[03:38] bvowk: ping - (assuming you read logs) I guess you know Steve Sutphen? [03:38] * asw waves to all [06:14] Join: fiveop joined #corewars [07:18] Join: brx joined #corewars [07:33] MSG: Quit: Lost terminal [08:20] Join: Mizcu joined #corewars [09:13] Join: Roy joined #corewars [10:21] Join: Core29 joined #corewars [11:14] Join: fiveop_ joined #corewars [11:17] MSG: Ping timeout: 252 seconds [11:39] Join: datagram joined #corewars [12:03] Part: Core29 left #corewars [12:04] Join: Fluffy joined #corewars [12:04] :) [12:05] hmm ... about 480 patches/hours ... "someone" seems to have access to piles of computers :-) [12:05] * hour [12:06] for simplicity let's call that "someone" bvowk ;-) [12:07] 480 patches per hour ... [12:08] It took me more than one hour to download all the generated data throught my little ISDN line [12:08] *through [12:09] I've already suggested to do a score surface for standard settings [12:10] It is roughly the equivalent of 250000 patches for tiny [12:10] But who would want to download 120 MiB PNG file to view that surface? [12:15] time to go ... I'm back later today ... maybe I'll have uploaded some score surfaces by then ... [12:15] * Fluffy waves [12:15] Part: Fluffy left #corewars [12:51] Fluff: Can't you just do a partial score surface for 94nop..? Locate the probable best score area(s) and patch them? Or you could run a whole benchmark in tiny :-) [14:01] MSG: Quit: Lost terminal [14:01] Join: fiveop joined #corewars [14:37] Join: bvowk joined #corewars [14:37] heya. [14:37] heya [14:37] hi [14:37] fluffy: are we almost done yet? should I submit another run? [14:38] asw: yes, I know him, but I don't have access to any of his gear [14:39] whatcha doing roy? [14:40] nothing, working.. [14:41] Join: Fluffy joined #corewars [14:41] bvowk: about 50 % done ... current info at http://corewars.jgutzeit.de/score_surfaces/dissc/index.en.html [14:41] how many blocks is that? [14:41] 8330 [14:42] during the last 24 hours: 5606 [14:42] hehe sweet :-) [14:42] yeah, I'll need to submit at least another little bit [14:42] no more patches until after you do some analysis right? [14:42] yes [14:42] But the partial images, that I've downloaded look interesting [14:43] thats good [14:43] hey.. update the stats :) [14:43] ok ... give me 15 mins to finish downloading [14:44] I thought you were a student roy.. [14:44] hehe, 5k+ from bvowk today I guess.. [14:44] bvowk > Doing my final internship now, next month I'll graduate (hopefully) [14:44] Then its time to face the real bad world.. [14:44] nice.. what will you have when you're doine? [14:44] er.. s/doine/done [14:45] No idea.. :-) [14:45] I meant what kind of a degree/diploma/ticket will you have? [14:45] Can't translate it to english I'm afriad.. [14:45] * Roy googles it [14:47] what is the topic of your education? [14:48] bvowk: You are producing patches way faster than I can download them. Once I've finished downloding there are already hundreds of new patches [14:48] :) [14:48] jens: you need better compression! [14:48] :P [14:49] fiveop: I could have told him to get a better isp [14:49] bvowk: No, I don't need better compression. I need a faster connection [14:49] he should get an adsl connection [14:49] I already have a faster connection, but not here [14:49] fiveop: I can't get that damn stupid DSL here [14:49] * Roy is going to be a BICT...Bachelor of ICT :-S [14:50] fiveop: And I'm living in BERLIN! [14:50] you live in berlin don't you? [14:50] what is ict? [14:50] .... [14:50] Just IT, Information Technology [14:50] Information and Computertechnologie or something? :P [14:50] yeah :) [14:50] ok :) [14:50] Information and Communication Technology to be precise [14:51] * Fluffy is still downloading patches ... [14:52] *sigh* [14:52] I wish I got to make computing hardware instead of them making crud for the masses.. [14:52] At least *I* can already have a look at the new data :) [14:53] I'd have a massively parallel nanotube on diamond wafer stacks.. [14:53] you could run that at massive temperatures, so you could just stack thousands of layers together and then wrap the outside in heatsink [14:53] Hehe, bvowk, its good to have dreams :) [14:54] bvowk: I want to rule the world :-) [14:54] * Roy wants to get rich (fast) [14:54] Waiting for somebody to say: "I just want to be happy.." [14:54] happy? [14:54] I