01:33 (join) gciolli
01:43 (quit) gciolli: Ping timeout: 240 seconds
01:45 (quit) realitygrill: Quit: realitygrill
02:05 (quit) jeapostrophe: Ping timeout: 248 seconds
02:09 (quit) yoklov: Quit: computer sleeping
02:24 ozzloy: how do i build the guide after making some change to it?
02:24 ozzloy: in the source distribution.  without doing make; make install
02:26 jonrafkind: raco setup scribblings/guide or something
02:26 jonrafkind: yea that should work
02:26 jonrafkind: it wont be super fast though
02:27 (quit) dnolen: Remote host closed the connection
02:30 ozzloy: will it rebuild the entire guide?
02:30 ozzloy goes to check
02:33 ozzloy: interesting.  is there a faster way to rebuild just the part affected by cond.scrbl, hypothetically speaking?
02:34 ozzloy: also thanks, this is at least better than make; make install
02:34 jonrafkind: well I think you can do 'scribble cond.scrbl' possibly
02:34 ozzloy: i tried that
02:34 jonrafkind: im not sure where the output goes
02:34 ozzloy: it makes a cond directory in the current directory
02:34 jonrafkind: hm
02:34 ozzloy: which is ok i guess
02:35 ozzloy: i could just do that and get over it
02:35 ozzloy: but i was hoping for a way to be reading through the guide, see a problem, fix it, continue reading
02:35 ozzloy: if i do it that way, it doesn't link with the rest of the guide
02:35 ozzloy: it also doesn't show up the same
02:35 ozzloy: the subsections each get there own page
02:36 ozzloy: which is interesting that it can change based on if it's part of a larger collection.  very neat
02:55 (quit) jonrafkind: Ping timeout: 260 seconds
03:02 veer: how do I customize what is printed in the repl , or is there a some kind of hook into print, write,etc
03:04 veer: I need something like prop:custom-write for tagged pointer
03:05 mithos28: veer: why cannot you write a struct to wrap the tagged pointer?
03:07 veer: mithos28: I can , but that adds another layer of wrapping and unwrapping , is there another way ?
03:07 mithos28: not that I know of
03:08 veer: ok np
03:17 Gertm: Is there a curses/ncurses library for Racket?
03:21 (join) sindoc
03:26 (quit) veer: Read error: Connection reset by peer
03:34 (join) Blkt
03:37 Blkt: good morning everyone
03:48 (join) Shvillr_
03:48 (quit) Shviller: Disconnected by services
03:48 (nick) Shvillr_ -> Shviller
04:18 (join) noelw
05:01 (join) q0tw4
05:07 (join) tfb
05:11 (join) karswell__
05:12 (quit) karswell_: Ping timeout: 240 seconds
05:19 (quit) mithos28: Quit: mithos28
05:29 (join) bluezenix
05:50 (join) veer
05:53 (quit) Frozenlock: Read error: Connection reset by peer
05:53 (join) Frozenlo`
06:44 (quit) Lajla: Ping timeout: 276 seconds
06:48 (quit) bluezenix: Ping timeout: 252 seconds
06:56 (join) masm
07:18 (quit) kvda: *.net *.split
07:23 (join) kvda
07:35 (join) yoklov
08:11 (join) jeapostrophe
08:20 (quit) veer: Quit: Leaving
08:28 (quit) yoklov: Quit: computer sleeping
08:40 (join) gciolli
08:48 (quit) gciolli: Ping timeout: 276 seconds
08:48 (quit) Shviller: Ping timeout: 252 seconds
08:50 (quit) jeapostrophe: Ping timeout: 276 seconds
08:59 bremner: hmm, I kindof miss implicit universal quantification of type variables (a la Haskell) in Typed Racket. wrapping everything in (All (A B) ... ) seems a bit boring.
