00:07:39 <int-e> zzo38: yes, the introdution looks good. reading between the lines, I can see that you want something human editable rather than just a data exchange format for programs, which explains the flexibility in the syntax
00:08:38 -!- ter2 has quit (Ping timeout: 240 seconds).
00:18:28 <zzo38> How well do you think it is compared to other formats and stuff? What one you use can depend on the application, but I would like to know other people's opinions on these comparison between formats too
00:24:24 <int-e> of the standard ones I tend to like JSON best but it's inconvenient for embedding binary data (since, as you write, it's tied to Unicode) I agrree that the standard text representation of XML is too verbose, and that binary formats will not be human readable nor editable.
00:26:06 -!- aretecode has quit (Remote host closed the connection).
00:26:13 <int-e> Oh and references are a nice touch (I've seen such before in lpmud save files, hmm :) ). I'm less convinced that macros should be part of the data format.
00:27:57 <int-e> (I'd be worried about somebody writing F(x) = [x;x] and then F(F(...F(42)...)) and crash my program.)
00:28:38 <int-e> but of course references have the same potential, if one isn't careful
00:29:25 <zzo38> Adding limits can avoid the F(F(...F(42)...)) problem. This is what SQLite does.
00:29:45 <int-e> is this right? [A.[A.[A.42,A],A],A]
00:30:47 <zzo38> int-e: Yes, it is valid, although may be ambiguous depending on the application.
00:30:49 <int-e> Or do I need 3 names for that?
00:31:08 <zzo38> Actually perhaps I should just specify you need 3 names for that.
00:31:12 <zzo38> It wasn't previously clear.
00:31:25 <zzo38> Needing 3 names is probably clearer and simpler and more sensible.
00:31:54 <zzo38> Note that references are their own datatype, not the type of whatever it is referencing.
00:31:55 <int-e> Oh. What about [A.[A]]?
00:32:04 <zzo38> Yes, that is valid.
00:32:25 <int-e> (why is this restricted to list items?)
00:32:26 <zzo38> It is a list containing a list containing a reference to the first item of the first list.
00:33:21 <zzo38> It is restricted to list items because of what references are and are not. They do not represent a copy of the data.
00:34:05 <int-e> it precludes {A=ref.[1;2;3];B=ref}
00:34:53 <zzo38> That wouldn't make sense though. If you want to use the list [1;2;3] in several places, you should want a macro, not a reference.
00:34:53 <int-e> which seems sensible. (whereas A.[A] is a bit silly)
00:35:07 <int-e> zzo38: but then it would be duplicated
00:36:21 <zzo38> References represent a pointer to the data, and are certainly not interchangeable with the data themself!
00:36:27 <int-e> It just seems strange to forbid that but allow [ref.[1,2,3],{A=ref,B=ref}]
00:36:52 <int-e> zzo38: yes. I want A and B to point to the same list.
00:37:00 <zzo38> Perhaps it does seem strange to you, but to me it doesn't
00:37:03 <int-e> not to two equal lists.
00:37:15 <elliott> int-e: wouldn't that make A the actual list [1;2;3] but B only a reference to it?
00:37:16 <zzo38> int-e: Well, references point to a position in a list.
00:37:18 <elliott> so it's not quite what you want
00:37:54 <elliott> you'd need, like, &ref.[1;2;3]
00:38:06 -!- ter2 has joined.
00:38:28 <int-e> Oh. [ref.[1;2;3];ref] and [ref,ref.[1;2;3]] are not the same?
00:38:42 <int-e> (sorry, I still write , for ;)
00:38:52 <zzo38> That is correct they are not the same at all.
00:39:02 <int-e> ok. that is strange to me.
00:39:03 <elliott> hence <zzo38> Note that references are their own datatype, not the type of whatever it is referencing.
00:39:48 <elliott> int-e: presumably foo.x is like ({ foo = &x; *foo }) for some imagined foo, so to speak, and foo is like... well, foo
00:40:09 <int-e> (then the lpmud reference (sic!) was invalid though, since that encoded actual sharing.)
00:40:47 <int-e> or encodes ... I know for a fact that a couple of those things are still running
00:47:52 -!- boily has quit (Quit: UNKNOWABLE CHICKEN).
00:48:24 -!- Phantom__Hoover has quit (Read error: Connection reset by peer).
00:48:48 <int-e> aaargh. I've been wondering for a week now why somebody would calle a function "runit" ... I read it as 'r-unit'...
00:59:01 -!- nooodl_ has quit (Quit: Ik ga weg).
01:00:43 -!- monotone has quit (Quit: "I can hear myself... I think I'm a bit afraid.").
01:01:02 -!- monotone has joined.
01:03:06 <int-e> heh. I found this comment in lambdabot: | not classFound -- can't trust those dodgy folk in #haskell
01:03:30 <lambdabot> Couldn't find class `Dodgy'. Try @instances-importing
01:13:23 -!- mhi^ has quit (Quit: Lost terminal).
01:14:41 <oerjan> why would you nit a ru
01:15:30 <oerjan> @instances-importing Extremely.Unsafe Dodgy
01:15:31 <lambdabot> Couldn't find class `Dodgy'. Try @instances-importing
01:30:25 -!- blitter64 has quit (Ping timeout: 276 seconds).
01:52:23 -!- blitter64 has joined.
02:08:23 -!- blitter64 has quit (Ping timeout: 252 seconds).
02:30:13 -!- BeingToDeath has quit (Ping timeout: 276 seconds).
02:31:33 -!- Sorella has quit (Quit: sleeeep).
02:45:09 -!- MoALTz has joined.
02:45:41 -!- MoALTz_ has quit (Ping timeout: 252 seconds).
02:51:27 -!- Tod-Autojoined has changed nick to TodPunk.
03:00:44 -!- MoALTz has quit (Read error: Connection reset by peer).
03:01:39 -!- MoALTz has joined.
03:51:19 -!- ter2 has quit (Ping timeout: 252 seconds).
04:53:03 -!- Vorpal has quit (Quit: ZNC - http://znc.sourceforge.net).
05:07:53 <kmc> hi sgeo, how goes it?
05:08:10 <Sgeo> Situation normal
05:15:52 <Sgeo> Sometimes I think I'm the sort of person who recognizes his mistakes, and then proceeds to do nothing about them.... how do I fix that mistake?
05:17:49 <zzo38> You did something (although not much) about that mistake, which is, to ask a question about it. That isn't doing much, of course!
05:29:10 <zzo38> O no, maybe ZCDSF is not LALR(1). I am trying to use it with the Lemon parser generator and get errors, which I think is because it has to look ahead a lot in order to figure out the difference of a macro invocation than a macro definition.
05:35:06 <zzo38> Perhaps it can be resolved in the tokenizer?
05:43:06 <kmc> Rust has this property that you can parse the source into a "token tree" before expanding macros
05:43:09 <kmc> that turns out to be very useful
05:45:14 <kmc> tt ::= '(' tt ')' | '{' tt '}' | '[' tt ']' | token tt | ''
05:45:16 <kmc> approximately
05:46:17 <kmc> macro expansion is sort of ugly though; there's an AST node type for each place a macro invocation can appear (ExprMac, PatMac, etc.) and these are expanded at some point
05:46:40 <Sgeo> Have macros changed since the last release, or from last release and release before then?
05:46:40 <kmc> and then all the code which runs after that point has like ExprMac(_) => tcx.sess.bug("unexpanded macro")
05:46:47 <Sgeo> I'm figuring maybe I should look at a stable macro system
05:46:47 <kmc> Sgeo: I don't pay attention to releases
05:46:56 <Sgeo> How much is it changing?
05:47:06 <kmc> as far as I'm concerned there are two versions of Rust: master, and whichever one Servo is using
05:47:11 <Bike> rust seems like a bad place to look for stability
05:47:13 <kmc> and the one Servo is using is almost never a release
05:47:19 <kmc> the macro system isn't changing that much, though
05:47:27 <kmc> I just added the ability to expand macros in pattern position
05:47:57 <kmc> and the mechanisms for "procedural macros", i.e. running arbitrary Rust code at compile time to produce an AST, are changing some
05:48:06 <kmc> and those are more exposed to other changes to compiler internals, of course
05:48:16 <kmc> but the plain Macros By Example macros haven't changed too much since I started using the language almost a year ago
05:50:30 <kmc> here is the graph of how far behind Rust master Servo has been over time: http://people.mozilla.org/~mbrubeck/servo-rust-updates.svg
05:51:58 <Bike> i like how when you do update it's not usually to master
05:52:42 <kmc> yeah because in the time between starting an update and finishing it, master will have moved forward
05:53:03 <Bike> by 10 days? hee.