want to get rich, building massive computers, which I'll use to take over the world! [14:54] You buy Happiness with enough money ;-) [14:55] bvowk: Too late, I'm already taking over the world [14:55] bvowk: Look for your own world. There should be plenty out there ;-) [14:55] eBay sells happiness for under a dollar! [14:55] Buy and play Spore (build a world!) man.. I just can't wait for that game! [14:56] roy: you should buy it in bulk, less waste and packaging :) [14:56] Roy: nice show the hype-movie was, yes. [14:56] anyone ever play the game stars!? [14:56] er.. stars! [14:57] I've played Tetris once [14:57] bvowk: tried quickly once, i guess it loses a lot against computer [14:57] miz: I really like that game :) [14:57] Very nice show, watched the movie again yesterday, all 35 minutes, its amazing he can still surprise you after 32minutes without showing anything more then once [14:57] as in space empires and vga planets [14:58] I can't lose to the computer anymore.. [14:58] Not even in Chess? you should consider a carrier change then [14:58] Roy: however, it might be quite heavy game, even the ultra-pimped presentation 'puter had slowdowns, and not to mention the save-game size [14:59] and they certainly hadnt finished all the graphics [15:00] Mizcu: The savegame size will be very small, those creaters are created on the fly using a very small fingerprint, thats the big power of the game and allows it to send lots of creatures over the internet....it will be heavy thats very true :( [15:00] i did however have a laugh with the multi-beaked bighead creature walking less than stable [15:01] Tweety :) very nice, and that Deadstar beam was also a good laugh [15:01] roy: crafty crushes me at chess.. [15:01] its really quite tragic [15:01] Mizcu: which game are you talking about? [15:01] EA's Spore [15:01] url [15:01] spore.ea.com? [15:02] http://video.google.com/videoplay?docid=8372603330420559198&q=spore [15:02] This is the movie, much better then the website of the game, release date is Q4 2006.. :( [15:04] "there goes my reputation as non-violent game developer" X) [15:04] hehehe [15:05] Is this the game developed under the supervision of the creater of 'the sims'? [15:06] http://corewars.jgutzeit.de/score_surfaces/dissc/statistics.en.html [15:06] Yeah, the The Sims/Simcity dude [15:08] bvowk: If you add one or two more piles of computers you might hit the 10000-patch-mark later today :) [15:09] lol [15:09] unlikely. [15:10] How many clients are you running at the moment, bvowk? [15:10] they've made a resource called a barkley, only 100 barkleys are in existance.. [15:10] so really.. only 100 clients can run [15:11] why only 100 barkleys and not 1000 :) [15:11] apparently the fact I could fill the whole thing so easily got some of the users a little worried :) [15:11] Maybe you should define, that such users don't exist and then "fill the whole thing" [15:12] It is quite common to me to define such things ... makes life easier ;-) [15:12] Join: Core29 joined #corewars [15:12] heh.. [15:12] they're not my computrons tho :) [15:13] bvowk: DEFINE them to be yours [15:13] It is easier! Trust me! [15:13] I'm sure it is :) [15:17] Anybody want a surprise? [15:17] Sure I'll have a dozen, with giftwrap! [15:17] I have only one. Do you want it or not? [15:18] Well... [15:18] * Roy has to think about that [15:19] Is it going to ruin my holliday? :) [15:19] I do! [15:19] No, it is a nice surprise [15:19] Then...I do! [15:20] Anybody else want it? [15:20] * Roy is going to France tomorrow...! [15:21] Somebody else is driving, but yesterday he suggested I should drive too, taking turns driving [15:21] he->a friend [15:21] Surpise can be found at http://corewars.jgutzeit.de/temp/surface.png [15:22] I think, that I'll call it the Scottish Score Surface ;-) [15:23] Weee! thats a cool surprise [15:23] BYAP is a modified YAP [15:23] boot spl 1 [15:23] mov.i -1, 0 [15:23] mov.i -1, 0 [15:23] silk1 spl @ silk1, < pStep1 [15:23] mov.i } silk1, > silk1 [15:23] mov.i > bStep1, { bStep2 [15:23] mov.i { silk1, < silk2 [15:23] silk2 djn.f @ silk2, < pStep2 [15:23] It even has a cross in the middle [15:23] where pStep1 and pStep2 are fixed to (440, 336) [15:23] hey.. thats interesting. [15:23] looks like a good tartan! [15:24] and bStep1 and bStep2 are tested, i.e. set to all possible