09:05 noelw: yeah
09:10 (join) sindoc1
09:14 (quit) sindoc: Ping timeout: 244 seconds
09:17 (join) yoklov
09:21 (join) bluezenix
09:24 bremner: I guess in Haskell the tradeoff is that you need to use lower case for type variables, iirc
09:28 (join) gciolli
09:33 (join) jrslepak
09:36 (join) jeapostrophe
09:39 RacketCommitBot: [racket] plt pushed 3 new commits to master: http://git.io/tgBMNg
09:39 RacketCommitBot: [racket/master] adjust a doc correction - Matthew Flatt
09:39 RacketCommitBot: [racket/master] change `ffi-lib' to not make library symbols global by default - Matthew Flatt
09:39 RacketCommitBot: [racket/master] improve Guide info on `PLTCOLLECTS' and `raco link -d' - Matthew Flatt
09:58 (join) karswell
10:00 (quit) karswell__: Ping timeout: 260 seconds
10:01 (quit) jeapostrophe: Ping timeout: 252 seconds
10:03 (quit) chemuduguntar: Read error: Connection reset by peer
10:04 (quit) rsimoes: Ping timeout: 240 seconds
10:09 (quit) Fulax: Read error: Connection reset by peer
10:12 (quit) gciolli: Quit: Leaving.
10:18 bremner: any typed racket expert can explain to me why this example (from eli's lectures) doesn't typecheck? Is it because the type of map is actually too general?
10:18 bremner: oops: this example http://paste.debian.net/153715/
10:19 ianjneu: Yes, you must always give type instantiations when giving a polymorphic function to another polymorphic function.
10:19 bremner: ah, right. foo.
10:19 bremner: ok, thanks
10:19 ianjneu: It's a limitation in the theory, not just Typed Racket.
10:23 (quit) yoklov: Quit: bye!
10:32 (join) dnolen
10:44 (quit) masm: Read error: Operation timed out
10:44 (quit) shadgregory: Quit: ERC Version 5.3 (IRC client for Emacs)
10:46 (join) masm
10:48 (join) shadgregory
10:48 (nick) chaozzbubi -> ChaozZBubi
10:55 (join) mceier
10:57 (quit) karswell: Ping timeout: 240 seconds
10:57 (join) karswell
10:58 eli: bremner: That example is only for intimidation.
11:03 (join) jeapostrophe
11:15 (quit) sindoc1: Quit: Leaving.
11:20 (quit) dsantiago: Quit: Computer has gone to sleep.
11:25 (join) sindoc
11:29 (join) realitygrill
11:54 (quit) jrslepak: Quit: This computer has gone to sleep
11:56 (quit) bluezenix: Quit: Leaving.
11:57 (quit) tfb: Quit: sleeping
12:02 (quit) realitygrill: Quit: realitygrill
12:05 bremner: haha
12:05 bremner: well, it worked on me ;)
12:07 (join) realitygrill
12:08 ASau: Good evening!
12:08 noelw: A great evening, I'd say!
12:08 ASau: I've got pair of questions.
12:09 ASau: 1. Will NetBSD patch land into 5.2.1?
12:09 ASau: Is there a branch or is 5.2.0.901 a snapshot?
12:09 ASau: 2. What needs to be done so that it hits trunk at least?
12:13 (nick) noelw -> noelw_away
12:15 (quit) Blkt: Read error: Connection reset by peer
12:22 (join) anRch
12:27 (join) mithos28
12:32 (quit) sindoc: Quit: Leaving.
12:51 (quit) anRch: Quit: anRch
12:58 (join) bluezenix
13:09 (join) jonrafkind
13:17 (quit) realitygrill: Quit: realitygrill
13:21 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/w4AJlg
13:21 RacketCommitBot: [racket/master] Revise commented code so it can replace exiting code without dependency issues. - Eli Barzilay
13:22 (join) realitygrill
13:34 (join) Shviller
13:51 (join) jrslepak
14:14 (join) anRch
14:15 (join) rsimoes
14:20 danlndi: anyone know an example of a function in the FFI return a Scheme_Object pointer (like how the *old* FFI worked)?
14:21 (quit) q0tw4: Remote host closed the connection
14:25 (join) sindoc
14:28 eli: danlndi: The old ffi?
14:28 eli: But in any case, grep the tree for `_scheme'.
14:30 ASau: eli: any chance my (NetBSD) fix lands in trunk at least?
14:30 (join) RyanFraise
14:31 RyanFraise: Hello
14:32 RyanFraise: Anyone here
14:33 (quit) RyanFraise: Client Quit
14:40 offby1: nobody here but us zombie processes.
14:40 (join) Fulax
14:47 (quit) dous__: Remote host closed the connection
15:13 (quit) ASau: Read error: Connection reset by peer
15:15 (quit) anRch: Quit: anRch
15:18 danlndi: eli: what should the ffi type _bytes look like on the C side?
15:18 danlndi: unsigned char * ?
