like Quickscanners - Part V - Part II [12:09] about papers with quickscanner and optimizing them [12:09] besides, JM has a lot of knowledge about qscans [12:10] Jens: optimization itself is an open question ... how to do it ... [12:10] that's why I'm writing the new text [12:11] I've had maybe even unexpectedly good results in Earthdragon's paper by using an approach that I haven't used before... it doesn't score well on the current hill, but there is not much food for it. [12:11] it speeded up the optimization a lot. [12:11] ? [12:12] well, it uses a 9instruction paper with 6processes and some antiimp power. [12:12] you probably know how that looks like.. [12:12] yes, I think I can imagine it [12:13] I optimaxed the 6line 6process bottom part first, and then just c/pasted it into the larger part and optimaxed the leading pstep separately [12:13] because I think [12:13] that the cooperation [12:13] of the last two psteps [12:13] is more important [12:14] of course, reducing the number of values in optimization (because there was one less in stage one) speeds up the process [12:14] I cannot agree. Look at the score surfaces of YAP and you will see, that both paper-steps are equally important [12:14] how many pspteps has YAP? [12:14] two [12:14] :) bu here there is 3 [12:14] *but [12:14] Yes, I know [12:14] and I'm talking about the last two [12:15] I'm already working on "Score Surfaces - The Movie" for 3 silks [12:15] But I don't expect the images to be that much different [12:15] why? you know that in math, as well as corewar, unexpected things may arise from smalles changes [12:16] btw, I think that Neo notices something similar in 3step silks [12:16] are you working on closed or open modules in 3step silks? [12:16] there may be a difference there, too [12:16] Not much [12:17] ? [12:17] I have calculated some patches for different versions of YAP [12:17] fsilk/fsilk, fsilk/bsilk, fsilk/fsilk + jmp, fsilk/fsilk + djn.f, fsilk/fsilk + jmz.a [12:17] consider this: [12:17] spl [12:17] mov [12:18] spl [12:18] mov [12:18] and this [12:18] spl mov [12:18] Apart from having different scores, the images are the same [12:18] mov [12:18] jmp [12:18] in the first case, [12:18] step2 doesn't influence the spread by step1 [12:18] while in the second case [12:18] copies copied step2 far [12:18] are doing copying by step1 as well [12:19] so, there has to be some difference. [12:19] only in score [12:19] but the structure of the score surface seems to stay the same [12:19] hm... [12:19] but the height changes? [12:20] Sascha and I have calculated several new score surfaces, but I haven't published the images/results. [12:20] I will work the next days on the text about optimizing papers and the I'll start to publish all remaining images/results for the score surfaces [12:21] I'll happy to talk about score surfaces then. [12:21] OK? [12:22] well, I guess that what i've said about optimization above can be interpreted like this: if you optimize the smaller component, you already know that that part is working correctly, and it is logical that the leading step can be changed so that the whole thing works nice. If you don't do that: it may happen that you'll generate good pstep2 and pstep3, and disregard that randry because of badly generated pstep1! [12:22] ok. [12:24] hm... maybe I should test my cuurent best randry of the s/p that is optimizing... [12:24] ... and age the hill a little bit :) [12:24] it has a decent optimax score of 145,36 ... after 2 hours of optimization [12:24] I want to see if it's better than EarthDragon... [12:25] so I'll dat the hill first to see its real score. [12:25] it bothers me that you need to dat the hill each time you want to see the score of first 20 warriors [12:25] I'm quite curious how many people will swear today after they have seen how many emails they've got from koth.org [12:25] why can't there be an option to see that score seperately... [12:26] :) well, there won't be that many [12:26] Roy was quite happy about it, because his warriors did well [12:26] and are a little bit older now [12:26] tiny hasn't aged for weeks :/ [12:26] besides, I've been resting from corewar for about a month. I need to catch up a little now :) [12:27] there was too much silence at the hills lately. [12:29] I'm downloading now all data from the SAL-hills to make some images about the