[00:06] hehe [00:06] im THIS close to making a replicator [00:06] uhoh [00:07] suppose, i should send another to the hill soon [00:12] is there a modifier which allows me to move without keeping the a and b feilds? [00:13] heheh [00:13] no [00:13] that sucks [00:14] overwrite the fields with the wanted values before? [00:14] ya i guess [00:14] but you cant change the adress modes [00:14] does an increment happen befor or after the operation? [00:15] pre-decrement, post-increment [00:15] huh? [00:15] decrements are allways done before the value is used, increments allways after [00:15] ok [00:15] sweet [00:18] can i use an increment and an at at the same time? [00:18] ie [00:18] it is the same, except for the increment [00:18] mov >@-2, 19 [00:18] mov >-1, 19 [00:18] -2 of course [00:19] is what i just did allowed? [00:19] yes [00:19] so i can use more then one at once? [00:19] everything is allowed except executing dat and division by zero [00:19] @ is indirect, > is indirect with post-increment [00:20] it gives an error when i do that [00:20] mov >@-2, 19 [00:20] this [00:20] causes an error [00:20] you can only use one adress mode at a time [00:20] kk [00:20] ty [00:20] but the meaning of @ is the same as >, except for > increments the pointer after execution [00:25] ya [00:25] ok [00:25] is there a way to get ares to run only one warrior and start it in pause mode? [00:25] ctrl-F9 [00:26] or start it from the editor [00:37] why does this fail horribly as a replicator? [00:37] spl 1, -2 [00:37] spl 20, 0 [00:37] mov >-2, 19 [00:37] jmp -1, >-1 [00:40] the "-2" in the first spl points to nirvana [00:41] the first call of mov >-2 uses the spl ,-2 and so the instruction 4 cells above the move is taken as source [00:44] really? [00:45] so what would it look like if it was written correctly? [00:45] gimme some minutes, phone [00:53] re [00:54] okay, lets have a look... you want a simple paper, right? [00:54] those are allways twisting my mind [00:57] what about [00:57] spl 1 [00:57] mov <1, {1 [00:57] jmp DISTANCE, 1 [00:58] MSG: Ping timeout: 255 seconds [01:27] ? [01:27] ! [01:44] schau mal rein, wie gefällt dir die neue startseite? [01:44] http://harald.ist.org/?site=ares [01:44] oopd [01:45] wrong win [07:40] Join: impomatic joined #corewars [07:40] Good morning :-) [07:40] morning [07:43] Hi Harald [07:47] what are you up to so early? [07:49] Just checking my email for the last time before Christmas :-( [07:50] Actually I've stayed in bed an extra 2 hours, I'm normally up at 4:30 [08:02] I've just opened the last day on my advent calendar :-) [08:02] :-) [08:02] you chose the right day for that :-) [08:02] (it's one with chocolates) [08:02] ooh [08:05] i am about to write an multiple choice core war test. Maybe you have some ideas for funny questions? (but the test is meant to be a real test) [08:05] Like an exam? [08:05] aye [08:08] With multiple choice answers? [08:11] Or you could have questions like: Write a program in Redcode to calculate the nth triangle number [08:12] could be difficult to be checked by the script [08:13] If the script could run redcode, it'd only have to check the results ;-) [08:13] there's a Java MARS [08:19] How about a small Redcode program, and the person taking the exam has to work out what it does [08:19] That could be multiple choice [08:20] yes, sort of "how many processes will this have after 4 cycles" and so on [08:20] Or even better a small Redcode progam with a bug they have to find :-) [08:20] and "which line contains a semantic error" [08:20] *g* [08:21] these kind of questions are easy to create... i'd like to have some serious questions, that are funny at the same time [08:22] how about an "obfuscated redcode" contest in rgc? *g* [08:27] I think it'd be better to have a contest for clearly written redcode :-) [08:28] heheh [08:28] your'e right [08:36] Join: Fluffy joined #corewars [08:36] :) [08:36] * Fluffy is quickvisiting .. [08:36] morgen [08:36] ares: a couple of hours ago maybe ;-) [08:36] Guten Morgen Fluffy [08:36] abend [08:37] ares: Remeber the nuke. I can do it again ;-) [08:38] * ares hides himself behind impomatic [08:38] :-( [08:38] * Fluffy starts to polish his small tactical nuclear weapon with a yield of about 2 kt [08:38] * impomatic kicks fluffy and attempts to defend ares [08:39] Hey, polish your shoes before kicking! [08:39] I'm wearing new slippers :-) [08:40] aye, folks [08:41] Hi Mizcu [08:41] wow - the first time, i see something else than idling redcoders here in the morning [08:42] ares: OK, then I can start to charge you for that quality time with me in #corewars. 