00:00:09 <oklopol> i hate most group activities that have a purpose
00:00:10 <ais523> you're supposed to moderate things up if you agree to them, and reply if you disagree
00:00:10 <AnMaster> <ehird> "Security researcher Jack Louis, who had discovered several serious security flaws in TCP software was killed in a fire on the ides of March, dealing a blow to efforts to repair the problem." <-- link
00:00:15 <ehird> http://it.slashdot.org/comments.pl?sid=1192793&cid=27509683 funny; shouldn't be
00:00:15 <ais523> downmods are only used against trolls, in theory
00:00:23 <ehird> AnMaster: http://it.slashdot.org/article.pl?sid=09/04/08/2010223
00:00:24 <ais523> in practice people downmod things they disagree with anyway
00:04:19 <AnMaster> so is this fixed in linux yet?
00:04:35 <AnMaster> and the details are not released?
00:05:47 -!- ais523 has quit (Read error: 104 (Connection reset by peer)).
00:08:14 <AnMaster> ehird, once linux fixes it it will be released in practise though...
00:13:02 <oklopol> fizzie: okay hierarchy looks quite interesting
00:13:35 <oklopol> the "a bit like chess" thing seems a bit far-fetched tho
00:14:01 <oklopol> i mean except for the fact the "pawns" can move twice on first turn :P
00:14:12 <oklopol> i guess that's kinda significant tho
00:16:09 <Sgeo_> oklopol, link to this game?
00:16:26 <oklopol> http://www.niksula.hut.fi/~svirpioj/hierarkia/rules_en.html
00:16:48 <oklopol> kinda bad english, hard to read imo
00:25:53 <oklopol> fizzie: that sounds like a very un-AI-zable game
00:27:11 <oklopol> of course i don't really know how games would go, so i can only guess
01:05:04 <Sgeo_> http://hof.povray.org/
01:07:37 <Sgeo_> psygnisf_, there are beautiful images
01:08:14 <psygnisf_> Sgeo_: very true. have you seen cgsociety's forums?
01:08:23 <psygnisf_> http://forums.cgsociety.org/forumdisplay.php?f=121
01:08:25 <Sgeo_> psygnisf_, no I haven't
01:08:44 <psygnisf_> http://forums.cgsociety.org/forumdisplay.php?f=137
01:09:22 <Sgeo_> Not enough outdoorsy or smooth stuff in the 3d stuff
01:10:25 <Sgeo_> Although I found a nice NSFW image >.>
01:11:31 <Sgeo_> And another that looks like a photo
01:11:39 -!- olsner has joined.
01:13:11 <psygnisf_> that utilizes some theory of grammar and semantics, plus maybe even pragmatics.
01:14:47 <oklopol> well umm i've always wanted to try that, sure
01:15:06 <psygnisf_> i guess ill have to teach you some linguistics then :p
01:15:14 <oklopol> but i don't really have that much free time, except for my idle time on irc, it's 3:24 and i'm reading electronics........................
01:16:16 <oklopol> and i only have like 4 exams in the summer.
01:17:59 <oklopol> easy leisurely exams i'm probably going to do standing on my head for shock value.,
01:25:01 -!- GregorR_ has joined.
01:25:39 -!- GregorR has quit (Nick collision from services.).
01:25:49 -!- GregorR_ has changed nick to GregorR.
01:26:59 -!- psygnisf_ has quit (Read error: 104 (Connection reset by peer)).
01:27:02 -!- psygnisfive has joined.
01:39:20 -!- M0ny has quit ("PEW PEW").
02:05:54 -!- FireFly has quit ("Later").
03:26:47 -!- Asztal_ has quit (Read error: 110 (Connection timed out)).
03:28:10 -!- zzo38 has joined.
03:28:35 <zzo38> I have a idea, which is, making Magic: the Gathering cards based on esoteric programming.
03:28:47 -!- Gracenotes has quit ("Leaving").
03:28:48 <zzo38> I have made cards with similar effects to SWAP command in CLC-INTERCAL
03:29:44 <zzo38> I have made a card that says "Swap the meaning of Flying and Trample."
03:29:48 -!- Gracenotes has joined.
03:31:43 <zzo38> Can you do something like, to check with greater probability that a quantum state is not very close to a particular state, to multiply the state by something like [1,0;0,40000000] is that possible?
03:33:14 <zzo38> I will be awaiting answer to all these things, please.
03:33:37 -!- zzo38 has quit.
03:59:57 -!- Sgeo_ has quit (Read error: 110 (Connection timed out)).
04:44:59 -!- kerlo has changed nick to ihope.
04:45:05 -!- ihope has changed nick to kerlo.
05:24:56 -!- neldoreth has quit (Read error: 113 (No route to host)).
05:25:09 -!- calamari has joined.
07:02:27 <fizzie> AnMaster: Yes, it's supposed to be a bit more difficult than chess, due to the larger branching factor.
07:02:35 <fizzie> Er, s/AnMaster/oklopol/
07:08:42 -!- psygnisfive has quit (Read error: 104 (Connection reset by peer)).
07:19:12 -!- psygnisfive has joined.
07:20:33 -!- oerjan has joined.
07:23:41 <oerjan> <oklopol> yeah oerjan is really mean always laughing at people's idiocy.
07:23:48 <oerjan> bwahaha what a stupid idea
07:27:36 <oerjan> <ais523> in practice people downmod things they disagree with anyway
07:28:16 <oerjan> i recall not too long ago reading a suggestion (or maybe it was actually applied somewhere) to have downmods lose a little bit of karma for the downmodder
07:28:31 <oerjan> so you would only do it when you really cared
07:29:06 <oerjan> well, and if your karma gets too low you cannot downmod at all
07:29:33 -!- calamari has quit (Read error: 60 (Operation timed out)).
07:29:51 <oerjan> i don't quite recall but by being upmodded by others presumably...
07:35:17 <oerjan> <zzo38> Can you do something like, to check with greater probability that a quantum state is not very close to a particular state, to multiply the state by something like [1,0;0,40000000] is that possible?
07:35:25 <oerjan> nope, that matrix is not unitary
07:36:37 <oerjan> all the things you can do with just 1 bit are more like rotations than like scalings, i think
07:39:48 <oerjan> unitary: all the row (equivalently, column) vectors in the matrix must have length 1 and be pairwise orthogonal
07:41:13 <oerjan> (as complex vectors, so you need to use conjugation in the scalar product)
07:41:52 <oerjan> <zzo38> I have made a card that says "Swap the meaning of Flying and Trample."
07:42:15 <oerjan> this could be the unholy child of Magic and Smetana/Smatiny...
07:44:54 -!- calamari has joined.
07:50:07 -!- neldoreth has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:28:18 -!- calamari has quit ("Leaving").
08:34:16 -!- KingOfKarlsruhe has joined.
09:05:57 <oerjan> qqqqqqqqqqqqqqqqqqqqqq
09:08:25 <oklopol> wait, fn? is that from the New Edition?
09:09:40 <oklopol> well need to brösh my töth.
09:09:43 <oerjan> fhtagnfhtagnfhtagnfhtagnfhtagn
09:12:19 <oklopol> new song in my head: "glio the safety conservative"
09:12:28 <oklopol> i think it's from one of my dreams
09:12:28 -!- sebbu has quit (Read error: 110 (Connection timed out)).
09:12:29 * oerjan swats psygnisfive. but not there. -----###
09:12:54 <psygnisfive> well i wouldnt want you to swap my anus anyway.
09:13:02 <oerjan> oklopol is clearly learning italian in his dreams.
09:13:30 <psygnisfive> while i, on the other hand, am learning italian on tuesdays and thursdays from 5:20p to 8:00p
09:14:08 <oklopol> err hmm kinda confusing dream, we were looking for some missing child, and when she was finally found, i was really disappointed 8|
09:14:27 <psygnisfive> thats because you were the person who kidnapped her.
09:14:29 <oerjan> yeah, those kids are so hard to get rid of
09:14:34 <oklopol> ...maybe i enjoyed the group effort
09:14:54 <psygnisfive> how horrible that we have such similar, sick ideas
09:14:58 <oklopol> yeah your answers are more probable
09:15:27 <psygnisfive> oklopol, didnt you once date a 13 year old or something like that?
09:15:31 <oklopol> that dude'll hit on anything that starts with an o
09:15:53 <oklopol> psygnisfive: yes; but when i was 11 dated an 11-year-old
09:16:18 <psygnisfive> yes but werent you like .. however old you are now minus a year or two when you were dating this 13 year old? :P
09:16:49 <oklopol> one of the 11-year-olds i dated when i was 11 had relationships with a 19-year-old and a 15-year-old, the latter kinda ended ours later on
09:17:27 <oklopol> psygnisfive: i don't remember that clearly
09:17:56 <psygnisfive> ehird, join me, my love! we can be free to express feelings for one another without the stares of police!
09:18:10 <oklopol> psygnisfive: no it's not legal, but you know she was a whore, who cares if it's mutual
09:18:41 <psygnisfive> surely paying a girl for sex is not "dating" in finland, is it?!
09:19:13 <oklopol> i meant you know whore like girl who likes giving.
09:20:29 <psygnisfive> she might even have been trying to enjoy the sex herself.
09:21:14 <oklopol> yeah, some things are definitely false
09:21:21 <oklopol> and no, there is such a thing.
09:24:22 <oklopol> psygnisfive: well yes of course she wanted to enjoy the sex, i basically just meant she liked sex, if you're an 11-year-old girl, that makes you a whore by some lesser definitions.
09:24:49 <psygnisfive> oklopol, if i travelled back in time, do you think i could get your 11 year old self to have sex with me?
09:25:06 <oklopol> i think i was very homophobic back then
09:25:36 <psygnisfive> given that homophobia is almost always repressed homosexual feelings.
09:25:48 <oklopol> no it's friends being homophobic
09:26:19 <oklopol> right, so i guess we were kind of a gay class.
09:26:36 <oklopol> think of all the wasted anal sex!
09:26:45 -!- olsner has quit ("Leaving").
09:27:08 <psygnisfive> oklopol: promise me youll make up for it with me some day
09:27:10 <psygnisfive> also: http://languagelog.ldc.upenn.edu/nll/?p=1309
09:28:15 <oklopol> but anyway, that "homophobic means gay" is freudian bullshit that's bullshit
09:28:26 <oklopol> so it's grammatically sensible.
09:29:14 <psygnisfive> that homophobes tend to be aroused by gay porn more than non-homophobes!
09:31:43 -!- M0ny has joined.
09:32:22 <oklopol> well that can be one motive, even the primary one
09:32:44 <oklopol> but i don't really believe in anything that reduces a human behavior in to one cause
09:33:35 <oklopol> psygnisfive: maybe when i get my master's, for phd stuff, if that works out
09:33:46 <psygnisfive> well, i might come to finland some time :o
09:34:40 <oklopol> shuuuuuuure i can tell you all about our generating grammars and stuff
09:38:22 <oklopol> <psygnisfive> i dont know if its freudian. <<< i use freudian synonymously to bullcrap
09:38:31 <oklopol> psygnisfive: earlier error
09:38:50 <psygnisfive> so lets drop the pretense and just start fucking.
09:39:02 <oklopol> i just have kinda multiple threads working on the talking
09:39:15 <oklopol> so corrections can come asynchronously occasionally
09:42:53 -!- BeholdMyGlory has joined.
09:53:08 <oklopol> okay now i beat the reversi ai in the endgame!
09:53:26 <psygnisfive> does that mean the world will end or something?! D:
09:53:35 <oklopol> (all my previous victories were killing it in its infancy, because the endgame is where it's good at)
09:53:37 <oerjan> this is bad. now it will get vengeful and... darnit psygnisfive
09:53:44 <oklopol> no it means i was lucky :<
09:53:51 <oklopol> it just made a very silly mistake.
09:56:33 <oklopol> blah now it beat my by one
09:57:22 * AnMaster wonders why this file include sys/time.h and sys/resource.h
09:57:24 <oklopol> the problem is i'm just too stupid for games, i'm not smart enough to actually think when playing
09:57:37 <AnMaster> as far as I can tell it doesn't need them. And I don't think it ever did
09:59:22 <oklopol> grrrrr i hate it, it always makes the same stupid mistake, and it's pretty much the only thing i can actually recognize as a mistake in that game
10:04:32 <oklopol> so umm turns out level 3's strategy is having almost no pieces on the board so that it can control my moves
10:05:16 * oklopol is suddenly reminded of the "it is generally recognized that humans are no match for computers in othello" mention in aima
10:06:53 <oklopol> i mean 2-player games where there is no absolute measure of success, i can do all kinds of puzzles quite well
10:07:30 <oklopol> my way to learn board games would probably be to learn more theory
10:08:29 <oklopol> that's how i learned puzzles, suddenly i realized a rigorous mathematical approach simply beats pretty much any puzzle you're going to find online
10:09:51 <oklopol> but for games all i can do is stare
10:10:04 <oklopol> i simply don't know how to think about my moves.
10:10:45 <oklopol> what the fuck is that thing
10:10:56 <oklopol> i mean a button that closes everything you have open
10:12:18 <AnMaster> some open one tab for each bookmark thing iirc
10:12:20 <oklopol> and because undo is not an os level feature, there's simply no way to reverse it
10:12:38 <AnMaster> or did it open the same page in all tabs
10:13:04 <oklopol> it opens two random tabs, one is bbc news, one is a getting started in mozilla page
10:13:07 <oklopol> both entirely useless to me
10:13:31 -!- oerjan has quit ("leaving").
10:13:32 <oklopol> well maybe set up by previous owner.
10:13:37 <AnMaster> oklopol, about closing tabs, undo exists as some addon in firefox
10:13:51 <AnMaster> or "tab mix plus" or "tabmix plus"
10:14:17 <oklopol> or i could just get a windows machine and use IE again
10:14:22 <AnMaster> oklopol, previous owner? Wouldn't you do a clean reinstall if you buy a computer second hand
10:14:24 <oklopol> usually i swap when i get annoyed with the other
10:14:38 <oklopol> AnMaster: no i had the previous owner do the installing for me
10:14:49 <oklopol> that's why i have ubuntu, you think i'd install an os
10:15:14 <oklopol> i'm not going to install anything that requires multiple clicks!
10:15:18 <AnMaster> oklopol, well what about disabling that button. iirc that is rather easy in firefox
10:15:43 <AnMaster> oklopol, just tell me where exactly it was, since there are several places like that iirc
10:15:59 <oklopol> just over the tabs, it's a whole bar
10:16:11 -!- FireFly has joined.