age-development of the hill [12:29] just like the image for the 94nop-hill at koth.org on my homepage [12:30] I haven't seen ti yet! [12:30] the url? [12:30] http://corewars.jgutzeit.de/history/index.en.html [12:30] you could also make some graphs about how prominent some name phrases like shot,scan,clear,silk, etc. have been [12:30] But the image is "old". I'll update the page soon [12:30] fiveop: good idea [12:31] fiveop: I think, John has already done that. Look at http://corewar.co.uk/logs.htm [12:31] Join: Mike joined #corewars [12:31] Hello [12:31] hi Mike [12:32] Hi Milke (still not convinced, you aren't a chat bot) [12:32] *Mike [12:32] :) [12:32] * Mike wonders [12:32] Jens: what I'm worried about is little or no activity at 94b [12:32] You need to find some beginners [12:32] Jens: have seen something like that but not as a graph over time [12:32] * Mike looks around [12:33] I have seen some submissions to the 94nop-hill, which looked like from beginners. Maybe we should have some prominent links to the beginners hill [12:33] It isn't that easy to find SAL [12:33] ... we should revive sourceforge first [12:33] Everyone should link to it. [12:33] imagine this: [12:33] you are a beginner [12:34] and go to sourceforge [12:34] and see that nothing works [12:34] you probably come to the conclusion [12:34] that people don't do corewar anymore [12:34] ... [12:34] so, we should either revive it or remove it [12:34] A message on the front page about it being down would be good [12:35] Mike: yes + links to LIVE hills [12:36] OK, first I'll finish the new images for the hill-development. Then I set some links to the hills [12:36] I think most people find Koth or corewar.info first [12:36] I liked sourceforge more than SAL as a beginner... [12:36] at first. [12:37] well, I'll probably have my corewar webpage ready in two weeks or something like that... [12:37] Oh :) [12:37] Anything ready yet? [12:38] I should have done it some time ago, but I was busy... no, nothing online yet. [12:38] Any plans yet what to go on it? [12:39] some warrior code, maybe some articles - I was planning to make a good oneshot tutorial, some info about me, corewar in general, links, etc. you'll see [12:40] I want to make a site too [12:40] Selling "Lucky" warrior names [12:40] Bvowk could be my best customer :) [12:41] No, I don't think. I've already tried to talk him into using better names. He didn't want to [12:42] ... I don't know why there aren't any new CoreWarrior issues.... [12:42] there was more than enough action to write about [12:42] keeping to 25 characters would be goof [12:43] *good [12:43] 2 rf rounds + hill reports + some article [12:43] maybe we should start a weekly journal or something like that... [12:43] Who has time to make it? [12:43] CoreWarrior? [12:44] something like ... a newsletter? :P [12:45] Choose 12 people and each person chooses a month and makes the newsletter for that month [12:45] Mike: are there 12 of us alive? [12:45] More [12:47] Post it to r.g.c. Maybe you'll find enough people :) [12:48] it seems that 94nop is quite young now, Jens. no more unbeatable warriors... well, maybe arrow... [12:48] but it's hardly unbeatable... [12:49] Yes, I'm just making a better image with some info about when the most important warriors have dies. [12:50] Schmidt, Metcalf, Neo, Jens, Gunnell, Mizcu, Rijn, Zapf, Nenad are all active and prob know enough about Corewar [12:50] Was is Corewar? [12:50] *what [12:50] ;-) [12:50] Sorry should I say Corewars? [12:51] HFB will probably be your first 100+ warrior, Jens [12:51] Officially it should be "Core War", but I was only expression my lack of knowledge about cw [12:51] ;-) [12:51] Nenad: No [12:51] Steve Gunnel... haven't seen him here in ages... [12:51] Nenad: +2600 ;-) [12:51] although he did submit something lately [12:51] I was looking at the hills to see who is active [12:51] :) I would be happy about it, too [12:52] it is my best food source for KL [12:52] which is now age = 200 [12:53] Why did Metcalf kill prototype? [12:53] ? what prototype? [12:53] when? [12:53] Sal nano today [12:54] ... no idea. (I dont' have any warriors there...) [12:56] :(. it scores a little worse than EarthDragon, apparently... I should probably kill it... [12:56] Is Arrow published? [12:56] no [12:56] of course not :) [12:57] Any hints what is in it? [12:57] scanner [12:57] Scissors obviously but are there any neat tricks [12:57] well, like the name says: zoom trick scanner. probably with a qscan [12:58] and triggering on fixed number of successfull scans [12:58] something like PanicAttack, I would guess. [12:58] but I haven't optimized it yet. [12:58] I understand zoom trick and qscan but not the rest [12:58] Fizmo probably spent several days improving Arrow [12:58] Crikey! [12:59] That's like having a full time job. [12:59] if you want to age beyond 1000... [12:59] How many lines of code, 100? [12:59] that doesn't matter. [12:59] 2.5 line of code per hour :) [13:00] I wonder what is the longest somebody used to make a nano warrior [13:00] Several months [13:00] Ask bvowk [13:01] :) [13:01] MachinesWillRule was made over a long period of time, too [13:01] He made lots in several months though and it is now autonomous? [13:02] He uses an evolver (+optimizer?) and the best versions are sent to nano automatically [13:02] Which hill is MachinesWillRule? [13:02] 94nop! [13:03] evolved, as well (I think) [13:03] and then added a qscan and some other stuff. [13:03] but the paper was originally evolved [13:04] hybrid cyborg code [13:06] New images for 94nop at http://corewars.jgutzeit.de/history/index.en.html [13:06] I hope, it is a little bit easier to "interpret" now [13:07] your page is the most update corewar site (except for koth.org hills) [13:07] All +1000-warriors are now marked [13:08] And look, who Google spits out, when you search for score surfaces :) [13:08] how old is the average thing that falls off [13:08] Give me 5 minutes and I can tell you [13:09] Maybe a silly question I figured it [13:09] 20 or 21 right? [13:10] no, that is the rank [13:10] I can't explain very well. [13:10] 20 on there [13:10] age goes up by 1 for all when something new goes on there [13:11] but one has to fall off [13:11] age probably gets to 20 on average [13:11] If I am right I choose your next warrior name :) [13:12] lol [13:12] Free of change of course [13:12] The average age of an warrior to fall of the hill is 21.71 (94nop, based on all available data, I could find, only successfull challenges counted) [13:12] But I will put a quote from you on my site selling "Lucky" warrior names saying how satisfied you are with your name. [13:13] Is that close enough? [13:13] close to what? [13:14] 20 [13:14] to 21.71 [13:14] Close enough to my guess of 20 or 21 [13:16] Jens is silent... he is trying to check if 21.71 == 21 by repeated attempts [13:16] Blimey, I've gotta go! [13:16] * Nenad waves to Mike [13:16] I'll be back with the name! [13:16] Part: Mike left #corewars [13:16] I'll have to go, as well, I'm afraid. time for lunch here. [13:16] * Nenad waves [13:16] Hmm ... I was thinking about why the average is about 20 [13:16] MSG: [13:16] * Jens waves too late [13:36] Join: Core29 joined #corewars [13:44] Hi Core29! [13:52] hey Jens [13:53] I'm really looking forward to my next holday .... (after 17. 10.=...hope to find some time to get started again with cw [13:53] :) [13:56] work just eating up too much time... [14:34] Join: Nenad joined #corewars [14:41] MSG: [15:04] In case someone is interested: I've updated http://corewars.jgutzeit.de/history/index.en.html again. Now the page included a short list of main events for 94nop. [15:05] * Jens is away for a while ... [16:26] MSG: Ping timeout: 252 seconds [16:29] MSG: Quit: Trillian (http://www.ceruleanstudios.com [17:17] Join: Nenad joined #corewars [17:29] Join: Mizcu joined #corewars [17:29] hi Mizcu [17:29] hi Nenad [17:30] I was just here for a quick visit to test my new scanner on 94nop [17:31] ... the hill is very hostile to scanners- lots of stone/x ... but it did ok. 15th place [17:31] it seems that ElvenKing will be the first to fall off [17:32] how are you? any progress with the paper? [17:32] not any redcode so far today [17:32] 2 et-matches and some reading done, not much else [17:33] well, not much action today in corewar, anyway. [17:34] it's mostly waiting for rf24 to be run and for the results to be announced [17:35] I think that results will be known in two days... [17:35] well, I've got to go now... [17:36] * Nenad waves [17:36] MSG: [17:56] Part: Jens left #corewars [18:32] Join: Fizmo joined #corewars [19:02] Join: Nenad joined #corewars [19:02] hi Fizmo [19:08] hi Nenad [19:10] how is