05:54:10 <kmc> yes, it can easily take 10 days to update several 100k LoC for a month or two of language changes
05:54:21 <kmc> and test it on all platforms and rebase it against the other changes happening in Servo etc.
05:54:44 <kmc> Mozilla has this "badges" thing (idk really) and you may be amused by the image for the relevant one: https://badges.mozilla.org/en-US/badges/badge/Servo-Rust-upgrade
05:54:50 <Bike> what platforms is servo targeting btw
05:55:06 <Bike> oh that's pretty good.
05:55:08 <kmc> right now, x86-64 Linux, x86-64 OS X, and ARM Android kinda
05:55:21 <Bike> firefox tells me its connection to mozilla.org is insufficiently secure.
05:55:39 <kmc> Bike: the Android testing is not very automated, so it tends to regress a lot
05:55:44 <Bike> This site does not supply identity information... who could they be.......
05:55:50 <kmc> I think currently we can't run JavaScript on Android.
05:55:55 <kmc> in fact people are puzzled as to how we were ever able to
05:55:58 * kmc -> afk -> ice cream
06:03:01 -!- password2 has joined.
06:04:33 <Sgeo> http://doc.x3dom.org/developer/x3dom/runtime/index.html
06:04:41 * Sgeo hits his head on a convenient wall
06:05:46 <oerjan> a very nice and clean bar
06:18:21 <ion> [TAS] GC Crazy Taxi by solarplex in 05:53.18: http://youtu.be/O2wnNYmpABg
06:22:18 <kmc> every channel I'm in becomes #rust
06:22:24 <kmc> I'm vaguely aware that this can be annoying
06:30:21 <kmc> you got that right
06:30:51 <Sgeo> Should I try to learn raw WebGL or some convenient library like Three.JS? Or both?
06:34:31 -!- shikhin has joined.
06:34:39 -!- MoALTz has quit (Quit: Leaving).
06:37:52 <kmc> what do you want to make?
06:38:22 <mcpherrin> Raw webgl is kinda annoying compared to shiny things like ThreeJS
06:38:34 <mcpherrin> so if you don't need it, use the abstractions
06:39:06 <kmc> OpenGL is pretty low level and WebGL / GLES remove more of the stuff you'd want for using OpenGL "by hand"
06:41:20 <Sgeo> Well, it could be interesting to learn how the low level stuff works
06:41:35 <Sgeo> And use the abstractions for anything interesting
06:42:13 <kmc> mcpherrin: just write an activex plugin
06:43:20 <Sgeo> Programming first 'clicked' in my mind when reading about ActiveX
06:45:33 <kmc> what about it?
06:47:04 <Sgeo> I think... just the concept of the computer being entirely controlled by code
06:47:08 <Sgeo> I'm not sure. I was a kid
06:58:08 -!- password2 has quit (Ping timeout: 255 seconds).
07:05:02 <oerjan> ah a classic "Users like this are like a mongoose backed into a corner: with its back to the wall and seeing certain death staring it in the face, it attacks frantically, because doing something has to be better than doing nothing. This is not well adapted to the type of problems computers produce."
07:05:31 <oerjan> http://www.chiark.greenend.org.uk/~sgtatham/bugs.html
07:22:50 <Sgeo> I have always loved metaballs
07:23:26 -!- conehead has quit (Quit: Computer has gone to sleep).
07:25:59 <kmc> what about meatballs?
07:26:04 <kmc> fungot: I found a food truck selling "fonduewurst"
07:26:05 <fungot> kmc: http://tinyurl.com/ fnord is a confusing little bit of code i've had to be crashed to the ground to touch your nose.
07:27:59 <Sgeo> kmc: they're awesome to look at
07:28:07 <Sgeo> Oh, you changed the word
07:29:24 -!- Tritonio has joined.
07:33:53 -!- pikhq has quit (Ping timeout: 264 seconds).
07:41:17 -!- slereah has joined.
07:41:22 <slereah> Hello, I would like to magic
07:41:29 <slereah> Can you teach me the secret arts
07:42:04 <oerjan> `` WeLcOmE slereah | rainwords
07:42:06 <HackEgo> SlErEaH: wElCoMe tO ThE InTeRnAtIoNaL HuB FoR EsOtErIc pRoGrAmMiNg lAnGuAgE DeSiGn aNd dEpLoYmEnT! fOr mOrE InFoRmAtIoN, cHeCk oUt oUr wIkI: <HtTp://eSoLaNgS.OrG/>. (FoR ThE OtHeR KiNd oF EsOtErIcA, tRy #EsOtErIc oN IrC.DaL.NeT.)
07:43:47 <zzo38> I do not think we are a school for any secret arts. (If we were, it wouldn't be much secret, isn't it?)
07:44:26 <slereah> Then why do you have the HackEgo golem
07:45:06 <zzo38> To write strange messages like that.
07:45:46 -!- MindlessDrone has joined.
07:46:16 <oerjan> slereah: you cannot command it without the secret art of syntagma
07:46:48 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: `help: not found
07:49:41 -!- oerjan has quit (Quit: Munus sigsigga ag bara ye).
07:53:15 -!- pikhq has joined.
07:56:29 <Taneb> I have an exam in like 4 hours
07:56:33 <Taneb> Time to clean my room
07:58:45 <HackEgo> crtcoin bracoin quycoin kellosophiloxicoin blindex.phpcoin bitbitzcoin killcoin neurcoin abccoin prestowcoin swellicoin thumseycoin aheltscoin rnatacoin rolcoin um-32coin ootcoin evilleravillcoin unlolanptcoin byacoin
08:21:52 <fizzie> crtcoin uses old, broken CRT monitors as physical coins.
08:25:33 <kmc> I helped my friend haul a 100 kg Trinitron flat-tube CRT television the other day
08:25:51 <kmc> flat tubes weigh even more than usual tubes
08:27:03 <fizzie> Is there a connection between Trinitron brand aperture grilles and trinitrotoluene?
08:28:22 -!- Patashu has joined.
08:28:54 <kmc> i hope not
08:28:57 <kmc> for my friend's sake
08:30:16 <olsner> it's probably trinity+tron ... trinity because of the three colors in RGB? perhaps
08:30:52 <Taneb> My bookshelf is getting increasingly full :D
08:31:20 <fizzie> olsner: It's kind of silly, since what it improved on also had three colors and elec-tron beams.
08:32:09 <olsner> apparently trinity is because they merged the three beams into one beam
08:32:33 <fizzie> Oh, I guess that makes sense.
08:32:41 <fizzie> For some values of sense, anyway.
08:35:49 <fizzie> (Except I guess it doesn't have holes?)
08:36:16 <HackEgo> [wiki] [[Talk:Rail]] http://esolangs.org/w/index.php?diff=39735&oldid=39722 * 84.174.111.48 * (+41)
08:36:33 <Taneb> ...is a bookshelf the right place for a book of stamps?
08:36:42 <slereah> "the Erlenmeyer flask was invented by Dr Julius Flask"
08:36:42 <kmc> i like to put them in the box of envelopes
08:37:30 <Taneb> Right, my book of stamps is now on my bookshelf
08:39:11 <slereah> It was from Uncyclopedia apparently but got deleted
08:40:02 <kmc> rustc has a flag that prints out the elapsed time for each compiler pass as they run (some 40 or so of them)
08:40:12 <kmc> this makes long compile times feel a lot less frustrating
08:40:25 <kmc> stuff is happening! borrows are being checked and freevars are being found!
08:40:28 <fizzie> Does it have a flag that lets you play tetris while it's compiling?
08:40:34 <kmc> fizzie: patches welcome
08:40:40 <kmc> though that might require an RFC first
08:41:01 <kmc> look through https://github.com/rust-lang/rfcs/blob/master/0000-template.md
08:41:20 <kmc> Why are we doing this? What use cases does it support? What is the expected outcome? Why should we *not* do this? What is the impact of not doing this?
08:41:36 <fizzie> Yes, those are the kind of questions I might have trouble with.
