[00:34] MSG: Remote host closed the connection [00:41] MSG: Quit: humhum [02:46] The way that SAL gives you results is somewhat confusing [02:55] Join: |caesar joined #corewars [03:03] <|caesar> halo [03:03] <|caesar> morning [03:03] Hello [03:04] <|caesar> hey, so u r have a good sleep:) [03:07] <|caesar> datagram,how long a HDD's life? [03:13] That depends on the HDD [03:13] But it is good to upgrade every two years or so if you need better performance [03:18] <|caesar> ok,thx [03:18] <|caesar> i think is to small for me now [05:00] Join: Mizcu joined #corewars [05:14] Hey Mizcu [05:15] hi gram [05:17] How goes it [05:17] My warrior got beat up on 94b [05:20] Yes, it might be because of its inability to face papers [05:20] But i think you can get higher without adding imps, and only by modifying the warrior [05:30] No, I did really well against papers [05:30] # %W/ %L/ %T Name Author Taken Given [05:30] 19 77.2/ 1.5/ 21.3 Break Down and Cry Jens Gutzeit 252.9 25.7 [05:30] 16 70.2/ 8.7/ 21.1 Barkosta[v0.3] inversed 231.7 47.1 [05:30] 4 66.0/ 9.6/ 24.4 D'n'B [v0.3] inversed 222.4 53.2 [05:30] 18 61.1/ 8.5/ 30.4 Biomass_v0.8 inversed 213.8 55.8 [05:30] 14 62.9/ 14.7/ 22.5 3[sm]md v0.2 inversed 211.1 66.5 [05:30] 9 62.2/ 14.2/ 23.6 Blotter inversed 210.2 66.2 [05:31] 2 64.9/ 19.9/ 15.2 Yatima v2.0.5 Jens Gutzeit 210.0 74.8 [05:31] 6 52.3/ 1.3/ 46.3 Snare_Rush_v0.7 Inversed 203.3 50.3 [05:31] 12 53.4/ 16.4/ 30.2 stealthbomb Fizmo and Neutrino 190.4 79.4 [05:31] 21 46.1/ 2.1/ 51.9 Everybody must get STONED madjester 190.1 58.1 [05:31] 15 51.9/ 16.7/ 31.3 Tom David Moore 187.1 81.5 [05:31] 8 51.5/ 16.9/ 31.6 The Silent Death Anonymous 186.0 82.4 [05:31] 24 45.2/ 6.7/ 48.1 Envane [v0.5] inversed 183.7 68.3 [05:31] 20 52.6/ 22.3/ 25.1 Enigma 2.1 brx/Roy 182.9 92.1 [05:31] 3 48.2/ 35.1/ 16.7 Svarog Nenad Tomasev 161.3 121.9 [05:31] 23 42.9/ 25.1/ 31.9 Spark_v0.4 inversed 160.7 107.3 [05:31] 17 41.3/ 27.5/ 31.1 Blindfolded Miz 155.1 113.7 [05:31] 7 40.6/ 45.2/ 14.2 Xenocitrum_v0.8 inversed 136.0 149.8 [05:31] 10 41.7/ 53.3/ 5.0 Star Sascha Zapf 130.2 164.8 [05:31] 26 35.4/ 41.2/ 23.4 Simon Gruber datagram 129.6 147.0 [05:31] 25 31.9/ 45.4/ 22.7 bloodhunter elkauka 118.3 158.9 [05:31] 22 33.2/ 48.6/ 18.2 dx42e inversed 117.8 164.0 [05:31] 5 32.3/ 48.4/ 19.3 test Andreas Scholta 116.3 164.5 [05:31] 13 32.3/ 58.9/ 8.8 think twice v.2 (optiMAXe el kauka 105.6 185.6 [05:31] 11 26.1/ 56.1/ 17.9 Unknown Neo 96.1 186.1 [05:31] 1 28.1/ 61.7/ 10.2 [05:31] 1 28.1/ 61.7/ 10.2 Dragonfly S.Fernandes 94.4 195.4 [05:31] That is assuming I am reading this right [05:31] The top 6-7 are all papers [05:31] You are that list the wrong way [05:32] That is how well they did against me? [05:32] Taken = Losees, Given = wins [05:32] Oh [05:32] So they I kicked the crap out of #1 [05:32] It seemed less logical that way [05:32] It beated some oneshots and scanners, but lost to papers [05:33] I see [05:37] Join: base_ joined #corewars [05:37] It's base_ [05:38] aye [05:46] and yet another prototype bytes the dust.. [05:47] Mizcu, ? [05:48] An idea that turned out to be that, scrapped it [05:49] I wonder were "that" came out, was about to write "bad" [05:51] I read it as bad anyways [05:55] MSG: <|caesar> Ping timeout: 252 seconds [06:10] Time needs to go faster so all the European guys log on : P [06:38] Join: base[64] joined #corewars [06:41] MSG: Ping timeout: 252 seconds [06:56] I'm benchmarking against Wilkies now [06:56] After I will do the newer benchmarks [07:02] Join: |caesar joined #corewars [07:11] Wilkies is a good starting point, though a bit depressing [07:16] Heh [07:16] I see alot of them listed as awesome in 88 [07:17] What is a good score to get vs wilkies? [07:18] I think you need about 120 to get to beginner-hill [07:19] Best score against Wilkies is, if i recall right 202 by Fire&Ice (a p-spacer) [07:22] But for beginner first goal would be to get over 100 [07:34] Blah, logitech's drivers suck [07:34] MSG: [07:38] Join: Mizcu joined #corewars [07:47] Back to cordless, no more that oldie-one anymore [07:49] 111.1 : / [07:49] Well, thats a start [07:51] The other stones beat my ass [07:51] Tornado has speed-advantage, bluefunk has imps, i dont know about firestorm [07:53] Cannon has imps too [07:53] Obvious solution would be to add imps too [07:53] Now I am trying vs wilbez [07:54] Will that help versus paper also? [07:54] Yes [07:54] Roy said all I needed was a quickscan added to the program [07:54] But I don't understand how that helps immensely [07:54] Roy didnt want to add another main component into the warrior [07:55] Quickscan does help a bit, but not immensely [07:55] (Unless you use Pyramid-style large qscan) [07:56] [10:24] and a qscan and you're done ;-) (its bad against papers...ssd clear would do better) [07:56] I dont believe myself that stone-> ssd clear could be good against papers [07:57] By the time you start the ssd-clear, the paper has already covered 50%+ of the core [07:57] Yeah that's what I figured [07:57] I figure I have to add scissor components in [07:57] And if you throw less bombs, thats not good against other stones [07:58] Scissor components are very tough to add to work together with other components [07:58] I know but I need that to help beat out paper [07:58] You cant beat everything [07:59] Yeah but my biggest problem on the hill is papers [07:59] You could do a stone/oneshot -hybrid, like pswing is on wilkies [08:00] But then again, the hill doesnt really need another oneshot or scanner [08:01] 7/10 warriors in beginner-hills top10 are scanners or oneshots [08:02] Which I find ironic because I did well against all the top ones, but got beat up by the lower [08:03] Have you been explained/do you understand how an imp-ring/spiral exactly differs from normal imps? [08:04] Yeha [08:04] yeah* [08:04] Adding an imp-spiral would give resistance against papers, but then there would be the problem of process-balancing [08:05] Because if i remember right, there was only one process looping in datagram's bombing-loop [08:05] A ring takes minimum of 3 processes (well 2, but thats not good for this situation) [08:05] Before the core clear, yes [08:06] and 1:3 is really overwhelming for the ring [08:08] So the bomber would need to be turned multi-process to add the ring [08:11] So I want to go from Bomber > core clear > imp [08:11] actually 3:3 would be too heavy too, 6:3 is probably the minimum where the stone keeps its usefulness (i dont know what is good rate to be exact) [08:11] The core clear portion creates a multi process warrior [08:13] i think i actually used 5:1 in my warrior Static, but thats not comparable because they were self-splitting components [08:36] Join: fiveop joined #corewars [08:52] *ping* <- got stupid idea [08:53] Well, atleast my creativity has returned after a while [09:13] ? [09:14] i had a 6 month phase of complete uncreativity redcode-wise [09:14] heh [09:14] now i've had weird ideas for some time [09:14] Anything effective? [09:14] (140 wilkies with 30min coding... not that bad.. but not competitive anyway) [09:15] Nothing really, except one paper [09:15] I realized that I benchmarked my warrior without his decoy code [09:15] So I'm doing it over [09:15] Decoy is 85 more lines, so yeah [09:16] And i dont know if that paper is competitive since its kinda in-between "normal" paper and heavily anti-imp paper [09:16] I see [09:16] Do you think it is plausible to get this warrior I have to #1 on 94b? [09:16] or would I have to use something else [09:17] At current stance, no. By adding and modifying, yes [09:17] Well that's what I mean heh [09:17] But you can go towards the top one component at a time [09:19] If i would send this thingie i have been working for last 30mins, it would probably go straight at 94b, or atleast high there [09:19] I see [09:19] But you're not a beginner [09:19] So yeah [09:20] But on the other hand it would take 3 or so days to get to 94b #1 with a paper or scanner [09:20] Heh [09:20] It seems like its more of a testing hill for higher rank players than it is a beginner hill [09:20] Its not that much ago when i coded a stone/imp on 15 minutes and it went 3rd or something like [09:21] The hill is really ripe for a S/I [09:21] (especially one that is good against oneshots) [09:21] Well