going? [19:10] nice. I was just testing some scanners at 94nop. it seems that the first version was better afterall. [19:10] I'll resubmit it in a minute. [19:10] a nice zoom trick scanner based on Dandelion [19:10] I've made a quickhack [19:11] ? [19:11] stone/scan into booted sd-clear/imps [19:11] still send it on 94nop [19:13] Did you use a different clear than Dandelion? [19:13] yes. [19:13] Dandelion uses Forgotten scanners, right? [19:13] yes [19:13] I used SOG, although it might not be the best idea at the current hill [19:14] but there are more clears which can beat paper/imps [19:14] and I've made some other changes: [19:14] different zoom trick + a very cool selfmutation for entering clear phase + different step, span, boot distances... [19:14] scores very different... [19:15] the problem is that there are no paper imps on the hill now, so... [19:15] well, they will appear [19:16] btw, is Arrow similar to Dandelion? [19:16] no [19:17] there are some similarity in the scanloop, but the rest is totally different [19:18] hmm... your quick hack didn't make it on... [19:18] heh, I hadn't expected that it enters [19:18] just score 110 against the fsh [19:19] well, I am quite satisfied with this new scanner, although it may fall off quickly - many hostile stones on the hill [19:19] it's only on half way to an sd-clear/imps oneshot [19:20] just replace undo_unsheathe_attempt ;-) [19:20] :) [19:21] seems quite similar to Starfall [19:21] well, it isn't... it is just a paper. [19:22] I see [19:23] I was mostly working on some new s/p's lately (EarthDragon, Forestdragon), but there isn't enough food (scanners) on the hill for them to do well. [19:23] and there are already several good s/p's there [19:23] Borgir is the only one in crisis [19:23] yes, it is [19:24] you should probably submit a couple of scanners to save it. [19:26] but it seems that the ElvenKing will die first... [19:26] Join: sascha joined #corewars [19:26] hi Sascha [19:26] Godd evening alltogether... [19:26] hi sascha :) [19:26] Hello Chris.... [19:27] I think I have some scanner never send onto the hill, just have a look [19:28] Hey, the hill doesn't need more of arrows kind...(Seems strong enough to be 1000+) [19:29] Arrow will be 1000+ only if KL becomes 600+ (look at the 94nop hill huge array) [19:29] without its points, Arrow has about as many points as RedSun... [19:30] but I wouldn't complain about that happening [19:30] :) [19:30] Maybe someone could challange some fooder for arrow... [19:30] Time for a real good stone... [19:31] unless enough s/p and pap are on the hill KL will be save [19:31] Save KL, long live arrow ;-) [19:32] heh :) [19:33] I send our rather unfinished teamwarrior, Sascha ;-) [19:33] Shame on me..sorry..but too many other projects running here.. [19:34] np :) [19:34] just checking now how good it scores so far [19:36] I was just wondering... when was the last time that a strong fixed carpet length scanner (like Win!) has been on the hill for more than just a couple of challenges [19:36] do you remember? [19:36] Oh, thanks Chris...So at least my name is on the 94nop so the poeple don't forget me ;-) [19:38] Well, at least I can say, that I pushed my own warriors [19:39] Oh no.... [19:39] Shit happens..didn't you check score against borgir before challenge ? [19:40] doesn't matter ;-) [19:40] and we can say, that we are on a good way with the bomber/imp [19:41] bomber/imp? you mean - s/i? [19:41] no [19:41] not a simple stone ;-) [19:41] incendiary + imp? [19:42] like airbag? [19:42] incendiary [19:42] :) yes. that is fun... loses far less to papers, I assume [19:43] but probably wins a little less against scanners, too [19:44] wow, I didn't remember that I've made such crazy thing: double blur-scanner [19:45] I think I'll give it a try on 94nop [19:47] blur scanner? [19:50] two blur scanner running independantly parallel [19:51] I think I should add a zooom-trick as well [19:52] to both independently? [19:52] or one to trigger them both? [19:52] wow, never expected that [19:52] just boot a copy away and launch both [19:52] no trigger [19:53] very straight forward, very simple [19:53] about adding a zoom trick... [19:54] if I look to the scores on 94nop I think I'll implement zoom-trick [19:54] should then score much better [19:59] sorry, it wasn't planned that it hurts KL [20:00] :( [20:01]