50 euro/minute should be fair enough. [08:43] i'll send you a scan of 50eur, when I have some [08:45] one PER minute, please [08:45] i'll write a script [08:45] that's good enough for me [08:47] so Ares and Fluffy will be spending some qualitytime together [08:48] * Fluffy plants his other small tactical nuclear weapon near Helsinki [08:49] * Fluffy sets its timer to 10 seconds [08:49] ok, problem solves [08:49] *solved [08:49] * ares places a SPL 0 on that nuke [08:50] Do whatever you want. Nobody will ever want to live near Helsinki now [08:50] for mizcu's sake, we have absolute adressing in the real world [08:52] I hope that the mushroom cloud looks a little bit like a christmas tree [08:52] put some glitter in the bomb [08:53] too late [08:54] its still SPL'ing 0, you have all time of the core [08:54] what's MAXCYCLES set to? [08:54] Zap it with a WRP 0 next! [08:55] thats a question for the astrophysics [08:56] ares: You should implement a "verbose" mode for ARES. [08:56] sth. like: cell 42 lauches an imp at position 100 [08:56] cell 107 retaliates by dropping a dat-bomb [08:56] ... [08:57] *g* nice idea [08:58] "Player B has no clue" [08:58] hmm ... and you should give each cell a name [08:58] like Peter, Sally, ... [08:58] yup [09:00] i will also include a coffe machine remote in ares [09:00] and a c++ to redcode compiler [09:00] eeek! [09:00] called "redscript" [09:02] Actually it should be to hard to compile to redcode [09:02] Though I'd suggest a simpler language like BCPL [09:03] *g* [09:04] impomatic: Do you really suggest us to use a languages, which uses keywords like "LET" or "GET"? [09:04] don't you do? #DEFINE LET /**/ [09:05] How about Brianfuck? [09:05] *Brainfuck [09:05] i saw a c program, that actually looked like pascal using #define [09:05] Sorry, Brian, whereevery you are ;-) [09:05] bf would be difficult, considering you can operate on instructions and modify them to new instructions [09:06] but a bf interpretter in redcode would be easy [09:06] I didn't say that it wouldn't be difficult, but it is ares, who will do all the work. So why should I care? ;-) [09:07] ares already wrote a Basic interpretter in redcode [09:08] hmm.... a c compiler written in redcode... [09:08] ares: You start to sound like John. [09:08] ? [09:08] John Metcal [09:08] f [09:08] this part was clear [09:09] He usually has strange ideas about what to implement in Redcode, too. [09:09] things like tetris? [09:09] Tetris would be tricky [09:10] not with A.R.E.S. [09:11] hmm... new opcmodifiers for 3d-grafics, like .mmx ? [09:12] java vm in redcode? [09:13] After Christmas I'll try to write a bf interpretter in Redcode [09:17] hmmm [09:17] How about an interpretter for p-code? http://en.wikipedia.org/wiki/P-Code_machine [09:18] what leads to a mars written in redcode [09:19] that would be the ultimative, i suppose [09:20] * Fluffy waves [09:21] MSG: Quit: fluffy.i < 1, # 42 [09:21] * impomatic waves [09:21] * ares waves back [09:21] a fluffy written in redcode would be funny, dont you think? [09:23] or not [09:23] We could certainly write a simulator [09:27] * impomatic now has to wrap presents :-( [09:27] * impomatic waves [09:27] merry xmas [09:28] Happy Christmas Ares [09:28] And Happy Christmas to all the other Corewar folk :-) [09:28] MSG: Quit: CGI:IRC 0.5.7 (2005/06/19) [13:22] Merry Christmas. [17:21] Part: fff left #corewars [18:39] merry [18:45] xmas [18:59] Join: brad joined #corewars [18:59] hi brad [19:00] hey [19:10] where is a good paper maing guide? [19:10] mom [19:11] www.corewar.info -> lexicon [19:11] http://www.corewar.info/lexicon/paper.htm [19:17] jeez [19:17] aye [19:17] i just cant get my head around how a silk works [19:18] thats the point of this tutorial: making the reader confused [19:18] can you give me one that has the phrase "for complete and total idiots" in it? [19:18] i have not written it yet *g* [19:18] keep trying, its not that hard after all [19:20] i