10:16:12 <oklopol> you could probably just close it completely somewhere
10:16:22 <oklopol> if only i knew the basics of this canonical interface...
10:16:53 <oklopol> it was the bookmarks toolbar
10:17:03 <oklopol> maybe it does open all bookmarked stuff.
10:17:08 <oklopol> i just haven't bookmarked anything
10:17:14 <AnMaster> oklopol, yes it is what it does
10:17:24 <AnMaster> and those two are probably default bookmarks
10:17:39 <AnMaster> anyway you mean the right-click menu for the bookmarks toolbar
10:17:44 <oklopol> i hear some people read the news
10:17:54 <oklopol> and yep that's what imeant
10:17:59 <AnMaster> oklopol, I spent half an hour reading the news paper this morning
10:18:07 <oklopol> AnMaster: i know you read it a lot
10:18:44 <oklopol> tbh i nowadays occasionally read the paper when waiting for the pizza at the place, if i forget to bring my own reads with me
10:18:58 <oklopol> it's a good reminder of why i don't read one at home
10:19:17 <oklopol> i meant there's like one interesting piece of news per ten papers
10:19:25 -!- asie[Virus] has joined.
10:19:38 <oklopol> and even that is usually in the comics section.
10:19:44 <asie[Virus]> Found a special tool for destroying the abomination that's called Win32.Virut.
10:20:00 <AnMaster> oklopol, I found the definition of it, let me figure out how to disable it
10:20:07 <asie[Virus]> Safe Mode + reinstall - maybe, but not exactly sure
10:20:14 <oklopol> AnMaster: i did it already!
10:20:18 <asie[Virus]> clean reinstall - sure, but i have too much stuff to remove
10:20:27 <AnMaster> oklopol, oh? by removing that toolbar or?
10:20:45 <AnMaster> oklopol, I was talking about just removing that single menu entry
10:20:48 <oklopol> because i don't use toolbars
10:21:00 <oklopol> would that have required like a compile?
10:21:26 <oklopol> i guess i use some toolbars.
10:22:16 <oklopol> well i've been working on this one lang
10:22:39 <asie[Virus]> what is it called and what's the "thing" it has that other ones don't
10:22:44 <oklopol> but i'm trying to keep it in the dark until ready, so no, not really
10:23:15 <oklopol> well the basic idea is guessing the function body.
10:23:32 <oklopol> the real ideas are in how this is done efficiently
10:23:34 <AnMaster> oklopol, this wouldn't need recompile
10:24:07 <fizzie> There's also that "history/recently closed tabs" thing, I'm not really sure if it records tab closed by the "open all in tabs" misfeature.
10:24:09 <oklopol> asie[Virus]: and called clue, atm
10:24:15 <AnMaster> oklopol, since most of firefox is written in javascript
10:24:28 <fizzie> If it does, you can "undo" the operation by using the "open all in tabs" option found in that recently closed tabs -menu.
10:24:42 <asie[Virus]> So what, do you write the function but it guesses what does the function take via examples?
10:24:56 <AnMaster> <fizzie> There's also that "history/recently closed tabs" thing, I'm not really sure if it records tab closed by the "open all in tabs" misfeature. <-- that one closes existing ones or?
10:24:57 <oklopol> fizzie: what? repeat that, that's useful to me
10:25:09 <asie[Virus]> or do you type only arguments AND it creates the body based on examples
10:25:54 <oklopol> asie[Virus]: you don't write function bodies
10:26:16 <oklopol> you give examples of input-output pairs, and a simple set of functions to build the new function out of
10:26:31 <AnMaster> oklopol, fizzie: http://jsbi.blogspot.com/2007/10/how-to-configure-open-all-in-tabs-in.html <-- solution for closing existing tabs it seems
10:26:35 <oklopol> and it brute-forces the body.
10:26:46 <oklopol> fizzie: thanks you saved both my lives
10:27:36 <oklopol> asie[Virus]: yes, but as i said, the real ideas are in how the brute forcing is made at least remotely doable
10:27:48 <fizzie> oklopol: Good if it helped. Anyway, according to that AnMaster link it's rather easy (one about:config property change) to configure the bookmarks "open all in tabs" thing not to kill existing tabs.
10:27:54 <asie[Virus]> oklopol: Oh, but still, I'd like to test it
10:28:07 <oklopol> there are multiple kinds of examples, used for different purposes, to make sure you never need to do recursion into a function you don't know is correct.
10:28:38 <oklopol> i mean when guessing the function
10:28:44 <oklopol> asie[Virus]: i can dig up factorial for you
10:28:49 <AnMaster> oklopol, fizzie, as for how to completely remove it (disclaimer: this may be outdated, I haven't tried it, it may be easy to mess up): http://codingforums.com/showthread.php?t=132383
10:28:53 <oklopol> or just write it now, i changed syntax a bit
10:29:53 <oklopol> factorial . 0>1 : 4>24 :. 5>120 :: 11>39916800; factorial ~ mul dec
10:30:04 -!- asiekierka has joined.
10:30:39 <oklopol> . is used for deducing base cases, : and :. are used in deducing the cody, :: is used for checking correctness
10:31:26 <oklopol> idea is it can't ever just take the least common denominator of all examples, because you can give it large examples not even used in the guessing process, except to test the end results
10:32:18 -!- asie[Virus] has quit ("http://www.mibbit.com ajax IRC Client").
10:32:46 <oklopol> there are some issues for functions that don't have quite as straightforward a recursion pattern, but they are probably doable, i just don't know how yet
10:36:34 <asiekierka> I'm thinking of an esolang while I have too much time on my hands xD
10:38:32 <asiekierka> One of the ideas I got is: Normal 6-sided cubes, a number on each side, on a map. Each cell on the 2-D map is mapped to an instruction, and you can map them yourself. So the cubes start on all sides as 0's, and you can move them left, right, up or down. The source code is the instruction map followed by the amount of cubes and instructions for each cube
10:38:47 <asiekierka> Each cube has a separate IP, which can be modified by an instruction on the map to create loops
10:39:05 <asiekierka> It's a neat idea and I think possible to implement by me
10:39:31 <asiekierka> well, the map could be implemented as a PNG file
10:40:00 <asiekierka> So the source code would consist of: x maps and 1 "cube instruction file"
10:41:08 <asiekierka> while rotating it 2 times south and 2 times east
10:41:58 <asiekierka> The cubes change the side on the axis they're moving
10:42:07 <oklopol> you have like a map and tons of cubes on it, and all cubes have a simple program controlling them?
10:42:52 <asiekierka> but the cube controlling programs have only 5 commands:
10:43:03 <oklopol> so you have tons of really simple entities and you need to use the interaction to get actual computation?
10:43:07 <asiekierka> N - Move North, S - Move South, E - Move East, W - Move West, P - Pause
10:43:21 <asiekierka> oklopol: Yep, interaction with the map
10:43:48 <asiekierka> oklopol: There'll be an instruction to dec/inc the IP
10:44:13 <asiekierka> and a double-move instruction "The next move will be carried out twice, ignoring the first block hit"
10:44:31 <oklopol> i'd probably prefer it if you could just have you know [NNSS] to loop that piece forever
10:45:03 <asiekierka> but I'd like to actually keep both ways
10:45:06 <oklopol> no actually i like the actual programs looping
10:45:27 <asiekierka> And the simple programs loop from beginning to end
10:45:28 <oklopol> well do what you wish, but i love the general idea
10:45:37 <asiekierka> until they hit a block causing the cube to "die"
10:45:58 <asiekierka> and the program ends either if a halt block is hit or if there are no cubes left
10:46:02 <oklopol> also maybe there could be commands in the map language to fill an area with cubes of a kind and such
10:46:14 <oklopol> i mean i definitely want like tons of cubes
10:46:29 <asiekierka> oklopol: Well, you would need to have preloaded programs
10:46:39 <asiekierka> but you can have as many cubes as your memory allows
10:47:17 <asiekierka> and remember, each cube has 6 local variables and a cell memory of 10000 8-bit cells
10:47:40 <oklopol> okay good was worried there for a sec
10:47:40 <asiekierka> This is a pain when you step on a command that needs a parameter
10:48:20 <asiekierka> But still, I think this is an awesome idea
10:48:29 <asiekierka> sadly, i can't do it until my PC is clean
10:49:00 <asiekierka> but I promise there will be at least the 4 commands from deadfish
10:49:03 <oklopol> i like the rolling dice idea, assuming current top number is used as a param to whatever instruction is stepped on
10:49:08 <oklopol> it's like wheel done right
10:49:40 <asiekierka> inc by 1, inc by the dice variable, dec by 1, dec by dice variable, swap cell and current top number, add current top number to cell...
10:49:47 <asiekierka> This one requires a lot of commands to be done easy
10:50:40 <asiekierka> oklopol: Selected by another two commands
10:50:51 <oklopol> cubes have their own cell in the global memory?
10:51:39 <oklopol> well no matter, as long as they are simple commands
10:51:56 <oklopol> just remember to keep the cubes dumb :P
10:51:59 <asiekierka> inc, dec, add, sub, move ptr left, move ptr right, skip command if blah, etc...
10:52:31 <oklopol> i need to make coffee now, and think about hordes of rolling dice
10:52:44 <asiekierka> oklopol: You mean, keep them with only 5 commands, NSEW and Wait 1 cycle?
10:53:04 <asiekierka> cuz the map will have a bunch of simple instruction
10:53:14 <asiekierka> while cubes will have some simpler commands
10:57:14 -!- asiekierka has quit ("http://www.mibbit.com ajax IRC Client").
10:57:19 -!- asiekierka has joined.
10:57:55 <oklopol> asiekierka: well point is it'd be cool if you actually had to use multiple cubes.
10:58:06 <oklopol> or that it actually was easier than just using one
10:58:15 <asiekierka> oklopol: Well, it IS easier if using more cubes
10:58:19 <oklopol> that's pretty hard to achieve ofc.
10:59:01 <oklopol> welllllll, just finish it and i can tell you whether i'd use just one cube for programs or multiple :|
10:59:11 <asiekierka> oklopol: One of the ways I could achieve this is that I only had 5 commands and different types of cubes
10:59:19 <asiekierka> For example, an addition/subtraction cube
10:59:27 <oklopol> i'm just saying usually it's easier to have you know an ip.
10:59:59 -!- tombom has joined.
11:00:00 -!- tombom has quit (Remote closed the connection).
11:00:08 -!- tombom has joined.
11:00:13 <asiekierka> oklopol: I plan to make text output doing just that
11:00:39 <oklopol> i mean like a turtle, something the programmer can focus on, "so okay it moves here, then it does this, then..."
11:01:10 <asiekierka> Where it outputs the map status, positions of cubes and what they're doing and their variables
11:01:55 <asiekierka> but it'll be much more fun using multiple cubes if I had 3D output
11:02:35 <asiekierka> but I think newbies would use a single cube
11:02:37 <oklopol> it's easier than you think
11:03:27 <asiekierka> oklopol: I can implement the interpreter
11:03:43 <oklopol> i'm just saying the hard parts of 3d'ing aren't present in such a simulation
11:04:17 <asiekierka> But still, I will do text output, and I will put the source code
11:04:29 <asiekierka> so anyone interested can ask me and help me adding 3D output
11:04:33 <oklopol> err what's there to get about gl, it's a library
11:04:50 <oklopol> but sure i'm fine with text output
11:05:06 <asiekierka> oklopol: As in, it will output the current map look
11:06:49 <asiekierka> I probably will given that you implement Clue
11:07:02 <asiekierka> and wonder how will I write apps for it xD
11:07:22 <asiekierka> I only need to craft the instruction set
11:08:26 <oklopol> i will implement clue if i have time for that before i realize some horrible defect in the idea...
11:08:45 <oklopol> i've realized many, but always found my way around them
11:08:53 <oklopol> it's just quite a different paradigm
11:09:05 <oklopol> which is ofc something i always aspire to create
11:10:33 <oklopol> the same paradigm as befunge presumably, depending on how relevant multithreading ends up being
11:11:10 <asiekierka> Probably "not exactly needed but useful for optimization and just having more fun doing an app in it"
11:11:14 <oklopol> i've been thinking of a similar language, except no instructions, just millions of cubes using a simple set of rules to move around
11:11:45 <oklopol> asiekierka: so probably the funge paradigm
11:13:03 <oklopol> dunno, depends on how memory is done
11:13:59 <asiekierka> which can be added to by cubes falling into holes
11:14:01 <oklopol> finite state var per cube probably
11:15:31 <asiekierka> I will show you how a one-cube cat could look like
11:15:41 <oklopol> anyway i need to start reading my book, would've started earlier, but it seems i don't have to do my exercises for tomorrow either so i'm kinda on holiday atm
11:19:56 <asiekierka> but it's a general outline of a 1-cube Cat
11:21:51 <oklopol> the NNSSSNSNSNSSN line is the cube's code?
11:22:03 <oklopol> what are ones and fives? where's the cube?
11:22:30 <asiekierka> The width, The height, The map, The cube amount, The cube instructions
11:23:38 <asiekierka> A hello world could be some cubes inputting their code
11:23:52 <asiekierka> then pausing and moving appropiately to output Hello, World!
11:23:55 <oklopol> okay then that looks about right
11:24:40 <oklopol> i think cubes having code has interesting implications at least for simple programs
11:25:23 <asiekierka> But I need to make the instruction set
11:25:36 <asiekierka> except if you want to take on with the project and finish it yourself
11:28:02 <oklopol> probably not, i'm not *that* interested
11:28:54 <oklopol> i just don't get this part so can't really get dragged into the book
11:29:19 <oklopol> but, i'll close the monitor now, so umm see you in a few hours maybe
11:33:31 -!- GiveMeMony has joined.
11:50:43 -!- M0ny has quit (Read error: 110 (Connection timed out)).
12:02:27 -!- GiveMeMony has changed nick to M0ny.
12:05:09 <AnMaster> anyone know of any program (on linux) to find out what notes were played in a piece of music. Hm...
12:05:57 <AnMaster> I can kind of hear several of them by listening, and can even play it on the piano from that, except some that I get confused by.