if I were to finish the addition of an imp into this it would be a stone/imp [09:21] yes [09:22] But you would first need to modify the stone a bit to get the imp-addition to work [09:22] Is it too slow to do bomb scattering > core clear > imp? [09:22] the core clear part is multiprocess [09:22] You see, the idea would be to run Imp the same time as bomb scattering [09:22] That early? [09:22] So you need multiprocess scattering [09:23] Imp(ring/spiral) isnt gonna do anything after 30000 cycles [09:23] You need to get it early on, because it is for defence [09:23] http://koti.mbnet.fi/mizcu/corewars/ [09:24] The 15min S/I is in bottom of Redcode [09:27] I got a package for myself and ill be fetching it; bb in 1 hour [09:56] Join: brx joined #corewars [10:19] ooh, 7 minutes less than in an hour [10:24] rm [10:40] MSG: <|caesar> Read error: Connection reset by peer [10:41] MSG: Quit: 12( www.nnscript.de 12:: NoNameScript 3.81 12:: www.XLhost.de 12) [10:42] brx: http://cs.helsinki.fi/u/janmatti/copyprot_readme.txt [11:08] Join: |caesar joined #corewars [11:08] <|caesar> Redcode... [11:17] ? [11:20] <|caesar> nothing [11:20] <|caesar> so,datagram is a beginning too? [11:22] yes [11:26] Yeah [11:26] If I understand you correctly, I am : ) [11:35] Mizcu, I could modify the warrior so that instead of a core clear it will become an imp [11:36] The warrior already overwrites itself so that it becomes multiprocess at that point [11:36] Mizcu: hrm [11:53] 13:51 < TimoT> cliini: just noticed a video of the tuesday demonstration, lot's of ponytailed geeks attending [11:53] 13:52 < cliini> Yeah. I guess among all this raving about cd copying the most worrisome aspects of the law [11:53] went entirely unnoticed. [11:54] 13:52 < cliini> I printed the list of the vote and put it on the coffee room wall. :) [11:54] 13:53 < cliini> KingNato, well, it's the coffee room for the people who count the votes in every election... [11:54] brx ... need urgently some tips :P [12:06] Join: elkauka joined #corewars [12:06] hi [12:11] Hello elkauka [12:12] What's up? [12:13] <|caesar> hi elkauka [12:13] i'm happy to be back here. yesterday mmy internet connection crushed and up to five minutes ago i had no idea why. never try to block the generic host process from win xp [12:13] Mizcu: 94b rep? [12:13] Heh [12:14] elkauka, svchost.exe controls most networking components [12:14] Even some services that don't necessarily need to be networked, too : x [12:15] Join: Jens joined #corewars [12:15] :) [12:15] <|caesar> what's that mean about"94b rep"? [12:15] It is Jens [12:15] Haven't talked to you in a few days. What's up [12:15] <|caesar> hay,jens [12:16] I'm working on the Core War Time Machine, which let's you view old standings for the 94nop-hill. [12:16] I hope it to be ready later today [12:17] hi jens [12:17] I made my run at 94b today [12:17] And got smacked down [12:18] I did do very well against the rank 1 warrior though. I found that ironic [12:18] name of the warrior? type? [12:18] Simon Gruber? [12:18] hi Jens [12:18] MSG: Read error: Connection reset by peer [12:19] datagram: It is often quite easy to be very good against one or two warriors, but to get good enough against the rest is the problem. [12:19] datagram: is it a stone? [12:19] Yes [12:19] datagram: well the hill is full of papers [12:20] http://sal.math.ualberta.ca/warrior.php?key=94b&id=467&t=467 [12:20] I know [12:20] I did well against 7, but the papers beat me up [12:20] 7 warriors, that is [12:20] test is my scanner [12:20] Jens paper kicked my ass [12:21] brx, Yeah I do well against scanners [12:21] Is #1 a scanner? [12:21] The source of Yatima 2.0.5 is published at Koenigstuhl. You can view its source there. [12:21] it is a oneshot [12:21] Ah [12:21] a oneshot is a scan->clear [12:22] datagram: Oh, I just see, that you probably talk about Break Down and Cry. [12:22] Join: elkauka joined #corewars [12:22] yeah [12:22] datagram: It is a p-spacer between paper and stone [12:22] What does p spacer mean [12:23] It switches between the to strategies [12:23] *two [12:23] if it turns out, that the used strategy doesn't work. [12:23] Oh [12:23] Yeah I also didn't see that it was P-Space enabled until after I submitted [12:24] http://www.corewar.info/lexicon/p-switcher.htm [12:24] I know 'p-warrior' [12:24] That is what the wiki has [12:24] But yes [12:25] Are many of the warriors there using p-space? [12:25] I don't think so. [12:26] even on the 94 with p space are only a few pspacer [12:26] hehe, I've just noticed, that the strategy line for BDaC is wrong, it says: paper [12:26] it doesn't have too much merit on a biased hill such as 94b [12:26] Why not use pspace if you have it? And just boot whatever part of code you have away [12:27] datagram: that adds overhead pure scanners will get points off [12:27] and since the hill is crowded with papers you'd probably get better scores with a oneshot [12:27] or a Moore-style-paper [12:28] http://www.corewar.info/lexicon/paper5.htm [12:29] I don't like papers [12:29] Too boring [12:30] That's stubborn, but I'd rather get my ass kicked with a bad stone that keeps me entertained : ) [12:30] Or a vampire. You know how I like vampires [12:30] elkauka, I was thinking about making and then I saw your 94b entry and felt very unoriginal [12:31] about making a vampire/scanner*** [12:32] hehe [12:32] But I like vampires [12:32] Simon actually started as a vampire, but then I liked keeping it's bombs small [12:33] So yeah [12:33] have you seen dracula2003 by roy? [12:33] I don't think so [12:33] that actually a good point to start on vamp/scan [12:34] I downloaded the source [12:35] All of the guides need to be updated to reflect new 94b entry levels : ) [12:35] I think you guide says you only need a score of 60, Jens [12:36] where? in a speccial benchmark? [12:36] Against Wilkies [12:36] I think... [12:36] to get onto the beginner hill? [12:36] datagram: Which guide? [12:36] yeah [12:36] Jens, Quickscanner, I think [12:36] I could be wrong [12:38] datagram: In "Qscans I" I say a little bit about sth. > 120 against Wilkies to enter 94b [12:38] my theory is: 90+ against fsh2.0 to get onto the beginner. 120+ for 94nop [12:39] elkauka: Probably depends on the warrior. My experience is (for papers, p/s): 130+ to enter, 145+ for koth [12:39] 145+ for koth anyway [12:39] How outdated is the nimbus spiral? [12:40] I am going to try taking out the core clear from my code and including an imp spiral [12:40] Several year, but depends on what you are using it with. [12:40] *years [12:40] or if I can figure it out I could try making it into all 3 [12:41] elkauka: 120? [12:41] elkauka: that number is too low. [12:41] I scored 112 against Wilkies [12:41] And was 20 points too low for 94b's lowest spot [12:42] my twoshot "thinktwice" had 119 against fsh2.0 and enter place #15.(but was pushed off at age 3) [12:42] elkauka, And I lost an assload of times to Jens paper and beat your vampire...so unfortunately I moved ya down, too [12:42] datagram: ignore all the ranting about minimum scores for specific hills. [12:43] datagram: and remember one thing instead: against any sufficiently balanced benchmark, a score of < 120 is *really really bad* [12:44] Heh [12:44] Well +100 was good enough for me for now [12:44] And my 2 days of playing [12:44] Will the addition of this nimbus spiral ruin my code? [12:44] but it also depents on the warriors on the hill. even a worse oneshot could be great if the hill is crowed with papers [12:44] Also, I am doing this: [12:45] spl 1, >-7 spl 1, >-9 spl 1, >-11 [12:45] To make more decoys behind my code [12:46] -7, ... is too close [12:47] -15, -35, -55, -75, -100, -200? [12:47] Or more [12:47] why aint you using 3c decoy maker? [12:47] lDEC equ 6 [12:47] sDEC equ -232 [12:47] start [12:47] i for lDEC [12:47] mov.i {sDEC+(i*2), rof [12:47] I just want it to get done while my spiral runs [12:47] more, > 200, < 7900 [12:47] MSG: Ping timeout: 252 seconds [12:48] much better than a little corecolouring [12:48] But I cannot do that while I am splitting [12:48] <7000 - most clears are running frontwards [12:49] >100 is ok [12:50] > 200! [12:50] At the positions 0 <= x <= 100 the opponent can't be [12:51] First for >= 200 the probability of hiting the