15:43 (join) dous
15:43 (quit) dous: Changing host
15:43 (join) dous
15:47 (quit) dous: Ping timeout: 240 seconds
15:57 RacketCommitBot: [racket] plt pushed 12 new commits to master: http://git.io/_5TFNA
15:57 RacketCommitBot: [racket/master] [honu] allow list comprehensions to follow expressions - Jon Rafkind
15:57 RacketCommitBot: [racket/master] [honu] then is optional in "if". allow multiple sequences in for. add for/fold - Jon Rafkind
15:57 RacketCommitBot: [racket/master] [honu] make "then" keyword required for if again. start to add more loop types to for - Jon Rafkind
15:59 eli: danlndi: Yes.
16:00 jonrafkind: eli, list comphrensions in honu are core to the parser because [] is also used as list-ref
16:00 eli: Someone should tell ASau that it would help to have a BSD build machine, but until then things will continue to be rough, probably.
16:01 eli Keeps staring intently at rudybot
16:01 eli: jonrafkind: Bah.
16:01 eli: gabot: slap rudybot
16:04 eli: jonrafkind: I also had about 37 other items.
16:04 (join) MayDaniel
16:05 (quit) kvda: Quit:  x___x
16:05 jonrafkind: i changed a few things, look at whats new!
16:06 jonrafkind: also blah(x){ 1 } is the same thing as (define (blah x) 1)
16:06 jonrafkind: i thought that would have been obvious but I guess it wasn't
16:07 jonrafkind: and its not 'when E then { E ... }' its 'when E then E', so { E ... } is just a block
16:08 eli: jonrafkind: Look where?  Your source is full of debugging stuffs, and even your commit messages...  (Replace `:=' with `=', oops, go back, no, try again, oh, two references there...)
16:08 jonrafkind: at asteroids
16:08 eli: And re `blah(x){ 1 }' -- yeah, I fugured that, it's exactly why I said that it's odd.
16:08 jonrafkind: ill file your sarcasm in my sarcasm folder to laugh at later
16:08 jonrafkind: why is it odd
16:08 jonrafkind: you want 'fun blah(x){ 1 }' ?
16:08 eli: Too short.
16:09 (quit) jrslepak: Ping timeout: 244 seconds
16:09 eli: Lacks clear visual indicator of a function start.
16:09 jonrafkind: its vestigal
16:09 eli: It's obscure.
16:10 jonrafkind: you know most of your comments make me thing you don't want anything to change -- you like the past 30 years of language syntax and you want it to stay that way
16:10 eli: foo(x){bar(y){x+y};bar}
16:10 (quit) karswell: Ping timeout: 252 seconds
16:11 eli: (a) I have no sympathy to N years of *that* kind of syntax.
16:11 jonrafkind: what is *that*
16:11 eli: {}-ish.
16:11 eli: (b) Respecting common syntax is -- AFAIK -- one of the main goals of this project.
16:12 eli: (c) Talking about `then' in `when' -- it *reeks* of old style syntax.  Mandatory `then's practically died with Pascal.
16:12 jonrafkind: its either 'if (x){}' or 'if x then {}'
16:12 jonrafkind: I like the latter
16:13 eli: Yes, that's definitely much more of a "I like the past 30 years".
16:13 eli: (See also bash.)
16:13 jonrafkind: well.. its actually more of trying to be like python
16:13 eli: (That's not a compliment.)
16:13 jonrafkind: if x == 2: ...
16:13 jonrafkind: no () in there
16:13 eli: Python is not a good syntactic goal to get the masses in.
16:13 jonrafkind: I just replaced : with then
16:14 eli: Bad choice then, since python uses it for more than just a `then'.
16:14 jonrafkind: im mixing up cultures according to my own preference
16:14 offby1: whachoo staring intently at rudybot for
16:15 jonrafkind: matthew said the same thing about () vs then
16:15 eli: jonrafkind: Again -- if the goal is to get average hackers into racket, then mixing cultures is the *absolute* best way to get people of all cultures to fart in your general direction.
16:16 eli: offby1: It's lack of "later tell" without the now-deceased minion.
16:16 eli: jonrafkind: Also, allowing people freedoms like "if (foo) E" and "if foo then E" is something that several languages tried, and I think that all cases ended up dead.
16:17 eli: People would just settle on one of them, forgetting the other that would be left for obfuscated code bits.