12:06:15 <AnMaster> oklopol, was that an answer to my question or just random
12:06:33 <oklopol> but yeah notes can be pretty confusing
12:07:50 <AnMaster> oklopol, there are two instruments (some stringed instrument and some sort of "no such instrument" from a synth I think) playing in the music file and I'm interested in one of them,
12:08:10 <AnMaster> and it is the latter instrument (much louder) that I'm interested in finding the notes for
12:08:39 <oklopol> well i can try listening, not that i'm especially good at it
12:08:43 <AnMaster> and of course I know with midi
12:09:15 <asiekierka> AnMaster: That's next to impossible with a computer program
12:09:27 <oklopol> i'm leaving soon tho, if i can't look now, i'll look later
12:09:51 <AnMaster> since this is from an open source game I guess I could try to contact the original author or something... but that sounds like more work
12:10:20 <AnMaster> oklopol, http://svn.gna.org/viewcvs/*checkout*/wesnoth/trunk/data/core/music/sad.ogg?rev=30628
12:11:44 <oklopol> hmph, not opening it seems
12:12:11 <oklopol> i mean the actual sound is not coming out, i get on the page
12:12:19 <fizzie> There's a lot of algorithms for that, but I'm not sure if there are very many applications.
12:12:21 <AnMaster> oklopol, that is a download link for the ogg
12:12:38 <oklopol> AnMaster: yeah but firefox just opens it in the browser
12:12:39 <AnMaster> so yes you can download it I guess, just wget it
12:12:56 <AnMaster> oklopol, meh it works with mplayer http://svn.gna.org/viewcvs/*checkout*/wesnoth/trunk/data/core/music/sad.ogg?rev=30628 on command line here
12:13:06 <oklopol> and there's "open with media player" in right-click menu, but that program doesn't really work
12:13:27 <oklopol> never tried wgetting, but okay let's try that
12:13:30 <AnMaster> fizzie, so no idea about any program at all
12:13:46 <AnMaster> oklopol, wget "http://svn.gna.org/viewcvs/*checkout*/wesnoth/trunk/data/core/music/sad.ogg?rev=30628"
12:13:57 <AnMaster> file will be named sad.ogg?rev=30628
12:14:09 <fizzie> AnMaster: Well, you could possibly abuse praat (that's in many package managers) but since it's really designed for speech processing, it'd be mostly manual pick-from-spectrogram stuff.
12:14:29 <AnMaster> fizzie, not in gentoo at least with that spelling
12:14:55 <fizzie> Well, it's in Debian. Anyway, it does not do "pick up notes", really.
12:15:27 <AnMaster> well I could pick from spectrogram I guess. If this app actually works well
12:16:02 <AnMaster> oklopol, yes I have parts of it already
12:16:29 <oklopol> 023...0...3...2.b.0.............023...0...3...5.2.0...............
12:16:42 <oklopol> wait a sec forgot the rest
12:16:58 <fizzie> I think I've seen at least one sampled-music-to-midi conversion application, but I think it probably didn't work very well. The easiest way would indeed have at least a bit music-oriented person do it.
12:17:12 <oklopol> 357...3...7...5.2.3.............357...3...7...5.2.3...............
12:17:21 <oklopol> AnMaster: notes are ba01234567 here
12:18:04 <AnMaster> I got A B C A C B G A as the start
12:18:49 <AnMaster> as for the timing I can figure out that myself much more easily
12:18:49 <oklopol> a change of 1 just means stepping one forward, in piano keys.
12:19:01 <oklopol> well you have the timing as well, unless i typoed dots
12:19:26 <oklopol> anyway i can't promise i can do the background melody, i'm not that good at listening
12:19:58 <AnMaster> lets see, I managed to get A B C A C B G A A B C A C D B A before I asked in here. Does that match... hm
12:20:02 <oklopol> it's somekinda weird instrument
12:20:25 <oklopol> the second part is just the first thing, one third up
12:20:27 <AnMaster> oklopol, then what after, lets see, I lost track of where I was in your line
12:20:48 <oklopol> but it's just repeating the first thing
12:21:29 <oklopol> btw we had much harder melodies as exercises at like 3rd grade
12:21:42 <fizzie> oklopol: Did it make you SAD, though? It is, after all, sad.ogg.
12:21:56 <AnMaster> oklopol, mhm, Swedish school system suck I guess.
12:22:05 <oklopol> fizzie: yes, sad that anyone would make suck a trivial piece
12:22:18 <oklopol> AnMaster: we were a special music class, lke
12:22:19 <fizzie> make: *** No rule to make target `suck'. Stop.
12:22:57 <AnMaster> oklopol, um I don't think our way to interpret first line match, since you are missing a -1 there?
12:23:08 <oklopol> AnMaster: i don't think there's any music "reverse-engineering" in most schools
12:23:26 <fizzie> There certainly wasn't music reverse-engineering in our school.
12:24:19 <oklopol> fizzie: that was pretty much the only thing i learned something from in elementary school
12:24:34 <oklopol> i mean the music stuff in general
12:24:49 <oklopol> problem is music is kind of a useless subject
12:25:14 <oklopol> i mean you can't use musical intuition for math.
12:25:54 <oklopol> AnMaster: it's in the correct place
12:25:59 <oklopol> so should be easy to find.
12:27:07 <AnMaster> <oklopol> 357...3...7...5.2.3.............357...3...7...5.2.3............... that would be CDE...C...E...D.B.C.............CDE...C...E...D.B.C............... right
12:31:57 <AnMaster> so rewritten as note names it end up as: ABC...A...C...B.G.A.............ABC...A...C...D.B.A...............CDE...C...E...D.B.C.............CDE...C...E...D.B.C............... indeed
12:32:05 <AnMaster> oklopol, anyway if you want something more complex...
12:32:36 <AnMaster> http://svn.gna.org/viewcvs/*checkout*/wesnoth/trunk/data/core/music/vengeful.ogg?rev=29785
12:32:45 <AnMaster> oklopol, it should be a bit less simple ;P
12:32:54 <AnMaster> (no I don't want a list of notes in it)
12:33:41 * AnMaster notes that in general wesnoth has very good in-game music
12:34:02 <fizzie> Is it just me, or was the overall volume level a lot higher for this latter song?
12:34:26 -!- asiekierka has quit ("http://www.mibbit.com ajax IRC Client").
12:34:42 <AnMaster> fizzie, that's odd, I noticed that it is higher in ogg123 but not in older mplayer versions (seems to be higher in new mplayer versions too)
12:35:04 <AnMaster> which makes no sense as far as I understood the ogg format
12:35:31 <fizzie> It's "MPlayer 1.0rc2-4.3.2" they've installed here.
12:35:45 -!- KingOfKarlsruhe has quit (Remote closed the connection).
12:36:12 <AnMaster> MPlayer 1.0rc1-4.1.1 (C) 2000-2006 MPlayer Team
12:36:25 <fizzie> The copyright note in this says "(C) 2000-2007".
12:36:29 <oklopol> AnMaster: that's kinda ambient, so yeah i can't say i remember it with one hearing
12:36:44 <AnMaster> oklopol, do you like it though?
12:37:41 <fizzie> Deewiant: I wonder if you get highlighted messages in a log or something. Anyway, about the amount of crashes in the tournament thing; it certainly is rather impressive: http://www.cs.hut.fi/Studies/T-93.4400/2009/results/
12:37:53 <AnMaster> oklopol, ok, depends on what type of music.
12:37:54 <oklopol> my longest pieces are like 20
12:38:15 <fizzie> "It stretches into distance like a 50-minute kraftwerk song", to quote one webcomic.
12:39:08 <AnMaster> well game music is a rather unusual genre really.
12:39:26 <AnMaster> close to certain types of film music.
12:40:59 * AnMaster doesn't usually like music with a lot of beat in it, the exception being film/game music where it fits the film or game.
12:41:50 <oklopol> fizzie: how did you score crashes?
12:41:58 <AnMaster> (easy to mix them up thanks to Swedish word film)
12:42:29 <AnMaster> Results 1 - 10 of about 7,720,000 for "film music". (0.14 seconds) Results 1 - 10 of about 6,640,000 for "movie music". (0.13 seconds)
12:42:34 <oklopol> so don't worry about that.
12:42:45 <AnMaster> or is THAT the real Swedishism
12:45:16 <oklopol> i don't know what film of oil means
12:46:03 <AnMaster> oklopol, I think either http://svn.gna.org/viewcvs/*checkout*/wesnoth/trunk/data/core/music/suspense.ogg?rev=32312 or http://svn.gna.org/viewcvs/*checkout*/wesnoth/trunk/data/core/music/heroes_rite.ogg?rev=30993 should be even harder.
12:46:15 <AnMaster> http://svn.gna.org/viewcvs/*checkout*/wesnoth/trunk/data/core/music/siege_of_laurelmor.ogg?rev=34075 sounds hard too to me. But for different reasons
12:46:45 <fizzie> oklopol: As a win for the non-crashing opponent.
12:46:48 <oklopol> well i'm leaving now, so can't listen
12:47:27 <AnMaster> fizzie, that bot boar or whatever seemed pretty stuipd
12:47:39 <AnMaster> and didn't some of them test this at all?
12:55:23 <fizzie> AnMaster: The boar bot is the one I mentioned last night: <fizzie> Deewiant: That "boar" bot there which has crashed all games has a "move()" method that has the form "do(); stuff(); and(); stuff(); /* something(); */ return null;"
12:56:13 <fizzie> I did try to tell them to test it under the tournament system, with the memory limits and such in place, but I guess not everyone bothered.
12:56:27 <AnMaster> fizzie, mhm, would it have worked otherwise?
12:56:45 <fizzie> Not that one, but there are some who've been crashing with OutOfMemoryErrors.
12:57:06 <AnMaster> fizzie, so what about the boar one, didn't they test it at all?
12:57:32 <fizzie> One submission only had the .java sources, not compiled class files at all. My personal guess for that is that they've been developing with Eclipse, and when you run the GUI thing under Eclipse it actually uses the eclipse-compiled classes no matter what's in the .jar file.
12:57:49 <fizzie> I guess the boar people could've simply accidentally returned the wrong .jar. Or something. Or maybe they didn't test it.
12:58:36 <fizzie> Haven't even announced the results officially yet, waiting for the "unofficial" participants (random-move-bot and last year's top 5) to finish so I get some sort of comparisons there.
12:59:58 <fizzie> Anyway, the instructions say they should be writing a couple of lines about the bot's tournament results in the final reports (due in two-three weeks), I'm sure they'll tell me there what went wrong.
13:02:38 <fizzie> The boar-bot results did crash my statistics-page-generation script; there is a silly "efficiency" measure -- log(score/totalcpu), basically "how good results achieved per CPU seconds of computation", in log-scale -- which didn't like the score == totalcpu == 0 case.
13:11:23 -!- andreou has quit (calvino.freenode.net irc.freenode.net).
13:11:23 -!- kerlo has quit (calvino.freenode.net irc.freenode.net).
13:11:49 -!- andreou has joined.
13:11:49 -!- kerlo has joined.
13:30:03 -!- KingOfKarlsruhe has joined.
13:41:35 -!- Asztal_ has joined.
13:44:44 -!- neldoret1 has joined.
13:56:18 -!- neldoreth has quit (No route to host).
14:03:40 <AnMaster> everything else thinks the file is either unreadable or size 0
14:03:49 <AnMaster> meh, /proc on freebsd is strange
14:07:55 <fizzie> Alternatively /proc on Linux is strange; they're just different, I think. At least on this FreeBSD the /proc/<pid>/map file of a random process is readable with cat, but maybe that was some special process?
14:09:21 <fizzie> Actually it was /proc/curproc/map... the ones in pid-dirs seem a bit uncattable, although with "Operation not permitted".
14:09:41 <fizzie> Must go catch a bus again.
14:27:34 <AnMaster> fizzie, it was a python process
14:27:58 <AnMaster> that was eating 50% of the 6 GB RAM in the server
14:32:01 -!- oerjan has joined.
14:33:54 <oerjan> <oklopol> my time is precious.
15:02:09 <ehird> 08:16 oklopol: one of the 11-year-olds i dated when i was 11 had relationships with a 19-year-old and a 15-year-old, the latter kinda ended ours later on
15:02:25 <ehird> i knew an idiotic 11 year old who simultaneously had a relationship with an 18 year old and a 19 year old, iirc
15:02:30 <ehird> well small values of "knew"
15:02:36 <ehird> more like "idly detested"
15:08:35 <Deewiant> fizzie: Heh, amusing results. And yes, I get highlighted messages in the awaylog.
15:09:37 <ehird> they seem slower than HDs?
15:10:10 <Deewiant> ehird: http://www.anandtech.com/printarticle.aspx?i=3531 - best article on the topic, ever. Read it.
15:11:16 <ehird> Deewiant: wanna give me a tl;dr summary while I read?
15:11:24 <ehird> Only it's kind of fucking hug
15:11:39 <oerjan> it hugs you and draws you into its maw
15:12:06 <oerjan> sorry, been reading about too many baby-eating aliens lately
15:12:18 <ehird> I got yall hooked on that story, Idid
15:12:47 <Deewiant> ehird: Well, it explains what SSDs are, why you want them, how they work, why most of the ones on the market are actually pretty crap, etc
15:13:05 <ehird> No Shit Sherlock(TM)
15:13:13 <ehird> I meant what is its essential results :-
15:13:47 <Deewiant> Intels rock but cost craploads, some of OCZ's new ones seem to be okay and aren't too expensive, all the rest suck
15:14:47 <ehird> Deewiant: All I want is a super-fast 1TB SSD for $3.
15:14:51 <ehird> That's not asking for much.
15:15:22 <oerjan> ehird: i'm sure only thing you need for that is a time machine
15:17:02 <ehird> Anyway, SSDs are really appealing to me atm as a silent replacement for the velociraptor
15:18:12 <Deewiant> Also much more expensive and less spacey
15:18:21 <ehird> oerjan: VelociRaptor. It's a 10K rpm drive.
15:18:42 <ehird> Deewiant: Expensive, yes, but a lot of things I'm doing for the silence is expensive. Less spacey, yep, that's irritating.
15:18:50 <oerjan> i was wondering if you were building onto my time machine joke
15:19:16 <Deewiant> ehird: Just get a 5.4K RPM drive if you want silence. :-P
15:19:30 <ehird> Deewiant: I'm also trying to get speed. :P
15:19:38 <ehird> Compromising is hard.
15:19:41 <ehird> Let's go shopping.
15:20:06 <oerjan> that would be in itself an awful compromise, since i hate shopping
15:20:19 <Deewiant> Well, put the two in priority order and use the lower one only to break ties :-P
15:20:49 <ehird> Deewiant: I'm ordering on the unordered tuple (power,silence).
15:20:54 <ehird> The ordering is decided at think-time.
15:21:10 <Deewiant> ehird: So think again until you get it the way I want you to get it.
15:21:36 <Deewiant> And then don't think about it again.
15:21:37 <oerjan> so if your thinking is confused you might end up comparing the power of one with the silence of the other?
