00:00:42 -!- FireFly has quit (Quit: swatted to death).
00:00:47 <Slereah> I tried coding in unlambda once
00:01:18 <Slereah> Then I just said fuck it and wrote a new language
00:01:20 -!- Mathnerd314 has joined.
00:01:24 <Slereah> And such is the story of Lazy Bird.
00:02:06 <oerjan> you got lazy and gave unlambda the bird
00:02:30 <olsner> in summary, unlambda is just another comninatory logic language
00:03:05 <Slereah> I didn't understand all that d and eager interpretation
00:03:14 <Slereah> Well, "just another" is a strange way to put it
00:03:24 <Slereah> It is the one most people think about.
00:04:10 -!- BeholdMyGlory has quit (Remote host closed the connection).
00:04:14 <Slereah> It's like saying brainfuck is just another brainfuck clone
00:05:50 <olsner> a bit... but probably more like saying brainfuck is just another P'' clone
00:06:48 <Slereah> INTERCAL is just another esolang~
00:07:36 <Slereah> I think the most awesome thing about INTERCAL is that it was written on punchcards
00:07:45 <cpressey> turing machines are just another turing machine
00:08:00 <oerjan> recursion is just another recursion
00:08:00 <Slereah> Actually, Turing machines are just another Post machine
00:18:52 <alise> Slereah is just another Slereah
00:22:56 -!- cpressey has quit (Quit: Leaving.).
00:37:40 <alise> death on a death stick
00:40:00 <Sgeo> Faith healing makes Baby Hyppocrates cry.
00:40:29 * Sgeo bibbles at that supposed to be Hippocrates
00:40:44 <Slereah> Hippocrates wasn't a paragon of science, either
00:40:52 <augur> hyppocrates is hypocritical hippocrates, obviously.
00:41:06 <Slereah> But well, it was a few milleuniums ago, so I'll let it slide
00:41:38 <Slereah> Back then people were big on things being true just because it's a nice idea.
00:41:49 <alise> stop saying bibbles
00:45:14 <alise> 14:04:45 <olsner> betwixt?
00:45:14 <alise> 14:07:39 <cpressey> it's weird when you hear it in a pop song
00:45:20 <olsner> bibble bobble blubble blebble
00:45:34 <alise> "OH BABY I WANNA BE BETWIXT YOU AND ANOTHER FEMALE"
00:45:46 <olsner> betwixt is weird enough in day-to-day conversation, but even weirder in a pop song
00:47:26 <olsner> oh, and it's way past bedtime
00:49:47 <alise> 14:20:19 <cpressey> oerjan: I know I've seen a paper trying to extend CA results to "infinitely small cells"
00:49:52 <alise> a continuous discrete automation, you might say
00:51:18 <Slereah> Is it above Turing completeness?
00:52:40 <Slereah> It sounds like the kind of thing that might be.
00:52:58 <augur> it almost certainly isnt.
00:53:05 <alise> 14:51:02 <oklofok> hard-on crime
00:53:23 <augur> oklofok: can i commit a hard-on crime against you
00:54:08 <Slereah> Two things told me it might be
00:54:20 <Slereah> And 2) you can't actually program it
00:54:22 <augur> well for one, such things are usually calculable with differential equations
00:54:30 <augur> infact, you have to define it using differential equations, do.
00:54:47 <Slereah> You could activate a non-measurable set of cells
00:54:49 <augur> well if you define it some other way, you've still defined it
00:55:08 <Slereah> Just because you can define it doesn't mean you can compute it!
00:55:22 <augur> if you can define the computational mechanism i think it does
00:55:22 <Slereah> Of course, it might not be, but well
00:55:38 <Slereah> There's many computational mechanisms that are >TC
00:55:46 <Slereah> They just can't be built or emulated
00:56:03 <augur> i think in this case you can both define the thing and calculate its behavior
00:56:44 -!- oerjan has quit (Quit: Good night).
00:56:59 <Slereah> I guess I'd have to see the article to know
01:02:33 <coppro> ugh... I should go out, but I'm so tired...
01:02:44 <coppro> oh well. sardines in the upper reaches of the math building should be fun
01:07:06 <alise> sardines are NEVER FUN
01:07:45 <coppro> it's a 10 minute walk and it starts in 8 minutes
01:20:50 -!- alise has quit (Quit: Leaving).
01:54:41 -!- Hiant has joined.
01:59:48 -!- Hiant has quit (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539]).
02:25:02 -!- Hiant has joined.
02:28:01 <Hiant> Hello everyone, I have come here for some suggestions about this esolang I have been creating. link: http://esoteric.voxelperfect.net/wiki/OTOH
02:28:03 <Hiant> I am curious as to whether adding a clone command, aka copies the value of one hand into another (non-destructively) would render the language too easy to use, rendering it a "bad" esolang.
02:34:39 <pikhq> Fucking motherfucking FUCK YOU SCIFI CHANNEL
02:35:05 <pikhq> "SyFy Channel teaming up with Billy Ray Cyrus for a new show where he investigates UFOs."
02:37:14 <Hiant> Why would they do that?
02:43:23 -!- Hiant has quit (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539]).
03:02:36 <Slereah> But turing tarpit ? esolang :o
03:02:57 <wareya_> turing tarpit implies esoteric, but esoteric does not imply turing tarpit
03:03:24 <Slereah> True, but the most popular ones are the tarpit.
03:05:06 <Slereah> At least if we were #tarpit, we wouldn't get so many wouldbe magicians :3
03:15:40 -!- lament has joined.
03:19:48 -!- olsner_ has joined.
03:22:42 -!- olsner has quit (Ping timeout: 240 seconds).
03:26:14 <Gregor> Slereah: No, but we'd get tar-fetishists.
04:03:05 -!- Gracenotes has quit (Ping timeout: 276 seconds).
04:18:59 -!- perdito has joined.
04:20:31 <Sgeo> Erm, have you ever been here before?
04:20:56 <Sgeo> Just to be sure: What do you think this channel is about?
04:21:01 <Sgeo> No offense, and you are welcome here
04:21:11 <Sgeo> It's just that some people come here with an incorrect idea
04:23:33 <lament> this channel is about worshipping Isis
04:24:35 <perdito> what are esoteric programming languages?
04:24:59 <coppro> perdito: Programming languages with little practical use, like INTERCAL
04:26:28 <coppro> in practice, we discuess a lot of random stuff in here
04:26:45 <coppro> about 40% of the time, it's related to computer science
04:26:56 <coppro> about 50% of the time, it's related to other computer stuff
04:27:00 <coppro> and most of the rest is politics
04:27:45 <perdito> i hope you'll excuse my poor english then.. but i'd like to stay and listen for a while
04:28:25 <pikhq> perdito: Your English doesn't seem poor to me.
04:28:39 <coppro> oh yes, I forgot about linguistics
04:28:44 <coppro> that's like 10% all on its own
04:28:47 <pikhq> Of course, that may just be the brutal assault that Japanese has had on my senses of bad English. :P
04:29:07 <coppro> other than the Internet grammar, it seems fine
04:29:17 <perdito> thx pikhq, but i think you'll change your mind soon enough
04:29:23 <coppro> and yours isn't particularly bad, perdito
04:29:38 <pikhq> perdito: Believe me, I've seen worse. From native speakers of English.
04:29:51 <coppro> I've read high school essays
04:30:15 <coppro> (Your English is better than that of the typical high school essay, believe it or not)
04:31:11 <perdito> k.. so most of you are programmers or sth?
04:31:13 <pikhq> Oh, definitely. The only things that differ from "correct" are just common traits of — what I shall now call — Internet Vernacular English.
04:31:15 -!- wareya_ has changed nick to wareya.
04:31:26 <pikhq> perdito: Pretty much all programmers here.
04:31:49 <pikhq> Not necessarily in a professional capacity, though.
04:32:46 <pikhq> Some of us are students, others are just people in different fields that enjoy bizarre programs. And then there's fungot.
04:32:47 <fungot> pikhq: right out right outside of philadelphia yeah so am i
04:33:23 <Sgeo> Scott Meyer is trying to kill me
04:34:04 <Sgeo> If laughter can be fatal...
04:34:09 <Sgeo> http://basicinstructions.net/
05:05:08 -!- TheProtomotor has changed nick to comex.
05:07:14 * Sgeo vaguely wonders if there's any languages that have the sort of libraries Factor has
05:07:37 <Sgeo> Factor has Erlang-style concurrency, some parser thing, call/cc
05:07:43 <Sgeo> Um, a lot of other stuff
05:23:27 * pikhq types in 変なコマーシャル (strange commercial) into Youtube to see what comes up
05:24:32 <pikhq> Yeah, that was pretty weird.
05:24:35 <pikhq> http://www.youtube.com/watch?v=y1jKAp9KEUs
05:24:41 <pikhq> Understanding it doesn't help.
05:25:17 -!- poiuy_qwert has joined.
05:30:08 -!- poiuy_qwert has quit (Client Quit).
06:32:18 -!- wareya has quit (Read error: Connection reset by peer).
06:33:20 -!- wareya has joined.
06:40:02 -!- augur has quit (Remote host closed the connection).