08:41:36 <Taneb> I've missed my chance this academic year to make lecture notes on a typewriter :(
08:41:56 <mcpherrin> fizzie: omg that's actually a hilarious idea
08:42:28 <kmc> mcpherrin: how about a nice game of global thermonuclear war
08:42:38 <Taneb> kmc, I watched that film last night! :)
08:42:41 <mcpherrin> don't let the build succeed until you can clear a line
08:42:56 <fizzie> kmc: Another suggestion: if the real passes don't seem to be enough, you can add fake "reticulating splines" kind of ones.
08:43:26 <Taneb> Not last night, the other night
08:45:16 <kmc> fizzie: yes, it reminds me very much of that kind of thing
08:45:25 <Taneb> Also Hackers, which is the most 90's film imaginable
08:45:43 <fizzie> Taneb: Hack the planet!
08:46:34 <kmc> Taneb: http://www.avclub.com/article/ihackersi-may-have-been-of-its-time-but-it-was-als-72249
08:48:46 -!- Patashu has quit (Disconnected by services).
08:48:46 -!- Patashu_ has joined.
08:48:47 <kmc> my current task involves converting a bunch of small functions to a different form
08:48:55 <kmc> which is nicely incremental work, and not too challenging
08:49:00 <kmc> which means I'll probably stay up all night doing it :/
08:50:55 <mcpherrin> rust is far too boring with what it'll accept as function names
08:51:28 <mcpherrin> Miscellaneous Symbols and Pictograms at least is needed
08:52:35 <Taneb> kmc, good read, thanks
08:55:43 <kmc> mcpherrin: but you can do https://gist.github.com/kmcallister/0b1c57d021e1899611a7 at least
09:01:16 <mcpherrin> hmm I should have a party where we get roller blades and 90s laptops and hack the gibson
09:01:22 <kmc> count me in
09:01:28 <kmc> I even have rollerblades!
09:01:36 <kmc> though I'm bad enough at using them sober...
09:02:06 <Taneb> I ought to get some
09:04:18 <kmc> "Rust input is interpreted as a sequence of Unicode codepoints encoded in UTF-8, normalized to Unicode normalization form NFKC"
09:04:21 <kmc> I have discovered that this is lies!
09:04:31 <kmc> python3 -c 'print("#![feature(non_ascii_idents)]\nfn sp\u00e4tzle() { }\nfn main() { spa\u0308tzle() }")' > foo.rs && rustc foo.rs
09:04:34 <kmc> foo.rs:3:13: 3:21 error: unresolved name `spätzle`.
09:05:06 <mcpherrin> kmc: lol, not really that surprising
09:05:16 <kmc> even in the future nothing works
09:05:55 <HackEgo> [U+222F SURFACE INTEGRAL]
09:05:59 <kmc> spätzle is delicious btw
09:06:14 <mcpherrin> or ∃ is a great one for collections
09:06:18 <kmc> mcpherrin: have you been to http://walzwerk.com/ ? "authentic and unique East German cuisine in the San Francisco Mission district"
09:06:39 <mcpherrin> tbh I haven't eaten many places in SF
09:06:56 <kmc> I am told that the German culinary divide is more north/south, and "East German cuisine" is more about decor and such
09:06:59 <mcpherrin> in the year I've been here, probably only a few, and half of them are burritos
09:06:59 <kmc> but still, tasty
09:07:28 <mcpherrin> so my previous city was very german
09:07:42 <kmc> you should eat there and then watch Good Bye Lenin!
09:07:51 <mcpherrin> (Kitchener was until some nastiness earlier this centry named Berlin)
09:08:10 <kmc> or last century?
09:08:42 <kmc> anyway I didn't know that
09:08:43 <kmc> that's funny
09:08:53 <kmc> there's a Madrid, Iowa but it's pronounced MA-drid
09:09:28 <kmc> mcpherrin: I found a food truck in Portland that offers currywurst and "fonduewurst"
09:10:18 <Taneb> I have not been to a single Yorkshire-style restaurant, which probably don't exist
09:11:03 -!- shikhin has quit (Ping timeout: 265 seconds).
09:11:09 <Taneb> This is despite living in York
09:11:25 -!- drdanmaku has quit (Quit: Connection closed for inactivity).
09:11:35 <kmc> I think non-existence is a pretty good excuse, tbh
09:12:38 <kmc> it's pretty easy to make at home
09:12:49 <kmc> less so if you're picky about the finer points of the sauce
09:13:38 <kmc> Rosamunde has it, at least the Mission location
09:13:56 <kmc> and there's a "King of Cürrywurst" [sic] food truck
09:15:44 <kmc> http://achewood.com/index.php?date=03222004 http://achewood.com/index.php?date=03252004
09:15:47 <mcpherrin> kmc: my attempts to bootstrap a rust compiler that uses the YOLO keyword aren't going well :p
09:17:33 <mcpherrin> is gonna need a lot of #[cfg(stage0)]
09:17:45 <kmc> why? as long as you don't use it in the compiler itself
09:25:05 <mcpherrin> kmc: 'cause I removed unsafe, haha
09:30:51 <kmc> well there's yer problem
09:31:14 <kmc> gotta sleep now, ttyl all
10:16:29 -!- boily has joined.
10:55:56 -!- pikhq has quit (Ping timeout: 265 seconds).
11:01:11 -!- Phantom_Hoover has joined.
11:01:46 -!- atehwa has quit (Ping timeout: 240 seconds).
11:02:45 -!- atehwa has joined.
11:04:31 -!- pikhq has joined.
11:06:47 -!- Sorella has joined.
11:07:00 -!- boily has quit (Quit: Poulet!).
11:16:56 -!- edwardk has quit (Quit: Computer has gone to sleep.).
11:17:17 -!- edwardk has joined.
11:30:46 -!- yorick has joined.
11:32:58 -!- scoofy has quit (Ping timeout: 276 seconds).
11:46:06 -!- scoofy has joined.
12:02:09 -!- Sgeo has quit (Read error: Connection reset by peer).
12:19:54 -!- blitter64 has joined.
12:42:34 -!- blitter64 has quit (Ping timeout: 240 seconds).
12:44:55 -!- M28 has quit (Ping timeout: 252 seconds).
13:06:21 -!- M28 has joined.
13:25:33 <slereah> Every part of my body is Turing complete
13:31:34 -!- Patashu_ has quit (Ping timeout: 265 seconds).
13:36:06 <nortti> are DNA processes turing-complete?
13:37:46 <slereah> I don't know how you could make a loop with DNA
13:44:20 <int-e> slereah: ask your kids
13:45:57 -!- spiette has joined.
13:47:00 <slereah> You don't need kids for that
13:47:34 <slereah> But how do you do conditions on that loop
13:51:54 <Bike> "DNA processes"?
13:52:29 <Bike> and it's easy to make a DNA loop! http://en.wikipedia.org/wiki/Plasmid lolol
13:53:19 <slereah> Be careful with plasmids man
13:53:29 <slereah> Last time I took some I had bees coming out of my hands
13:57:22 <int-e> that sounds excruciatingly painful
13:58:00 <slereah> http://fc09.deviantart.net/fs70/f/2011/063/b/2/ttdi_bioshock_1__bees_by_anniezard-d3awuto.jpg
13:58:06 <slereah> http://mimg.ugo.com/201002/36219/cuts/insect-swarm-3_480x360.png
14:05:59 <impomatic> There something called HDNA (H for Hofstadter). I think it's also know as typogenetics or something.
14:06:17 -!- Phantom_Hoover has quit (Ping timeout: 264 seconds).
14:07:14 <Bike> http://en.wikipedia.org/wiki/Triple-stranded_DNA ?
14:07:16 <impomatic> Consecutive base pairs are interpreted and do various things like copy a base, cut a strand etc, loop, etc. It appears to be Turing complete.
14:07:19 -!- Sprocklem has quit (Ping timeout: 252 seconds).
14:07:34 <Bike> hoogsteen is a ridiculous name, though
14:08:00 <Quintopia> impomatic: yes it's described in GEB. i once considered writing a wiki article on it
14:08:32 <Bike> but it's not real?
14:08:33 <impomatic> I think he wrote a bit about it in Goedel, Escher, Bach. I read about it in the January 1982 issue of Practical Computing which has a HDNA interpreter written in Basic.
