[00:55] anyone here? [01:46] Join: Haubitze joined #corewars [01:46] Part: Haubitze left #corewars [01:58] Join: brad_ joined #corewars [02:00] MSG: Read error: Operation timed out [08:15] Join: pak21 joined #corewars [10:45] hey [10:45] pak21 [10:48] Hi. [10:50] hows it going? [10:51] i just put my firt warrior on the noobie hill [10:51] it got it's ass kicked [10:52] so i tweaked it abit and now it should be faster. [10:52] only two processes , but if tha dosnt work i have a new idea [10:54] I haven't really played for years. I just lurk around here these days :-) [10:54] why dont you create a new warrior? its really fun and theyre always changing from what i can see [10:56] No time... [10:56] oohhh [11:06] MSG: Read error: Operation timed out [11:08] Join: fiveop joined #corewars [13:32] Join: ENinja joined #corewars [13:35] hi [13:39] hi [13:39] how may i be of assistance? [13:47] MSG: Quit: Download Gaim: http://gaim.sourceforge.net/ [13:47] -_- [16:50] Join: John joined #corewars [16:51] Hi :-) [17:00] MSG: [17:32] Join: John joined #corewars [17:43] Anyone know why the colours work differently in Firefox to IE? http://numbers.atspace.org [17:48] Join: Fluffy joined #corewars [17:48] :) [17:49] Hi Jens [17:49] Haven't seen you for a while (apart from when I've been checking the logs) [17:51] :) [17:51] I've been busy and now I've finally found the time to write a warrior [17:52] What kind? [17:53] It is already on 94nop [17:53] Hey, don't push me off :-( [17:55] I've solved my CSS colours problem. [17:55] Firefox accepts "grey" but IE only accepts "gray" :-( [17:55] hehe ... and I've just started to go into it [17:56] John: You are always safe with the hex-values [17:56] Something's wrong with the hill [17:56] I should have remembered an older version of IE used to render "grey" as green! [17:57] hehe [17:57] Have you broke Koth? [17:57] I've sent a replacement for Pandaemonium [17:57] It was in the queue, but it isn't on the hill [17:57] and I've DATed the hill, too [17:57] Not sure why that happens sometimes [17:57] and I see nothing of it either [17:58] I thought I saw it mentioned in the newsgroup or the logs [17:58] Something about submitting too quickly with the same name [17:59] ok, I'll retry [17:59] I'd better submit before I'm pushed off! [17:59] hehe [18:02] John: Hurry up or the hill won't accept mails from you any longer [18:02] :-/ [18:03] I can't hurry if I'm submitting something new [18:03] ok, now the replacement has been accepted [18:04] Why do you want to hurt my scanner? :-( [18:04] I don't want to [18:05] I've been forced to write the warrior by the daemons ;-) [18:05] * John summons bvowk to carry out an exorcism [18:05] It is too late. The warrior is already on the hill [18:06] Is that the original OoS? [18:06] or a little bit beefed up? [18:06] Yes, but with a decoy [18:08] hmm [18:08] 8th ... at least better than my last test warrior [18:12] Join: Nenad joined #corewars [18:12] hi [18:12] :) [18:12] Hi Nenad! [18:12] Hi Nenad :-) [18:12] seems that there is some people here after all :) [18:13] John: Do you really think, that that name will help? [18:13] it always does [18:13] Yes, I'm here 3/4 times per week. [18:13] is that 3..4 times or 0.75 times? :) [18:13] Fluffy: not sure it's the first one I thought of [18:14] 3 to 4 times :-P [18:14] hehe [18:14] I'm getting back into Corewar now I've got more time [18:14] hmm, Pandaemonium doesn't like '94draft [18:14] JM: did you find the time to answer those interview questions? [18:14] I was working too many hours for a long time [18:15] John: ok, just keep on sending new scanners to '94nop [18:15] Nenad: don't you want to do a video interview or something? ;-) [18:15] :) [18:15] well, if you come over here, I'm sure we could find a camera somewhere [18:15] :) [18:16] Join: inversed joined #corewars [18:16] Hi inversed :) [18:16] hI [18:16] hi inversed [18:16] You could arrange an interview in a Paris hotel, all expenses paid by CoreExplorer! [18:16] Hi Inversed [18:16] !hi_JM [18:17] John: almost :) [18:17] I'm just starting to update my website if anyone has specific requests [18:17] Update HoF! [18:17] Good idea, inversed [18:17] well, you