15:23:20 <oerjan> you might order by silence^power, then you get a math pun out of it
15:24:20 <ehird> "SSDs have +5 armor immunity to random access latency (that’s got to be the single most geeky-sounding thing I’ve ever written, and I use words like latency a lot)"
15:24:30 <oerjan> hm, this makes me wonder if evil = money^power
15:25:09 <ehird> so it follows that evil = money^2
15:25:17 <ehird> evil = money^power
15:25:25 <oerjan> um roots are not necessarily square roots
15:25:52 <oerjan> which was my thinking exactly
15:26:46 <ehird> It has to be said that I'm fucking crazy (I spent yesterday chasing up anything hinting at a fanless i7 cooler...)
15:27:54 <oerjan> i've read somewhere that liquid nitrogen is cheaper than beer, but the rest of the necessary equipment might be a bit more
15:28:14 <oerjan> (beer in the US, i think, which is probably pretty cheap)
15:28:22 <ehird> watercooling would work it's just that no fucking way man
15:28:36 <ehird> i'm a new member of the Huge Fucking Heatsink church
15:29:03 <oerjan> fucking is not generally considered a heatsink
15:30:45 * ehird lols at the hypothetical 20KB drive
15:38:29 <ehird> well, the erasing issue and degrading performance is sad. OTOH they are still faster than the raptors
15:38:39 <ehird> I'd go with the intel x25-e, I think
15:39:10 <ehird> only goes up to 64gb
15:39:19 -!- neldoret1 has quit (Read error: 113 (No route to host)).
15:39:26 <ehird> well the -M would be fine too :P
15:43:07 <ehird> but srsly, $700 more than a 300gb velociraptor for a 150gb x25-m
15:43:15 <ehird> that's just redonkulous
15:44:50 <ehird> SSDs' form factor is nice though
15:45:02 <ehird> I hate pluralization
15:46:02 <Deewiant> Well, if it's 50x faster and 0.5x as big then 10x the price is a good deal? :-P
15:46:16 <ehird> "I told him I’d need an average response time in the sub-1ms range and a max latency no worse than Intel’s 94ms. I didn’t think it would be possible. I was prepared for OCZ to hate me once more. He told me to give him a couple of days."
15:46:20 <ehird> this guy is a bastard, I love him
15:46:35 <ehird> Deewiant: i simply can't afford to pay that much for a drive :-P
15:48:38 <ehird> when will he quote the price
15:49:20 -!- Asztal_ has quit (Read error: 110 (Connection timed out)).
15:52:20 <ehird> http://images.anandtech.com/graphs/thessdanthology_031809001858/18643.png
15:53:15 <oklopol> fizzie: what's wrong with the form do();stuff();and();stuff();? i mean i don't know what you mean by that form
15:56:44 <ehird> the vertex looks nice
15:57:50 <oklopol> <oerjan> <oklopol> my time is precious. ||| <oerjan> k <<< :D
15:59:34 -!- neldoreth has joined.
16:12:32 <ehird> http://www.research.att.com/~njas/sequences/A104175
16:25:37 <Deewiant> ehird: Your Velociraptor doesn't seem so fast after looking at graphs like that does it? :-P
16:25:52 <ehird> Deewiant: Yeah but $s.
16:26:46 <Deewiant> The X25-M doesn't seem /that/ expensive actually
16:26:59 <Deewiant> The -E is, but the -M is almost purchaseable
16:27:11 <ehird> $700 more than a raptor, Deewiant.
16:28:00 <ehird> Deewiant: 80GB for all OSery?
16:28:51 <ehird> I can't imagine dualbooting and having a good collection of apps with 80gb
16:29:01 <Deewiant> I have a 50G partition on which Vista lives — it's using 34
16:29:11 <Deewiant> Linux also has a 50G partition, it's using 21
16:29:19 <Deewiant> (And has way more apps installed)
16:29:22 <ehird> Maybe I generate more shit than you
16:29:44 <Deewiant> My data partition is using 459G currently
16:29:47 <ehird> Deewiant: ask $PKG_MANAGER how many packages lunix has?
16:30:09 <Deewiant> Of which 308 are dependencies and 309 not
16:30:21 <ehird> I dunno, 80GB seems stifling.
16:30:47 <Deewiant> Like said I'm essentially working with 100 and I've got 35 to spare
16:31:32 <Deewiant> If you're worried, get two ;-P
16:32:04 <ehird> Deewiant: Two would be more expensive than a bigger one, I imagine...
16:32:27 <ehird> Anyway, I'd have to RAID them
16:32:34 <Deewiant> There are many here with differing prices
16:32:44 <ehird> How much is the vertex
16:32:48 <Deewiant> SSDSA2MH160G1 and SSDSA2MH160G1C5
16:33:13 <ehird> that's the only difference? ;)
16:33:14 <Deewiant> I wonder what that 2.5 mm brings :-P
16:33:20 <Deewiant> ehird: According to the shop's product pages, yes
16:33:26 <Deewiant> Of course they don't have much info
16:33:36 <Deewiant> Amusingly the height is even in bold
16:33:47 <Deewiant> Like it's the most important thing :-P
16:34:19 <ehird> So £372 real money
16:54:22 <ehird> doing other things
16:58:51 <AnMaster> just I needed him due to IFFI..
17:24:32 -!- KingOfKarlsruhe has quit (Remote closed the connection).
17:24:32 <AnMaster> anyway someone tell ais when he is here next time that he need to pull from my darcs repo for ick since I had to change API of one of the functions he use in cfunge (IFFI should work with both old and new now)
17:24:51 <AnMaster> ("had to" as in part of code cleanup)
17:25:15 -!- sebbu has joined.
17:29:54 -!- oerjan has quit ("Sproing!").
17:31:07 -!- BeholdMyGlory has quit (Remote closed the connection).
17:35:06 -!- ais523 has joined.
17:35:42 <ais523> I've been getting C-INTERCAL running on clang
17:35:51 <AnMaster> ais523, pull from my darcs repo to get a fix
17:35:51 <ais523> AnMaster: what, again?
17:36:07 <ais523> that's the great thing about having a repo
17:36:09 <AnMaster> ais523, the new code checks cfunge api version and thus works on 0.4.0 and last
17:36:10 <ehird> somehow I don't think I'll get a totally fanless i7
17:36:16 <ehird> stupid intel and their stupid hot.
17:36:34 <AnMaster> ehird, AMD cpus are generally cooler, even the high end ones I hear
17:36:54 <ehird> AnMaster: the amd system meeting my requirements doesn't have the fanless cooler option but I could get my own cooler I guess
17:37:09 <Deewiant> ehird: Aren't you getting your own cooler anyway? :-P
17:37:24 <ehird> Deewiant: endpcnoise.com offers the ones I was considerng
17:37:37 <Deewiant> Oh, you're using some kinda prebuilt mess
17:37:46 <ais523> C-INTERCAL now builds on K&R C with unproto, and cross-compiles to ARM without trouble
17:37:46 <AnMaster> ehird, wouldn't it be simpler to get the components separate and build your own
17:37:51 <ehird> define 'mess', all of spcr's reviews are glowing
17:38:02 <ehird> AnMaster: no, endpcnoise gets it right apart from that
17:38:02 <ais523> it builds on both llvm-gcc and clang with a bit of build system fiddlery
17:38:09 <ehird> eg HD enclosure, acoustipak
17:38:32 <AnMaster> ais523, oh? So what about the wrong collect2 thingy
17:38:46 <ais523> AnMaster: I got around that different ways for the two builds
17:38:59 <ehird> I'm not sure about replacing the cooler myself anyway
17:39:05 <ehird> Me and thermal paste is a recipe for disaster
17:39:06 <ais523> on llvm-gcc I just added RANLIB=ranlib to the configure line, that builds two indexes for the .a files so that either native or llvm collect2 works
17:39:33 <AnMaster> ais523, um, but isn't the actual object file format different too
17:39:35 <ais523> on clang, I use CFLAGS=-emit-llvm LINK='llvm-ld -o $@'
17:40:06 <ais523> AnMaster: what do you mean by that?
17:40:23 <AnMaster> was thinking two things at once. heh
17:40:25 <ais523> that's the through-native build
17:40:32 <ais523> whereas clang is via-bytecode
17:40:37 <ais523> and builds to bytecode in the end
17:40:38 <AnMaster> ais523, also did you push your updated ick yet, since I pulled shortly before you joined, no new changes
17:40:52 <ais523> in the last couple of minuts
17:41:14 <ehird> I wonder if you can buy pre-watercooled sytsems
17:41:40 <AnMaster> ehird, no one would be insane enough to provide the warranty...
17:41:54 <ehird> the g5 mac pro was watercooled i think
17:41:57 <ehird> I guess apple are insane enough
17:42:36 <AnMaster> yeah I remember hearing about that
17:42:53 <AnMaster> ehird, I don't think all G5 were, just some of the final models of it.
17:42:55 <ehird> it wasn't called a mac pro
17:42:59 <ehird> AnMaster: power mac g5
17:43:17 <ehird> 2004 June: 90 nm DP 1.8, DP 2.0 and DP 2.5 GHz replace all previous models. The 2.5 GHz model is notable as the first major PC with liquid cooling included as stock.
17:43:23 <ehird> so it went a year without watercooling
17:43:24 <AnMaster> ehird, well I meant the model you can do more than just replace ram inside
17:43:31 <ehird> http://en.wikipedia.org/wiki/Power_Mac_G5
17:43:40 <ehird> from its second year on it had watercooling
17:43:56 <AnMaster> ais523, btw you may need to update pull path for cfunge itself
17:44:22 <AnMaster> ais523, some days ago it changed to http://rage.kuonet.org/~anmaster/bzr/cfunge/trunk (added trunk at end while reorganizing things)
17:44:29 <ais523> oh, the version I'm currently using is a release version, not dev
17:44:35 <ais523> and was downloaded via tarball
17:44:37 <AnMaster> ais523, well it should work too
17:46:11 <AnMaster> ais523, btw you may want to adjust IFFI a bit to take full advantage of the change.
17:46:40 <AnMaster> ais523, http://bzr.kuonet.org/cfunge/trunk/annotate/head%3A/doc/API_CHANGES
17:47:47 <AnMaster> ais523, fungespace_load_string was just a thin wrapper that did strlen() and passed it on to the same code cfunge used internally in the last few releases
17:48:04 <ais523> oh, handling embedded NUL
17:48:18 <ais523> I may modify the compilation technique for Befunge, in that case
17:48:34 <AnMaster> ais523, well atm it does strlen() if the newer API version is detected.
17:48:44 <AnMaster> but yeah you might want to replace that.
17:49:17 <AnMaster> ais523, point is that the external API supports it now too.
17:50:40 <AnMaster> ais523, the ick side has issues with embedded 0-bytes?
17:50:49 <ehird> maybe i'll just not cool anything and be careful with load :-D
17:50:53 <Deewiant> ais523: Mycology has an embedded NUL :-)
17:51:13 <ehird> quiet and cheap though ;)
17:51:33 <AnMaster> Deewiant, and internally cfunge handled this for ages. Just the external code for IFFI didn't handle it
17:51:36 <ais523> AnMaster: not really, but it doesn't store any info on the true length of the string
17:51:50 <ais523> so there's no way to record whether it continues past the NUL or not
17:51:54 <ehird> http://bilder.wibla.net/albums/monster/DSC_1820.sized.jpg
17:51:57 <ehird> Wwwwwwwwwwwowwwwwwwwwwww.
17:52:20 <AnMaster> Deewiant, anyway I suggest adding an embedded form feed, if you haven't already. I think it still breaks rc/funge, unless you contacted him about it or such.
17:52:30 <Slereah> What are you even going to do with 11TB?
17:52:43 <Slereah> I have like 1.3 and I still have plenty of space
17:53:23 <AnMaster> Deewiant, I think it continues loading rest of program like if it was trefunge: incrementing z
17:53:30 <ehird> I'm just going to have 130gb+1TB, heh.
17:53:37 <Deewiant> AnMaster: I guess it's a good thing to test anyway
17:53:38 <Slereah> But then again, it's under 1000
17:53:45 <Slereah> So it might actually be 11.
17:54:10 <ehird> Slereah: Well, that's real-bytes.
17:54:16 <ehird> HDs use marketing-bytes.
17:54:26 <AnMaster> Deewiant, it certainly doesn't ignore it like newline is ignored in unefunge at least, nor does it store it literal into funge space and just continue loading.
17:54:34 <ais523> marketing uses metric kilobytes, not binary kibibytes
17:54:35 <AnMaster> Deewiant, there is some test case included with cfunge for it btw
17:54:50 <ehird> "Silencing your scroll mouse."
17:55:01 <ehird> now that's just ridiculous
17:55:04 <ais523> I like it when I can hear my mouse and keyboard
17:55:14 <ehird> "Not because the silencing effect but because the feel the mouse gets That it gets more silent is just a bonus! "
17:55:16 <AnMaster> ehird, I had one ages ago that was very loud, but that was an old PS/2 logitech one
17:55:19 <ehird> well, okay, that's slightly less ridiculous
17:55:20 <ais523> otherwise it's less obvious whether I actually managed to press the button or not
17:55:27 <ehird> ais523: you'd like a model m/das keyboard
17:55:35 <ehird> BAM BAM BAM WORLD ENDING CLATTER NUCLEAR REACTION CRASH BANG
17:55:38 <ais523> the model m keyboards are world-famous as being the best ever
17:55:45 <ais523> although I've never seen one
17:55:47 <ehird> but I broke one of the arrow keys
17:55:50 <ehird> and the power supply fucked up
17:56:02 <AnMaster> ehird, the keyboard had a separate power supply?
17:56:39 <AnMaster> I thought keyboards used power from PS/2 (or for modern ones USB, which wouldn't be relevant in this case)
17:56:51 <ehird> ah, the mod makes the scrollwheel smooth
17:57:05 <ehird> I meant the ps2 cable
17:57:17 * AnMaster is typing on a PS/2 keyboard atm
17:57:35 <ehird> ais523: http://upload.wikimedia.org/wikipedia/commons/5/57/ModelM.jpg
17:57:36 <AnMaster> works like a charm, after many many years
17:57:50 <ehird> I'd give you a video but it's youtube
17:58:03 <ehird> http://www.youtube.com/watch?v=CoWXuVdlKZw&fmt=18 anyway
17:58:05 <AnMaster> ehird, used one once. Very nice feeling
17:58:13 <ehird> (http://www.youtube.com/watch?v=SUzReqjyNfQ&fmt=18 das keyboard)
17:58:19 <ais523> well, this keyboard is part of my laptop
17:58:21 <ehird> AnMaster: hard to type on
17:58:55 <ehird> weak fingers AnMaster
17:58:58 <AnMaster> <ehird> http://www.youtube.com/watch?v=CoWXuVdlKZw&fmt=18 anyway <-- fast typer though
17:59:33 <AnMaster> ehird, did it even make sense or was it just random garbage on screen
17:59:41 <ehird> it looked like senseful typing
17:59:48 <ehird> it was meant as an example of real world typing too
17:59:48 <AnMaster> it was almost movie typing (see tv troupes)
17:59:53 <ehird> the guy says he's used a model m for 15 years, IIRC
18:00:08 <ais523> given that they're so famous, why did IBM stop making them?