14:08:49 <Bike> i mean real DNA has plenty of stuff about alternative splicing, god knows
14:08:55 <Bike> also pseudoknots which are pretty wack
14:08:58 <slereah> But does it have a basic interpreter written in HDNA
14:09:13 <impomatic> Quintopia: it's definitely esoteric. You should add it to the wiki :-)
14:09:16 <Bike> "so there's NP completeness, at least"
14:09:44 <Quintopia> impomatic: oh it's not that it's not esoteric that i haven't done it. it's that i'm too lazy.
14:10:22 <int-e> slereah: this is cute though: http://bioshock.wikia.com/wiki/File:Insect_swarm_plasmid_by_tinamin1-d4jh31s.jpg
14:10:51 <slereah> The bee plasmid was pretty weak but I used it a lot
14:10:58 <slereah> Because it's fucking BEES FROM MY HANDS
14:11:17 <slereah> That is enough to make a whole game out of
14:11:33 <Bike> oh and Turing wrote a nice paper about symmetry breaking in devbio, imo read it.
14:12:08 <impomatic> Another one which ought to be on the wiki. The language of Tierra. (I don't mind doing this one, I just need to research it a bit more. I haven't seem Ray naming it anywhere)
14:13:28 <Quintopia> is tierra the one that's just code in a large block of memory with program pointers passing over it repeatedly
14:13:46 <Quintopia> or is the one with actual moving organisms in 3d space
14:14:04 <impomatic> It's all in code, no actual organisms.
14:15:27 <Quintopia> yeah i saw something about the language used in that once. no jumps, right? just reading and writing to addresses
14:19:08 <Quintopia> i guess i don't remember it too well
14:19:11 <impomatic> No operands to instructions. You'd have jmpb / nop1 / nop0 / nop1 / something else.
14:19:48 <impomatic> That would jump back to the first template matching the complement of the three nops.
14:22:19 <Quintopia> ah. so you can jump, but you can't be sure where you're jumping
14:36:43 <impomatic> Only if the matching templates is in your own code. Otherwise you might jump into another program and run it's code (like a parasite)
14:39:20 -!- mihow has joined.
14:43:34 -!- jix has quit (Remote host closed the connection).
14:44:03 -!- jix has joined.
14:45:10 -!- edwardk has quit (Quit: Computer has gone to sleep.).
14:56:35 -!- Gregor`` has changed nick to Gregor.
15:14:36 <Taneb> Help I need to sort out my life
15:16:05 <Taneb> Had a room inspection and the result was "We are worried about your health"
15:25:52 -!- nooodl has joined.
15:30:55 -!- Tritonio has quit (Ping timeout: 252 seconds).
15:36:11 -!- ter2 has joined.
15:46:26 -!- oerjan has joined.
15:49:21 -!- drdanmaku has joined.
15:51:05 <impomatic> Taneb: why exactly? Assuming it's messy, just hire a cleaner (or find a girlfriend who's clean / tidy). Assuming it's full of alcohol etc just print off some healthy looking labels to disguise it.
15:51:38 <Taneb> impomatic, evidently there's an oppressive smell
15:53:08 <Bike> that's... extreme
16:01:19 -!- nooodl has quit (Quit: Ik ga weg).
16:06:14 -!- slereah has quit (Quit: Leaving).
16:06:29 <Taneb> Anyway, you can imagine my emotional state right now
16:09:50 <oerjan> maybe there's a capitalist swine hiding in there
16:10:10 <Taneb> oerjan, don't think so, wrong smell
16:10:45 <oerjan> and this smell wasn't there when you moved in?
16:11:19 <Taneb> No, the smell is almost certainly my fault
16:11:59 <oerjan> i suppose only you can tell the reason.
16:12:35 <Taneb> Because I normally keep the window closed I don't know how to turn the radiator off
16:12:36 <FireFly> what kind of smell could give such a judgement...
16:12:51 <oerjan> and whether you need to clean or visit a doctor
16:12:55 -!- edwardk has joined.
16:13:34 -!- Zuu has quit (Remote host closed the connection).
16:14:26 -!- edwardk_ has joined.
16:15:11 <oerjan> hm maybe you have fungus in the room, i hear that can be awful.
16:16:19 <Taneb> It's a messy smell rather than a nasty smell
16:16:23 <Taneb> It's just a loud messy smell
16:16:56 <oerjan> well the solution is obvious but i would be a hypocrite to say it.
16:17:22 -!- edwardk has quit (Ping timeout: 240 seconds).
16:17:46 <Taneb> Yes, I know the solution to the primary problem very well
16:18:38 -!- edwardk_ has quit (Ping timeout: 245 seconds).
16:20:25 <Taneb> In fact, I suspect the majority of the smell comes from one source
16:24:08 -!- MindlessDrone has quit (Quit: MindlessDrone).
16:25:29 -!- edwardk has joined.
16:39:18 <kmc> which source is that
16:39:38 <kmc> impomatic: I like that "hire a cleaner" and "find a girlfriend" are alternatives on equal footing
16:39:40 <kmc> like, or something
16:42:08 <Taneb> They... I have not done the right thing with my bedsheets.
16:44:42 -!- conehead has joined.
16:52:48 -!- MindlessDrone has joined.
16:54:47 -!- MoALTz has joined.
17:02:34 <HackEgo> [wiki] [[Twocode]] N http://esolangs.org/w/index.php?oldid=39736 * GermanyBoy * (+4719) Created page with "'''Twocode''' is a [[:Category:Two-dimensional languages|two-dimensional]] [[esoteric programming language]] created by [[User:GermanyBoy]] in 2014. It mixes two-dimensional '..."
17:03:05 <HackEgo> [wiki] [[Language list]] http://esolangs.org/w/index.php?diff=39737&oldid=39713 * GermanyBoy * (+14) /* T */
17:03:41 <HackEgo> [wiki] [[User:GermanyBoy]] http://esolangs.org/w/index.php?diff=39738&oldid=39648 * GermanyBoy * (+14) twocode
17:05:11 -!- Sprocklem has joined.
17:15:55 <Taneb> `run echo bin/coins
17:16:00 <Taneb> `run cat bin/coins
17:16:01 <HackEgo> words ${1---eng-1M --esolangs 20} | sed -re 's/( |$)/coin\1/g' | rainwords
17:16:03 <zzo38> Then wash your bedsheets
17:16:16 <Taneb> That is the logical thing to do
17:16:43 <Taneb> `run cat bin/rainwords
17:16:44 <HackEgo> #!/usr/bin/python \ import random; w=[l.split() for l in open("/dev/stdin").read().split("\n")]; r=[4,7,8,9,2,6,13]; print "\n".join((lambda s: " ".join(chr(3) + "%02d"%r[(i+s)%len(r)] + l[i] for i in range(len(l))))(random.randrange(0, len(r))) for l in w)
17:16:53 <HackEgo> anollcoin pheycoin alagirdiminiscoin aubatwaldcoin clearecoin hlietcoin hoto++coin pcoin nestcoin intfcoin grocoin jetzschecoin membecoin sheedcoin stificcoin wudiffelliicketlangcoin brycoin tetcoin con-of-unbcoin truecoin
17:20:10 <zzo38> I read some book where they didn't write "Allah"; they wrote "al-Lah", I don't know if it is supposed to be the same word but it seem like it from the context. Do you know about this to answer it?
17:20:25 <kmc> Taneb: you can wash them every so often
17:20:36 <kmc> i don't have more specific advice
17:20:43 <kmc> wash them when they smell or feel weird or have disconcerting stains
17:20:56 <Taneb> I think not doing what I did would be a good start.
17:21:05 <Taneb> These haven't been off my bed since the end of September :(
17:21:15 <zzo38> What did you did at first?
17:21:20 <kmc> that is sub-optimal
17:21:25 <kmc> <kmc> rustc has a flag that prints out the elapsed time for each compiler pass as they run (some 40 or so of them)
17:21:28 <kmc> here's what that looks like: https://gist.github.com/kmcallister/acb1df198dbac070263f
17:21:33 <zzo38> You should probably clean it
17:21:34 <kmc> for building rustc itself
17:22:47 -!- Sprocklem has quit (Quit: Reconnecting).
17:23:08 -!- Sprocklem has joined.
17:30:14 -!- mihow has quit (Quit: mihow).
17:33:47 -!- mihow has joined.
17:36:41 -!- Sprocklem has quit (Ping timeout: 252 seconds).
17:37:22 <Melvar> zzo38: “al-” is the definite article, “allah” is more literally “the god”.