values [15:24] The clan MacByap [15:24] Another surprise is the lack of diagonal lines going from up-right to bottom-left...? [15:24] I wanted to see the effect of the "bombing line" "mov.i > bstep1, { bstep2" on the paper [15:27] Seems to me, that there is an "easy" way to predict good values for bStep1 and bStep2 if you know the values for pStep1 and pStep2 :-) [15:29] yes, but the BEST still requires a lot of tinkering.. [15:30] But its cool to have a generalized formula for bombsteps in that kind of paper :) [15:30] Yes, maybe, but if I can reduced the number of values to check by an large amount, I could make optimax (or sth. similar) run faster [15:30] *reduce [15:31] * asw waves [15:31] * Fluffy waves [15:31] True, and good idea [15:31] * Roy waves [15:31] hey asw [15:32] Any other comments on that Scottish Score Surface? [15:32] No, I just want to print it and turn in into a kilt.. [15:33] CW-Kilt, the perfect birthday gift [15:34] If I remember correctly I can include a gift-shop into my homepage for no extra money [15:34] with t-shirts and cups [15:35] hehe, www.coffeepress.com ..? [15:35] hrm, not the right url [15:36] I meant http://www.cafepress.com [15:38] Hmm ... it seems to be equisto [15:38] available "items" can be found at http://www.equisto.de/perl/product_source_list.pl?SESID=4055327jzxcv944185052 [15:38] Everything can have a score surface on it :) [15:38] hmm ... I wonder, whether there is enough space on that "Girly Tanga" [15:39] oooh! [15:39] * Roy wants the Classic shorts with "Replicator" on it! [15:40] lol [15:42] when I was in highschool I had a pair that said something like "black goat that spawned 10000 young" [15:42] or something to that effect, had a little cthulu on the leg [15:42] :) [15:42] I still wear my black T-Shirt with "MILK" on it [15:42] We had a little hype here with girls having their thongs very high, and they printed they mobile numbers on it :) [15:43] hmm [15:43] I hope, that those girls have been intelligent and used numbers, where you have to pay 10 Euro/min [15:44] lol [15:44] I ordered a black tshirt that says "white tshirt" on it.. [15:44] Nice idea [15:44] so now when my wife complains about me only wearing black tshirts, I can say "but I wore my white tshirt yesterday!" [15:45] hmm ... white t-shirt with score surface on front and back only 21 Euro [15:45] probably + shipping [15:46] To expensive... :( [15:47] Roy: You've said, that it might be a good idea to calculate parts of a big score surface (for standard stettings). How am I supposed to know, where interesting parts are? [15:47] Yesterday I saw somebody with a blackshirt and just the word: "Butter" [15:48] That was probably the next version of my "Milk"-shirt ;-) [15:48] Studying other surfaces (smaller) or stepping with gaps [15:48] Join: Fizmo joined #corewars [15:48] hiii [15:48] hello fizmo [15:48] Maybe I should ask our anonymous donor with that many computrons about calculating one big score surface :) [15:48] Hi Fizmo [15:49] holy crap. [15:49] I thought fizmo was dead or something! [15:49] Fizmo: Now we are in the Scottish era for score surfaces [15:51] http://www.funny-tshirts.com/prodimages/large__22_08_2005_18_11_parti_large.jpg [15:51] Does "parti" mean anything special? [15:51] party? [15:52] No idea, I just think party.. [15:52] Actually, I want to have a normal Atari shirt, those are cool [15:53] fluffy: you need that made into a hawaiin style shirt. [15:53] that'd be great :) [15:54] bvowk: The scoure surface? [15:54] yeah [15:55] I'd wear that home cause my dad hates my hawaiian shirts.. that'd be perfect. [15:55] But shipping to Canada might be expensive [15:55] bah. [15:55] well worth it. [15:55] Maybe I should really open a Core War Shop [15:56] * Roy owns a WARNING: Grumpy! shirt [15:57] hahaha, sorry. Bit silent [15:57] will talk later [15:58] Hehe "Hard work has a future payoff Laziness pays off now." [15:59] hmm ... yet another 200 new patches ... [16:02] Yessss! [16:02] 17:00, work done, time to ski!!! [16:02] :) [16:02] * Roy waves [16:02] MSG: [16:03] Hey, did I mention, that we can bvowk now officially call the first "Score Surface Imp" ;-) [16:03] * Fluffy waves too late [16:18] Strange ... almost all new score surfaces still have some unfinished patches [16:37] Part: Core29 left #corewars [16:54] * Fluffy is back in a