06:59:35 -!- oklofok has quit (Ping timeout: 255 seconds).
07:15:20 -!- FireFly has joined.
07:59:59 -!- clog has quit (ended).
08:00:00 -!- clog has joined.
08:02:01 -!- relet has joined.
08:06:18 -!- cheater00 has joined.
08:10:05 -!- cheater99 has quit (Ping timeout: 276 seconds).
08:10:15 -!- perdito has quit (Ping timeout: 240 seconds).
08:23:11 -!- MigoMipo has joined.
08:24:42 -!- relet has quit (Quit: Leaving.).
08:26:59 -!- wareya has quit (Ping timeout: 276 seconds).
08:36:57 <fizzie> http://golf.shinh.org/p.rb?C+style+constants -- ha, Befunge is objectively a better language than Java.
08:37:43 -!- MigoMipo has quit (Read error: Connection reset by peer).
09:22:39 -!- atrapado has joined.
09:32:56 -!- relet has joined.
09:36:00 -!- lament has quit (Ping timeout: 252 seconds).
09:37:40 -!- augur has joined.
09:38:10 -!- oerjan has joined.
09:40:37 <oerjan> <pikhq> "SyFy Channel teaming up with Billy Ray Cyrus for a new show where he investigates UFOs."
09:40:47 <oerjan> perfectly good science fiction, that...
09:42:59 <fizzie> "SyFy" is still a silly name.
09:43:32 * oerjan is commenting on nothing but this particular event
09:45:00 -!- tombom has joined.
09:52:18 -!- relet has left (?).
10:16:44 <myndzi\> if you don't have cable, does that mean you are... SyFy-less?
10:17:25 <oerjan> ...is there a pun in there somewhere?
10:17:40 <myndzi\> it's buried deep, but if you dig you can find it
10:17:46 -!- myndzi\ has changed nick to myndzi.
10:18:46 <myndzi> at least, that's what comes to mind whenever i read SyFy
10:19:02 <myndzi> ironic seeing as how i use a y in my nick for a long i sound
10:19:14 <myndzi> but i see SyFy and read "siphee"
10:22:19 <fizzie> Officially it's pronounced just like "sci-fi", except that now they can trademark it.
10:22:49 <fizzie> I did catch the syphilis pun, but couldn't come up with a STD counter-pun in time.
10:29:38 <oerjan> i guess you didn't have the right aids
10:39:18 <fizzie> I have to applaud you for that: clap, clap, clap.
10:42:06 * oerjan has a near-whoosh experience
10:48:33 <fizzie> 2. gonorrhea, gonorrhoea, clap -- (a common venereal disease caused by the bacterium Neisseria gonorrhoeae; symptoms are painful urination and pain around the urethra) -- just in case it whooshed by someone completely.
10:49:05 <fizzie> Gahh. Stop erroring, you stupid script!
10:50:02 <fizzie> "exception: PreModule: The file has invalid dimension". Graa, why can't you say which file is "the file", and what is this "dimension" it has and what it should have.
10:52:22 <fizzie> The exception came from an ugly C++ blob.
10:54:46 <fizzie> The file I think it should be reading does have the right dimension, but since it doesn't bother saying the file name, I can't quite be certain it's reading the right file.
11:00:25 <fizzie> Oh, it's the *other* file that's wrong.
11:00:54 <oerjan> well files contain strings, which is clearly related to string theory, which has extra dimensions. you must have got an invalid one into the mix.
11:02:08 <oerjan> try and see if you can extract it, it might contain the secret to FTL travel
11:06:15 <fizzie> According to the thing, it was supposed to have 21 dimensions, but it had 4290576399 of them instead.
11:12:19 -!- nooga has joined.
11:25:54 -!- oerjan has quit (Quit: Later).
11:27:26 -!- BeholdMyGlory has joined.
12:03:31 -!- relet has joined.
12:13:08 -!- wareya has joined.
12:14:59 -!- Zuu has quit (Read error: Connection reset by peer).
12:17:15 -!- augur has quit (Ping timeout: 272 seconds).
12:18:15 -!- relet has quit (Quit: Leaving.).
12:19:21 -!- Zuu has joined.
12:19:31 -!- Zuu has quit (Changing host).
12:19:31 -!- Zuu has joined.
12:26:19 -!- Akash has joined.
12:46:06 -!- augur has joined.
12:50:50 -!- Akash has quit (Ping timeout: 276 seconds).
12:57:36 -!- Akash has joined.
12:58:18 -!- augur has quit (Remote host closed the connection).
12:58:52 -!- sftp has joined.
13:03:11 -!- Akash has quit (Ping timeout: 276 seconds).
13:19:44 -!- MigoMipo has joined.
13:20:57 -!- Akash has joined.
13:50:22 -!- augur has joined.
14:08:42 -!- Akash has quit (Ping timeout: 240 seconds).
14:22:40 <fizzie> There are bird-like things called Zuu in many of the Final Fantasy games, but that might be completely irrelevant.
14:28:14 <Zuu> nooga, im certainly not a wtf
14:31:38 -!- MigoMipo has quit (Read error: Connection reset by peer).
14:33:04 -!- sftp has quit (Remote host closed the connection).
14:33:29 -!- sftp has joined.
15:03:38 <nooga> wtf -> Who the f**k
15:04:57 -!- augur has quit (Remote host closed the connection).
15:04:58 -!- sftp_ has joined.
15:05:06 -!- sftp has quit (Ping timeout: 240 seconds).
15:07:26 -!- augur has joined.
15:08:28 -!- derdon has joined.
15:18:40 -!- cpressey has joined.
15:30:26 -!- relet has joined.
15:35:35 <cpressey> Well, I started writing up my apply/ylppa language, and while the two operations have some nice properties (like (x/y)*y=x, and x/x=1,) I don't think I can make it into one of those "mainstream" algebraic structures (like a group) because both operations are non-associative, and I can't see any way offhand to make them associative.
15:36:53 -!- distant_figure has joined.
15:42:21 <fizzie> "Ylppö" is a Finnish male name (c.f. http://en.wikipedia.org/wiki/Arvo_Ylppö); I can't not think of it every time you say "ylppa".
15:42:28 <fizzie> Sorry, surname, I mean.
15:44:17 -!- augur has quit (Ping timeout: 265 seconds).
15:56:36 -!- relet has left (?).
16:00:36 -!- Phantom_Hoover has joined.
16:01:21 <nooga> Hi Phantom_Hoover!
16:03:21 <Phantom_Hoover> 15:50:29 <cpressey> unlambda in unlambda 15:51:21 <cpressey> that's just not right <-- Unlambda in Lazy K is even shorter, I think.
16:15:39 <cpressey> fizzie: i'm in ur irc channel, makin u think names
16:16:48 -!- Wamanuz2 has quit (Ping timeout: 272 seconds).
16:17:07 -!- Wamanuz2 has joined.
16:23:20 -!- Wamanuz2 has quit (Read error: Connection reset by peer).
16:23:38 -!- Wamanuz2 has joined.
16:29:09 -!- Wamanuz2 has quit (Read error: Connection reset by peer).
16:29:42 -!- Wamanuz2 has joined.
16:34:36 <cpressey> Isn't /tmp kind of an "antipattern" (though I despise that word)? ~/tmp would be better, except for the doesn't-count-for-quota/could-get-deleted-at-any-time rules, but those could be implemented some other way.
16:35:02 <cpressey> symlink into a private subdir of /tmp, for example
16:51:01 <cpressey> "TOO LATE" meaning, god only knows how many programs already write to /tmp and use kludgey ways of compenstating for its suckage
16:52:55 -!- comex has quit (Ping timeout: 265 seconds).
16:54:06 * Phantom_Hoover wonders how much of the esolangs wiki is Turing tarpits.
16:54:25 <Phantom_Hoover> And how much is languages with single-character syntax.
16:55:30 <cpressey> Not *too* many "unary languages", I hope
16:58:29 <cpressey> Oh hah, yeah, not what I was thinking.
16:58:44 <cpressey> Program in a single-character language, to me: ||||||||||||||||||||||||||||||
16:59:08 <cpressey> Single-character tokens in syntax, I get ya.
17:01:12 -!- calamari has joined.
17:02:41 <Phantom_Hoover> I wonder how many of them are like that because the designer had an idea but couldn't be bothered writing a parser?
17:05:58 <Vorpal> <cpressey> Isn't /tmp kind of an "antipattern" (though I despise that word)? ~/tmp would be better, except for the doesn't-count-for-quota/could-get-deleted-at-any-time rules, but those could be implemented some other way. <-- what if /home is on, for example, nfs?
17:06:07 <Vorpal> while /tmp is then presumably local
17:06:21 <Vorpal> cpressey, of course you could make login scripts mount /home/whatever/tmp
17:06:35 <Vorpal> still, complicates things
17:07:45 -!- Gracenotes has joined.
17:07:59 <Vorpal> Phantom_Hoover, well you need a parser for brainfuck
17:08:25 <Vorpal> Phantom_Hoover, while for befunge... you don't
17:08:35 <Vorpal> but for brainfuck you need to match up [ and ]
17:08:49 <Vorpal> Phantom_Hoover, indeed. A simple recursive one works very nicely
17:10:38 <Phantom_Hoover> (You can write a BF interpreter sans parser, but it's ugly.)