17:40:22 <olsner> Taneb: now you have an upper bound of how long you can use the same bedsheets, now wash them and keep them for half as long and see if you pass an inspection
17:40:50 <olsner> (if you do, take the average, and so on until you have found the maximum acceptable bedsheet age)
17:41:32 <olsner> this can also be a way to make it fun to switch bedsheets regularly
17:42:15 <Taneb> olsner, I feel like 1 month is possibly too long
17:42:26 <Taneb> 9 months is really bad
17:45:56 <olsner> it depends on how dirty you are and how much of it you rub into the sheets, I think
17:46:56 <olsner> and on how dusty the room is, and whether you make the bed with a cover or not (so that the dust settles in the sheets), etc
17:47:13 <olsner> if you wet the bed you may also need to switch more often
17:48:43 -!- oerjan has quit (Quit: Sleep tight and don't let the bedbugs evolve sapience).
17:56:25 <kmc> "WebGL 2 is based on what we think will be in OpenGL ES 3, which is the stuff in OpenGL 4"
17:56:30 <kmc> can't they just skip some to make the numbers match up
17:58:56 -!- Sprocklem has joined.
17:59:11 -!- Bike_ has joined.
17:59:14 -!- mihow has quit (Quit: mihow).
17:59:41 <fizzie> They don't want to confuse people.
17:59:46 <olsner> or just have a single "OpenGL" that includes the ES and Web stuff
18:00:35 -!- Bike_ has changed nick to Bicyclidine.
18:06:02 <zzo38> Melvar: O, OK thanks
18:07:28 <Bicyclidine> when did C get // comments because shouldn't those work
18:07:59 <Bicyclidine> god knows why a C program is including an .hpp. C/C++ i guess
18:08:37 <fizzie> (As an extension probably before that, too.)
18:08:40 <Bicyclidine> Do any gcc flags turn that off? -ansi or something maybe?
18:08:54 <fizzie> Well, -ansi is the same as -std=c89 which does not include them.
18:09:10 <fowl> this would be a good source for fungot http://www.scribd.com/doc/225960813/Elliot-Rodger-Santa-Barbara-mass-shooting-suspect-My-Twisted-World-manifesto
18:09:11 <fungot> fowl: what's a chatterbot then. no way sir, that i do
18:09:40 <Bicyclidine> fungot already quotes, what, nixon? that's gross enough already
18:09:40 <fungot> Bicyclidine: perhaps we should celebrate these days with guys like simon peyton-jones, simon fnord, etc
18:09:59 <Bicyclidine> not really soon so much as that the rodger thing is just gross
18:10:18 <Bicyclidine> the manifesto i mean, by itself, in addition to the shooting
18:10:55 <Bicyclidine> i guess if you wanted fungot to talk like a gross murderer you could go with Industrial Society and Its Future or something
18:10:55 <fungot> Bicyclidine: later tell riastradh http://tmp.barzilay.org/ r.txt: (
18:11:25 <fowl> Bicyclidine, ok :)
18:12:02 <Bicyclidine> fungot what's your support for made-up scripts like
18:12:02 <fungot> Bicyclidine: what's a respectable typing speed? i'm getting tired. must sleep... hiking trip tomorrow... have a flex question for you
18:13:54 <Bicyclidine> the worst part about the C thing is that the file is the complex.hpp that (re)implements complex arithmetic. it uses no C++ features, other than // comments I guess.
18:15:15 <Bicyclidine> yes, but - hilarity continuing here - this has to be compiled with -ansi, probably because for matlab's ffi, which also complains that my gcc is too new
18:17:10 <olsner> I wonder if that fnord was marlow
18:18:24 -!- Sprocklem has quit (Quit: Class).
18:18:41 <fizzie> olsner: You are correct. You win ten (10) FunPoints.
18:19:57 <fizzie> You may redeem your FunPoints for fabulous prizes at THIS FEATURE NOT FINISHED YET.
18:20:23 <olsner> Bicyclidine: what part of -ansi does it require?
18:20:30 <olsner> if it relies on somehting that changed in c99 vs c90 (whatever that might be) you could try something like gnu90 to get comments and useful stuff but otherwise c90
18:30:05 <Bicyclidine> still not sure what the hell "#include <math.h> /* Added for MS Visual C++ Compatibility 1999 */" is supposed to mean
18:33:13 -!- ter2 has quit (Ping timeout: 245 seconds).
18:33:40 -!- ter2 has joined.
18:34:14 <FireFly> Bicyclidine: // comments predate C
18:34:31 <FireFly> At least according to the Wikipedia article on BCPL
18:41:46 <mroman> Bicyclidine: Maybe in the compiler he used some stdlib header also included math
18:42:13 <kmc> `addquote <fungot> Bicyclidine: perhaps we should celebrate these days with guys like simon peyton-jones, simon fnord, etc
18:42:14 <fungot> kmc: srfi-7 doesn't standardize that, so you can't extend arbitrarly your plugin, if you half-ass something but provide the right social cues then you get 4
18:42:15 <HackEgo> 1200) <fungot> Bicyclidine: perhaps we should celebrate these days with guys like simon peyton-jones, simon fnord, etc
18:42:26 <kmc> fungot: I love halfassing social cues
18:42:26 <fungot> kmc: install-info does exist internally." page 7, under heading 5, evaluation.
18:42:35 <Bicyclidine> if noiseType == 0 randn('seed', 37) <-- chosen by fair die, guaranteed random
18:43:50 <Bicyclidine> "The latter method requires recursion and is computationally very expensive."
18:43:52 -!- mihow has joined.
18:50:18 -!- ter2 has quit (Ping timeout: 240 seconds).
18:58:35 <zzo38> /* ... */ comments in C are actually a bit somewhat problematic if you want to divide by a value that a variable points to; you would then need a space or parenthses.
18:59:13 <olsner> you should have spaces around binary operators anyway
19:03:23 <FireFly> The slashes situation is worsened in JS due to the existence of slash-delimited regex literals
19:03:59 -!- nooodl has joined.
19:04:03 <lifthrasiir> except that a valid regex cannot start with a Kleene star
19:04:04 <zzo38> FireFly: That is only if you are commenting out a regex literal though, since a regex doesn't start with /*
19:04:19 <FireFly> the empty string is a valid regex though
19:04:28 <FireFly> which interferes with single-line comments
19:04:56 <lifthrasiir> AFAIK Perl 6 does not allow an empty pattern as well
19:05:09 <Bicyclidine> damn i was about to ask how perl dealt with that since it was no doubt funny
19:05:40 <FireFly> I think it's possible to have "//" not trigger a line-comment in JS actully
19:05:59 <FireFly> if you divide a regex literal by something
19:06:06 <Bicyclidine> elseviers puts their impact factors and it is the funniest shit
19:06:25 <zzo38> I don't know why you would want to divide a regex literal by anything, but a regex might end with \//
19:06:52 -!- ter2 has joined.
19:10:02 <lambdabot> Maybe you meant: tell thank you thanks thesaurus thx tic-tac-toe ticker time todo todo-add todo-delete type v @ ? .
19:10:56 <lambdabot> parse error (possibly incorrect indentation or mismatched brackets)
19:11:31 -!- Phantom_Hoover has joined.
19:12:02 <lambdabot> parse error (possibly incorrect indentation or mismatched brackets)
19:15:10 <olsner> -- starts a comment btw, hth
19:15:36 <Taneb> Yeah, took me too long to realise
19:17:11 <int-e> > let (+)--+--(-) = 42 in (-)--+--(+)--but not always
19:17:37 <olsner> > let 2 --++ 2 = 5 in 2 --++ 2
19:17:59 <int-e> (confusingly, ---- does start a comment)
19:18:23 <int-e> Haskell has some weird corners
19:18:27 <olsner> I think any sequence of (more than two) -s is a comment?
19:19:22 -!- ter2 has quit (Ping timeout: 240 seconds).
19:19:41 -!- ter2 has joined.
19:20:33 <zzo38> Can you play pot-limit hold'em? How commonly do you know if chess clubs will have poker and other games too instead of only chess?
19:22:45 -!- edwardk has quit (Ping timeout: 252 seconds).
19:23:26 <kmc> how about chess boxing
19:33:23 <Taneb> I am worried that I am one of those people who is "good at maths", and I've never really had to push myself
19:33:33 -!- ter2 has changed nick to tertu.
19:33:44 <Taneb> And now I don't know how to learn
19:37:06 -!- spiette has quit (Quit: :qa!).