couple of hours ... [16:54] Part: Fluffy left #corewars [17:04] hi bvowk [17:04] bvowk: It'S the return of the living dead ;-) [17:06] ok, gotta go now :-/ [17:06] MSG: [18:30] hrm. [18:30] fluffy: its not strange, remember, the jobs are uncerimoniously terminated periodically, so some work units are dropped on the floor and have timeout for recycling. [19:00] Join: brx joined #corewars [19:06] Join: sascha joined #corewars [19:06] MSG: Remote host closed the connection [19:28] Join: me joined #corewars [19:29] Nick Change: me changed nick to Fluffy [19:29] :) [19:33] try stumpwm fluff :) [19:35] Dear brx, if you could please point out at least ONE reason, why I should do that, I'd immediately switch over to that stupidwm of yours. [19:36] Fluffy: it's good? [19:36] Dear brx, that's a question. [19:36] Fluffy: it's good. [19:36] nope, it isn't. [19:37] that's highly speculative on your part. [19:37] I'm still waiting for a reason ... [19:37] but you are right. I wanted to waste your time and bathe in your frustration :) [19:38] I'm not frustrated. Today I will have more than 10000 completed patches :-) [19:39] If I'd want to see a frustrated person, I'd point at you and tell you one link. [19:39] bvowk has unloaded his full arsenal [19:39] fiveop: No, only 100 barkleys ;-) [19:39] Fluffy: what link? :) [19:40] brx: http://corewars.jgutzeit.de/score_surfaces/dissc/statistics.en.html [19:40] but you are mistaken fluff, that does not frustrate me, instead I find it highly amusing! [19:41] "Screw you guys, I'm going home!" ;-) [19:43] Did I mention, that PyCorewar now has an official homepage? [19:43] http://corewars.jgutzeit.de/pycorewar/index.en.html [19:44] 9892 patches done ... [19:45] bvowk: I could use two or three more piles of computers running DiSSC! [19:48] brx: Do you one or two piles of computers left? [19:48] ;-) [19:48] Fluffy: I currently have 3 computers at my disposal. [19:48] No, that's not even one pile [19:48] 4 if I do something nasty [19:49] I suppose I was aware of that [19:49] your standards have risen, you need no brx cycles anymore :( [19:49] brx: You don't need to do sth. nasty. Remember: it is only a question of definition! [19:49] brx: Of course I need brx-cycles [19:50] I still don't understand, why you've stopped submitting patches [19:50] You could have easily been the first one with 1000 [19:50] I think I mentioned needing the cycles for something else [19:50] I never thought of it as a contest anyway [19:51] Are you really suggesting, that there are more important things than score surfaces? [19:51] Fluffy: like running the glmatrix screensaver on my root window? :) [19:52] hmm ... not I don't like that [19:53] but http://electricsheep.org is nice [19:53] ah yes, I know that [19:55] brx: Would you rather like a score surface for standard settings or a score surface movie? [20:06] greets fluffinator. [20:08] bvowk: Do you have time? [20:08] what sort of time/ [20:09] to talk about an idea [20:09] sure [20:09] as long as I don't have to do it.. I've already got myself oversubscribed.. [20:09] Sascha has just reminded me of was John did for the first score surface [20:10] You don't have to do anything [20:10] only one tiny thing [20:10] Anyway ... to get a quick overview John calculated only every 5th point or of the score surface to get an overview [20:11] I think, that He-Who-Must-Not-Be-Mentioned did the same later [20:11] I really like to do one score surface for standard settings. [20:12] Would you mind, if I would write a little script, which would do that and then run it on you computers? [20:13] It would be about the equivalent of 10000 patches for tiny? [20:13] why do you think? Is that possible? [20:14] it would certainly be possible.. [20:14] "possible" as in "I'd run that script, if I would have it"? [20:14] judging from the tartan you posted this morning, I expect you could run a only a few thousand patches and extrapolate the rest. [20:14] the pattern is pretty regular. [20:14] by the way ... we have just passed the 10000-patch-mark :-) [20:15] bvowk: The score surface, that I have in mind would be less regular [20:15] more like the first score surface, but still a little bit different [20:15] or if you wanted to get interesting, you could run some sort of montecarlo, estimate the areas of "solidish" colour and only calculate the