18:00:33 <ehird> ais523: they weren't famous when they stopped
18:00:38 <ehird> also: expensive to make
18:00:41 <ehird> not everyone likes the loudness
18:00:49 <ehird> and they're VERY BIG
18:00:55 <AnMaster> I would like the feeling without the loudness
18:01:04 <ehird> a bit less feeling
18:01:06 <ehird> AnMaster: I mean, non-keys
18:01:10 <ehird> it has huge padding around it
18:01:21 <AnMaster> ehird, true, but that is not much of an issue
18:01:28 <ehird> too big for a desk
18:01:33 -!- tombom has quit ("Peace and Protection 4.22.2").
18:01:35 <AnMaster> ehird, I would prefer a good hand rest
18:01:41 <ehird> I rest my hands on the desk :P
18:02:01 <ehird> http://www.youtube.com/watch?v=H-WdHYoHEDk&NR=1&fmt=18 ← hthis is what a model m sounds like really
18:02:04 <ehird> the other ones are too loud
18:02:12 <ehird> the twangy sound sucks
18:02:45 <ehird> http://www.youtube.com/watch?v=befAQ6BVNGM&NR=1&fmt=18
18:02:49 <Deewiant> I'm annoyed that pretty much all semi-good keyboards don't come in 105-key layouts :-/
18:03:31 * ehird browses cherry.com's range of keyboards
18:03:50 <Deewiant> Only in 104 or occasionally in 108
18:04:03 -!- M0ny has quit ("PEW PEW \(#_é)/").
18:04:08 <AnMaster> Deewiant, which is the 105 one
18:04:31 <AnMaster> Deewiant, which is the extra key
18:05:04 * AnMaster has whatever the normal Swedish full size keyboard layout is called
18:06:07 <Deewiant> 104 is http://www.cooltoyzph.com/image/US_Keyboard_layout.jpg
18:06:54 <AnMaster> Deewiant, what about F-keys and so on
18:07:07 <AnMaster> aren't they counted as part of the layout
18:07:24 <AnMaster> Deewiant, then what about laptop keyboards with the fn thingy instead of numerical keyboard
18:07:35 <Deewiant> They're not over 100 keys now are they
18:07:47 <AnMaster> well I never heard them called 80-something
18:07:53 <ehird> Hello, world! I am typing to show how fast my typing is and this is important for me so that I can show how fast my typing is, actually I'm doing it to show the noise level of my keyboard but that's how it goes isn't it? Yes indeed it is and thusly I end this typing (can you tell I'm recording? I bet you can. Blah blah blah blah qwerty.) Fake error. Etc.
18:08:37 <AnMaster> ehird, was that was the person typed or what?
18:08:44 <ehird> It was what I was typing.
18:10:09 <Deewiant> AnMaster: Anyway, http://upload.wikimedia.org/wikipedia/commons/thumb/b/bc/KB_Japanese.svg/800px-KB_Japanese.svg.png is 108-key
18:10:58 <AnMaster> Deewiant, what I really want is 105 but with separate Meta, Super, Alt an Ctrl keys on each side
18:11:17 <AnMaster> well one alt could be altgr I guess
18:11:55 <AnMaster> Deewiant, some programs act as if meta and alt were same, and some programs act as if they were separate. emacs is an example of the former
18:12:28 <Deewiant> Be careful that you don't run out of room for a space bar :-P
18:12:56 <ais523> AnMaster: emacs knows the difference between meta and alt, but if you don't have a meta it maps alt to meta
18:13:00 <AnMaster> Deewiant, I would make up for it by not having any Fn. If laptops manage to fit it in...
18:13:36 <AnMaster> xmodmap didn't manage to do the trick in X for me, and of course it doesn't solve it outside X at all
18:14:44 <ais523> you can put control where capslock is
18:14:53 <ais523> I don't, but lots of AnMaster-attitude people do
18:15:11 <Deewiant> I put backspace where capslock is
18:15:18 <ais523> Deewiant: what do you put where backspace is?
18:15:20 <AnMaster> well I don't often use capslock so maybe
18:15:42 <ehird> ssd\sad\'a\s'\as\sad\'asd\asd\'a\sd'\as'da[]f;l[wlfpawfgkqfjeiorafgioafjlaf jklbfjlqwfjnkjtklqwrhjakwfhjklrjkrf krjlzfaklgdsjklg jdlkjhdltk;jnfkrj lafhdkr hlgdfkjlg fdjql gios;hgl j;l jfogjh o;s jgio;e jpgos;oi jrg;js;gj ;oj g;l lwk el; gj;klgj ar;ogj gkom ;ops ps0'gk eorkg [p'sjfg ag
18:15:44 <AnMaster> anyway how do you do this outside X
18:15:45 <ehird> Super fast typing.
18:16:16 <Deewiant> The stupid groove on the caps lock key annoys me very much, I might add
18:16:18 <AnMaster> or even inside X (so that it works)
18:16:22 <ehird> I have my model m lying around
18:16:30 <AnMaster> Deewiant, ah yes, why does it have that
18:16:53 <ais523> PS/2 isn't hotpluggable safely
18:16:54 <Deewiant> AnMaster: Presumably since people were pressing caps lock accidentally... so instead of solving the problem they make the key harder to press
18:16:59 <ehird> http://www.dansdata.com/images/clicky2/ergo1280.jpg Ergoclick
18:17:09 <AnMaster> <ais523> PS/2 isn't hotpluggable safely <-- I know, it caused system resent when I tried it once
18:17:15 <AnMaster> Deewiant, and what does that mean exactly
18:17:26 <ais523> AnMaster: you were lucky, in theory it can burn out the motherboard
18:17:29 <Deewiant> Where sec = 10 mins actually ->
18:17:42 <AnMaster> ais523, it was on an old computer
18:18:02 <AnMaster> <Deewiant> AnMaster: n-key rollover <-- what does that mean, any idea ais523
18:18:28 <ehird> http://en.wikipedia.org/wiki/Rollover_(key)
18:18:35 <ehird> Certain high-end keyboards have "n-key rollover". This means that each key is scanned completely independently by the keyboard hardware, so that each keypress is correctly detected regardless of how many other keys are being pressed or held down at the time. [3]
18:18:44 <ehird> yeah cuz i press 50 keys at once all the time
18:18:50 <AnMaster> well... you don't get that in PS/2
18:18:56 <ais523> well, I actually did some experiments on that a while back
18:18:59 <ehird> the most you need is 5 keys held at once
18:19:10 <ais523> most keyboards can distinguish either 2 or 3 keys at once, depending on which they are
18:19:11 <ehird> windows-menu-alt-control-shift-<key>
18:19:17 <ais523> and any number of modifier keys
18:19:22 <AnMaster> I have a PS/2 and it can't handle more than like 2 normal keys, and a few modifiers
18:19:50 <AnMaster> and it can't handle shift, up left
18:19:58 <AnMaster> which was needed in some game (forgot which)
18:20:17 <ehird> i love the idea of getting an expensive silent pc then using a model m
18:20:38 <fizzie> oklopol: The wrong part is that the function ends with a commented "more_stuff();" call and a fixed "return null;", while it should return the move the bot wants to make.
18:20:43 <AnMaster> ehird, that is a rather different noise, not a constant noise in the bg
18:25:00 -!- Gracenotes has quit (calvino.freenode.net irc.freenode.net).
18:25:00 -!- ineiros has quit (calvino.freenode.net irc.freenode.net).
18:25:02 <AnMaster> <ehird> http://www.dansdata.com/images/clicky2/ergo1280.jpg Ergoclick <-- interesting. How does it work
18:25:13 <ehird> like any other ergonomic keyboard
18:25:31 <AnMaster> ehird, you take it apart in two parts?
18:26:13 <AnMaster> what is the round thing at the top
18:28:28 <fizzie> My hypervisor at work has this split-at-the-middle in-two-parts keyboard, and self-built plywood-or-something meter-long sticks-of-sorts taped into them, so that he can just keep his hands down on each side of his chair, and the keyboards are sort-of like ___/H\___ where H is the chair, ____ is the floor, / and \ are the keyboard halves with the sticks, and this picture is from the front (or behind) the chair.
18:28:31 <ehird> this model m is love
18:29:59 <Deewiant> AnMaster: You /need/ PS/2 for NKRO
18:30:14 <Deewiant> The USB protocol can only handle 6 keys at once, it's an arbitrary limitation
18:30:30 <Deewiant> So even if your keyboard can handle it your OS can't unless you write your own keyboard driver
18:30:35 -!- ineiros has joined.
18:31:49 <ehird> no ps 2 on this computer
18:31:49 -!- neldoreth has quit ("Lost terminal").
18:31:51 <ehird> so I can't use my model m
18:31:57 -!- neldoreth has joined.
18:32:17 <ais523> what about through a PS/2 to USB adapter?
18:32:21 <ehird> pet peeve: no Windows key or equivalence
18:32:23 <ehird> ais523: don't have on
18:32:28 <ehird> and the computer shop is closed
18:32:48 * ehird bashes the model m excessively
18:33:44 <ehird> he was talking to me
18:34:50 <ehird> from rolling my hands like a maniac I conclude that the model m can support typing speeds up to 500wp
18:36:48 -!- Asztal_ has joined.
18:37:56 * ehird records the keyboard-rapage
18:40:28 <ehird> aw shit i snapped a nail
18:44:39 <AnMaster> fizzie, keyboard on the side sounds crazy
18:45:12 <ehird> what, one keyboard piece at each side?
18:45:33 <ehird> AnMaster: citation:
18:45:41 <ehird> http://jwz.livejournal.com/493388.html
18:45:48 <fizzie> I don't see how it wouldn't be ergonomic; certainly it's more natural to keep your hands on your sides than in front of you in a dog-begging-for-food pose.
18:46:37 <fizzie> The ones he uses aren't bolted on the chair, but other than that I guess the principle is rather similar.
18:46:37 <AnMaster> I thought you mean arms hanging straight down the sides and the keyboard in vertical position
18:47:34 * ehird wonders alterantives to ps2→usb conversion
18:48:57 <AnMaster> ais523, is serial cable safe to hot plug assuming device at other end is turned off? Just wondering
18:49:15 <AnMaster> (since I assume turning on the device later is safe, it wouldn't make sense if it wasn't
18:49:22 <ais523> in fact, it's safe to hot-plug even if the device is turned on
18:49:28 <ais523> RS232 is very tolerant to all sorts of things
18:49:52 <ais523> I don't know whether they're meant to work in theory, but IME they do in practice
18:50:25 <AnMaster> same, and it saved me a few times when some headless computer oopsed
18:50:45 <AnMaster> actually not oopsed last time, just silent death of anything network related
18:52:54 <AnMaster> <ehird> http://jwz.livejournal.com/493388.html <-- that's a CRT at the top of the pic isn't it
19:03:10 <ehird> http://www.evertype.com/standards/csur/seuss.html
19:04:00 * ais523 ponders the idea of C-INTERCAL as a kernel module
19:04:19 <ais523> so you can cat INTERCAL programs to /proc/compile-intercal, and read the equivalent C back from it
19:04:44 <AnMaster> ais523, does darcs have issue with moving one file and then adding a different file with the same name as the old name of the first file being done in a single commit?
19:04:59 <ais523> AnMaster: use darcs mv to move the one file
19:05:11 <AnMaster> ais523, I mean both being done in one commit
19:05:14 <ais523> darcs has no trouble with that
19:05:17 <ais523> as long as you told it about the move
19:05:26 <AnMaster> like darcs mv a b; touch a; darcs add a
19:05:32 <ais523> yep, that should be fine
19:05:43 <AnMaster> ais523, right, svn seems to get confused by it
19:06:11 <AnMaster> ais523, well I did it recently in bzr with no issues either for bzr or me
19:06:18 -!- Alicce has joined.
19:07:28 <ehird> esolangs or esoterica
19:07:33 <ais523> fungot: give me some nonsense
19:07:35 <fungot> ais523: installing now. :) just had to check the source addr there. since i didn't bother packing any necessities ( sleeping bag or anything).
19:07:36 <ehird> if it's the latter you've come to the wrong place
19:07:54 <ais523> wow, fungot that almost made sense...
19:07:55 <AnMaster> ais523, was when double including a c file (named as .h) to create two versions of the code.
19:07:55 <fungot> ais523: everyone invents the game when learning about portal culling your e-mail address,
19:08:38 <ais523> and yes, if you play too much Portal your email stops working, that's why millions of people all across the world have made their own versions
19:09:05 -!- Gracenotes has joined.
19:09:13 <ehird> i thought he meant everyone invents the think-about-it-and-lose gmae
19:09:19 <ehird> as an alternative to Portal, maybe
19:09:37 <ehird> Gracenotes: ps, you lost
19:09:53 <ais523> ehird: you know, that connotation never crossed my mind
19:09:59 <ais523> my The Game defences seem pretty good
19:10:00 <AnMaster> ais523, http://rafb.net/p/3ptqcK43.html
19:10:01 <ais523> even though I don't play it
19:10:12 <ehird> $ bzr log -v -r656
19:10:13 <ehird> ------------------------------------------------------------
19:10:16 <ehird> timestamp: Fri 2009-04-03 23:20:20 +0200
19:10:21 <ehird> Add support for multiple hash libraries.