17:13:16 -!- oerjan has joined.
17:15:16 -!- cal153 has joined.
17:15:37 <oerjan> <Phantom_Hoover> 15:50:29 <cpressey> unlambda in unlambda 15:51:21 <cpressey> that's just not right <-- Unlambda in Lazy K is even shorter, I think.
17:16:10 <oerjan> i would be surprised if that were the case, after stripping comments and whitespace at least
17:19:53 -!- augur has joined.
17:20:17 <fizzie> It wouldn't even necessarily need a symlink in ~, just something to generate /tmp/foo for each user, and set TMPDIR to that. Though I'm sure there's an uncountable number of apps that have "/tmp" hardcoded.
17:21:19 <oerjan> approx 1928 stripped (approx because it might also strip a few chars from the character table)
17:25:13 <nooga> i can't figure out how to zip lists
17:25:31 <nooga> it was easier even in Haskell
17:26:29 <Vorpal> <Phantom_Hoover> (You can write a BF interpreter sans parser, but it's ugly.) <-- one that interprets directly and then seeks on the source to find matching [ and ] ?
17:28:11 -!- Phantom_Hoover has quit (Ping timeout: 252 seconds).
17:28:17 <oerjan> something like zip([X|XS],[Y|YS],[Z|ZS]) :- tuple(X,Y,Z), zip(XS,YS,ZS). for the main case, assuming i remember syntax right
17:28:45 <Vorpal> fizzie, the most straight forward way to find such would be to grep in /usr/bin for the string /tmp and then in that set grep for those that doesn't also contain TMPDIR (in case they check TMPDIR then fall back to /tmp if missing)
17:28:49 <fizzie> zip([], [], []). zip([A|As], [B|Bs], [[A,B]|Cs]) :- zip(As, Bs, Cs). -- if you want to handle just two lists, and want "tuples" made out of two-element lists.
17:29:46 <Vorpal> hm environment is rather inefficient, as far as I know it is stored as an unsorted array of char*
17:29:59 <Vorpal> would be better if it used some sort of binary search tree
17:30:47 <nooga> yesterday alise came up with something pretty simmilar but then it didn't work
17:31:11 <oerjan> <cpressey> Isn't /tmp kind of an "antipattern" (though I despise that word)? ~/tmp would be better, except for the doesn't-count-for-quota/could-get-deleted-at-any-time rules, but those could be implemented some other way.
17:31:36 <oerjan> um isn't there an actual function for getting a temp file...
17:31:42 <fizzie> I don't think an implementation is forbidden to maintain some sort of a lookup-helper structure for the environment, but I don't know any that do.
17:31:50 <fizzie> oerjan: Yes, several even.
17:32:26 <oerjan> and don't they look at the $TEMP variable or something, or was it $TMPDIR
17:32:29 <fizzie> mkstemp, mktemp, tempnam, tmpnam and tmpfile, at least.
17:32:41 <fizzie> Many do look for TMPDIR.
17:32:54 <oerjan> so the antipattern would be rolling your own, wouldn't it
17:35:04 <Vorpal> fizzie, mkstemp iirc only fills in the Xes at the end of the string?
17:35:27 <fizzie> The functions, or at least those that actually create the file instead of just returning a name for it, also to some degree handle awful race conditions where you choose a nonexisting file name, then try to open it, and meanwhile some other process goes and creates that file.
17:36:04 <fizzie> Yes, I guess mkstemp/mktemp only use a template you provide.
17:36:56 <fizzie> tmpfile() in glibc seems to do this: "[POSIX] does not specify the directory that tmpfile() will use. Glibc will try the path prefix P_tmpdir defined in <stdio.h>, and if that fails the directory /tmp."
17:37:06 <fizzie> So it doesn't even look at the environment.
17:37:29 * oerjan finds today's mezzacotta somewhat amusing
17:41:35 -!- calamari has quit (Quit: Bye).
17:42:59 <fizzie> A common (read: seen in more than one place) Linux-demoscene trick to make a "transparently" decompressing "executable" is to put "x=/tmp/Q;tail -n+2 $0|bzcat>$x;chmod +x $x;$x;rm $x;exit" on one line, and concatenate a bzip2'd (or lzcat+lzma, or xzcat+xz) executable after that; then you just hope that the app managed to make a /tmp/Q and execute it.
17:43:47 <fizzie> It's two bytes shorter than the arguably a tiny tiny bit cleaner version which does "x=`mktemp`;tail ..." instead.
17:44:52 <fizzie> Some also leave out the "rm $x" part, that's six more bytes saved.
17:46:47 <oerjan> thus _ensuring_ that running it twice will fail?
17:47:13 <fizzie> Not if you're the only person who keeps running it, which is quite reasonable.
17:47:22 <fizzie> Not so many people run demos on so very multiusery systems.
17:49:04 <fizzie> If you chmod it "a+rx" instead of just "+x", then those other people will end up running your version, also.
17:49:41 <fizzie> I guess with some umasks plain old "+x" is enough for that.
17:50:10 <oerjan> um wouldn't it bail out when trying to write to it
17:50:31 <fizzie> I don't think so, it's just commands separated with ; after all.
17:50:37 <nooga> crap, i failed discrete maths
17:50:54 <fizzie> fis@eris:~$ false;false;false;echo "ha, I don't mind the failures"
17:50:54 <fizzie> ha, I don't mind the failures
17:51:22 <nooga> because of goddamn linear algebra
17:51:45 <fizzie> (The tail ...>$x would fail, and the chmod would fail, but the plain "$x;" would still try to run it.)
17:56:18 <Vorpal> fizzie, isn't +x enough for a+x with the standard 022 umask?
17:56:58 <fizzie> I don't know how standard 022 is, vs. the 077.
17:57:11 <fizzie> But yes, "with some umasks".
17:57:50 <Vorpal> fizzie, I can't remember last seeing 077. Pretty sure stock arch and stock ubuntu at least both use 022
17:58:36 <fizzie> Possibly. I've been setting it consistently to 077 everywhere for myself, I may have just been assuming that other people were also.
17:59:16 <fizzie> Right, /etc/login.defs "# 022 is the "historical" value in Debian for UMASK when it was used" and /etc/profile "umask 022" for this Ubuntu.
18:00:35 <Vorpal> fizzie, 077 messes up for public_html
18:00:37 -!- atrapado has quit (Quit: Abandonando).
18:00:56 <fizzie> I prefer to make public only things I explicitly want to make public.
18:01:28 <Vorpal> fizzie, yeah but scp file webserver:public_html and you likely end up getting 403 for it
18:01:39 <Vorpal> and well, desktop and laptop are single user systems.
18:02:49 <fizzie> I don't do that, though. I have ~/www mounted (with the gvfs nonsense nowadays) over Samba from the webserver, and have set it to override permissions there so that everything copied there goes public.
18:04:38 <fizzie> I may have started the umask 077 habit from my university shell account; though admittedly it's a bit paranoidical, having 0700 for ~ is likely to be enough to keep snooping people out.
18:05:46 <fizzie> Heh, the http://golf.shinh.org/p.rb?big+number+arithmetic problem had a hilariously misleading name.
18:07:36 -!- distant_figure has changed nick to sleg.
18:10:14 <Vorpal> fizzie, that is very tricky without cheating
18:10:43 <fizzie> Yes, and based on the lengths everyone *is* cheating, with just three testcases. Still, I liked the... pun, if you can call it that.
18:11:18 <Vorpal> fizzie, most are not marked as cheats though
18:11:47 <fizzie> I'm not sure what's the usual way of marking them is.
18:12:10 <Vorpal> iirc "<username>(cheat)" or "<username>(embed)" or similar
18:12:36 -!- sftp_ has quit (Remote host closed the connection).
18:12:52 -!- sftp_ has joined.
18:12:53 <nooga> http://golf.shinh.org/p.rb?big+number+arithmetic this one is hard
18:13:52 <Vorpal> nooga, see comments just above
18:14:45 <Vorpal> nooga, ... isn't it obvious?
18:14:54 <fizzie> Cheating by just looking at two characters of the input, and deciding, based on those, whether to output one of three existing constants.
18:15:10 <Vorpal> fizzie, 3, unless you mean those actively looked at
18:15:31 <Vorpal> (as opposed to those you have to read in)
18:16:02 <nooga> i thought there are more input datasets
18:17:07 <fizzie> I meant "actively looked at"; and you have to read in 4, because both the latter ones start with three spaces.
18:17:24 <nooga> normally they don't use example data shown with the problem to actually test programs
18:17:46 <fizzie> anagolf doesn't have any hidden testsets; it does say it's not serious golfing on the front page, though.
18:19:03 <fizzie> In the interests of promoting cheatingship, I put in a (marked-as-such) Befunge version too, though it's a bit on the long side.
18:23:05 <cpressey> oerjan: you expect programmers to use an *API*? bah!
18:24:25 <Vorpal> cpressey, where did he say that?
18:24:49 <cpressey> oerjan: um isn't there an actual function for getting a temp file...