19:37:14 -!- spiette has joined.
19:40:03 <Melvar> I don’t think it’s impostor syndrome he’s describing …
19:42:26 -!- MindlessDrone has quit (Quit: MindlessDrone).
19:43:03 <kmc> that would depend on a bunch of information you don't have
19:43:15 <kmc> it's a valid concern and I don't want to dismiss it as a delusion
19:48:42 <Bicyclidine> you could get a really hard math book. i was "good at math" and i got a bunch of books that i can't read through and just kind of look at sometimes, slowly getting a bit better
19:54:43 -!- Sprocklem has joined.
19:59:45 -!- mihow has quit (Quit: mihow).
20:00:20 -!- mihow has joined.
20:05:40 <Taneb> Bicyclidine, this is probably separate from my imposter syndrome
20:08:18 <olsner> if you had it excessively easy before, it makes sense if you're lazy now
20:14:21 <Taneb> Bicyclidine, any you'd recommend?
20:15:33 <Taneb> But it's expeeeensiiiiive
20:15:48 <mcpherrin> I had a dream last night I found a copy of TAOCP at a used bookstore for $10
20:15:52 <Taneb> But it's raaaaaaining
20:16:26 <Bicyclidine> Taneb: http://www.math.upenn.edu/~wilf/AeqB.html this one's free for your computer. there is no escape.
20:17:55 <Taneb> I ought to get a copy of Euclid's Elements
20:18:47 <Bicyclidine> honestly i think reading classics is overrated
20:19:10 <Bicyclidine> their methods and culture are just so different from modern math
20:19:11 <Taneb> Even if it is just as a bookshelf decoration
20:19:20 <fizzie> They sold extra copies of SICP at the university bookstore (after our Introduction to Programming course switched from Scheme to Java or Python or something) for something like 5 EUR.
20:19:28 <fizzie> Some people bought lots, to give away to people.
20:19:49 <Bicyclidine> Taneb: also, you can get books really cheap on amazon, like five bucks USD (so probably a pound cos our economy's shit), such as Counterexamples in Topology
20:20:01 <Bicyclidine> which, iirc, was recommended by some esolang guy? cpresseY?
20:20:22 <Bicyclidine> http://esolangs.org/wiki/User:Chris_Pressey#Esoteric_Reading_List.21 yeah
20:21:32 <Bicyclidine> other than that there's like, that book on kolmogorov complexity that uses the taocp answer system
20:21:52 <Bicyclidine> or like, you like haskell probably, so one or two of pierce's books i guess
20:23:12 <Taneb> Types and Programming Languages costs £50 :(
20:25:06 <Bicyclidine> maybe get some cheapish ones... dover has a bunch of old (like, 60s?) good texts, like Introductory Real Analysis by Kolmogorov, for like ten bucks
20:25:30 <Bicyclidine> the cheapness because they're not hardcover :P
20:26:52 <Bicyclidine> i mean maybe you odn't give a shit about real analysis though, in which case you could just find a good used bookstore, or wnader through your university library
20:27:10 <Taneb> My university does have a library, hmm
20:27:22 <Bicyclidine> have you ever wanted three hundred years of research on generalized continued fractions? ~it's in the library~
20:27:45 <Bicyclidine> i mean you go like, york, right? it's not some puny community college, and even those are usually pretty good
20:27:51 <Taneb> I might wander in after my grand beginning-of-June adventure
20:28:34 <Bicyclidine> honestly if i have work or something to do i usually just do it in the library, and goof off by reading books instead of by reading the internet
20:29:52 <Taneb> Probably a good idea
20:30:21 <Bicyclidine> just a state school. Washington State University
20:30:28 <Bicyclidine> presently i am sitting here putting off reading about ears
20:31:21 <Bicyclidine> but, the engineering library is six stories, and has everything from Bifurcation Theory to The Soviet Anti-Plague System
20:31:41 <Taneb> We only have the one library
20:31:45 <shachaf> six stories? that's not a v. big library
20:31:49 <Taneb> Except for a number of tiny libraries dotted about
20:31:55 <Taneb> Which I have never seen
20:31:58 <shachaf> are they at least split across six books
20:32:13 <Taneb> Despite both the maths and computer science departments having a tiny library
20:32:18 <Bicyclidine> well, we have this six story library, and there's the main library that's two buildings with like three floors (but it's more history, social sorta stuff), and a veterinary library that's a few rooms
20:32:37 <Bicyclidine> oh, and are you a math student? i don't know how research works in math but you might try that
20:33:07 <Taneb> Maths and Computer Science
20:33:13 <Bicyclidine> the library at this campus is small by comparison, it has World of Warcraft Programming Second Edition
20:33:42 <Bicyclidine> well i don't know how that works but you might try looking at faculty profiles and asking somebody that seems interesting for work
20:33:51 <Bicyclidine> if you want to be a mathematologist, anyway, i don't know
20:34:57 -!- ais523 has joined.
20:34:58 <Taneb> Right, I've found the Computer Science library on Google Maps
20:35:00 <Bicyclidine> in my experience uni professors are more than willing to talk at somebody who displays any interest, even if they're a dirty undergrad
20:35:07 <fizzie> I've told this before, but there was a maths journal on the "recent issues" table at the university library, and it had a fascinating picture in the inner side of the cover, and it was titled: "Fig. 1: A fascinating picture."
20:36:05 <fizzie> I think it was some sort of a graph.
20:39:49 -!- ais523 has quit (Client Quit).
20:39:56 -!- callforjudgement has joined.
20:40:58 -!- callforjudgement has changed nick to ais523.
20:43:06 -!- vyv has joined.
20:45:08 <int-e> so true. "People get unhappy when a computer blinks its lights for"
20:45:09 <int-e> a while and then announces a result, if people cannot easily check the truth of the
20:45:12 <int-e> result for themselves.
20:47:25 -!- Patashu has joined.
20:49:14 <olsner> huh, I never check the results
20:50:22 <int-e> olsner: the audience is mathematicians. they prefer humanly checkable proofs. see also http://star.psy.ohio-state.edu/coglab/Miracle.html
20:53:48 <int-e> (Requiring too much work to check is also an issue, as can be seen in the criticism of the Appel & Haaken proof of the four color theorem, which was partly justified.)
20:54:56 -!- Patashu_ has joined.
20:54:57 -!- Patashu has quit (Disconnected by services).
20:55:09 -!- ais523 has quit (Ping timeout: 252 seconds).
20:56:40 -!- ais523_ has joined.
21:01:18 -!- Slereah_ has quit (Remote host closed the connection).
21:01:34 -!- Slereah_ has joined.
21:03:48 -!- Patashu_ has quit (Read error: Connection timed out).
21:04:28 -!- Patashu has joined.
21:05:29 -!- Bike has quit (Ping timeout: 255 seconds).
21:07:15 -!- Bike has joined.
21:15:03 -!- Sprocklem has quit (Quit: Leaving).
21:23:39 -!- tertu has quit (Ping timeout: 252 seconds).
21:26:19 <ais523_> huh, someone proved a false statement in Agda (presumably due to a compiler bug): https://proofmarket.org/problem/viewa/58
21:26:29 <ais523_> don't really know enough Agda to follow the proof, though
21:27:05 -!- Patashu has quit (Ping timeout: 255 seconds).
21:33:42 <kmc> working with a bunch of PL nerds is so much fun
21:36:54 <Taneb> ais523_, #agda has "We last proved false on <date>" in the topic
21:37:03 <Taneb> Currently 11th of May
21:37:29 -!- Frooxius has quit (Quit: *bubbles away*).
21:37:44 <HackEgo> Today is Pungenday, the 7th day of Confusion in the YOLD 3180
21:37:45 <ais523_> writing sound type systems is hard
21:38:00 <HackEgo> /home/hackbot/hackbot.hg/multibot_cmds/lib/limits: line 5: exec: sdate: not found
21:39:10 <kmc> I think somebody is working on a proof of soundness for Rust this summer
21:39:13 <kmc> non-mechanized :/
21:39:48 <zzo38> I also wrote a implementation of that ddate in ifMUD too, although the output format is a bit different: "Today is Pungenday, Confusion 7, YOLD 3180." Still, I am glad that these two dates are in agreement!