19:10:25 <ehird> lib/libghthash_fspace/cfunge_mempool.c
19:10:27 <ehird> lib/libghthash_fspace/ght_hash_table.h
19:10:29 <ehird> lib/libghthash_fspace/hash_functions.c
19:10:31 <ehird> lib/libghthash_fspace/hash_table.c
19:10:35 <ehird> lib/libghthash_fspace/cfunge_mempool.c => lib/libghthash_fspace/cfunge_mempool_priv.h
19:10:35 <AnMaster> ais523, bzr didn't like me trying to renaming the directory they were in at the same time
19:10:37 <ehird> lib/libghthash_fspace/ght_hash_table.h => lib/libghthash_fspace/ght_hash_table_priv.h
19:10:38 <ehird> lib/libghthash_fspace/hash_functions.c => lib/libghthash_fspace/hash_functions_priv.h
19:10:41 <ehird> lib/libghthash_fspace/hash_table.c => lib/libghthash_fspace/hash_table_priv.h
19:10:45 <ehird> lib/libghthash_fspace/cfunge_mempool.h
19:10:47 <ehird> src/funge-space/funge-space.c
19:10:48 <ehird> lib/libghthash_fspace/cfunge_mempool_priv.h
19:10:51 <ehird> lib/libghthash_fspace/ght_hash_table_priv.h
19:10:52 <AnMaster> like lib/libghthash_fspace -> libghthash
19:10:53 <ehird> lib/libghthash_fspace/hash_functions_priv.h
19:10:54 <ehird> lib/libghthash_fspace/hash_table_priv.h
19:10:57 <ehird> (my new anti-rafb strategy)
19:11:02 <ais523> ehird: I hope pasting AnMaster's entire commit message in-channel was an accident
19:11:10 <ehird> 19:10 ehird: (my new anti-rafb strategy)
19:11:14 -!- Alicce has left (?).
19:11:19 <AnMaster> ais523, it refused to do that with modified files in the directory
19:11:28 <AnMaster> ehird, well let me rafb my xorg.conf...
19:11:44 <AnMaster> ehird, it's 480 lines with comments
19:11:56 <ais523> AnMaster: go for it, probably it's a good idea to not put it somewhere permanent because it isn't of permanent interest
19:11:59 <AnMaster> well it is you who spammed then not me
19:12:06 <ehird> you spammed my inner logreader.
19:12:09 <ais523> hmm... wait for lament to get back first, though, so we can kick you
19:12:19 <ehird> lament never kicks anyone just threatens <3
19:12:20 <AnMaster> ais523, you mean kick ehird indeed
19:12:24 <ehird> (now he'll kick me to prove me wrong)
19:12:44 <ais523> well, I'm pretty sure fizzie /has/ kicked ehird in the past
19:12:46 <ehird> fizzie AnMaster's provoking me kick him.
19:12:50 <ehird> ais523: I requested once
19:12:57 <ehird> ais523: Dunno if he's done it again
19:13:21 <AnMaster> I just pastebinnned a file, not my fault ehird decides to paste it in channel then
19:13:32 <ehird> MOD SUCKUP FIGHT FIGHT FIGHT
19:14:13 <AnMaster> http://rafb.net/p/WkUSTs47.html
19:15:04 <ais523> AnMaster: why not take screenshots of all the relevant GUI configuration tools instead/
19:15:13 <ehird> rafb doesn't take screenshots
19:15:22 <ehird> and most image hosters leave images around
19:15:22 <AnMaster> ehird, there are other temporary services
19:15:32 <ais523> ehird: I was busy trying to combine two flamewars...
19:15:55 <AnMaster> I will convert the screenshot to a dataurl
19:18:19 -!- neldoret1 has joined.
19:18:43 <ehird> yes, because we didn't talk about magick.
19:18:44 <ais523> maybe she came to the wrong channel
19:18:59 <AnMaster> he or she, you can't know on irc
19:19:09 <ehird> alice is a female name.
19:19:20 <ais523> AnMaster: you can have a female nick but a male name
19:19:24 <ais523> and "she" is often correct then
19:19:30 -!- ehird has changed nick to VANESSA.
19:19:32 <ais523> in #nethack, I use people's character genders when determining pronouns
19:19:33 <VANESSA> AnMaster: hostname was alice
19:19:46 -!- VANESSA has changed nick to ehird.
19:20:22 <AnMaster> ais523, that would different between different games
19:20:59 <oklopol> please refer to me as an it from now on
19:21:06 <AnMaster> ais523, and how do you handle a male chaotic valkyrie (damn amulet of changing...)
19:21:11 <ehird> it says we should call it an it
19:21:15 <ehird> should we call it an it?
19:21:35 <ais523> AnMaster: why does the chaotic matter?
19:21:45 <AnMaster> ais523, no, but that was another mishap
19:21:46 <ais523> also, ais523 refers to themself as "they" whenever possible
19:22:06 <ehird> a friend suggested "fat tub of lard" as a generic pronoun
19:22:18 <ehird> fat tub of lard says we should call fat tub of lard a fat tub of lard
19:27:57 <oklopol> racecar is a palindrome and no one told me!
19:28:28 <oklopol> well that would've been one expensive book
19:28:56 <AnMaster> oklopol, make a script which searches for them all and lists that
19:30:35 <oklopol> i think that's a very hard problem, algorithmically
19:31:33 * oklopol write palindrome checker in clue
19:31:41 -!- neldoreth has quit (Read error: 113 (No route to host)).
19:32:54 <ais523> oklopol: do you have a clue interp yet?
19:33:24 <oklopol> no, just increasingly more confidence in the concept
19:33:42 <oklopol> although i still feel like it could explode any second
19:34:10 <oklopol> that some things simply require either somehow brute forcing the exact definition, or aren't doable at all.
19:34:53 <oklopol> brute forcing as in forcing an exact definition by making all expressions in the body into separate functions
19:35:20 <ehird> $ perl -ne'$_ = lc $_; chomp; print $_, "\n" if $_ eq reverse($_)' /usr/share/dict/words
19:35:53 <ais523> my /usr/share/dict/words isn't all that good
19:37:03 <ehird> $ perl -ne'$_ = lc $_; chomp; print $_, "\n" if $_ eq reverse($_) && length($_) > 3' words
19:37:22 <ehird> $ perl -ne'$_ = lc $_; chomp; print $_, "\n" if $_ eq reverse($_) && length($_) > 3' words | uniq
19:37:27 <ehird> to handle things only differing in case
19:39:00 <ais523> riff-raffless isn't a palindrome
19:39:08 <oklopol> http://www.vjn.fi/pb/p255166222.txt
19:39:14 <oklopol> that is one goddamn hard language
19:39:39 <oklopol> but i love the syntax, and the concept
19:39:49 <ais523> oklopol and I have been working on it secretly for ages
19:39:56 <ais523> oklopol's been doing the work, I've been shouting encouragement
19:39:58 <ehird> he's talked about it a lot
19:40:19 <ehird> http://www.vjn.fi/oklopol/clue.txt
19:40:21 <ehird> i asked about this
19:40:23 <oklopol> when i suddenly decided to publish it
19:40:54 <oklopol> yeah okay, but anyway, mostly it's been secret, since i felt like changing my vaporware image :P
19:41:10 <ehird> http://esolangs.org/wiki/Talk:%22The_most_important_thing_in_the_programming_language_is_the_name._A_language_will_not_succeed_without_a_good_name._I_have_recently_invented_a_very_good_name_and_now_I_am_looking_for_a_suitable_language.%22
19:41:10 <oklopol> but, meh, taking too long, need the small fame of vapor ;)
19:41:14 <ais523> I have the best mix, I think, real langs with real implementations /and/ vaporware
19:41:21 <ais523> ehird: yes, it was spammy
19:41:26 <ais523> and lots of people didn't want it there
19:41:28 <ehird> ais523: it was being worked on :-(
19:41:30 <ais523> especially as it had no useful content
19:41:39 <ais523> and no it wasn't, obviously
19:41:43 <oklopol> ais523: i have real implementations and real langs, although very few
19:41:44 <ais523> you said a few minutes, it was there for weeks
19:41:59 <ehird> it was a bung-attractor for hype and ridicule that would simmer it down into a language
19:42:05 <oklopol> i'm such a perfectionist when it comes to languages
19:42:09 <ehird> now i have to recreate it to restart the naturamachine
19:42:16 <ais523> oklopol: well, I can be too
19:42:40 <ais523> the trick is to release an imperfect lang and have a vaporware version which is meant to be perfect
19:42:43 <ais523> which is why it's vaporware
19:43:16 <AnMaster> oklopol, any interpreter for it
19:43:27 <AnMaster> also I rememeber ais talking about a lang like that
19:43:30 <oklopol> AnMaster: no, but i'll try speccing on sunday
19:44:00 <AnMaster> both is brute forcing definition
19:44:35 <oklopol> the concept of guessing programs is a very general concept
19:44:57 <oklopol> it just hasn't been done a lot, so just doing it twice seems like the same thing
19:45:17 <AnMaster> I can't think of any other langs like that
19:45:31 <oklopol> of course the reason it hasn't been done is it's a really stupid idea pretty much anywhere except here :P
19:46:58 <ehird> oklopol: so how does clue infer
19:47:03 <ehird> please tell me it uses an ai
19:48:14 <oklopol> i try to put my languages between "pure declarativity", that is, giving absolutely no clue about what to do, and "imperativity", where you tell the interp exactly what to do, kind of a bad term, since haskell would be imperative too
19:48:21 <oklopol> that's the least explored area afaik
19:48:43 <ehird> make a language based on inferring infers
19:49:05 <oklopol> ehird: basically you can just brute force the bodies, there should be no need for making an intelligent implementation.
19:49:30 <oklopol> if you need speed, optimization must be done manually, it's kinda like prolog in this sense
19:50:01 <ehird> make a lang based on optimizing a program that does everything
19:50:12 <ehird> dangerous optimizations, by reducing everything that you don't want to do into a nop
19:50:33 <oklopol> i have many ideas about languages where you teach the deduction engine how to infer functions based on io-pairs, but that's still very experimental
19:51:11 <oklopol> this is more about giving the correct io-pairs
19:52:02 <oklopol> giving just random palindromes and non-palindromes will not work, even if the body of a palindrome checker was accidentally generated, it would be discarded.
19:52:25 <oklopol> because it doesn't work by generating bodies and then trying them.
19:53:24 <oklopol> it works by generating code that links all clues on level 3 to a clue on level 2, then running those bodies, making sure you reach a clue on level 1, then level four is tried for checking correctness
19:53:53 <oklopol> the running phase runs the whole recursion, some predefined amount of steps, until base cases, or failure
19:55:24 <oklopol> it works as long as your functions are really simple, but the beauty is unlike with a system that requires intelligence from the interp, here your functions just need to be simple in function body, not in semantics
19:55:49 <ehird> so you can't do complex stuff
19:56:03 <oklopol> so enough modularity is necessary, and thus also guaranteed, pretty much making this a perfect language
19:56:21 <oklopol> ehird: you can, i just said it in a very philosophical and mysterious way.
19:56:22 <ais523> the language does not allow any large complex chunk
19:56:32 <ais523> everything has to be broken down into small chunks
19:56:33 <ais523> but they can have complex interactions
19:56:39 <ehird> oklopol: so do you have an interp :P
19:56:52 <ais523> the way in which they interact also has to be simple, but it can be structured over several layers so it becomes complex overall
19:56:57 <oklopol> spec on sunday, imp on X-day :)
20:01:00 <oklopol> the problem is while it can do something like a static complex splicing and dicing of the arguments easily, it is very sucky at having any kinds of actual control structures
20:01:19 <ehird> oklopol: how about making examples have 'aliases' like
20:01:27 <ehird> "abc"->poop('ab')+poop('bc')
20:01:40 <ehird> where it acts just like you put the result in, but uses them as hints, sort of
20:03:00 <oklopol> try explaining that again, maybe
20:03:36 <ehird> okay in the example sections instead of just input->constant you can do input->expr, now this expr can call other functions, and it just subtitutes the value BUT it remembers the structure you gave it, so that its end implementation will call the functions etc in the same way
20:04:00 <ehird> so "abc"->poop('ab')+poop('bc') would be "abc"->7, but the end function coming out of it would make those calls to poop and + when you give it that input, and others, when it generalizes
20:04:01 <oklopol> ehird: do realize i'm a bit wary of your suggestions, you usually have ideas i don't really like that much but which you consider much better than mine ;)
20:04:22 <ehird> this gives you function calls and control structure and etc while still not breaking the inferring stuff
20:05:06 <oklopol> actually i have that already, as syntactic sugar
20:05:16 <ehird> but mine takes it beyond that
20:05:23 <ehird> since it remembers the example structure and applies it to its generated function
20:06:26 <oklopol> what do you mean applies it to its generated function
20:06:46 <ehird> oklopol: let's say we have a reverse() function which reverses a string
20:07:06 <ehird> "abcd"->reverse("ab")+reverse("cd")
20:07:16 <oklopol> interesting reverse, but go on
20:07:26 <ehird> "abcd"->reverse("cd")+reverse("ab")
20:07:32 <ehird> "xyzf" -> reverse("zf")+reverse("xy")
20:07:34 <oklopol> less interesting, but go on
20:07:35 <ehird> this would be the same as
20:07:51 <ehird> it would remember the code structure of your example
20:07:55 <ehird> when it generates the final function
20:07:58 <ehird> it would call reverse and +
20:08:05 <ehird> instead of just treating them as constants
20:08:13 <ehird> then generalize this to multiple examples with diff structures etc
20:08:21 <ehird> can't do it in all cases, but it's a fun unification thingy
20:11:13 <oklopol> clue's level system is basically for doing that, except it doesn't rely on somehow extending the pattern
20:11:37 <oklopol> i mean i get it as in i understand it, probably i don't understand it well enough for you to acknowledge i understand it, as usual
20:11:50 <ehird> so I take that means you don't like it
20:12:04 <oklopol> i like it, i just feel like that's the problem and clue is the solution :)
20:12:18 <oklopol> i mean for the "generalize this pattern" part
20:12:22 <ehird> oklopol: semantically it's identical modulo side effects
20:12:35 <ehird> oklopol: it just lets you optimize with control structures etc without having a separate layer for that
20:12:37 <oklopol> basically that'd just be a hint for the body?
20:12:54 <ehird> it's purer than having a separate thing for optimizing the implementation imo
20:13:31 <oklopol> err clue doesn't have that
20:13:41 <ehird> if you want to optimize
20:13:53 <ehird> is palindrome ~ take first, take last, drop first, drop last
20:13:58 <ehird> that seems very non-example to me
20:14:13 <oklopol> no no i meant you just need to know what order to give examples in, and stuff, to make it fast.
20:14:22 <ehird> "is palindrome ~ take first, take last, drop first, drop last"
20:14:30 <oklopol> why it's the function bag!
20:14:34 <ehird> also, even so, I like my thingy
20:14:50 <oklopol> you need to tell it what it's allowed to use for the body
20:15:05 <ehird> it should figur that out
20:15:18 <ehird> yeah it should, and it fits in with my idea too
20:15:21 <ehird> since it can use what you use
20:15:24 <ehird> if you need to guide it
20:15:29 <ehird> but apart from that it should figure it out
20:16:55 <oklopol> ehird: can you write palindrome in yours?