could work on the cw people list... [18:18] I'm going to start with the logs because a lot of stuff depends on info from them [18:18] but simply ignore our suggestions and update everything :) [18:19] yes and keep on sending scanners [18:19] I need to change the navigation somehow. [18:19] Stuff is too hard to find [18:19] write a search engine ;-) [18:19] What, it redcode? [18:20] :-) [18:20] Join: Core_old joined #corewars [18:20] it EQU in [18:20] John: of course [18:20] Hi Core_old [18:20] Hi old Core! [18:20] Join: codgeriff joined #corewars [18:20] hi Core_old [18:20] Wow, it is getting crowded :) [18:21] omg this is an invasion :) [18:21] Hi codgeriff :) [18:21] hi Paul [18:21] hi! congrats Fluffy! [18:21] Thanks [18:21] Hi Paul [18:21] i count 14 connected [18:21] But I didn't write the warrior [18:21] Join: Core_old_ joined #corewars [18:22] John: You should use a name that every daemon loves [18:23] Looks like I'm not going to get on the hill with this one :-( [18:23] depends on the hill :) [18:23] Any other requests for updates on corewar.co.uk? [18:23] yes, update everything! [18:24] especially everything that you don't think to update [18:24] err ... you know, what I mean [18:24] MSG: Ping timeout: 255 seconds [18:31] I was here for the IRC tournament a couple of weeks ago Nenad ;-) [18:31] there is a meeting here and nobody told? [18:31] ? [18:31] whee [18:31] been playing freespace 2 for a while [18:31] Hi Mizcu [18:32] [20:01] [14:13] to the log readers: come to a sunday irc minitournament at 20h GMT this sunday, if you can. :) [18:32] ? I was there [18:32] [20:01] That's now! [18:32] wait... [18:32] One hour early: [18:32] [18:51] seems noone came yet :( [18:32] :-P [18:32] Logs are in GMT ;-) [18:33] what? did I miscalculate the time again... happens to me all the time [18:33] :) [18:33] but we could have one... now :) [18:33] time to go [18:33] * Fluffy waves [18:33] Happens to my girlfriend too, but she's normally late [18:33] * John waves [18:33] MSG: Quit: fluffy.i < 1, # 42 [18:33] * Nenad waves [18:36] I've been trying out some new scanner ideas but nothing is working :-( [18:36] I like Dark Crystal, I'll try to improve it [18:37] Did I publish it? [18:37] :) [18:37] You've pasted it in IRC [18:38] Oh, right [18:38] I suppose I'd better send it to Koenigstuhl then [18:38] yes. you should :) [18:43] Found it http://www.koth.org/irc-logs/2006-02-19.txt [18:47] Just waiting to see if it gets on the hill this time round [18:50] I'll start a topic in rgc about the upcoming tourney, so if you want to take part, reply to it in rgc. [18:51] inversed: is izum = invention in russian? [18:51] No, izum means raisins :) [18:52] :) [18:52] well, it seems that some words have quite different meaning here then [18:52] how do you say invention? [18:53] izobretenie [18:54] hmm... [19:02] MSG: Ping timeout: 255 seconds [19:03] My warrior keeps disappearing without challenging the hill :-( [19:04] What's going on? I've sent izum test 164, koth reported that it entered the hill, but it is not on the hill now! [19:04] I'll try sending it with another name [19:05] inversed: I send a warrior with the secret time reversal opcode [19:05] Metcalf: you broke the hill again! :) [19:06] i think there are many pieces to each page displayed on koth and [19:06] they are not all updated at the same time. refreshing seems to bring in [19:06] new bits [19:06] maybe it is a shared server and something else keeps it busy. [19:07] or stuff gets cached and replayed [19:07] maybe it is so... but strange things are known to happen on 94nop... [19:07] I broke it once, you know :) [19:07] many of the parts are referenced indirectly, by a number and every time the hill [19:07] Yes, it suddenly turned into a multi-warrior 94nop hill one day [19:07] changes so do the numbers, but not all the cached bits get updated right away. [19:07] Then changed back to normal a few days later [19:08] and of course .... it is buggy. [19:08] This time it's fortunate for me, I don't want HalucScan to be pushed off [19:08] I'll be back later [19:08] MSG: [19:08] * John waves [19:08] * Nenad waves to inversed [19:09] hallucination is hanging on by