16:17 jonrafkind: if C had 'if foo then ...' we would all be using that
16:17 jonrafkind: its hard to know what syntactic lessons have been learned over the years
16:18 eli: jonrafkind: See, that's the sentimentalism that you accused me of.
16:18 eli: The thing is that I've been using sexpr languages for long enough that I have no sentiments whatsoever.
16:18 jonrafkind: aha yes, thats true. one the one hand you have 30 years of knowing what works and what doesn't, but at the same time I feel that maybe you don't like new things
16:18 jonrafkind: oh
16:18 eli: The only point that matters is making a familiar language.  Period.
16:19 eli: I have no aspirations of doing the right thing with `then', and no desire to discuss whether C was right in omitting it.
16:19 jonrafkind: so then you are arguing by proxy for other people who will have an opinion about C-style syntax
16:19 eli: The only thing that matters is that people just don't use that -- the rest is just {}-style bikeshedding, and since I don't like {}-ish languages it's not that interesting...
16:20 eli: Yes, of course.
16:20 eli: But more like Javascript these days than C.
16:20 jonrafkind: ah well, I thought you wanted to use the language for something
16:20 (quit) MayDaniel: Read error: Connection reset by peer
16:20 jonrafkind: you want/wanted to use it in your class?
16:30 (join) LeNsTR
16:52 em: Is there a racket package in Ubuntu 11.04 ?
16:53 jonrafkind: yea, there is a ppa
16:54 jonrafkind: https://launchpad.net/~plt/+archive/racket
16:54 jonrafkind: do you know how to install ppa's?
16:57 (join) francisl
17:07 eli: jonrafkind: Sure -- I'd love to use it in my class, and exactly the same arguments hold with people knowing JS syntax; if I give them a "mix of cultures" they'd definitely treat it as a weird toy.
17:07 eli disappears
17:08 jonrafkind: so the question is which syntax to choose. C, javascript, ruby, python, scala, something else
17:09 eli: In order of what I see in class: javascript, java (which you didn't include), C, python, ruby.
17:09 eli: (And scala is so far, it's even less known than racket.)
17:09 eli disappears for real now
17:16 (join) jao
17:25 RacketCommitBot: [racket] plt pushed 1 new commit to master: http://git.io/0oxGsQ
17:25 RacketCommitBot: [racket/master] [honu] fix tests - Jon Rafkind
17:25 jaimef ponders converting some clojure code to racket
17:33 (quit) dnolen: Ping timeout: 240 seconds
17:35 offby1 ought to go the other way
17:36 offby1: just as a way to learn clojure, you understand
17:37 jonrafkind: i started doing that clojure 'solve a puzzle' thing
17:37 jonrafkind: its a pretty good way to learn a language imo
17:39 offby1: I should do that
17:43 (join) jrslepak
17:44 (quit) jrslepak: Client Quit
17:45 (join) jrslepak
17:45 em: jonrafkind: yep I can do the ppa thanks.
17:48 (quit) francisl: Quit: francisl
17:52 (join) Lajla
17:56 (join) sindoc1
17:58 (quit) mceier: Quit: leaving
17:59 (quit) sindoc: Ping timeout: 255 seconds
18:00 (quit) realitygrill: Quit: realitygrill
18:00 (part) sindoc1
18:01 (join) dous
18:03 (quit) jrslepak: Quit: This computer has gone to sleep
18:10 (quit) jeapostrophe: Read error: Operation timed out
18:41 (nick) offby1 -> offby1-otron
18:41 (nick) offby1-otron -> offby1
18:42 (join) karswell
18:53 (join) chemuduguntar
19:03 (join) karswell_
19:05 (quit) karswell: Ping timeout: 252 seconds
19:25 (quit) bluezenix: Quit: Leaving.
19:30 (join) jeapostrophe
19:32 (join) karswell__
19:34 (quit) karswell_: Ping timeout: 252 seconds
19:41 (quit) tauntaun`: Remote host closed the connection
19:43 (join) zdeus
19:43 (quit) zdeus: Remote host closed the connection
19:43 (join) realitygrill
19:43 (join) tauntaun
19:43 (nick) tauntaun -> Guest55743
19:44 em: What is the definition of 'contracts' in Scheme. Can someone give me a brief explanation for a layperson?