20:17:19 <ehird> is palindrome :: "saippuasammakkokokkammasauppias">true
20:17:20 <ehird> :. "abba">true, "acba">false, "abbc">false, "abcba">true
20:17:22 <ehird> : "bb">true, "cb">false, "bcb">true
20:17:24 <ehird> . "c">true, "">true
20:17:28 <ehird> it'd take 6 billion years to find a correct impl
20:17:30 <ehird> so you can forget that
20:19:14 <Asztal_> are the ::, :., :, and . significant, or just decorative?
20:19:25 <oklopol> Asztal: significant, levels
20:19:43 <ais523> heh, it's onespot, twospot, threespot, fourspot
20:20:13 <ehird> is palindrome :: "saippuasammakkokokkammasauppias">reverse("saippuasammakko")=="saippuasammakko"&&reverse("k")=="k"&&reverse("okkammasauppias")=="okkammasauppias"
20:20:13 <Asztal_> so it should satisfy the first level before moving onto the next one? Sort of like a prune for obviously-bad implementations?
20:20:14 <ehird> :. "abba">reverse("abba")=="abba", "acba">reverse("acba")=="acba", "abbc">reverse("abbc")=="abbc", "abcba">reverse("abcba")=="abcba"
20:20:17 <ehird> : "bb">true, "cb">false, "bcb">true
20:20:19 <ehird> . "c">true, "">true
20:27:29 <oklopol> if reverse is defined, then palindrome only requires two level 1 clues
20:27:50 <ehird> I'm just saying that mine eliminates the ?
20:28:47 <oklopol> right. people have a tendency to want to drop it.
20:29:23 <ehird> ais523: BAND WITH ME
20:29:41 <ais523> ehird: I wasn't paying attention
20:29:48 <oklopol> but he wanted to take it down the purity lane, you seem to want a bit more concrete
20:30:11 <ehird> imo mine increases purity
20:30:15 <ehird> and also serves as optimization
20:30:46 <oklopol> whellll, i disagree, but you're welcome to make your own langer
20:31:32 <oklopol> yours is just a half-baked aardappel ;;;;;)
20:33:51 <ehird> BUT WHO KNOWS BUTTS?
20:34:18 <oklopol> he who knows butts, does not be know to have one.
20:35:08 <oklopol> well, gotta be going now, i've successfully pissed away the whole 10 or so hours i've been awake
20:38:47 -!- psygnisfive has quit (Read error: 104 (Connection reset by peer)).
20:39:06 -!- psygnisfive has joined.
20:41:58 <AnMaster> and the next few lines make no sense to me either...
20:44:26 <AnMaster> as for aardappel it can't be just "apple"
20:46:03 <AnMaster> for both potato and potatoe I guess.
20:46:11 <AnMaster> well I don't know for the latter
20:46:58 <AnMaster> psygnisfive, it looked like you were saying potatoe == potatis
20:47:23 <AnMaster> psygnisfive, <AnMaster> that looks like typoed potato <psygnisfive> nah
20:47:40 <psygnisfive> oh, i thought you were talking about something else when you said that
20:50:34 * AnMaster wonders how sane upgrading between major X releases while it is running is...
20:50:59 * ehird makes Squeak look nice
20:51:36 <AnMaster> xorg 1.3.0.0 -> 1.5.3 while xorg and KDE is running
20:51:37 <ais523> AnMaster: you'd probably need to restart X afterwards
20:51:49 <AnMaster> ais523, yes but will stuff break horribly while it is going say
20:51:58 <ais523> but I've upgraded Ubuntu from major release to major release before now with X running
20:52:06 <ais523> and the version of X probably changed in at least one of those
20:52:13 <AnMaster> ais523, say I want to start a new app using X while it has recompiled half of the libraries
20:52:45 <ais523> in an existing X session, or a new one?
20:52:45 <ehird> this squeak could look passable if you gave it native window creation
20:53:11 <AnMaster> ehird, that would be less closed world
20:53:19 <ehird> Pharo is less closed world.
20:54:00 <ehird> http://www.pharo-project.org/home
20:54:16 <AnMaster> ehird, well it looks like it has native OS X windows
20:54:25 <AnMaster> but does it look as good on *nix
20:54:31 <ehird> just lookalikes, I said
20:54:33 <ehird> there are other themes
20:54:39 <ehird> AnMaster: IT DOESN'T USE THE OS X THEME EITHER
20:54:44 <ehird> it's just a theme made to look like os x
20:55:07 <AnMaster> I hope Apple didn't copyright the GUI...
20:55:38 <ais523> it seems Apple patented automatic updates
20:55:50 <ais523> stupid software patents FTW?
20:56:01 <AnMaster> ais523, antivirus apps had it for way longer
20:56:11 <ehird> ais523: no, it's different
20:56:21 <ehird> i don't like apple any more and even I can see through the hyperbole
20:56:23 <ehird> yes it's a shit patent
20:56:29 <ehird> no it's not a blanket patent on everything that updates itself
20:56:37 <ehird> AnMaster: of course they did
20:56:41 <ehird> but they lost in court
20:56:59 <ehird> [which is bullshit, design is one of the few things that copyright can probably reasonably apply to]
20:57:03 <AnMaster> <ehird> AnMaster: of course they did <ehird> but they lost in court <ehird> in the 80s <-- who wut
20:57:14 <ehird> 20:55 AnMaster: I hope Apple didn't copyright the GUI...
20:57:42 <ais523> well, it's a patent on automatic updates without notifying the user
20:58:00 <AnMaster> ais523, prior art: some backdoors
21:01:07 <ehird> Monticello is nice
21:01:54 <Gracenotes> ehird: actually, no, I didn't lose when you said I did
21:02:10 * Gracenotes was away with college shopping and walking the dog
21:03:51 * AnMaster should make a bot that prints such a message at random times in this channel
21:04:50 <Gracenotes> no, people would be under the constant shadow of ulostgaemlol bot
21:05:07 <Gracenotes> and so reminding them of a situation they already knew doesn't have much effect
21:05:22 * ais523 wants to get a widespread portability survey of C-INTERCAL done
21:05:50 <ais523> I know it works on: gcc, bcc (real-mode K&R C compiler), arm-linux-gcc (cross-compiler), clang, llvm-gcc
21:05:57 <ais523> on x86 except where stated
21:06:57 <ais523> I don't have a 64-bit system to test on
21:07:03 <ais523> even though I do have an ARM emulator
21:07:05 <AnMaster> ais523, open64 works on 32-bit
21:07:13 <AnMaster> in fact I only have it on my pentium3
21:07:27 <ais523> apt-cache turns up nothing
21:07:53 <AnMaster> ais523, frontend is gcc-based, backend is not
21:08:26 <AnMaster> ais523, http://www.open64.net/
21:08:53 <AnMaster> ais523, btw which gcc versions have you tested ick on
21:09:10 <ais523> I haven't checked on old versions, although that's a good idea
21:10:05 <ais523> gcc-3.4's the oldest in the repo, I'll check on that
21:10:17 <AnMaster> ais523, cfunge is tested on gcc 3.4.6, 4.1.2, 4.2.1, 4.3.2 and 4.3.3 (the latter two only because I happen to have both around, usually I only bother with one from each x.y
21:10:51 <AnMaster> ais523, and icc 10.1, open64 4.1, and svn head of clang
21:11:03 <AnMaster> and I tested pcc and tcc and they fail totally
21:11:11 <ais523> pcc failing is bad bad bad
21:11:15 <ais523> I should definitely test gcc on that
21:11:19 <ais523> I'm not sure if I have it handy, though
21:11:25 <AnMaster> why is pcc failing bad bad bad?
21:11:39 <ais523> hmm... pcc isn't in my repo
21:11:46 <ais523> and it's bad because it's the default compiler on various systems
21:11:47 <AnMaster> ais523, indeed, it is *bsd only iirc
21:11:55 <ais523> why wouldn't it run on Linux?
21:12:00 <AnMaster> I never even got it to compile hello world on linux
21:12:04 <ais523> does it depend on BCD behaviour?
21:12:07 <AnMaster> it fails at glibc system headers
21:12:24 <ehird> binary coded decimal
21:12:27 <ais523> why doesn't it use its own
21:12:44 <ais523> ehird: x86 has binary coded decimal instructions, nobody ever uses them
21:12:46 <AnMaster> ais523, a bit hard to not use system signal.h
21:12:58 <ais523> AnMaster: bcc seems to manage it
21:13:01 <AnMaster> I mean limits.h and some stdint and such sure
21:13:21 <AnMaster> ais523, well one is real mode, the other is cross-compiler
21:13:44 <AnMaster> while if you want to compile native normal apps for a *nix you want to use system headers + a few compiler specific
21:13:45 <ais523> admittedly, gcc-bf has problems with multiplication
21:13:51 <ais523> which is much easier than signals
21:14:55 <AnMaster> ais523, anyway pcc isn't default on any OS I know
21:14:57 <ais523> http://www.open64.net/faq.html is not very encouraging
21:15:29 <AnMaster> ais523, indeed. open64 sucks when it comes to docs and so on
21:15:54 <AnMaster> ais523, this may be a better place to read: http://en.wikipedia.org/wiki/Open64
21:15:56 <ais523> C-INTERCAL doesn't even depend on C89 nowadays
21:16:11 <AnMaster> ais523, so what do you do for vsnprintf() and such?
21:16:34 <ais523> there's a config option
21:16:51 <AnMaster> ais523, but since there is no standard for these pre-ANSI ones
21:17:00 <AnMaster> how the heck do you know what to use
21:17:22 <AnMaster> I mean, what functions are guaranteed to exist in pre-ANSI C? Any at all?
21:17:24 <ais523> you do it via experimentation
21:17:26 <fizzie> Z80 has even more BCD instructions, and they're very much used by the TI calculators, since the TI float format is a BCD one.
21:17:31 <ais523> and all the ones in K&R1 are likely to exist
21:17:44 <ais523> autoconf's great, though, it does the experimentation all by itself
21:18:01 <AnMaster> ais523, iirc the BCD instructions are illegal under x86_64
21:18:47 <AnMaster> AAA ASCII Adjust After Addition
21:18:48 <AnMaster> Adjusts the value in the AL register to an unpacked BCD value. Use the AAA instruction after using
21:18:48 <AnMaster> the ADD instruction to add two unpacked BCD numbers.
21:18:48 <AnMaster> Using this instruction in 64-bit mode generates an invalid-opcode exception.
21:19:03 <AnMaster> that last line seems to figure in several of the BCD instructions
21:20:38 <fizzie> Aw, no love for BCD. Maybe they could add some SSE-whatever instructions that operate on 32-digit BCD numbers.
21:21:08 <AnMaster> BOUND (check array bound) is also invalid in 64-bit mode
21:21:25 <AnMaster> but it seems a rather silly bounds checking one indeed
21:24:10 <AnMaster> not a 2-word CAS though :/ it seems to require the two words to be after each other
21:26:12 -!- Mony has joined.
21:26:41 -!- Mony has changed nick to Guest34814.
21:32:15 <AnMaster> To avoid an invalid-opcode exception (#UD) on those processor implementations that do not support the CPUID instruction, software must first test to determine if the CPUID instruction is supported.
21:32:21 <oklopol> ah it's you, Guest34814, you shouldn't join with confusing nicks like Mony.
21:33:03 <AnMaster> oklopol, isn't mony someone we have seen here before iirc?
21:33:17 <oklopol> but Guest34814 is a regular
21:33:39 <AnMaster> oklopol, the number is random, when you don't identify
21:33:51 -!- ais523 has quit ("rebooting after kernel upgrade, I'll be back soon").
21:33:57 <AnMaster> services auto-changes your nick after a timeout
21:34:04 <AnMaster> there is some option to turn it on
21:35:42 -!- Guest34814 has changed nick to M0ny.
21:37:28 -!- M0ny has quit ("test").
21:37:56 -!- M0ny has joined.
21:38:20 -!- ais523 has joined.
21:38:51 <M0ny> autochange nick
21:40:41 <AnMaster> M0ny, just set your client to use M0ny by default then
21:41:43 <AnMaster> -M0ny- VERSION clientIRC 0.1 monyOS [Intel/480,00GHz]
21:43:37 <fizzie> Is it possible to dump a web page as a nicely formatted plain-text file with no word-wrapping (one paragraph per line) but simultaneously so that <hr> tags don't turn to absurdly long lines? Both "lynx -dump" and "w3m -dump" do word-wrapping, and both have options to control column width (-width for lynx, -cols for w3m) but *both* turn make absurdly long hr lines if I specify an absurd width to avoid wrapping. (At least lynx only writes a thousand _s, w3m see
21:43:37 <fizzie> ms to do something like 20000.
21:45:59 <fizzie> I guess I can just undo the wrapping, that'll be easier.
21:47:08 <fizzie> But that would mess things like lists where there are newlines that I actually want.
21:48:47 <fizzie> "elinks -dump 1 -dump-width <a large number>" does an equally silly thing about <hr>.
21:49:12 <AnMaster> "but *both* turn make absurdly long hr lines"
21:49:26 <fizzie> Just remove the "turn" or something.
21:49:44 <fizzie> I guess I could strip hr tags and dump. Stripping all tags leaves me with the suboptimal whitespace in the HTML file.
21:49:51 <Deewiant> fizzie: If hr is your problem then... yeah
21:50:09 <fizzie> Well, hr is the only problem with "absurd width output" I've seen so far in this.
21:50:15 <AnMaster> fizzie, why not dump and strip the _____
21:50:39 <fizzie> Well, I guess I could do that too. Or rather, normalize it to some sensible size.
21:50:52 <M0ny> [22:41] <AnMaster> M0ny, just set your client to use M0ny by default then <-- I prefer Mony with a o instead of the 0
21:53:03 <AnMaster> fizzie, crazy idea: curl http://example.com/ | html2latex | latex2lyx | lyx -e 'buffer-export "Plain text"'
21:53:57 <AnMaster> and wouldn't work with pipes but need to load from files
21:56:44 <fizzie> Simple w3m -cols 10000 -dump ... | sed -re 's/^━+$/--------------------/' seems to do a reasonably sensible output.
21:58:12 <fizzie> There's something weird about w3m spacing in that there's a table of contents done (ugly) with <table> and lines "<tr><td>N.</td><td>Nth chapter title</td></tr>" and if I do -cols 1000000 it outputs a thousand spaces between "N." and "Nth chapter title".
21:58:49 <Deewiant> I guess it makes tables 100% wide
21:59:33 <fizzie> Probably, although with -cols 10000 there's just a few spaces in there; I mean, there's no need to allocate much extra width to that first column.