its fingernails [19:09] Is it time to send the Halloween themed warriors yet? [19:09] dang! that is a perfect theme for Shamu ;-) [19:10] :) [19:10] maybe i'll publish it Halloween eve [19:10] you won't like it :-))) [19:10] :) [19:10] * Nenad has to go [19:10] We already have a couple of monsters and a ghoul on the hill :-) [19:10] * Nenad waves [19:10] * John waves [19:10] MSG: [19:11] and a D3vilstick and a Gabble'n [19:12] Hallucination Scanner was pushed off the hill, I saw it in 21st place myself. [19:12] 21 40/ 51/ 9 Hallucination Scanner inversed 129 201 [19:13] it's still there [19:13] doesn't want to stay dead :-( [19:17] i think that was a caching problem, warrior #21 was actually "another name!" [19:17] Yes, submitted after the warrior which killed Hallucination Scanner [19:18] And the email with the results show Hallucination Scanner still on the hill [19:18] I'll try another warrior to see what happens [19:20] It's dead again! [19:20] (but for how long?) [19:20] ok, now hs is off with a score of 128 and deliverance is on with 134 [19:21] and 4x2k thanks you for your support, (but Shamu slowly sinks...) [19:21] My oneshot doesn't work very well [19:21] Trying now [19:23] it's sitting in the q, ... sitting in the q,... sitting in the q... [19:24] * John is watching Hallucination Scanner out the corner of his eye [19:29] I've now got two challenge results showing Hallucination Scanner being pushed off the hill by two different warriors. [19:30] ok, so the email results suffer from the same problems as the web displays [19:30] I hope this hasn't happened often before, there might be a right mess in my hill logs [19:33] Well my oneshot is officially rubbish! [19:35] bomb dat #1, 8 [19:35] for step-4 [19:35] dat 0,0 [19:35] rof [19:35] loop sub inc, scan [19:35] scan sne f+i, f [19:35] djn.f loop, *scan [19:35] inc mov.i -step, #-step [19:35] clr spl #24, inc+5-scan [19:35] mov *inc, >scan [19:35] bptr djn.f -1, }clr [19:35] :-( [19:38] with only one mov, it is destined to be [19:39] I tried two mov and it scored 1 point less [19:40] too easily decoyed I guess [19:40] Yes :-( [19:42] i have come to suspect that a bomber-scanner will perform better as a one-shot [19:42] Hmmm... [19:42] if your opponent is paper you don't have to scan all that much to find him :-) [19:42] and if he's not, then a dat-bomb may do as well [19:42] certainly against other scanners [19:44] time to resend OOS, eh? [19:44] :-) [19:45] i can't remember what that is... what day is it anyway? [19:45] Tuesday [19:45] thanks, my anniversary is Wednesday [19:45] http://corewar.co.uk/infitiny/OOS.RED [19:48] Yes, bomber/scanner seems to work ;-) [19:50] congrats, it is the only program to score above 125 points against 4by2k [19:50] Thanks [19:51] but it gives up too many points against the monsters, please kill if off asap! [20:03] I'm placed 18/19/20th :-( [20:17] Join: brad joined #corewars [20:17] hey [20:17] Hi Brad [20:17] hi codegriff hi John [20:18] hi pak21 if your still here [20:18] well i put my first bot up against the beginner hill today [20:18] I notice you've made it onto the beginner hill, congrats [20:18] john, please note that there are 25 slots and im at the bottom as 23 ;p [20:19] Hmmm... [20:19] but now im gonna improve [20:19] i think that maybe self replicator would do better [20:19] Yes, replicators do well [20:20] the one that i did this morning is three carpet bombers, two dats and one spl, they bomb every four squares so they do not hit themselves or their friends [20:20] i think they my bomb just as bad [20:20] may* [20:21] and im going to make a new email adress to submit bots from(and a new account) since the one i have is so spammed. [20:22] does the bomber idea sound good enough to get me up past last place or should I scrap it and go with something else? [20:22] John? [20:23] Sounds fine, but it's better if you arrange it so it bombs almost every location [20:23] It doesn't matter if your warrior eventually kills itself, as long as it's had a good go at killing everything else first [20:23] right, im just not quite sure how to do that and still retain the safety of my three programs [20:23] oh ok. [20:24] also [20:24] Just make sure one of the three survives :-) [20:24] i could do five space intervals as well as four since the core is also divisible by five [20:25] the idea behind the three is that they will still bomb at the same speed as one(just spread out diferently) and if one dies the other two speed up and continue to bomb at optimal speed [20:25] :-) [20:26] but you probably knew that. [20:26] so brad is back [20:26] A simple 3 line bomber like dwarf can bomb 1/2 of the core before killing itself. [20:26] what did you think i would give up? think i was two dumb to write warriors? [20:27] i think i could almost hit all of it [20:27] nah, just that first time i am awake enough to say something to you [20:27] if they all bomb at diferent intervals while still avoiding themselves. [20:27] Mizcu oh ok. [20:28] Correct, with 2 you could bomb almost all locations, you just need to get the spacing right [20:29] so which is better? hitting slightly less and staying alive to go for another run or hitting slightly more but dying on first or second run? [20:31] Hit as much as possible and find some way to stop before you kill yourself. [20:32] E.g. a timed look - djn loop, #1000 [20:32] repeats a look 1000 times then drops through to some different code [20:32] look EQU loop [20:33] right [20:33] a loop would be best [20:33] especialy if you know the corezize beforhand [20:34] that way you could just bomb everyspace until you end up where you are then do it all over again [20:35] That would work well [20:35] Another trick is to overwrite a pointer [20:35] or better yet you could just change the value of the dat directly behind the program and then if that value is returned to zero then you know the bomb is right behind you and you start over again [20:35] For example if you are bombing with DAT 8 [20:36] and the pointer to where you are bombing is in PTR: DAT somewhere [20:36] oh [20:36] then you overwrite it and it restarts [20:36] thats genious [20:36] You could arrange for one of the bombs to overwrite the pointer and reset it. [20:37] hence just jumping over your code. [20:37] It could even be reset with a different value to the original, or you could change other values too ;-) [20:38] im gonna write that just a simple like six liner or so and see how it does [20:39] just gotta finish registering my new account [20:39] :-) [20:40] this one is just for Cw [20:40] Right, just going to see how my new monster does, then I'm off home! [20:41] where are you now? [20:41] don't you have a comp at home? [20:43] No internet at home [20:44] oh [20:44] what's the strategy of ur new monster? [20:44] bomber -> clear [20:45] i dont get it lol [20:45] im new and dumb [20:45] It bombs for a while until it's bombed every 5 locations [20:46] Then it switches to a clear, which sequentially bombs every location [20:46] so a quick sweep then a slow clean sweep? [20:46] Yes [20:47] cool [20:47] By the way have you discovered it is sometime better to bomb every 12th location instead of every 4th location [20:48] no. [20:48] why is that? [20:48] The bomber will wrap around when it gets to the other end of core so it will still bomb every 4th location [20:48] However it will hit bigger programs sooner [20:48] ohhhh [20:49] good idea [20:49] Who said ghoul won't be on the hill much longer? ;-) [20:49] 2nd place now [20:50] cool [20:50] did you write it? [20:50] No, sf did [20:50] i havnt been playing long so i dont know much of whats been going on [20:50] who is sf? [20:50] i havn't seen any sf in here [20:50] S.Fernandes, or FatalC [20:50] Normally here as sf_ghoul [20:50] also sf_ghoul [20:51] hmm [20:51] so hes a good coder? [20:51] Join: Fizmo joined #corewars [20:51] hi hi :-P [20:52] Hi Christian :-) [20:52] hi John :) [20:52] hi! [20:52] how is it going? [20:52] 94-nop is jumping [20:53] hi John :) [20:53] Nice to see you here [20:53] almost can't remember when I was the last time here :-/ [20:53] :-( [20:53] I guess you're busy with work too [20:54] yes [20:54] and maybe a new job [20:54] does mov move relative the the object being moved or the mover? [20:55] depends on the modifyer [20:55] right [20:55] so which modifier will make