interesting bits. [20:16] (being the edges, troughs and peaks) [20:19] hmm ... montecarlo ... hmm [20:21] Do you have any requirements for the script? [20:22] not really.. [20:22] but if you do it in python.. why not use fmars [20:22] yes ... I could use my hacked version [20:22] its stupidly fast, and if you're only running two warriors there's no reason you can't make a custom mars.. [20:22] This version supports -P :-) [20:22] heh.. [20:22] I have my own -P fmars version for python too :) [20:23] I expect you could get almost 3x more battles done in the same period. [20:23] My version calculate several pairs at onces with parsing both warriors every time [20:23] *with->without [20:23] yeah, you won't need to parse the warriors more than once. [20:24] The script, that I've used then only parsed the output, compressed it and sent it to the server [20:24] BUT: [20:25] Can you really compile fmars on those computers without problems? [20:25] yeah [20:25] alrady tried? [20:25] fmars runs on pretty much anything now.. [20:25] yeah.. [20:25] Do you use an unpublished version? I mean, I've hacked the 0.0.207-version [20:26] thats the one I'm using.. but my hacks are unpublished [20:26] as are mine :) [20:26] I believe I sent them to michal tho [20:26] I should do that ... I've only deleted all the parts, that I don't need [20:27] and adjusted it to do -P [20:28] hmm ... the current patches will probably finished tomorrow, right? [20:28] *be [20:30] And how long can my script run without being killed? [20:30] I want to minimize the data loss AND the load on my server [20:30] Or can I store the generated data locally? [20:31] sorry.. I'm back. [20:32] well.. if you gave me a copy of the server I could run locally.. you wouldn't have to worry [20:32] as a) you'd be connected by at least 100Mbit ethernet, and b) I've got way more server computrons than you do. [20:32] jobs, I can only schedule them for an hour.. [20:32] longer than that and they get turfed. [20:35] server? [20:35] It is a bunch of PHP-scripts with MySQL-db [20:36] I've randomly arranged code lines until they worked ;-) [20:36] why don't you make it python and a stand alone process so I can run it locally [20:36] hmm .. never done that, but shouldn't be that hard (I hope) [20:36] its not hard at all.. [20:37] if you want a suggestion for a starting point, take a peek at my huge hill scoring widget [20:37] http://www.math.ualberta.ca/~bvowk/files/hill-srv.tar.gz [20:38] will do :) [20:39] that includes my -P fmars too [20:39] OK, I write the server and client and mail everything to you [20:39] ... one they are done [20:39] *once [20:39] ok [20:39] Yippie [20:51] MSG: Quit: leaving [21:00] http://corewars.jgutzeit.de/score_surfaces/byap/index.en.html [21:00] Finally I have all patches for that surface [21:02] At least I could calculate ONE patch for that surface [21:02] Someone else did all the rest :-) [21:12] lol [21:14] are we done yet? [21:17] no [21:17] about 5000 patches still to do [21:18] You would have been done, if you had listened to me and added two or three more piles of computers ;-) [21:18] brb [21:18] MSG: Quit: Lost terminal [21:21] Join: me joined #corewars [21:21] Nick Change: me changed nick to Fluffy [21:21] hrm.. 10 hours then.. [21:21] thats not quite as quick as I had expected. [21:21] Join: brx joined #corewars [21:22] what did you expect, bvowk? [21:22] I've just uploaded another score surface (BYAP vs. Muskrat) [21:22] I had been expecting to be done by about now according to this envelope. [21:22] must be the difference in the round times.. [21:23] my estimated average time per block looks low [21:23] It is hard to estimate such things as soon as you go beyond one score surface [21:24] thats right.. [21:24] different score surface now.. thats why [21:24] I guess I better schedule some more [21:24] some more clients? [21:25] heh [21:25] no, some more jobs [21:25] there's only 4 hours scheduled in the queue [21:29] I think I write a short summary about those scottish surfaces and then get some sleep [21:37] ... done [21:39] hmm ... found some mistaken in my new pages [21:39] It'll have to wait until tomorrow. [21:40] Sascha/Fizmo/inversed/brx: Read rgc :-) [21:40] * Fluffy waves [21:40] MSG: Quit: leaving [22:23] MSG: [22:41] MSG: Quit: humhum [22:48] MSG: Quit: using sirc version 2.211+ssfe