22:02:04 <ehird> debian bug tracking sux
22:08:54 <ehird> ais523: I just ran into an actual write-only thing
22:09:19 <ehird> when we last discussed those we couldn't think of any
22:10:02 <fizzie> Okay, the w3m output is acceptable. (I'm turning some html files into plaintext for reading on the DS; have to leave tomorrow at around 06am to an Easter trip to Eastern Finland.)
22:10:52 <ehird> fizzie: can't you get an html viewery thinger for it
22:11:18 <fizzie> Probably, but I've habitualized myself with the moonshell text-viewer, which is a bit idiosyncratztic but otherwise okay.
22:11:29 <ehird> port w3m to it ^_^
22:11:44 <fizzie> I think DSOrganize already has some sort of HTML renderer, but it's less nice.
22:11:59 <fizzie> Besides, I think it corrupted the FAT once.
22:12:06 <fizzie> Could've been some other program too, though.
22:13:21 <AnMaster> <ehird> ais523: I just ran into an actual write-only thing <-- the debian bug tracker?
22:13:48 <ehird> a monticello repository
22:14:08 <AnMaster> "Monticello (pronounced [mɒntəˈtʃɛloʊ]), located near Charlottesville, Virginia, was the estate of Thomas Jefferson"
22:14:11 <ehird> http://www.wiresong.ca/Monticello/
22:15:48 <AnMaster> or a specific such repo being that
22:16:57 <ehird> sm://, which is just a protocol for copying monticello repos to make a SqueakMap releas
22:17:40 <ehird> "Now please type 'make' to compile. Good luck."
22:17:52 <ehird> i appreciate your thoughtfulness!
22:18:06 <ais523> who puts that at the end of a configure, I wonder?
22:18:15 <ais523> it's only one extra line in autoconf, although you have to write it by hand
22:19:11 <ehird> ais523: this thing has bolded section headings for different parts of the configure
22:19:14 <ehird> so it's quite tweaked
22:19:18 <AnMaster> it would be two extra lines in cmake (one to import a standard module for summary at end, another for the message)
22:19:24 <ehird> the first line in the .ac:
22:19:24 <ehird> dnl Hey Emacs, I want this in -*- Autoconf -*- mode, please.
22:19:31 <ais523> ehird: that's not really tweaked
22:19:40 <ais523> it's just a case of adding extra commands, which is easy
22:20:01 <ehird> AM_CONDITIONAL(WITH_EMACS, test "$EMACS" != no)
22:20:10 <ehird> if you don't have it it compiles 10x slower because it doesn't like you.
22:20:46 <ais523> WITH_EMACS? in /Automake/?
22:20:56 <ais523> does it build elcs, or something?
22:21:04 <ais523> or maybe install .el files, that's saner
22:21:34 <ehird> it has an emacs mode
22:21:45 <ais523> ehird: AM_ implies it's using automake
22:21:50 <ais523> automake requires some hooks in the autoconf file
22:21:55 <ais523> which is what you're seeing there
22:23:18 <ehird> http://libsigsegv.sourceforge.net/
22:23:39 <ehird> has portability considerations
22:23:58 <ehird> Copyright 1998-1999, 2002-2008 Bruno Haible <bruno@clisp.org>
22:23:59 <ehird> Copyright 2002-2005 Paolo Bonzini <bonzini@gnu.org>
22:24:05 <ehird> the bullied-by-rms guy and the guy who wrote this autoconf
22:25:52 -!- akiross has joined.
22:26:36 <akiross> hey, ehird :) long time no see
22:26:49 <akiross> i've been busy with exams for a while :S
22:27:15 <ehird> ais523: he's been in here before
22:27:18 <akiross> ow, i joined some weeks ago...
22:27:32 <ehird> sorry, we haven't forgotten you :P
22:27:49 <akiross> :D thanks i'm not really present on irc
22:28:40 <AnMaster> ehird, interesting this libsigsegv
22:28:42 <ehird> st> true become: false
22:28:42 <ehird> Object: true error: Invalid value nil: object is read-only
22:29:02 <akiross> actually i think i presented myself only to few ones :S
22:29:14 <AnMaster> ehird, I much prefer the feather way of making true always have been false, all the time along
22:29:26 <ehird> AnMaster: true havenBecome: false
22:29:34 <ehird> (yes, that's grammatically correct)
22:29:51 <AnMaster> ehird, what would the "haven" mean
22:30:11 <akiross> i'm developing a sort of assembly language for oo message passing and i'm here to learn something :) that's all
22:30:15 <ehird> "My brother and I, having destroyed the nuclear reactor,"
22:30:19 <ehird> haven X is in the sense of having there
22:30:23 <ehird> maybe havingBecome: would be better
22:31:05 <ehird> ais523: hasBeen: surely
22:31:10 <ehird> but that implies too recent imo
22:31:32 <ehird> true hasAlwaysBeen: false
22:31:44 <AnMaster> ehird, even better, would be this: "if (a == 1) a haveBeen: 1; else a haveBeen 0;
22:31:58 <AnMaster> ais523, Feather is TC to parse right
22:32:25 <ais523> AnMaster: actaully, no, it parses very simply
22:32:36 <ais523> it just recursively parses itself several times after it's been parsed once
22:32:40 <AnMaster> ais523, couldn't you do same parse-tc thing as in perl really
22:32:45 <fizzie> Well, it's nice that it's someone else who's had to think about how to port that sigsegv handling; there seem to be some specialization involved for almost all (os, architecture) pairs in it. Also it doesn't do what I want in jitfunge (not surprisingly, since it's such a peculiar need), which is to fake the stack underflow so that it skips the faulting instruction, "returns" a zero, and counteracts the effect of any instructions that moved the stack pointer.
22:32:45 <fizzie> It seems that a libsigsegv handler can only resume execution if it manages to make that location of memory usable. (It could be used for growing the stack, though.)
22:34:16 <fizzie> (I guess you could use it for stack underflow by allowing the stack to grow to the "wrong" direction too, and add some mapped pages with zeroes.)
22:35:26 <AnMaster> but it seems useful for other stuff
22:35:39 <AnMaster> I might end up doing something like that in cfunge some day
22:36:16 <ehird> that's why i mentioned it
22:36:22 <AnMaster> since popping a zero has high overhead then
22:36:30 <AnMaster> and pop on empty stack is very common
22:36:41 <AnMaster> at least those I write and in mycology
22:36:50 <ehird> you talked about it yesterday
22:36:55 <ehird> and I said I knew of a portable lib
22:37:07 <fizzie> Common, schmommon; I guess it's common with the weenies who use {}. :p
22:37:42 <fizzie> If you don't use {} and it is a real program, there's always something relevant on the stack and you're not going to be underflowizing it. At least that's what happens to me always.
22:39:22 <fizzie> Yes, well, if you do n in a long-living program, either you are storing in funge-space something that you could as well keep on stack, or you're inside a {} block. Well, again, that's what I feel, anyway.
22:39:26 <ais523> AnMaster: what I mean with Feather is that the first pass is always trivial, because it's just tokenising into letters
22:39:28 <ais523> the rest happens at runtime
22:40:44 * ehird types 2+2, middle clicks, Print it, gets 2+24
22:44:18 <AnMaster> ais523, write a first stage parser then. I mean do you ever plan to spec, or even implement feather?
22:44:24 <AnMaster> btw why the name feather for it
22:44:37 <AnMaster> ehird, that's a silly way to print it
22:44:43 <ehird> AnMaster: not really
22:44:48 <ehird> it's how smalltalks work
22:44:49 -!- akiross has quit ("Leaving").
22:44:52 <ehird> albeit most print a space before the output
22:45:11 <AnMaster> ehird, which smalltalk is this btw
22:45:19 <ehird> http://smalltalk.gnu.org/
22:45:23 <ehird> conclusion: lamer than squeak
22:46:14 <AnMaster> which does have certain advantages though
22:46:42 <AnMaster> anyway I'm not comfortable with IDEs. I prefer a text editor
22:46:59 <ehird> yes, well, that's an excellent way to get completely nothing out of smalltalk
22:47:11 <ehird> its world, its OS is its IDE; just like Unix is an IDE
22:47:14 <AnMaster> ehird, and sure if you have an extensive class library, say Java or .NET then an IDE is nice to keep track of things
22:47:23 <ehird> i don't think you understand
22:47:50 -!- KingOfKarlsruhe has joined.
22:47:53 <AnMaster> ehird, but do you want an IDE when coding C for example
22:48:03 <ehird> yes, I want my IDE to be unix
22:48:07 <AnMaster> or just a reasonable text editor with some syntax highlighting
22:48:16 <ehird> my editor is vim and it integrates into my IDE called unix
22:48:17 <AnMaster> ehird, that's an interesting way of describing it
22:48:29 <ehird> AnMaster: it describes the unix philosophy perfectly
22:48:37 <ehird> emacs is its own IDE
22:48:41 <ehird> not as good as unix, I find
22:48:58 <AnMaster> ehird, actually I have been using kate for editor a lot recently
22:49:06 <ehird> does it have an equivalent of vim's :! ?
22:49:13 <ehird> if so, it's an editor for the ide unix
22:49:30 <AnMaster> ehird, clicking "yes I'm sure I want to quit"? wait no that was :q! hm
22:49:44 <ehird> :! filters the selection/document through a unix command
22:49:56 <AnMaster> ehird, no clue, nothing I used anyway
22:50:00 <ehird> reverse the document? :!tac
22:50:15 <ehird> admittedly, you can do that with /
22:50:21 <ehird> but sed is smaller for some things
22:50:45 <AnMaster> ehird, it does have a built in mini terminal emulator, konsole lite
22:51:01 <ehird> that isn't really the same thing
22:53:29 <AnMaster> ehird, yes there seems to be such a feature, but rather hard to use
22:53:46 <ehird> then it's not much of a unix app :-)
22:54:08 <AnMaster> ehird, well Ctrl-AltGr-+ to get Ctrl-\ is bulky IMO
22:54:17 <AnMaster> I guess that depends on keyboard layout
22:54:37 <AnMaster> ehird, Ctrl-\ enter command hit enter
22:54:45 <AnMaster> that is how you do the filter thing in kate
22:55:24 <AnMaster> ehird, it does have one nice feature: autocompletion based on existing words in the document
22:55:45 <ehird> you can even rebind <TAB> to be smart: completes or indents based on your context
22:55:45 <AnMaster> ehird, true, but as a pop up as you type
22:55:52 <ehird> it comes as a dropdown
22:55:57 <ehird> it's all just a copy of IntelliSense
22:55:59 <AnMaster> ehird, that looks like KDE style
22:56:06 <ehird> AnMaster: if you have a kde vim gui, yes.
22:56:16 <ehird> also, vim can have language-specific completion files too
22:56:25 <ehird> so you can complete builtins
22:56:35 <AnMaster> ehird, vim can't have sane keybindings thogh
22:56:43 <ehird> sure it can, you can rebind anything
22:56:48 <ehird> yes, it's modal, that's a feature
22:56:52 <AnMaster> ehird, to remove the two-mode bit
22:56:56 <ehird> the composition of commands is beautiful and expressive
22:57:00 <ehird> visual mode for selection.
22:57:07 <ehird> also, I have some insert-mode bindings
22:57:11 <ehird> like ^A/^E from emacs/bash
22:57:18 <ehird> AnMaster: no because it defeats the point
22:57:27 <ehird> "Can I run IE on Linux? Also I want Outlook and viruses."
22:57:29 <AnMaster> ehird, then vim isn't for me, I want a single mode editor
22:57:39 <AnMaster> ehird, you can run IE on linux
22:57:47 <ehird> Whooooooooooooooosh!
22:57:52 <Deewiant> Normal, insert, replace, reverse insert, visual, visual line, visual block, operator-pending, command-line
22:58:17 <AnMaster> anyway I like emacs and kate better
22:58:28 <ehird> AnMaster: your loss
22:58:33 <ehird> t ell me when you switch to unix
22:58:38 <ehird> instead of the bad copies
22:58:59 <AnMaster> anyway I think you can rebind in kate
22:59:49 <AnMaster> there is a "google selection" heh
23:00:10 <AnMaster> (you can add your own commands btw, either simple script or full-fledged plugins)
23:00:18 <AnMaster> ehird, iirc plugins is hard in vim
23:00:27 <ehird> there are many vim plugins
23:00:33 <ehird> cd ~/.vim/plugin; wget foo
23:00:38 <ehird> load in your .vimrc
23:00:40 <AnMaster> as in say auto completion is a plugin you can select in some configuration location
23:00:58 <ehird> Deewiant: not really
23:01:02 <ehird> not if you don't use vimscript
23:01:08 <AnMaster> well for emacs it is easy to make them, a bit harder for kate
23:01:16 <Deewiant> ehird: I've heard the APIs for non-vimscript are crippled
23:01:17 <AnMaster> ehird, vim still seems rather monolithic
23:01:24 <ehird> Deewiant: Beats using vimscript
23:01:27 <ehird> AnMaster: It's not, in practic
23:01:38 <Deewiant> ehird: Point being you have no choice since you can't do stuff without vimscript
23:01:47 <ehird> Deewiant: Do bits in vimscript and call out to it
23:02:20 <ehird> AnMaster: If there isn't one of many plugins that do what you want you have obscure needs anyway
23:02:35 <Deewiant> ehird: And yeah, that's already making it something other than 'easy' IMHO
23:02:45 <ehird> Deewiant: Well, then writing emacs plugins isn't easy either.
23:03:04 <Deewiant> http://items.sjbach.com/560/extensibility-in-vim-and-emacs
23:03:19 <ehird> I disagree with most its points
23:04:43 <Deewiant> It's mostly the points in 2 that seem problematic
23:05:19 <Deewiant> Although not all of them are quite pointful
23:05:42 <Deewiant> (Point vi in particular is a bit of a joke)
23:06:59 -!- neldoret1 has quit (Client Quit).
23:07:06 -!- neldoreth has joined.
23:08:06 -!- BeholdMyGlory has joined.
23:08:30 -!- neldoreth has quit (Client Quit).
23:08:41 -!- neldoreth has joined.
23:11:46 -!- KingOfKarlsruhe has quit (Remote closed the connection).
23:12:15 <fizzie> Heh, 01am and the alarm clocksies are set for 05am to have enough time for packing and getting to the train place. Night, and habby easter-time; probably won't be much here until late Sunday-night. (Although there should be wlan in the hotel; but wouldn't count on that.)
23:19:13 -!- ais523 has quit (Remote closed the connection).