it move relative to the object being moved? [20:55] mov 123, somewhere ; moves relative to the current instruction [20:56] any kind of indirection { } * < @ > will move relative to the instruction containing the actual pointer [20:57] Part: codgeriff left #corewars [20:57] * John waves [20:57] I've got to go too [20:57] so will move 3, @3 move whatever is three spaces ahead three more spaces ahead or will it just put it where it is? [20:58] It depends on the value 3 spaces ahead [20:58] oh [20:58] ya [20:58] assuming thats three lol [20:58] just caught that [20:58] if that is a DAT 10 then it will move it 13 spaces ahead [20:58] right [20:58] i got ya [20:58] Anyone use stumbleupon here? [20:59] nope [20:59] dont even know what it is. [20:59] Like a channel flicker for the internet [21:00] weird [21:00] i dont get it. [21:01] another one of bvowk's evil warriors off the nano hill ;-) [21:01] 51 45.2 49.3 5.5 2439-2216-sandpile-nano-eve79 bvowk 141.1 219 [21:03] who is bvowk? [21:03] evil man, slaves dozens of computers to churn warriors for him [21:04] I'm working on this http://corewar.co.uk/whoswho.htm [21:04] he uses genetic algorithms? [21:04] Bvowk's wesite is here http://www.math.ualberta.ca/~bvowk/corewar.html [21:04] Yes, genetic algorithms. [21:04] whats the fun of that? [21:05] And they're performing very well. [21:05] can't he be like...banned? [21:05] Because the search space is huge and it's a challenge to find the best algorithm to search it! [21:05] looool [21:05] he keeps sal -server running, so banning might be a bad idea [21:05] I second that, let's ban bvowk! [21:05] :-/ [21:06] It's okay, his warriors aren't a serious threat! :-P [21:06] * John waves at bvowk if he's reading the logs [21:06] Okay, got to go [21:06] this is all very confusing. [21:06] * John waves [21:06] he hosts the server just to test his genetic algorithms? [21:06] Good luck brad [21:07] MSG: Quit: mov.i #1,1 [21:12] he has been a little quiet recently, better for us [21:12] Join: inversed joined #corewars [21:13] hi hi [21:13] hi Fizmo [21:13] hI_Fiz [21:14] hmm [21:15] should i try out my new bomber on the nano hill as well as newbie? [21:15] it would fit in... [21:15] nano has a different kind of warrior-philisophy [21:16] philosophy, or however it is written [21:17] phallusophy [21:17] :-P [21:19] how so? [21:19] hmm [21:19] my emails dont seem to be getting through [21:20] i hate that [21:20] how long does it usually take them to get into the que? [21:20] the smaller the core, the more powerful coreclears are [21:21] oh ya ur right. [21:21] on queue of which server? Sal rarely over 5 minutes, koth usually 5-15 minutes [21:21] nano is frenzy [21:23] sal [21:23] beginners [21:24] ok then whats wrong with this email?: [21:24] ;Redcode-94b [21:24] ;name Carpet_v_0.1 [21:24] ;author Bradman69 [21:24] mov 3, @3 [21:24] add #12, 2 [21:24] so it should work with no problems, unless the mail gets caught in spam-filter [21:24] jmp -2, <-4 [21:24] dat #0, #12 [21:24] end [21:24] @end [21:24] add ;assert 1 in front? [21:25] what is ;assert 1? [21:25] a mars makes a check if a program is meant for the current setting, the assert -line is used to mention what kind of setting the warrior is meant on [21:26] ok [21:26] never had to do that befor [21:26] ;assert 1 means that the check always comes clean [21:27] pmars gives a warning if there is no ;assert -line [21:27] right, i used org s for my other programs does that do the same thing? [21:27] nope [21:27] hmm [21:27] then there must be something else wrong [21:27] because ive sent two without needing that. [21:28] nothing wrong in the code, the mail must've got stuck somewhere [21:29] though sal can be slow to answer is something is wrong in the code [21:30] oh [21:30] yahoo parsed the caps out of my adress when i created it [21:31] it should not matter.. [21:31] i think it did. [21:31] well see [21:32] is it yahoo.ca if i live in canada or still yahoo.com? [21:33] i sent a test and it came in 10 secs, and the challenge after otehr 10 secs (suicide-warrior) [21:33] ok [21:33] ill try .com [21:33] cool im no longer last on the hill [21:34] someone put a bot below me [21:34] * Fizmo