18:25:02 <cpressey> sorry should have been <oerjan>
18:25:05 <Vorpal> cpressey, on *nix: several
18:25:24 <cpressey> cpressey: I wasn't asking, I was quoting oerjan.
18:25:51 <Vorpal> cpressey, and those doesn't put the file in any sort of $TMPDIR or such afaik
18:25:55 <fizzie> fizzie: I like that "talking to myself" angle of that last comment.
18:26:56 <fizzie> Vorpal: tempnam in glibc uses the $TMPDIR directory as its first priority, if set.
18:26:58 <fungot> Vorpal: as a as a religious fnord or not
18:27:25 <Vorpal> fizzie, downside: not secure
18:27:56 <fizzie> Vorpal: Meh, it's just not secure if you don't use it right... but I guess that's a reasonable concern.
18:27:58 <oerjan> fungot: let's hope not
18:27:58 <fungot> oerjan: ah when i was going with you know like
18:28:24 <fizzie> fungot: No, I don't really know. Like what?
18:28:25 <fungot> fizzie: a lot of mean l._a. it the whole experience you know so
18:28:26 <Vorpal> fizzie, well you have to add logic to try again if it fails.
18:28:34 <Vorpal> while mkstemp iirc does that for it
18:28:59 <fungot> Available: agora alice c64 ct darwin discworld europarl ff7 fisher* ic irc jargon lovecraft nethack pa speeches ss wp youtube
18:29:17 <fizzie> It's a conversational telephone speech corpus thing.
18:29:21 <fungot> Selected style: fisher (Fisher corpus of transcribed telephone conversations)
18:29:31 <fizzie> http://www.ldc.upenn.edu/Catalog/CatalogEntry.jsp?catalogId=LDC2004T19
18:29:50 <fizzie> (I'm not exactly sure if it's that particular set.)
18:30:21 <cpressey> I suppose if I really cared, I could make a distro without a /tmp dir, which complains when you try to make a dir called /tmp, and which wires up all the API's/defaults/env vars sanely, and which says "oh boo hoo" if your script breaks trying to get at /tmp...
18:31:27 <fizzie> You're a really careful person if you really care that much.
18:32:41 -!- sleg has changed nick to hiato.
18:33:31 <fizzie> And yes, I forgot to actually reply, but mktemp/mkstemp/tmpfile all handle the actual file creation for you, but none of them -- in glibc -- look at the environment for the directory. (Okay, mktemp/mkstemp can't very well, since they just use a template with substituted characters, but tmpfile could.)
18:34:24 <Vorpal> cpressey, issue: breaks X11 amongst other things
18:34:25 <fizzie> GNU coreutils "mktemp" binary does use $TMPDIR, though.
18:34:56 <Vorpal> /tmp.X11-unix /tmp.X0-lock and /tmp.ICE-unix
18:35:02 <Vorpal> all have very good reasons for /tmp
18:35:14 <Vorpal> /tmp/.X11-unix /tmp/.X0-lock and /tmp/.ICE-unix
18:36:46 <fizzie> If you're rolling your own distro, I'm sure those *can* be moved, it might just be bit of work.
18:39:08 <Vorpal> fizzie, and it will be quite a pain for users who wish to compile programs you haven't made a package for
18:39:36 -!- kar8nga has joined.
18:40:04 <Vorpal> cpressey, an alternative idea that I seen in some hardened kernel patch is to add support for uid-dependant symlinks. So you make /tmp such a symlink and then it ends up pointing to different directories for different users
18:40:37 <fizzie> How does *that* work with X's root-owned /tmp/ socket-dirs?
18:41:42 <Vorpal> fizzie, maybe symlink those into the tmp dir for each user?
18:42:48 <Vorpal> fizzie, this would be easy to do on linux if only /proc/self/ had some magic $HOME symlink or such
18:43:57 <Vorpal> /proc/self/cwd/proc/self/cwd $
18:44:14 <Vorpal> only works by starting from / obviously
18:44:29 -!- hiato has quit (Ping timeout: 276 seconds).
18:44:58 <Vorpal> and it is just a trick "confusing" bash, I think the symlinks are actually resolved when changing directory
18:45:33 <fizzie> /proc/self/root/proc/self/root would always work, but it's perhaps less confusing-looking.
18:45:50 -!- hiato has joined.
18:46:12 <Vorpal> fizzie, with cwd you have to remember that cd affects it
18:46:35 <Vorpal> and thus introduces a element of confusion that the root pointer doesn't
18:46:46 <Vorpal> fizzie, now, chrooting to /proc/self/root/proc/self/root might do the trick
18:47:13 <fizzie> In most cases bash's symlink-handling works reasonably well, but it's sometimes annoying how "cd .." moves "logically" how the path-in-the-prompt would suggest, while "cp blah .." or "ls .." of course won't.
18:50:07 <Vorpal> speaking of annoyance with paths: rsync on two directories (recursively)
18:50:17 <Vorpal> iirc, if one ends with / and the other argument doesn't
18:50:20 <Vorpal> you get strange results
18:50:47 <Vorpal> I tend to add it on both sides thus
18:52:32 <Vorpal> iirc the issue was that with something like rsync -r foo/ bar, it would actually sync foo/ to bar/foo/
18:52:34 <Vorpal> or something like that
18:52:53 <fizzie> Yes, it had some sort of logic in it that wasn't exactly obvious.
18:53:31 <fizzie> "Had" meaning I ran into rsync directory-creation strangeness too before.
18:53:37 <fizzie> I'm sure it technically speaking still has it.
18:53:43 <Vorpal> fizzie, ah, not as in "they fixed it" then
18:54:00 <fizzie> Well, it's not *broken*, it's just different.
18:54:58 <Vorpal> fizzie, I seem to remember that cp(1) on openbsd or netbsd or something such had some rule that cp -r foo/ bar would do cp -r foo/* bar/
18:55:06 -!- oklofok has joined.
18:55:18 <Vorpal> fizzie, might not have been that exact way to write it, as in it could have been / on the other argument or such
18:55:28 <Vorpal> but yeah, some strangeness on one of the *bsds
18:55:39 <Vorpal> it might even have been mv
18:55:41 <oklofok> indeed, completely fucked up
18:56:00 <fizzie> The man page says that "rsync -av /src/foo /dest" and "rsync -av /src/foo/ /dest/foo" do the same thing; i.e. copy /src/foo's contents into /dest/foo.
18:56:18 <oerjan> oklofok: even the rabbits?
18:56:28 <Vorpal> oerjan, rabbits always are
18:56:41 <Vorpal> fizzie, so / on second argument?
18:56:46 <fizzie> Also, they have some sort of a mnemonic that a trailing slash in the source means "copy the contents of the directory", as opposed to "copy the directory itself".
18:57:28 <Vorpal> fizzie, yeah I tend to do trailing on both ends
18:57:32 <fizzie> It doesn't mention (in this section I'm reading) how (or if) a trailing slash on the second argument would change the behaviour.
18:58:10 <Vorpal> <fizzie> The man page says that "rsync -av /src/foo /dest" and "rsync -av /src/foo/ /dest/foo" do the same thing; i.e. copy /src/foo's contents into /dest/foo.
18:58:11 -!- kar8nga has quit (Remote host closed the connection).
18:58:12 <Vorpal> <fizzie> Also, they have some sort of a mnemonic that a trailing slash in the source means "copy the contents of the directory", as opposed to "copy the directory itself".
18:58:19 <Vorpal> contradicting each other?
19:01:06 * Sgeo sneezes bullets
19:01:44 <Vorpal> hm how does moulting work? I mean, shouldn't it get smaller over time? Like Russian dolls. Rather than larger?
19:03:00 <Vorpal> I mean for snakes and such
19:03:29 <fizzie> In contrast, shouldn't all non-moulting species always say the exact same size?
19:05:36 <Vorpal> fizzie, well no, the issue with moulting to me is that the reason for doing it, seems to be that they can't expand the old skin? Right? So how can they store a larger one underneath the smaller old one? For snakes you could presumably make it elastic or fold it or such. But for some insects the exoskeleton that they shed is rigid
19:05:48 <Vorpal> which makes it a lot trickier
19:06:06 -!- Ilari_an1rcomp has changed nick to Ilari_antrcomp.
19:06:26 <Vorpal> oh wikipedia claims the new one hardens over time
19:08:36 <cpressey> <fizzie> fizzie: I like that "talking to myself" angle of that last comment.
19:09:08 <cpressey> i am one dangerously confused mofo!
19:10:22 <cpressey> <Vorpal> cpressey, issue: breaks X11 amongst other things
19:11:12 <Vorpal> cpressey, well, as a distro maintainer you would presumably have to patch the most popular packages
19:11:51 <fizzie> Or at least pay some alimony.
19:11:55 <cpressey> <Vorpal> cpressey, an alternative idea that I seen in some hardened kernel patch is to add support for uid-dependant symlinks. <-- yeah, DragonFlyBSD has "varsyms" that do this sort of thing
19:12:39 <Vorpal> cpressey, hm haven't used that one
19:13:01 <Vorpal> cpressey, how is it done though? Must be out of band data for it to work
19:13:20 <Vorpal> out of band wrt file name it points to
19:13:38 <cpressey> there's a context of substitions, like env vars, but only substituted inside link destinations, which look like: /usr/${var}/etc/whatever
19:13:39 <Vorpal> which would incur a checking overhead
19:13:45 <Vorpal> no clue how large that overhead would be
19:14:08 <cpressey> and you as a user can always alter the varsym vars, I think, so it might not even work well for this
19:14:13 <Vorpal> cpressey, what if I created a file named ${blargh} and wanted to symlink to it?