opponent is maximal [12:51] Well technically it'd put it at like 105 [12:51] Heh [12:51] (relative to the start of your warrior) [12:52] Ok, partly wrong, maximal for warriors with a length of 100, which is quite usual [12:53] Mine is length 99 [12:53] But I boot away from it... [12:58] Join: Nenad joined #corewars [12:58] Hi [12:58] Hello [12:59] I see that you're haveing some trouble with papers. [12:59] *having [12:59] MSG: Read error: Connection reset by peer [12:59] why don't you try a oneshot? [12:59] you can make a really simple one, if you like. [13:00] It's okay [13:00] ? [13:01] what's ok? [13:01] I like my warrior; I will just change his design a bit to handle papers better [13:01] (without the use of a oneshot) [13:01] oh... [13:01] :) [13:01] Yes [13:01] I am stubborn, and think oneshots/papers/scanners (sorta) are fairly boring ;) [13:02] :) [13:02] well, you'll probably change your mind later on. [13:02] Maybe [13:02] * Jens is away for a while [13:02] You guys could always not make papers [13:02] That'd work, too [13:03] did you try your warrior with a decoy? [13:05] Yes [13:05] It has a bootstrap and a decoy [13:09] Join: Roy joined #corewars [13:09] Hi all [13:09] * Roy got a new idea, not competative but cool [13:09] Hello [13:09] Hi Roy [13:11] so, do you want to tell us that idea? [13:11] Yeah, was working on something for a second [13:11] I'm thinking about making a self-programming warrior :-) [13:12] Or maybe a self-optimising warrior [13:12] :) [13:12] I've had some ideas of that sort. [13:12] Using p-space to store variables [13:12] but why do you say non-competative? [13:13] I believe that it can actually work [13:13] Well, I wanted to start out with nothing, I just thought about a self-tweaking warrior [13:13] Even more, I was convinced that someone already made stuff like that [13:13] thats could be done.. [13:13] Someone did? damn.. [13:13] well, Combatra is something similar, right [13:14] but with a little extra intelligence [13:14] No idea.. never analysed it [13:14] For ex. a smart-oneshot, the steps can make or brake the scores [13:14] I will, of course, explore this idea to its full potential when I start writing that chapter of my oneshot tutorial :) [13:15] I already thought about that. [13:15] I am going to have a whole chapter about adaptive oneshots [13:15] I hope... [13:15] someday, when I find the time to write it. [13:15] :) [13:15] Hehe, when can we expect the first part? [13:15] no first parts. [13:15] I'll publish it when it's done [13:16] on my web page [13:16] which will be online as soon as I finish the tutorial [13:16] :) [13:16] 16 chapters planned. [13:16] Roy, Did you email me? My inbox was empty when I woke up [13:16] 2 written (in 2 days) [13:17] data: I think I did.. [13:20] Roy: Jens helped us about that paper warrior part yesterday, but it couldn't enter the hill without a qscan ... and some additional weaponry. so the score is a little disappointing. We'll probably have to work on it in the next issue, as well - to improve it [13:22] Yup, good idea [13:22] If Miz sends us his 94b report today, we can finish this issue soon [13:22] Wow :-) yay! [13:22] I'll print it and put it in a frame above my bed [13:22] ? [13:23] * Roy is proud of the ascii drawing [13:23] I didn't get any hill reports yet. what happend? [13:23] oh. [13:23] :) [13:23] I thought your Maelstrom entered high. [13:23] No, Gargan2an sucks :( [13:24] And my maelstrom is bad too [13:27] Hi Roy! [13:27] Hi Jens: Good idea about the time machine, I like it :-) [13:27] yes, it would be very cool [13:28] It isn't ready yet, at first I need to weed out the bugs [13:37] Join: brx joined #corewars [13:37] hi brx. [13:39] hi [14:07] Haha I got 0.2 better this time [14:07] Spot #25 better watch it's ass [14:08] lol [14:08] Because in 18.2 more points I'm going to rock his ass [14:08] lol [14:08] Roy: What is Work!? [14:08] Well, its [14:08] working :-) [14:08] * datagram goes back to trying to add the imp spiral to his code [14:09] Take a guess Jens [14:09] s/i [14:09] ? [14:10] I'm still not good at guessing [14:10] No, wrong :-) [14:10] Its a pwi+pap [14:11]