waves [21:34] just the test i sent [21:34] MSG: [21:34] now the one im sending will make it and even 25 and people will start getting pushed off, YAY [21:34] oh sure, take away from the moment. [21:36] darned, the silly test aged the whole hill [21:37] whats so bad about that? [21:38] if something reaches age of over 50, it gets automatically kicked down on the beginner hill [21:38] oh [21:38] this is to make sure that the hill stays somewhat easy to get on [21:38] thats no big deal is it? [21:39] is sal down? [21:39] not really, yeah, but on the other hand i want to keep a probe on the hill, and everything extra is a step towards making another probe [21:39] MSG: Ping timeout: 255 seconds [21:39] works fine with me [21:40] Join: inversed joined #corewars [21:40] hey inversed [21:41] heY [21:42] hey inversed [21:42] im trying to get my warrior to challenge a hill but it wont [21:44] the is only a redirect-address, i would give the direct address but i dont remember it exactly now [21:44] it dosnt really matter [21:44] sal keeps timing out on me [21:45] wtf is this? [21:45] yes, now with me too [21:45] and now normal again [21:46] so you are Canadian? [21:46] ya [21:46] yes [21:46] even stranges since the server is close to you [21:48] where is it? [21:48] us of a? [21:49] take another look at the sal [21:49] 's address [21:49] oh lol [21:49] silly me [21:49] wtf is wrong wit this email: [21:49] ;redcode-94b [21:49] ;name carpetV0.1 [21:49] ;author Bradman69 [21:49] org s [21:49] s: [21:49] mov 3, @3 [21:49] add #12, 2 [21:49] jmp -2, <-4 [21:49] dat #0, #12 [21:49] end [21:49] @end [21:50] just got a new message from sal [21:50] saying what? [21:50] I've just sent a stone and it worked fine [21:50] yes i know [21:51] and you beat over half the newbie hill lol [21:51] can you show me the exact email you used less code? [21:52] ;redcode-94b [21:52] ;name SWS-test [21:52] ;author inversed [21:52] ;strategy Stone test [21:52] ;strategy To see if the hill works [21:52] ;assert CORESIZE == 8000 [21:52] pmars says the code has nothing wrong (except missing assert) [21:53] ok ill use inversed assert then [21:53] (very, very obvious that one was..) [21:55] yup, the challenge is going on [21:55] what challenge? [21:55] your warrior is fighting the hill [21:55] oh god [21:55] i never sent it yet [21:56] thats one heck of a lag [21:56] so i bet that when it was dropping out yahoo kept the mail to send later and now the hills gonna get covered in carpets [21:56] maybe yahoo is having a bad day [21:56] just so u guys know its gonna suck [21:56] there is only 1 in queue now, though [21:56] ill prob have to improve on it [21:57] ya so maybe the last one i sent befor this one got through [21:57] or like u said its just massive lag [21:57] yahoo isnt known for being slow either [21:57] hmm not bad for vo 0.1 [21:57] *V0.1 [21:59] thats the first warrior ive put up that got anywhere but last [21:59] oh darned, just remembered that my stone-improvement article is still on pause [21:59] MSG: Remote host closed the connection [21:59] i would've continued on it now, but the latest version is not on my 'puter [22:02] (well, it is on my 'puter and that is the problem..) [22:04] Join: brad joined #corewars [22:04] sorry about that guys [22:04] my power ran out [22:05] and you had to put the cables on new set of batteries? [22:07] acht, lame attempt on a joke [22:08] (to joke would be the right form..) [22:10] funny story to tell [22:10] gotta fix keyboard first [22:12] `````````````````````````````````````````` [22:12] ok [22:12] sorry about that [22:12] wow i feel like an idiot [22:14] im on an apple laptop running debian, so when i accidentily hit the numblock key it locked the function key, this meant that some keys gave numbers and others nothing, me being a computer nerd and not noticing the odvious decided that the pins on the keyboard must be misaligned, so after opening 'er up and checking the pins i finally discover the real problem. [22:14] and have to close it all up for nothing [22:15] and you have to remember this is a laptop not a desktop so the keyboard is built in [22:15] mizcu [22:15] Mizcu [22:15] do tell [22:15] i did [22:15] look up [22:16] yes, but why mention my name afterwards [22:16] to get your attention [22:16] most irc clients flash when someones name is mentioned [22:17] just switched irc-client some time ago and dont know how to make it give a sound on a highlight [22:17] and i dont really even care [22:18] ok. [22:18] im sorry. [22:19] no need to be sorry [22:21] ok [22:22] im finding myself more and more meek these days, but thats ok because aparently ill inherit the earth :) ;p [22:22] and this channel has priority over others when there is activity, unless i am wasting time on Civ4 or something else fullscreen [22:24] ooh i love that game [22:25] corewars is addictive though [22:25] newbies are rare nowadays so i have to keep an eye on them [22:25] especially since i just put my first warrior up anywhere then last on the hill [22:25] and now im(hopefully) improving it. [22:27] as i mentioned just after you dropped, i have an article-on-make for improving stone-type warriors, but it is geared for those having a little more knowledge on CW [22:27] cool. [22:28] im trying to make it cover more of the core but it's a little tricky. [22:28] nvm [22:28] just a little typo [22:29] just that i am a lazy git that never remembers to improve the code during spare time [22:29] not, not the code, the article [22:29] i do remember to improve code every once in a while [22:30] drn [22:30] darn [22:30] it may not do to well but ill try it on the hill [22:30] or better [22:30] but whatever questions you get, feel free to ask anyone on this channel [22:30] nvm [22:30] didnt save it's former self DOPE [22:31] i do suggest you build up an offline-benchmark for warriors [22:32] lot easier to compare the effect of a small change [22:32] yes that's what im trying to do. [22:32] it was a mistake [22:33] it does seem to cover alot of core quite quickly though so. [22:34] one bomb in three lines, .33c [22:34] what is that? [22:34] optimal? [22:34] ill check mine. [22:34] we use "C" as a unit of speed-comparement [22:35] if you do one thing on a one line, that is 1C [22:35] two in four lines .5c etc [22:36] most competitive stones today are <0.5c (because a trick they use they take a while to reach ~full speed) [22:36] tornado is 0.6c [22:37] ohhh [22:37] most common scanning-speeds are .5c .66c and .8c [22:37] mine is 0.33c and it covers more core than just a quarter [22:38] tablescan is 1c, quickscan is 2c , but both only for a limited time [22:38] If we cound decrements, Spooky Wench stone is .66c, and stone from Frantic is .75c [22:38] *count [22:39] do decrements take up a cycle? [22:39] usually not [22:39] when do they? [22:39] they are almost always used if/when there is a chance for free addition [22:40] eg. if you have a free b-field in an instruction, you can make it something xxx, which gives a free decrement [22:41] ok let me put it this way, would jmp -3, <-10 take up an instruction? [22:41] it is a single instruction, with a "free" decrement added in it [22:41] a cycle that is [22:41] since jmp -instruction doesnt use its b-field [22:42] yes, a single cycle [22:42] ok [22:42] (thats my just-in-case imp gate eh, it should nail most spirals just-in-case [22:44] MSG: Quit: Lost terminal [22:46] Even though higher speed is better theory-wise, that doesnt mean that components that reach higher speed are better [22:46] Tornado was .6c in time where .33c and .5c were most common, and it never fared well by itself [22:47] When making Reepicheep, very good stone/paper, John and Grabun experimented different stones to use with the paper [22:48] ok i put my new stone on the hill [22:48] lets see how it does [22:49] they ended up with .5c stone; the .6c stone just didnt work well [22:49] now this is just experimental, it may totally suck worse the v 0.1 [22:49] But dont let my speech with .c 's mess you up, it is just an instruction or few difference in real code [22:50] no, i get it [22:50] darn the new carpet did worse then the first [22:51] GoingToSleepNow() [22:51] ok [22:51] good night [22:52] MSG: [22:56] id better go too [22:56] nooooooo [22:56] ok [22:57] till tomorrow [23:16] anyone still here? [23:48] MSG: Ping timeout: 255 seconds