21:40:19 <Taneb> ais523_, the weird bit in that is the definition of eq
21:40:35 <Taneb> Which asserts that foo and bar are different
21:41:00 <Taneb> But they are both values of the unit type defined in slightly different ways
21:41:56 <ais523_> hmm... ((), 1), ((), ()) is different from ((), 0), ((), ()) in most programming languages
21:42:22 <Taneb> Yes, but the first element of the second element of those two tuples are the same in most programming languages
21:42:46 <Bicyclidine> so i guess the difference between 0 and 1 spreads somehow.
21:43:11 <zzo38> ais523_: Are they the people who wrote Agda they payment in case of the bug like Knuth does with TeX and METAFONT, or is it different?
21:43:45 <ais523_> zzo38: it's a bit different; the Agda devs don't pay for things like that, but the link in question indicates that someone did pay for the proof, anonymously
21:45:38 -!- edwardk has joined.
21:46:05 <zzo38> I think how the HOPE conference accepts payment by bitcoins too
21:51:23 <zzo38> That proof does not have any = line for eq; is that wrong?
21:51:54 <Taneb> zzo38, no, the "eq ()" means "There are no cases to match against"
21:52:10 -!- mihow has quit (Quit: mihow).
21:52:40 <Bicyclidine> it turns a proof that foo = bar into bottom?
21:53:03 <Taneb> Yes, because the type checker thinks foo != bar
21:53:10 <zzo38> How do you know that there are no cases to match against? How does the equivalence sign work anyways?
21:53:34 <Bicyclidine> zzo38: well the return type is bottom so it's not like it could return anything
21:58:12 -!- mihow has joined.
22:00:33 -!- ais523_ has quit (Quit: Page closed).
22:10:20 -!- Slereah_ has quit (Remote host closed the connection).
22:10:36 -!- Slereah_ has joined.
22:16:36 <idris-bot> Data type = : ({A0} : Type) -> ({B0} : Type) -> A -> B -> Type
22:16:36 <idris-bot> The propositional equality type. A proof that x = y.
22:16:36 <idris-bot> To use such a proof, pattern-match on it, and the two equal things will then need to be the same pattern.
22:16:36 <idris-bot> Note: Idris's equality type is heterogeneous, which means that it is possible to state equalities between values of potentially different types. This is
22:16:36 <idris-bot> sometimes referred to in the literature as "John Major" equality.↵…
22:19:40 <elliott> “It is now time to reveal the definition of ≃, the ‘John Major’ equality relation. (Footnote 2: John Major was the last ever leader of the Conservative Party to be Prime Minister (1990 to 1997) of the United Kingdom, in case he has slipped your mind.) John Major’s ‘classless society’ widened people’s aspirations to equality, but also the gap between rich and poor. After all, aspiring to be equal to others than oneself is the ...
22:19:46 <elliott> ... politics of envy. In much the same way, ≃ forms equations between members of any type, but they cannot be treated as equals (ie substituted) unless they are of the same type. Just as before, each thing is only equal to itself.”
22:21:14 <Taneb> Phantom_Hoover, I'd presume it's an old quote
22:21:16 <elliott> Phantom_Hoover: he was an optimist at the time
22:22:14 <Phantom_Hoover> i thought it was some kind of pedantic joke, like with labour/new labour
22:22:39 <Bicyclidine> this is the view from the left guy, right? some kind of "Left ist"
22:23:49 <shachaf> whoa, I never ended up reading /The view from the left/. :-(
22:25:35 <Bicyclidine> i'm not in the right mindset to find mcbride's jokes funny, tho :/
22:28:58 <zzo38> I have used a kind of equality that works between different types, in Haskell, by using (Eq, Typeable)
22:31:27 -!- Phantom__Hoover has joined.
22:31:31 <zzo38> Do you know much about LALR(1) parsing?
22:34:07 <lambdabot> <hint>:1:1: parse error on input ‘:’
22:34:16 <idris-bot> A proof that x in fact equals x. To construct this, you must have already shown that both sides are in fact equal.
22:34:16 <idris-bot> (implicit) {A0} : Type -- the type at which the equality is proven
22:34:16 <idris-bot> (implicit) {x0} : A -- the element shown to be equal to itself.
22:34:47 -!- Phantom_Hoover has quit (Ping timeout: 252 seconds).
22:35:47 <Melvar> ( :t vectNilRightNeutral
22:35:47 <idris-bot> Prelude.Vect.vectNilRightNeutral : (xs : Vect n a) -> xs ++ [] = xs
22:36:09 <Melvar> ↑ use of heterogeneous equality
22:37:10 <Bicyclidine> it's funny how i keep thinking of things in terms of taking and returning arguments, when that's clearly fucking hopeless
22:38:31 <Bicyclidine> basically i have no idea what any of this means
22:38:58 <Melvar> ( :t \n,a, xs : Vect n a => xs ++ []
22:38:58 <idris-bot> \n => \a => \xs => xs ++ [] : (n : Nat) -> (a : Type) -> Vect n a -> Vect (n + 0) a
22:40:51 <Melvar> The left side of the equals is a Vect (n + 0) a, the right side is a Vect n a, which don’t immediately unify (you can prove they’re equal though), so if the equality were homogeneous, you’d have to rewrite one side within the type of vectNilRightNeutral to state the equality.
22:40:51 <elliott> Bicyclidine: think of it as a forall
22:43:20 <shachaf> "taking and returning arguments" is a good way to think about it.
22:43:45 <lambdabot> <hint>:1:1: parse error on input ‘:’
22:43:49 <idris-bot> Prelude.Basics.the : (a : Type) -> a -> a
22:44:17 * Melvar kicks self for mixing up his prefixes for the second time.
22:46:29 <zzo38> Why is Bison parser generator so much larger (when compressed) than Lemon parser generator (when uncompressed)?
22:46:53 <Melvar> But anyway, the is the way to do type annotations in idris.
22:46:57 <idris-bot> Control.Category.id : Category cat => cat a a
22:47:26 <Melvar> id is the same, but with an implicit argument (hence the a is underlined).
22:47:32 -!- mihow has quit (Quit: mihow).
22:48:30 <Taneb> https://scontent-b.xx.fbcdn.net/hphotos-xfa1/t1.0-9/s480x480/282298_10150894076356344_561739999_n.jpg
22:48:45 -!- mihow has joined.
22:50:37 <kmc> o_O https://github.com/mozilla/rust/blob/master/src/librustc/middle/ty.rs#L1091-L1142
22:51:22 -!- Sgeo has joined.
22:53:00 -!- aretecode has joined.
22:53:08 <int-e> kmc: that should not come as a surprise after seeing lines 240 to 259.
22:53:50 <Sgeo> http://www.babylonjs.com/ not so sure about the library, but I love these demos
22:54:03 <Sgeo> I don't know why people are saying it's a Microsoft clone of Three.JS
22:54:19 <kmc> pub trait_impls: RefCell<DefIdMap<Rc<RefCell<Vec<ast::DefId>>>>>
22:54:23 <kmc> that's quite the type
22:57:20 <int-e> what's the Rc<> thing?
22:57:34 <mcpherrin> int-e: Rc<T> is a reference counted T
22:57:37 <kmc> Rc<T> is a refcounted box containing T
22:57:48 <kmc> even if T isn't cloneable, you can clone Rc<T>; you get another pointer to the same underlying object
22:57:53 <kmc> it's much like std::shared_ptr<T> in C++
22:58:12 <kmc> RefCell<T> is "interior mutability"; it basically moves Rust's alias checking to runtime
22:58:14 <Sgeo> You can clone, but can you send? (I'd hope not... I think?)
22:58:20 <kmc> Sgeo: nope, but you can send Arc<T>
22:58:23 <kmc> ("atomic reference counting")
22:58:38 <kmc> if you have a no-mutation reference to an object, you can still mutate stuff inside a RefCell
22:58:49 <kmc> i.e. you can go from &RefCell<T> to &mut T, approximately
22:59:24 <kmc> doing so sets a flag at runtime saying "this thing is borrowed mutably", and what it gives you is actually a smart pointer which un-sets that flag on destruction
22:59:30 -!- mihow has quit (Quit: mihow).
22:59:44 <kmc> if the value has already been borrowed, it's a task failure
23:00:07 <kmc> DefId and DefIdMap are rustc internal types for definitions and maps keyed on definitions
23:00:22 <Melvar> Why does that need a special Map?