19:44 em: Like what is their primary purpose
19:49 offby1: for checking that you've passed the right sorts of values into a function
19:49 offby1: and similarly checking the function's return value
19:49 offby1: kinda like Eiffel's "programming-by-contract", if that means anything
19:50 offby1: have you ever written assertions at the start of your function, so that it'll croak if you accidentally pass (say) a string where you should have passed a number?
19:50 offby1: That's all they are; they're just a special syntax for doing that
19:54 (nick) Guest55743 -> tauntaun
19:54 (nick) tauntaun -> Guest98150
19:56 (nick) Guest98150 -> tauntaun
20:01 (quit) karswell__: Read error: Operation timed out
20:02 (join) karswell__
20:02 em: offby1: so is it like a type-checking system then?
20:03 em: offby1: basically it's a way to get the advantages of static type?
20:08 (quit) tauntaun: Quit: ERC Version 5.3 (IRC client for Emacs)
20:14 danlndi: ok, can anyone help with a 'link: reference (phase 0) to a variable in module "blahblah" that is uninitialized (phase level 0); reference appears in module "blahblah2"' error?
20:14 danlndi: the first module does require the second (by file name)
20:15 offby1: em: well, it's a little like type-checking, but it all happens at runtime
20:15 offby1: whereas the usual static type checker happens as compile time, which to me is nicer
20:16 offby1: since at compile time, it can check _all_ your functions; whereas at runtime, only those functions that actually get called will get checked.
20:16 offby1: So if you're a test-driven-development die-hard, the benefits will be similar ...
20:16 offby1: ... oh and contracts slow down your program.
20:16 offby1: So this is why we have typed racket :)
20:21 Frozenlo`: I see there's a `first' and `rest' function. Should I stop using `car' and `cdr'?
20:21 jonrafkind: yes
20:22 danlndi: car cdr are kind of the slang of scheme/racket... right?
20:22 offby1: "slang"?
20:22 jonrafkind: they are 50 year old procedures
20:22 Frozenlo`: What about cadr? :P
20:22 jonrafkind: second
20:22 danlndi: although there's no substitude for cadr cdar cadar etc etc etc
20:22 jonrafkind: oh well..
20:23 jonrafkind: yes thats true
20:23 danlndi: actually, the substitude is actually to not organize your data that way. bad racket programmer!
20:23 jonrafkind: yes, exactly :p
20:23 jonrafkind: lists are a terrible ad-hoc data structure
20:24 danlndi: substitude => substitute
20:24 Frozenlo`: How so? Lists are really easy to understand. When you really need more power, you can switch to hashtables later.
20:25 Frozenlo` loves lists of lists of lists....
20:25 jonrafkind: most people make that transition too late
20:25 jonrafkind: and then end up with spaghetti code
20:26 danlndi: actually, lists are a pretty good ad-hoc data structure, but using ad-hoc data structures should be confined to prototyping if used at all
20:26 jonrafkind: hashtables are the best adhoc datastructure
20:27 danlndi: yup. only drawback has been poor standardization
20:28 jonrafkind: in what, scheme?
20:30 danlndi: yeah...
20:30 danlndi: see srfi-69 and srfi-13
20:31 (join) karswell
20:32 (quit) jeapostrophe: Ping timeout: 255 seconds
20:33 (quit) karswell__: Ping timeout: 276 seconds
20:34 (quit) jonrafkind: Ping timeout: 240 seconds
20:41 (join) jeapostrophe
20:42 (join) dsantiago
20:44 (quit) Lajla: Ping timeout: 252 seconds
20:55 (join) 16WAAN2SU
20:59 Frozenlo`: "while a base-2 IEEE floating-point number can represent 1/2 exactly, it can only approximate 1/10:"
20:59 Frozenlo`: Oh
20:59 (nick) ChaozZBubi -> chaozzbubi
20:59 Frozenlo`: ooooooooooooooooooh!
20:59 Frozenlo`: :D
21:02 (join) karswell_
21:03 (quit) karswell: Ping timeout: 255 seconds
21:06 danlndi: I bet it can represent 1/4, 1/8, 1/16, etc exactly as well
21:07 bremner: except the etc.. only goes so far.\
21:11 danlndi: i wonder how for for single precision vs double
21:11 danlndi: how far
21:13 chandler: an IEEE float is represented as a tuple of (sign, exponent, fraction)
21:14 chandler: logically the exponent determines what values 1/(2^n) can be stored exactly...