19:14:40 <cpressey> there's no var expansion in filenames, only in symbolic link destinations
19:15:12 <Vorpal> cpressey, well yes but consider: touch '${HOME}' && ln -s './${HOME}' ./foo
19:15:17 <cpressey> a file named ${blargh} is just like a file named ${blargh} on any other unixoid
19:15:35 <Vorpal> cpressey, see the issue here?
19:15:44 <cpressey> Vorpal: then ./foo can point to different things...? no, i don't see the issue
19:16:16 <cpressey> set varsym var HOME to '${HOME}' and it'll point to the file you just touched
19:16:19 <Vorpal> cpressey, well it could in theory break some programs that create a symlink to a file path containing such a file or directory. Unlikely yes
19:16:52 <Vorpal> cpressey, I *think* this might make it non-POSIX compliant :P
19:16:59 <cpressey> if you are creating files named './${HOME}' you're getting what you pay for
19:17:03 <Vorpal> not completely sure but likely
19:17:14 <cpressey> Vorpal: OMG NO SOMEONE PROTECT THE GEESE QUICKLY
19:17:28 <Vorpal> cpressey, which geese? *looks around*
19:17:35 <Vorpal> wait, why did cpressey turn into a goose?
19:17:54 <Vorpal> and why would he need protection?
19:22:09 -!- Gracenotes has quit (Quit: ...).
19:24:08 <oerjan> he just wanted to send you on a wild goose chase, is all
19:45:02 -!- augur has quit (Ping timeout: 255 seconds).
19:48:17 <cpressey> The thing with X11, IIUC, is that it is following a *different* antipattern (Man! I can't stand that word! Why do I keep using it?): it is exploiting the fact that /tmp is shared. It should instead be creating those things in /shared. Which should exist.
19:48:48 <cpressey> alise will probably read all this in scrollback and say, cpressey: Plan 9!
19:49:07 <oerjan> cpressey, stuck in the antipattern of using "antipattern"
19:49:34 <cpressey> antipattern is just another antipattern
19:51:12 <fizzie> cpressey: Windows 3.11 for Workgroups! (It doesn't put sockets in /tmp, I'm pretty sure.)
19:54:04 * cpressey sends his thoughts on that to fizzie using DDE
19:57:02 <cpressey> actually, I thought DDE was a heck of a lot more coherent as an IPC mechanism than OLE was
19:57:21 <cpressey> or is? what do they use now? i don't want to know.
19:57:37 -!- Hiant has joined.
19:57:54 -!- Hiant has quit (Client Quit).
19:58:10 <fizzie> I think it is (or was) something COM-based now.
19:59:59 <cpressey> just because they have COM (and used to have .COM) and .NET doesn't mean microsoft is the internet
20:00:23 <fizzie> Yes, because org-mode still holds court at .org files.
20:01:08 <fizzie> Actually I think now-now you're supposed to use some sort of .NET IPC channels, in fact.
20:02:22 <fizzie> .NET Framework 2 IPC channels have been deprecated in .NET Framework 4, in favour of using Windows Communication Foundation (WCF).
20:02:40 <fizzie> "Windows Communication Foundation (WCF) is Microsoft’s unified programming model for building service-oriented applications. It enables developers to build secure, reliable, transacted solutions that integrate across platforms and interoperate with existing investments."
20:02:50 <fizzie> See there, interoperate with existing investments.
20:05:01 -!- augur has joined.
20:05:48 <fizzie> Incidentally, I have that Linux-based minimal module player code, and its stakeholder (ha, I've heard of these terms too!) wanted a Windows executable, so I quick-and-dirty ported it to use winapi "waveOutFoo" functions and linked with cross-mingw's binutils' linker; it works fine in Wine, but apparently no longer in Windows 7.
20:06:22 <fizzie> I tried to take a look at how audio should be done nowadays (the multimedia functions are from the 16-bit era, after all) but haven't yet bothered to find out how to call through COM interfaces from assembly.
20:09:55 <fizzie> I guess it's just about fetching a vtable and calling via it, but gah.
20:35:21 <coppro> hey, can someone explain to me when I would write a monad in Haskell?
20:36:33 <pikhq> coppro: When you realise that you have already written >>=.
20:43:38 * Sgeo wonders if he should learn orn
20:45:21 <Vorpal> <cpressey> The thing with X11, IIUC, is that it is following a *different* antipattern (Man! I can't stand that word! Why do I keep using it?): it is exploiting the fact that /tmp is shared. It should instead be creating those things in /shared. Which should exist.
20:45:30 <Vorpal> /var/run or similar probably
20:46:04 <Vorpal> cpressey, I think that would probably be best
20:48:50 <Vorpal> fizzie, "stakeholder"?
20:49:01 <Vorpal> I have not heard of that word. Or if I have, I forgot about it
20:52:03 <wareya> welcome to #unix_and_tarpits
20:52:36 <Vorpal> wareya, what? We discuss befunge here too. Very non-tarpitty
20:52:49 <Vorpal> and also about everything else
20:53:00 <Slereah> How many commands does befunge have?
20:53:03 <Vorpal> Slereah, yes, especially befunge98
20:53:38 <Vorpal> Slereah, not sure about 93. But 98 has all printable chars in ASCII mapped. A-Z are mapped to loadable extensions. Of which there exist many
20:53:54 <Vorpal> 93 has a subset of these. But even it has a pretty large instruction set
20:54:27 <wareya> It's rediculously easy to do simple things in X ways; lim(x->inf)
20:54:40 <wareya> Damned american school failed me again!
20:54:45 <Vorpal> oh wait, I read that as C first XD
20:55:25 <Vorpal> Slereah, consider the fingerprint (loadable extension) SOCK for example. IIRC CLC-INTERCAL has some networking support, but only to other CLC-INTERCAL implementations. fungot uses SOCK to connect to IRC
20:55:26 <fungot> Vorpal: ( ( this is)) did we look at historical things or did we look at each other no i'm right laughter you know so
20:55:40 <Vorpal> and FILE to treat a file as an open stream
20:56:00 <Vorpal> (befunge98 has built in load/write for files, FILE is however more advanced)
21:03:33 <fungot> Available: agora alice c64 ct darwin discworld europarl ff7 fisher* ic irc jargon lovecraft nethack pa speeches ss wp youtube
21:03:42 <fungot> Selected style: darwin (Books by Charles Darwin -- you know, that evilution guy)
21:03:46 <fungot> Vorpal:/ second volume/ causes and means :) temperature for corresponding times in other years.
21:03:56 <fungot> Vorpal: pebbles :), strewed over patagonia. --of patagonia, formation :).
21:05:46 <fizzie> Vorpal: "A corporate stakeholder is a party that can affect or be affected by the actions of the business as a whole."
21:06:26 <fizzie> ("The term has been broadened to include anyone who has an interest in a matter.")
21:06:39 <cpressey> <Vorpal> /var/run or similar probably <--- yeah, that would be appropriate
21:12:04 <cpressey> fizzie: transacted solutions that interoperate. imagine that! the stakeholders will be well pleased
21:12:34 <cpressey> <coppro> hey, can someone explain to me when I would write a monad in Haskell?
21:12:41 <cpressey> coppro: AT ANY AND ALL OCCASIONS
21:12:41 -!- Phantom_Hoover has joined.
21:14:41 <Vorpal> coppro, I would suggest "never: if you need a monad you are probably doing something practical"
21:15:08 <Vorpal> coppro, try instead staying completely pure.
21:16:44 <cpressey> there's a certain satisfaction one can get out of writing completely monadless code, but, monads have a place. and i wouldn't call them impure.
21:18:29 <Sgeo> Vorpal thinks they're impure
21:18:29 <Vorpal> Phantom_Hoover, see log?
21:18:34 <cpressey> Phantom_Hoover: <coppro> hey, can someone explain to me when I would write a monad in Haskell?
21:19:00 -!- comex has joined.
21:19:47 <Sgeo> Well, maybe I mischaracterized what Vorpal said. I hope I mischaracterized what Vorpal said
21:20:00 <Sgeo> alise's habits are rubbing off on me :(
21:20:02 <Vorpal> Phantom_Hoover, of course. I was making a *joke*
21:21:28 -!- nooga has quit (Quit: Lost terminal).
21:22:11 <Phantom_Hoover> coppro, for an answer, you could look at the use cases for most of the standard monads on the Haskell wiki.
21:25:07 <Vorpal> Phantom_Hoover, about mathematicians being too theoretical I think
21:25:32 -!- augur has quit (Ping timeout: 245 seconds).
21:25:43 <Phantom_Hoover> Vorpal, it fails hugely because noöne who understands monads whatsoever thinks they are "impure".
21:29:40 <Phantom_Hoover> OK, general question: I have a copy of "Gödel, Escher, Bach" lying around. Do I read it?