23:00:38 <kmc> well it's a typedef: http://doc.rust-lang.org/rustc/util/nodemap/type.DefIdMap.html
23:00:58 <kmc> and FnvHashMap is another typedef for a standard collections::hashmap::HashMap
23:01:29 <kmc> "FnvHasher: ... A speedy hash algorithm for node ids and def ids. The hashmap in libcollections by default uses SipHash which isn't quite as speedy as we want. In the compiler we're not really worried about DOS attempts, so we just default to a non-cryptographic hash."
23:03:08 <kmc> whoa, TIL rust has default type parameters
23:03:12 -!- madbr has joined.
23:03:16 <kmc> pub struct HashMap<K, V, H = sip::SipHasher> { ...
23:03:40 <mcpherrin> kmc: haha yeah they're pretty shiny
23:04:02 <mcpherrin> Hashmap was the first user, and I'd bet $1 still the only major one :p
23:04:13 <zzo38> Is this how a LALR(1) parser is supposed to work? list(A) ::= beginlist(B) listitems RBRACKET. { A=doc->group; doc->group=B; } beginlist(A) ::= LBRACKET. { A=doc->group; doc->group=zcdsf_value_empty_list(0); }
23:04:13 <mcpherrin> (allocators were/are going to do the same
23:04:36 <zzo38> I don't quite understand it if it is OK or not
23:05:15 <fowl> in my progrmaming language " will be the ditto operator, it refers to the operator or word above it
23:05:36 <mcpherrin> fowl: immediately above, as in line-column wise?
23:06:08 <Melvar> kmc: How does a Hasher work independently of the key type?
23:06:19 <olsner> something like a default initialization would be pretty great for that initializer of doom
23:06:33 <mcpherrin> Melvar: The key type implements a trait
23:06:50 <Bicyclidine> Can you define hash functions for your own types?
23:07:29 <Melvar> mcpherrin: And this is not noted on the type declaration, or is kmc’s above incorrect?
23:07:34 <Bicyclidine> assuming you don't need other equality, i guess
23:08:44 <mcpherrin> Melvar: the Hashmap itself doesn't care, just the hasher
23:10:16 <mcpherrin> The traits implemented on the HashMap do, though
23:10:40 <mcpherrin> impl<K: Eq + Hash<S>, V, S, H: Hasher<S>> HashMap<K, V, H>
23:11:22 <kmc> well that's not a trait for HashMap, it's just the plain non-trait methods of HashMap, but otherwise yeah
23:12:04 <mcpherrin> yeah I just copied the first relevant line from rustdoc:P
23:12:08 <mcpherrin> impl<K: Eq + Hash<S>, V, S, H: Hasher<S>> MutableMap<K, V> for HashMap<K, V, H>
23:12:08 <Melvar> Ah, so the constraint is not on the data structure, but on the functions. Thank you.
23:12:22 <kmc> I don't think we even support constraints on data structures?
23:12:35 <kmc> I am not sure
23:12:52 <Taneb> I am not very good at person
23:13:26 <Taneb> I feel that that should have a winky face somewhere
23:13:47 <idris-bot> BotPrelude.ShowGlue : a -> String => Show a
23:13:47 <mcpherrin> kmc: Uh, apparently not trait bounds though :P
23:13:59 <kmc> mcpherrin: what other kind of constraints are there?
23:14:06 <kmc> Taneb: what aspect(s)?
23:14:23 <kmc> you mean like struct Foo<T: 'static> { ... } ?
23:14:39 <Taneb> kmc, currently, putting bedsheets back on beds
23:14:40 <kmc> 'static is the only lifetime which can be used that way, right?
23:14:50 <kmc> I still don't totally undersand what it means
23:15:04 <mcpherrin> I don't know if I've ever seen a struct templated that way
23:15:09 <Bicyclidine> oh no is rust continuing the proud tradition of static being wonky as hell
23:15:12 -!- yorick has quit (Read error: Connection reset by peer).
23:15:35 <mcpherrin> Bicyclidine: 'static isn't C's whatever-we-want-today keyword :-)
23:15:49 <mcpherrin> It's the lifetime of data that lives "forever"
23:16:10 <Bicyclidine> oh so like the kind of thing that is serialized into a binary, ok.
23:16:35 <mcpherrin> you can have static mutable data, but it's unsafe
23:16:40 <Melvar> I think idris has a static keyword somehow … it indicates that some argument must be fully known at compile time I think.
23:16:42 <kmc> because it gives you task-shared mutability
23:16:51 <kmc> so you can only use it in the unsafe dialect
23:17:08 <kmc> it's fine, it's just like "let mut"
23:17:24 <mcpherrin> kmc: were you around for the great "what to call static" debates?
23:17:39 <Bicyclidine> it's stored in the dynamic library binary, so, call it dynamic
23:18:12 <kmc> a 'static bound on a type asserts that all refs in the type have static lifetime, I think, but I'm not sure
23:18:49 <kmc> mcpherrin: no
23:18:52 <kmc> Bicyclidine: :3
23:19:58 <kmc> std::cell defines Cell and RefCell and Ref and RefMut B|
23:20:04 <kmc> I know what they all need, but the naming is a bit confusing
23:20:29 <kmc> what they all mean*
23:23:25 <mcpherrin> "`static mut` looks weird" was a common opinion
23:28:49 <kmc> "life is weird, get used to it"
23:29:06 -!- nooodl has quit (Quit: Ik ga weg).
23:32:07 <Bicyclidine> more C questions: if i calloc an array of doubles, does the zero memory have to equal numeric zero?
23:33:02 <zzo38> The zero memory corresponds to integer zero. I have not had problems using it as floating point, though.
23:33:41 <scp> Bicyclidine: yes
23:33:42 <fowl> ieee floats are
23:34:39 <scp> this is because of the (useful/planned) coincidence that double(+0.0f) has the same representation as a uint64_t(0)
23:34:52 <scp> calloc doesn't know anything about floats/doubles
23:35:29 <Bicyclidine> i'm imagining the ol' fuckyoutron 8000 having a 0.0f not made of zero bytes
23:35:36 <HackEgo> 847) <zzo38> What is portable way of load/save floating points in files, using a C code? <kmc> #ifndef __STDC_IEC_559__ #error Here's a nickel, kid. Buy a real computer. #endif
23:36:02 <Bicyclidine> i mean, i don't know the ins and outs of floats.
23:36:14 <elliott> I think this might be one of the things you could DS9k
23:37:04 <Bicyclidine> this code is bullshit conformance-wise so i'm asking out of curiosity more than anything :P
23:38:55 <olsner> the boring part about actually building a ds9k is that no existing software would build and work on it
23:39:14 <zzo38> olsner: I doubt it; some software may work.
23:39:51 <zzo38> Also, some software might compile but fail to run properly, while others will result in compile errors since the program contains things to make the compiler to check.
23:40:52 <olsner> I doubt OpenSSL is actually valid enough C, it rather has a wide collection of differently non-conformant hacks, and would need a completely new set to work on ds9k
23:41:31 <zzo38> For example you use the negative-array-size trick.
23:43:37 <Sgeo> So many abstractions in the WebGL world
23:43:56 <Sgeo> tQuery = abstraction of ThreeJS = abstraction of abstraction of WebGL
23:45:01 <olsner> Bike: the fuckyoutron8k could be the budget model of the ds9k
23:45:58 <Bicyclidine> i don't know why i've never seen static variables used for defining a filter, that i can remember... outside of that one OS i guess
23:46:54 <zzo38> I have used negative-array-size trick, which is like: struct { char unused[sizeof(long)>=sizeof(void*)?1:-5]; } and similar things like that.
23:46:57 <kmc> http://shitgonutssays.tumblr.com/
23:47:14 <zzo38> A variable of this structure type is not declared, therefore don't waste memory.
23:47:53 <zzo38> How common is stuff like this?
23:48:58 <Bicyclidine> pike seems like kind of an ass from this, o well
23:49:59 <kmc> just a little ;_;
23:50:38 <Bicyclidine> that childish things quote was so much improved by lewis
23:52:08 <Taneb> I am also not very good at booking train tickets
23:53:41 <Phantom__Hoover> i have lost an awful lot of money to my own stupidity in and around trains
23:53:42 <Bicyclidine> this code is rapidly making me think C is not good for writing signal processing in
23:54:19 <Bicyclidine> i don't know what could possess something to define a basic looking linear filter on the last few intermediates, and then allocating storage for intermediates over the entire input signal
23:54:29 <Bicyclidine> possess someone*, i imagine this isn't computer generator