21:14 (join) vkz
21:31 (quit) karswell_: Read error: Connection reset by peer
21:31 (join) karswell_
21:32 danlndi: yeah it depends on the number of bits in the exponent
21:41 (part) vkz
21:42 (join) vkz
21:42 (part) vkz
21:43 (join) vkz
21:59 Frozenlo`: I see a patern here... set! and hash-set!. Is the exclamation mark for when you are modifying a variable?
22:01 bremner: yup
22:09 (quit) masm: Quit: Leaving.
22:11 (quit) vkz: Quit: vkz
22:19 (quit) karswell_: 
22:55 (join) jrslepak
23:02 (quit) jrslepak: Ping timeout: 245 seconds
23:07 (quit) Fulax: Ping timeout: 252 seconds
23:11 (join) jonrafkind
23:12 (join) jrslepak
23:16 eli: jonrafkind: I know what wanting something yesterday means.
23:16 eli: I wanted it a year ago.
23:16 jonrafkind: heh ok
23:16 eli: (By the same scale)
23:16 eli: (And it's not like ubuntu is the only place you'd want it, btw)
23:17 jonrafkind: righyt
23:17 jonrafkind: i could do it myself but I hate messing with the package stuff
23:18 (quit) jrslepak: Quit: This computer has gone to sleep
23:18 eli: jonrafkind: Not only that, I think that it's still easy to fall into platform-specific traps.
23:19 eli: For example, I don't know what's that `update-icon-caches' is -- I don't have it.
23:19 jonrafkind: ah, i have that but i have no idea what it is
23:19 eli: I do have the update-mime-database, but I don't know if it's reliable.
23:21 jonrafkind: ok so eli 'function foo(x){ }' == (define (foo x) ...) and 'function (x){ ... }' == (lambda (x) ...) ?
23:25 eli: jonrafkind: Yes, just like javascript is fine.  But just make it `fun' like everybody actually wants it to be.
23:25 jonrafkind: ffffffffffffffffffffffffffffff
23:25 eli: Yes, this is earth shuttering, I know.
23:26 jonrafkind: javascript uses function!
23:26 jonrafkind: everyone loves it
23:26 eli: s/u/a/
23:26 eli: Most hate it.
23:26 jonrafkind: people make songs and dance around camp fires in praise of 'function'
23:26 eli: It makes quick lambda-like expressions verbose.
23:26 eli: (Which is what JS people discovered they can do, and therefore want it to be a shorter thing.)
23:26 jonrafkind: dont you mean 'it mks qk lmba-lk exprs vrbs'
23:26 jonrafkind: look at all the time wasted writing english!
23:27 eli: No, function is *far* more important than all these all these other keywords.
23:27 jonrafkind: i dont mind supporting both 'fun' and 'function'
23:27 jonrafkind: I can just provide it as both
23:27 eli: See `λ'.
23:27 jonrafkind: yea lambda would be cool if I knew how to type it
23:27 offby1: rudybot: λ
23:27 rudybot: *offby1: error: #<undefined>:1:0: λ: bad syntax in: λ
23:28 offby1: rudybot: (λ (λ) (λ λ))
23:28 rudybot: *offby1: ; Value: #<procedure>
23:28 offby1: rudybot: ((λ (λ) (λ λ)) λ)
23:28 rudybot: *offby1: error: #<undefined>:1:15: λ: bad syntax in: λ
23:28 jonrafkind: eli, you should try agda, people go nuts with unicode in that language
23:28 eli: If it's easy to change it, then whatever works for me.  But for the core language, I'd have a quick poll if I really wanted to be careful.
23:28 offby1: rudybot: ((buffalo (buffalo) (buffalo buffalo buffalo) (buffalo (buffalo buffalo))))
23:28 rudybot: *offby1: error: reference to an identifier before its definition: buffalo in module: 'program
23:29 eli: Re unicode, I could help you if you'd use Emacs
23:29 eli: offby1: What do you use for λ?
23:30 jonrafkind: in vim i mapped \-l to that lambda symbol, so i could theoretically use it, but i just never bother
23:30 jonrafkind: but its still hard to type outside of vim
23:30 jonrafkind: i really have no problem at all writing out 'lambda'
23:30 eli: A solution to just λ is not enough, which is why I didn't do it for years.
23:31 eli: Now I get to write things like Γ⊢τ₂
23:31 jonrafkind: well how do you type arbitrary unicode characters
23:32 jonrafkind: I think theres a vim mode for writing \gamma which turns into the gamma symbol
23:32 eli: I have a key that toggles the Emacs input method to one where I dumped a shitload of short notations, and does that for just one "key".