21:31:18 <Vorpal> Phantom_Hoover, isn't that one of those book that have a low ratio of readers to owners?
21:31:35 <Vorpal> Phantom_Hoover, so it is up to you I guess
21:31:52 <Vorpal> Phantom_Hoover, that's nasty
21:32:04 <Vorpal> Phantom_Hoover, stealing in other words
21:32:15 <Vorpal> Phantom_Hoover, he gave it to you?
21:33:57 <cpressey> Phantom_Hoover: Do not read it.
21:34:46 <cpressey> No. I have respect for the Necronomicon.
21:35:10 <cpressey> GEB is not evil, it's just... spinny.
21:35:27 <cpressey> WOW MATHEMATICS IS SUCH A TRIP
21:36:27 <cpressey> Phantom_Hoover: How do I explain "spinny"?
21:38:08 <cpressey> Phantom_Hoover: I changed my mind. Read it. I think your system will reject it, like mine did. I mean, I can only hope.
21:38:23 <cpressey> Reject it like a donor's kidney.
21:38:54 <Phantom_Hoover> What was the reason your mental immune system rejected it?
21:39:22 <cpressey> Phantom_Hoover: I think it's because I don't consider mathematics to be a religious experience.
21:39:43 <Vorpal> cpressey, with the right dose of LSD and some luck that would probably change?
21:39:53 <cpressey> Vorpal: *Bad* luck, I hope you mean
21:40:02 <cpressey> Phantom_Hoover: Well, not explicitly AFAIR.
21:40:06 <Vorpal> cpressey, "some" does not indicate which direction
21:40:12 <cpressey> Phantom_Hoover: But... well, read it.
21:40:14 <Vorpal> cpressey, well it does in everyday speech
21:40:21 <Vorpal> cpressey, but not if you actually look at what the words say
21:41:05 <cpressey> My only interest is that you'll become far more obnoxious in this channel if you become one of ... *them*.
21:41:06 <Phantom_Hoover> cpressey, it's too big to embark upon without serious and irrefutable reason to do so.
21:41:43 <cpressey> Phantom_Hoover: Well, I only mean the preface, or first chapter, or whatnot. (The one where he explains to the reader what a "strange loop" is.)
21:42:21 -!- iGO has joined.
21:42:31 <cpressey> That's Hofstaderese for OMG RECURSION BLEW MY MIND
21:43:38 <cpressey> Note that I tend to be rather anti-Pop-Mathematics, and anti-Pop-Science, and anti-Pop-* in general.
21:43:52 <cpressey> Phantom_Hoover: That's quite nice.
21:44:29 <cpressey> Phantom_Hoover: There are a few small exceptions -- I said "in general". I may or may not know Ian Stewart
21:45:40 <cpressey> No, nor have I read Discworld anything.
21:46:19 <cpressey> anyone have an oxygen cannister? PH seems to be... subventilating
21:46:55 <Vorpal> oerjan, there? A question about Norwegian. What does "Faen" mean?
21:47:03 <Vorpal> oerjan, I guess "the devil" based on sv:fan
21:47:33 <Vorpal> oerjan, google translate claims it means "fuck" for some strange reason
21:47:47 <Vorpal> oerjan, and it also does sv:fan -> no:fuck which seems completely broken
21:47:53 <oerjan> it's a contraction of "fanden", i don't think it's anything to do with fauns
21:48:39 <oerjan> Vorpal: well in modern norwegian it's mainly a swearword
21:49:11 <Vorpal> oerjan, "damn" seems more relevant
21:49:39 <oerjan> you'd use "djevelen" or "satan" if you wanted to talk about the religious matters. of course those _are_ also swearwords :)
21:50:08 <oerjan> well, "dæven" rather than "djevelen" as a swearword, mostly
21:50:33 <oerjan> and "damn" is pretty close in meaning, yeah
21:50:59 <Vorpal> "dæven" I can't think of anything like in Swedish
21:52:00 <fizzie> oerjan: However, Google translate no:faen -> fi gives "naida", which is the first infinitive form of the verb, "to fuck", and is not really something that's be very commonly used as a swearword. (It's a bit on the crude side, but not actually really swearwordy.)
21:52:34 <oerjan> (not the same for swearing, though)
21:52:42 <Vorpal> fizzie, yeah google fails
21:52:55 -!- Fuzz3r has joined.
21:52:58 <Vorpal> at translating that is
21:53:26 <oerjan> "fuck" is a perfectly valid swearword in colloquial norwegian, me thinks
21:54:10 <Vorpal> I think it is that here too. It can't be used in the English non-swearword sense though in Swedish.
21:54:52 <oerjan> while the norwegian translation "pule" would be extremely crude in comparison
21:54:52 <Vorpal> hm I wonder what sort of maximal tarpit you could make
21:55:13 <Vorpal> get as many instructions as possible in a language while making sure removing any single one would make it non-TC
21:56:01 <fizzie> Google does correctly translate en:"fuck you" to the probably closest-in-sense Finnish all-purpose swearword, even though it has a different literal meaning (female genitalia, basically).
21:59:32 <oerjan> hm "Jäkel" would appear to be a german surname
21:59:32 <Vorpal> oerjan, Jäkel and Mr Hyde?
21:59:32 <oerjan> yeah run and hyde, you
21:59:32 <fizzie> I used to live on "Jääkärinkatu" (Jäger street).
21:59:32 <fizzie> (Which is also a German surname.)
21:59:32 <Vorpal> fizzie, is Jäger related to jägare? That is en:hunter
21:59:32 <oerjan> Vorpal: hm i imagine a language in which the command to increment a value depends on the value and you need to use all of them to avoid all cells ending up with the value you left out the increment for
21:59:32 <Vorpal> hm where have I heard "jäger" spelled like that. Somewhere the last few months. And no I don't speak German...
21:59:32 <fizzie> It's German for hunter, but among English speakers the military meaning might be more well-known.
22:04:35 <oerjan> Vorpal: for your maximal tarpit
22:04:35 <Vorpal> oerjan, but I don't see what you mean with "to avoid all cells ending up with the value you left out the increment for"
22:04:35 <Vorpal> oerjan, or why you would need all to do that
22:04:35 <fizzie> (Or possibly the drink that starts with the word.)
22:04:35 <oerjan> well if you left out the command to increment a 42, and there was no way to get back down without it - i'm assuming the largest value wraps here, and no decrement command
22:04:35 <oerjan> then every cell that got incremented enough would end up stuck as 42
22:04:36 <Vorpal> oerjan, hm. Doesn't sound terribly interesting
22:04:36 -!- Fuzz3r has quit (Client Quit).
22:04:36 <Vorpal> oerjan, also I suspect you could just keep moving along the tape forever
22:04:36 <Vorpal> to copy your data to a few range, at a lower value
22:04:36 <oerjan> a subtlety i see is how do you avoid just ignoring those cells and moving along the tape (assuming it's bf inspired in other respects)
22:04:36 <Vorpal> oerjan, hah beat you to it!
22:04:36 <Vorpal> oerjan, anyway, it doesn't need to be bf based
22:04:36 <Vorpal> could be lambda calculus based
22:04:36 <Vorpal> oerjan, but your suggestion is a good start
22:04:36 -!- hiato_ has joined.
22:04:36 -!- Jonty_ has joined.
22:04:36 <Vorpal> Phantom_Hoover, why does Thursday matter here?
22:04:40 <Phantom_Hoover> Because I have to do other things and I'm not going to use free time protesting
22:05:23 -!- GreaseMonkey has joined.
22:05:23 -!- hiato has quit (Ping timeout: 276 seconds).
22:05:23 -!- GreaseMonkey has quit (Changing host).
22:05:23 -!- GreaseMonkey has joined.
22:05:29 -!- Jonty has quit (Ping timeout: 276 seconds).
22:06:07 <fizzie> Phantom_Hoover: But is he going to be wearing the pope hat?
22:06:41 <oerjan> and what about the popemobile
22:06:47 <fizzie> Speaking of pope hats: http://www.craigslist.org/about/best/phx/1258405496.html (a very old thing, but still)
22:07:05 <fizzie> "I have pope hats in every closet, pope hats under the sing, pope hats full of other pope hats."
22:07:15 <fizzie> (I suspect that was supposed to be "sink".)
22:12:39 <Vorpal> Phantom_Hoover, grapnel?
22:13:02 -!- Dionisia has joined.
22:13:10 <fizzie> It's a big target, easy to hit.
22:14:06 -!- madbrain has joined.
22:14:21 <madbrain> would a 12 bit computer be intersting :D
22:15:13 <Dionisia> is this a channel about drawing?
22:15:41 -!- Dionisia has changed nick to dionisia.
22:15:46 <oerjan> http://ed-thelen.org/comp-hist/pdp-5.html
22:16:09 <oerjan> dionisia: definitely not
22:16:35 <madbrain> oerjan: except that's a poring computer for banks and governments
22:16:59 -!- dionisia has left (?).
22:21:00 <Phantom_Hoover> madbrain, pretty much any number of bits can be interesting
22:30:52 <madbrain> phantom: yeah... but what about on a game console-like system?
22:32:38 <madbrain> well, ever seen demoscene demos?