just dont get it [19:20] i know that you have to branch a process which is just about to execute a dat when another process sticks a spl in there right? [19:22] watch it in the debugger and keep the task queue in your eye. Its possible, because new process entries are added at the bottom of the queue. read the icws 94 draft carefully, then you should have the necessary knowledge [19:22] i know how it works [19:22] then forget the last sentence [19:22] i just cant seem to put all the peices together [19:23] keep going. tomorrow you will kick me of the beginners hill. [19:23] i have a carpet bomber on the beginners hill [19:23] i think [19:23] whats the link? [19:23] using ares? look in the bookmarks, SAL [19:27] nope [19:27] it was kicked off [19:29] well, feel free to ask me, or to give upgrade-hints about your warriors [19:29] also use benchmark for checking out, if your warrior is worth uploading. Note, that the resulting score is only a relation to the warriors of the benchmark, and the score on the hill will be different [19:30] Mizcu: i am still not giving up on writing a selfrepair "immortal" [19:30] wilkies enough to start with, even though it will regardless curb your enthusianism with your test-warriors [19:31] my best score with wilkies is currently about 82 pts [19:31] ares: good luck with the selfrepairer [19:32] i think mine is.. 180 or like [19:32] jaja [19:32] with record being something like 210 [19:32] argl :-) [19:32] (pspacer) [19:35] you can meet all sorts of weird scores over time when testing [19:35] ever had a negative score? :-)=) [19:36] nope, i didnt make the benchmark-script myself :) [19:36] heheh [19:40] hmm, maybe i should put together some sort of beginner-benchmark [19:41] since wilkies is somewhat unbalanced [19:41] i would appreciate that very much [19:45] hmm [19:45] * brad is pulling his hair out over papers [19:45] old-style paper or silk-paper? [19:46] ehm... how does a typical oldstyle-nonsikl-paper look like? [19:47] yes? [19:47] how? [19:47] top: mov.ab #lenght, #0 | mov <-1, no, jmn.b -1, @-2 [19:48] hmm, thats a crappy one, bad quickcode, but i hope you get the idea [19:49] i wonder about the #switch... [19:49] should be also put to the beginning mov? [19:50] http://www.ociw.edu/~birk/COREWAR/LP/HILL/gelablob.red [19:51] oldstyle-papers have use on the LP-hill where you just cant spread aggressively [19:53] what is this whole thing about top and bottom? [19:54] the copy routine needs to know, how much instructions to relocate [19:54] hmm [19:54] there might be some more code as "payload" within the self replicator [19:54] im trying to build this warrior right now that copies itself ahead and then starts dat bombing [19:54] but it wont work [19:55] spl 2, 0 [19:55] spl 100, 0 [19:55] mov.i >-2, 99 [19:55] jmp -1, >-1 [19:55] only one copy? [19:55] ya [19:55] it copies itself ahead then starts dat bombing then the one ahead copies itself ahead then starts dat bombing [19:56] or at least its supposed too [19:56] spl 100, 0 [19:56] mov >-1, }-1 [19:56] jmn.a -1, @-2 [19:56] goddamnit, screwed again [19:57] well, not. jmn.a will work [19:57] what did u wanna use? [19:58] jmn.f [19:58] hmm [19:58] taht dosnt work [19:59] damn, i am now not in coding mood [19:59] thats ok [20:00] i think, the problem ist, that the code above needs as many processes as instructions are to be moved [20:00] so just add another split? [20:00] nope [20:00] hmm... [20:01] it splits to +100, the copies a new split there, then jmz's, then executes a dat att +101 [20:01] hmm [20:01] ill bbl [20:02] if you come up with the solution keep it in the back of your head for me [20:02] MSG: [20:02] just add spl 1 at the top of the program [20:03] spl 100, 0 [20:03] mov >-1, }-1 [20:03] damnit [20:03] it works with an added spl 1 [20:03] but i suppose the jmn is not necessary [20:03] well, i got a solution here [20:04] how strong are silks in gerneral? [20:04] very [20:05] pretty much every paper nowadays uses either silks or moore-papers (in 94 and 'nop) [20:29] i request Fiz, Roy or Met to arrive here and now [20:36] kinda impossible.. [21:43] Join: sascha joined #corewars [21:43] Part: sascha left #corewars [22:15] http://www.ociw.edu/~birk/COREWAR/94/HILL/fatallure.red <- errmm.. is this pre or post HSA?