23:32 eli: The fact that it's an input method makes it work in all contexts, even in interactive searches etc.
23:32 eli: 〈〉
23:32 eli: «»
23:32 jonrafkind: yea, I guess if you can remember all the notations its ok
23:32 eli: ←→
23:32 jonrafkind: but somehow you have to choose a good subset of unicode to remember
23:33 Frozenlo`: Wait what... you can use λ instead of lambda?
23:33 jonrafkind: ya
23:33 eli: ❮❯
23:33 jonrafkind: do a u with a umlot on it
23:33 eli: ü
23:33 jonrafkind: how about ae
23:33 eli: ae
23:33 jonrafkind: no the single ae character
23:34 jonrafkind: aether
23:34 eli: æ
23:34 eli: What's the name of that thing/
23:34 eli: ?
23:34 jonrafkind: uh i dunno
23:34 eli: Æ
23:34 jonrafkind: yea that
23:34 jonrafkind: so how many of these did you commit to memory?
23:34 eli: So I was right.
23:35 eli: Not so much.
23:35 jonrafkind: you just type a logical notation and the mode figures it out?
23:35 eli: But I included common sgml/html names, so I get an instant big-ish pool of names.
23:35 eli: Like  
23:35 eli: (that was an nbsp there)
23:35 eli: — –
23:36 eli: Then I have a bunch of "graphic" names that I remember, like "|-": ⊢
23:36 eli: And a bunch of latex-like things, like "_3" ₃
23:36 eli: ☺
23:36 eli: ☻
23:37 eli: ♥♦♣
23:37 eli: What's the name of the last one?
23:37 (join) dnolen
23:37 jonrafkind: clove
23:38 jonrafkind: oh the missing one
23:38 eli: No, the one that's missing
23:38 jonrafkind: spade
23:38 eli: ♠
23:38 jonrafkind: can you do big pi
23:38 eli: Obviously.
23:38 eli: π Π
23:38 eli: σ Σ
23:38 jonrafkind: what do you type for those, sigma?
23:38 jonrafkind: or just 'o' and 'E'
23:39 eli: sigma and SIGMA
23:39 jonrafkind: so how do you get lambda?
23:39 eli: s/SIGMA/Sigma/
23:39 jonrafkind: do you type 'lambda' ?
23:39 eli: I have that (from the sgml thing), but I also added \
23:39 jonrafkind: and hopefully you use the \ often right
23:40 eli: Yeah, that's why I added it.
23:40 eli: Here: https://github.com/elibarzilay/eliemacs/blob/master/extras/special-chars.el
23:40 rudybot: http://tinyurl.com/6o6z35m
23:40 jonrafkind: ok just doing a sanity check :p
23:41 eli: The first few blocks are mine, then there's the sgml names.
23:41 jonrafkind: lrarr2 of course!
23:41 eli: ⇆
23:41 jonrafkind: there4 rofl
23:41 eli: ∴
23:42 eli: ¾⅞
23:42 jonrafkind: so you must love APL huh
23:42 eli: Not really.
23:42 jonrafkind: a single line of code is worth 50 volumes of C
23:43 eli: I've seen a discussion recently where someone said that APL programmers like to write functions explicitly.
23:43 eli: The claim is that being short means that it's easy to just write an explicit function rather than define a name for it.
23:44 eli: That's true in Racket/Scheme up to the extreme complexity of (lambda (x) (* x 120))
23:44 eli: I find that using `λ' rasies the bar a little.
23:45 eli: But in APL it goes way high.
23:45 (quit) jeapostrophe: Ping timeout: 244 seconds
23:49 eli: jonrafkind: Here -- this is apparently average: {(+/x)%(#x)}
23:50 eli: It was described as equivalent to sum(x)/len(x) in python
23:50 eli: But apparently that's the long APL-ish syntax, the shorter one is +/%#
23:51 eli: So APL programmers apparently see that and immediately see "average".
23:52 offby1: *shrug*
23:52 offby1: they say there are people who understand Chinese, too
23:55 eli: Max slice sum: |/0(0|+)\
23:55 eli: Flatten: ,//
23:56 jonrafkind: rofl, looks like perl
23:56 eli: No, it makes perl looks like pearls.
23:57 offby1: y'all talkin' 'bout J ?