22:34:44 <madbrain> something that's not underpowered so that you can do realtime gfx
22:34:53 <Phantom_Hoover> Well, 12-bit doesn't seem to have any particular advantages.;
22:36:28 <madbrain> well, 6-bit or 12-bit color modes could be interesting
22:37:31 <Vorpal> madbrain, I seem to remember some PIC (SoC thingy) that has a 12-bit memory for program
22:37:35 <madbrain> but they're the best things :/
22:37:44 <Vorpal> it being a Harvard architecture
22:37:57 <Vorpal> it had 8-bit words for data iirc
22:38:02 -!- derdon has quit (Ping timeout: 245 seconds).
22:38:11 <madbrain> vorpal: ah... cool but if I do harvard I'll need some data mem->cpu mem transfer instruction
22:38:15 -!- Gregor has changed nick to Gregor-L.
22:38:32 -!- Gregor-L has changed nick to Gregor.
22:38:36 <madbrain> I'm not sure separate data and instruction buses are cool
22:39:08 <Vorpal> madbrain, note a PIC12F629 (I think it was that one I coded for, and it had 12 or 14 bits for the program memory) has is an 8 leg DIP thingy. 1024 words flash for program
22:39:10 <fizzie> The C64 graphics chip, VIC-II, is 12-bit (but not a CPU, of course) in a data-bus-width sense; the low bits are to the cpu/memory bus, and the four high ones are to the vic-only color RAM.
22:39:41 <Vorpal> and iirc 64 bytes SRAM + 128 bytes data EEPROM
22:39:46 <Vorpal> or something like that
22:40:03 <madbrain> yeah I'm really more like thinking about a dram based system
22:40:08 <Vorpal> madbrain, PIC12* are accumulator machines iirc
22:40:19 <Vorpal> madbrain, making them painful to program for
22:40:34 <madbrain> ideally 1 bus shared by cpu and gfx and sfx
22:40:47 <madbrain> and good for making screwy c64 style effects
22:40:55 -!- tombom has quit (Quit: Leaving).
22:41:16 <Vorpal> madbrain, and I too avoid sound and display
22:41:47 <Vorpal> madbrain, the most interesting thing in programming for that PIC12F629 was writing the interrupt handler for the serial port communication
22:42:36 <fizzie> At least 12 divides equally by 3, which might be nice for packed-pixel RGB444 pushing, but that sounds a bit memory-intensive.
22:43:42 <Vorpal> 16 bits per channel at least. Or even better: 32-bit floating point per channeö
22:44:21 <madbrain> but it's presumably something you'd do on a 24bit computer with 12 bit words
22:44:40 <Vorpal> madbrain, how many different colours can you get hm... 64?
22:44:45 <Vorpal> too limited to be interesting
22:45:23 <Vorpal> madbrain, well 4*4*4 is 64
22:45:28 <madbrain> 6bit is 64 colors, ie probably paletted and amiga
22:46:08 <Vorpal> still extremely limited
22:46:19 <fizzie> I seem to recall a 4k color screen in some handheld.
22:46:23 <Vorpal> photos certainly won't look very nice
22:46:38 <Vorpal> especially any gradients in them
22:47:01 <Vorpal> madbrain, dithering looks horrible on all but extremely high-DPI monitors
22:47:18 <Vorpal> madbrain, well I guess it looks okay on CRTs, but only because they are naturally blurry
22:47:28 <Vorpal> used a CRT today. UGH I hate that
22:47:40 <fizzie> You and your "horribles"; earlier we had 16-color palettes and that was plenty.
22:47:41 <Vorpal> flickering, blurry crap
22:48:04 <madbrain> yeah, 256 colors is plenty from a 2d game pov
22:48:07 <Vorpal> even at a refresh rate of 75 Hz
22:48:22 <Vorpal> and of course it didn't have any auto adjust button
22:48:42 <Vorpal> I had to move the image about with the buttons when I changed to 75 Hz
22:49:42 <Vorpal> gave me a headache that monitor
22:49:58 <fizzie> Anyway, full-"byte" pixels (whether 8 or 12 bits; I guess for a 12-bit arch you'd make the memory bus wider) in something that has a framebuffer in a slow-ish main memory shared by graphics and CPU doesn't sound like it'd necessarily work.
22:50:19 <madbrain> like, compare ega (16 colors with a fixed bad palette) with vga (256 color with variable palette)
22:50:33 <madbrain> vga just looks incredibly better
22:50:44 -!- Mathnerd314 has left (?).
22:50:51 <GreaseMonkey> cga is fixed pal, ega is (mildly) variable pal
22:51:10 <Vorpal> is that cable? Or something else?
22:51:12 <madbrain> yeah but I've almost never seen games using ega's variable colors
22:51:22 <Vorpal> nowdays VGA, SVGA and so on seem pretty meaningless
22:51:42 <madbrain> I'm referring to specific pc hardware
22:52:01 <Vorpal> madbrain, right. today most people think "the blue cable" when you say VGA
22:52:13 <fizzie> EGA monitor connector is a 4-bit digital RGBI thing, I don't see how you get anything else than the 16 specified colors out of it. (Or did it actually have a six-bit digital thing?)
22:52:50 <fizzie> No, EGA definitely is a digital thing.
22:52:54 <Vorpal> <GreaseMonkey> cga is fixed pal, ega is (mildly) variable pal <-- so... interlacing?
22:52:58 <Vorpal> since PAL uses interlacing
22:53:11 <madbrain> fizzie: dunno how ega connected, probably 6bit digital yeah, also it had incredibly weird timing (22khz vertical retrace???????)
22:53:48 <Vorpal> madbrain, why is 22 khz so bad?
22:53:49 <madbrain> yeah that's short for palette, not the tv standard
22:53:50 <fizzie> "EGA produces a display of 16 simultaneous colors from a palette of 64 --"
22:54:13 <madbrain> vorpal: also good luck getting anything to sync to it
22:54:25 <Vorpal> madbrain, besides... khz? which direction is it drawing in?
22:54:28 <fizzie> Right, a palette of 64 does sound like the six-bit digital one, where there's RGB, and then another set of "brighter" RGB ones.
22:54:58 <madbrain> vorpal: well, ega has different horiz refresh in 200 line 70hz, and 350 line 70hz modes
22:55:17 <Vorpal> madbrain, 70 hz doesn't seem too bad
22:55:31 <Vorpal> madbrain, even though you need 75 hz for reasonably low flickering
22:55:38 <fizzie> Anyway, you can hook up 4-bit RGBI (like the C128 VDC) into a EGA monitor by wiring all three "intensity" pins to the I.
22:55:49 <Vorpal> 90 hz or above is to be preferred to make a CRT even remotely usable for more than a short duration
22:56:14 <madbrain> but yeah you'll probably never get a pic out of a computer with an EGA or CGA unless you specifically have an EGA or CGA monitor
22:56:24 <madbrain> since anything you'll find is VGA
22:56:29 <Vorpal> madbrain, couldn't you build a capture device?
22:56:46 <madbrain> vorpal: dunno if that exists for CGA or EGA
22:57:03 <madbrain> vorpal: yeah, that could do it
22:57:15 <Vorpal> fizzie, also did you say ega was digital? Unlike vga?
22:57:21 <madbrain> vorpal: still a lot more easy to use a computer with a VGA
22:57:22 <fizzie> Anyway, you can hook up 4-bit RGBI (like the C128 VDC) into a EGA monitor by wiring all three "intensity" pins to the I.
22:57:31 <Vorpal> madbrain, I prefer a DPI
22:57:51 <fizzie> "I think I've seen some schematics. They were pretty hacky." is what I was supposed to say.
22:58:05 <fizzie> For an EGA-VGA box, that is.
22:59:15 <Vorpal> madbrain, why on earth?
22:59:28 <Vorpal> madbrain, absolutely no one uses analog for anything new
22:59:29 <madbrain> secret: analog DVI is actually just VGA remapped to a different connector :D
22:59:40 <Vorpal> what monitor interface did the old lisp machines use?
22:59:56 <madbrain> vorpal: they switched to analog for VGA!
23:00:06 <Vorpal> what monitor interface did the old lisp machines use?
23:00:12 <madbrain> CGA and EGA had digital cables, VGA had analog
23:00:28 <madbrain> and VGA was standard for like over 10 years
23:00:48 <Vorpal> madbrain, again: what monitor interface did the old lisp machines use?
23:01:07 <madbrain> nobody has hardware that will interface old lisp machines
23:01:17 <madbrain> but vga is still really common
23:01:19 <Vorpal> madbrain, unless you have a monitor still left
23:01:36 <Vorpal> madbrain, but presumably you could construct a conversion box
23:01:44 <Vorpal> what monitor interface they used
23:02:05 <Vorpal> madbrain, lisp machines...
23:02:25 <Vorpal> madbrain, well it is much more interesting than IBM PC history
23:02:39 <Vorpal> and I know I'm not alone in that
23:02:41 <madbrain> it's not like you'll actually output something neat to look at out of a lisp machine
23:03:01 <fizzie> The Sun box I have in the basement has Sun's own extremely silly 13W3 connector; it's shaped like the usual n-pin D connector, but inside there's three tiny coaxial-type connectors in addition to 13 regular pins.
23:03:04 <Vorpal> madbrain, you are trolling now
23:03:34 <madbrain> fizzie: is that the one that's actually exactly like VGA but with sync on green? :D
23:04:10 <Vorpal> madbrain, of course you can output neat stuff with it
23:04:15 <fizzie> madbrain: No, it's like VGA but there's a combined H/V sync. (Instead of VGA's separate.)
23:04:20 <Vorpal> such as lisp code. Or drawing stuff on screen
23:04:36 <fizzie> madbrain: SGI's boxes, I think, mostly did sync-on-green.
23:05:02 <Vorpal> why on earth you only care about graphics demos I can't fathom
23:05:33 -!- Hiant has joined.
23:05:36 <fizzie> I have a 13W3/(4xBNC) cable that worked in a CRT with the usual 5xBNC inputs; according to the monitor manuals it would've done sync-on-green too.
23:05:40 <Vorpal> madbrain, so is Gregor? And?
23:05:54 <madbrain> computers that don't output music bore me
23:05:59 <Vorpal> madbrain, there is more to computing than graphics demos. Especially in this channels
23:06:12 <Gregor> Pianos that don't output music bore me. Computers that don't output music are useful for computational tasks.
23:06:34 <Gregor> Also, http://codu.org/tmp/existentialism.png
23:06:59 <madbrain> vorpal: but I have to admit that sound and gfx are my personal favorite topics
23:07:37 <Vorpal> Gregor, pokemon joke right?
23:07:52 <Gregor> Via someone else, but yes :P
23:08:28 -!- Phantom_Hoover has quit (Ping timeout: 265 seconds).
23:08:33 <Vorpal> Gregor, it is Super Effective of course
23:09:31 <Vorpal> Gregor, what exactly _is_ existentialism?
23:09:31 <Gregor> Cropped from http://a.imageshack.us/img843/9525/confusion.png
23:10:04 -!- FireFly has quit (Quit: swatted to death).
23:10:33 <cpressey> madbrain: i think you should have a relatively powerful co-processor to do the gfx and/or sound
23:10:55 <Vorpal> Or: just pour more resources into the CPU
23:11:12 <cpressey> not as cool... you don't get to worry about bus contention that way!
23:11:55 -!- Hiant has quit (Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100401080539]).
23:12:06 <Vorpal> cpressey, iirc there exists dual-access-port-thingy RAM?
23:12:19 <Vorpal> not sure if that allows concurrent access
23:13:12 <Vorpal> cpressey, anyway I programmed on a thing recently where the CPU (a SoC) had a 1 bit serial databus to the LCD controller
23:13:16 <madbrain> vorpal: more cpu probably translates into at least having CPU cache
23:13:16 <fizzie> You can get dual-access memory; some boards based on that TI DSP I keep talking about have some. Then you can do two reads/clock.
23:13:46 <madbrain> plus that sort of architecture makes everything look like 1996 pc demos :D
23:14:06 <Vorpal> fizzie, couldn't you do different devices doing the different reads/writes?
23:14:40 <Vorpal> madbrain, need to add an advanced GPU then
23:14:49 <Vorpal> madbrain, then you get an interesting 3D demo
23:15:17 <Vorpal> there needs to be an FPGA demo scene
23:15:29 -!- GreaseMonkey has quit (Quit: Welcome honored guest. I got the key you want! would you like onderves. of Yourself).
23:15:30 <Vorpal> now that would be cool
23:15:33 <madbrain> it's a great platform and they did tons of stuff on the psx but it's also kinda complicated
23:15:49 <Vorpal> madbrain, I was thinking along the lines of PS3 rather
23:16:13 <madbrain> vorpal: too bad there's not enough overlap between the demoscene and hardware people
23:16:36 <Vorpal> madbrain, hardware people probably find it more interesting to do other stuff than demoscene
23:16:44 <Vorpal> I completely understand them
23:17:25 <madbrain> dunno, one guy I know in this audio dsp channel complained that the hw guy projects were usually nothing except interfacing mp3 player chips
23:17:26 <Vorpal> madbrain, ASIC demoscene would be even better
23:17:33 <Vorpal> too bad it would cost a fortune
23:17:51 <fizzie> Last year's altparty had a Cray competition -- Cray bought one of their tiny "desktop supercomputers" there -- but the participation was pretty low, and the results not so fancy. And anyway, the box was mostly unremarkable; it was just a cluster of 8 multicore Linix nodes, with a faster-than-usual network in-between.
23:17:56 <cpressey> CPU cache and dual-port RAM sound waaay too sophisticated for an 8(12?)-bit-esque gaming machine
23:18:26 <Vorpal> cpressey, yes we moved onto a more interesting topic
23:18:33 <madbrain> (1) no cache. cache is too complicated
23:18:41 <Vorpal> madbrain, read the scrollback
23:18:47 <Vorpal> fizzie, awesome idea though
23:18:52 <cpressey> madbrain: the computer I built couldn't play music. It only had 8 LEDs with which to impress the world of its powers. In my defense, though, the next output device I was going to build was going to be a sound-output-thingy
23:19:00 <madbrain> dram = cannot do multiple operations on same cycle
23:19:03 <cpressey> then i gave the computer away to a friend...
23:19:38 <madbrain> cpressey: I'm more like working on a project where I'll do an emulator but it has to be implementable IRL
23:19:40 <cpressey> Vorpal: I suspect madbrain is still on the topic of their project
23:19:55 <Vorpal> madbrain, use drum memory then
23:20:01 <fizzie> Vorpal: There was one demo with a nice-ish fluid-simulation flowy thing; that's sort of a natural match to the environment.
23:20:01 <Vorpal> then you get back to cool
23:20:22 <madbrain> naah I want to do the job of the guy who designed the atari jaguar
23:20:55 <madbrain> or most 80s/90s computer or console systems really
23:21:04 <madbrain> they're ALL designed around DRAM access cycles
23:21:17 <Vorpal> I prefer more exotic systems
23:21:40 <cpressey> custom amiga coppers back when routing was done *by humans*... sweet
23:22:05 <madbrain> like, I could design a nice system around 4mb of 32bit or 64bit fast page DRAM
23:22:12 <Vorpal> cpressey, err, I'm pretty sure routing been done my machines longer than amiga existed
23:22:17 <cpressey> madbrain: well, around DRAM access cycles and/or some convenient multiple of the video signal :)
23:22:25 <Vorpal> cpressey, I read about machines doing wire routing
23:22:38 <cpressey> Vorpal: the custom amiga chips were routed by humans.
23:22:57 <madbrain> cressey: yeah, what they did is round dram access time to nearest multiple of the NTSC color carrier :D
23:23:02 <Vorpal> cpressey, but mechanical wire-wrapping-and-routing tools existed for longer
23:23:14 <madbrain> that's why the same clock rates appear in the NES, SNES and amiga
23:23:19 <cpressey> there's a whole story about how HAM mode was on the chopping block, but if they took it out, there'd just be a big hole in the middle of the chip where it was anyway!
23:23:50 <cpressey> Vorpal: I'm not sure what this has to do with wire-wrapping... I'm talking about the chip itself
23:24:05 <Vorpal> cpressey, oh on the IC level?
23:25:00 <Vorpal> cpressey, still what is HAM mode?
23:25:14 <cpressey> Vorpal: I refer you to wikipedia page about the amiga computer
23:25:39 <Vorpal> cpressey, will do later then. I'd hate to do it in w3m
23:25:46 <Vorpal> and I don't have X running atm
23:26:26 <fizzie> HAM's very uniqueness makes it awesome; I'm not very sure I'd like to be writing code for it though.
23:28:19 -!- madbrain has quit (Ping timeout: 252 seconds).
23:30:35 <fizzie> If you want a short summary: it's a mode where the the pixels either select a color from a palette (as usual), but alternatively they can also mean "set only R (or G/B) value and keep the others from the previous pixel", so that you get sort-of n-bit color in something like n/2 bits per pixel. At the expense of making abrupt color-changes a bit messy.
23:31:12 <fizzie> HAM being short for "hold-and-modify".
23:40:53 <oklofok> "<cpressey> Note that I tend to be rather anti-Pop-Mathematics, and anti-Pop-Science, and anti-Pop-* in general." <<< pop = bullshit
23:42:06 <oklofok> "<oerjan> "fuck" is a perfectly valid swearword in colloquial norwegian, me thinks" <<< i tend to use it a lot, was a bit more embarrassing when i went to scotland
23:42:58 -!- nooga has joined.
23:43:08 <nooga> anyone played such thing as Colobot? :D
23:45:07 <oklofok> "<nooga> anyone played such thing as Colobot? :D" <<< no
23:47:27 <oklofok> colon is where your ass comes from right
23:48:09 <oklofok> bots are automatical humans
23:52:51 * oerjan read that as bots are automatically humans
23:54:42 <Gregor> `addquote <oklofok> colon is where your ass comes from right
23:54:59 <HackEgo> 219|<oklofok> colon is where your ass comes from right
23:55:50 -!- cheater00 has quit (Ping timeout: 276 seconds).
23:56:31 -!- cheater00 has joined.