#clojure – 2012-01-15

2012-01-15

03:34	bitops	has anyone done any 2D animation with sprite maps in Clojure? any libraries or pointers?
03:35	amalloy	bitops: just the java ones, i think
03:39	bitops	amalloy: makes sense...tooling around I've found the clj-processing library. looks like it could be a good one to use.
04:39	notsonerdysunny	is there a good clojure library to deal with ".zip/.z/.tar.gz/.tar/.gz" files?
04:40	Raynes	notsonerdysunny: https://github.com/Raynes/fs has some functions for decompressing and unarchiving files.
04:40	notsonerdysunny	preferably without depending on the system commands.. purely based on JVM stuff..
04:40	hiredman	yes
04:40	Raynes	They are implemented entirely in Clojure backed by core and apache libraries.. None of them shell out.
04:42	notsonerdysunny	thanks Raynes .. that is exactly what I was looking for.
05:09	PntBlnk	Hi all, would someone be able to let me know how to structure a macro so that I can use the 'var' special form on an argument? Or is there an example somewhere I could study? Thanks in advance if you can help.
05:13	amalloy	PntBlnk: i'm not 100% sure what you mean, but binding is the only example i can think of at the moment
05:14	amalloy	it's a little complicated, though. a simple answer would be (defmacro my-var [x] `(var ~x)), for example
05:14	PntBlnk	Thanks, I'll have a go... :)
05:15	PntBlnk	Unfortunately that doesn't work for me.
05:16	amalloy	that's not really an actionable statement. what did you do; what did you hope would happen; what happened instead?
05:17	PntBlnk	I've created a modified version of the robert-hooke code with a few extra features, and I'm currently storing hooks as a vector: [target-var <string representation of hook-var that I want to turn to a real var> hook]
05:18	PntBlnk	So a hook might be [#'korma.core/exec #'my.ns/my-fn #<Atom......>].
05:20	PntBlnk	The macro in question is to create the hooks. It works okay, but I'm unhappy that the hook function is currently (.replace (.getName (class f)) \$ \/), which is pretty painful.
05:22	PntBlnk	The macro takes the target-var and the hook function, and tries to do everything in a syntax-quoted let: `(let [~'fn-var `(var ~~f) ~'fn-atom (atom ~f)](vector ~target-var ~'fn-var ~'fn-atom)))
05:23	PntBlnk	It's getting the fn-var value which is proving challenging for my tiny noob mind...
05:26	amalloy	yeah, if you're trying to manage nested syntax-quotes you're just going to be sad overall
05:27	PntBlnk	I feel very much that way.
05:27	amalloy	but i don't think you need them, really
05:28	amalloy	`[~target-var (var ~f) (atom ~f)] is it, right?
05:28	PntBlnk	Well, var is a special form, and can't be treated like a function...
05:28	amalloy	that seems to be the less-confused version of what you seem to be trying to do
05:29	PntBlnk	It is very much nicer, but still doesn't work :)
05:30	PntBlnk	I get #'my.ns/f ..
05:31	amalloy	&(let [target-var #'inc, f 'swap!] `[~target-var (var ~f) (atom ~f)]
05:31	lazybot	java.lang.RuntimeException: EOF while reading, starting at line 1
05:32	amalloy	&(let [target-var #'inc, f 'swap!] `[~target-var (var ~f) (atom ~f)])
05:32	lazybot	⇒ [#'clojure.core/inc (var swap!) (clojure.core/atom swap!)]
05:32	amalloy	point is, you have still not really said what you're doing. you provided a snippet of what you think is the relevant code, but without an idea of how your macro is getting called i can't really write it
05:35	PntBlnk	Sure. Are you familiar with the robert-hook lib? I'm using a vector to hold the hooks in the form I described earlier. With this code, I can list existing hooks, remove all of them, and will be able to add them to entire namespaces if that's at all useful.
05:36	amalloy	i know vaguely what robert.hooke does, but nothing about the structure of the lib
05:36	PntBlnk	The main problem for me was that hooks weren't playing nice when compiling my code.
05:38	PntBlnk	Aside from the overall functionality my problem is quite simple. I want to store two vars and an atom in a vector. The first var (the target) is fine, because you provide that in #'namespace.qualified/form, but the second is not (unqualified function name), and I'm keen to keep the usage the same as the original code.
05:40	PntBlnk	The var special form is not available at compile time, so you don't get the same results as in the REPL.
05:40	amalloy	what do you want an invocation of your macro to look like, and what do you want it to expand to
05:41	amalloy	those are the important questions. once you have those answered, you generally can write the macro (or determine it's impossible)
05:44	PntBlnk	Here's one I was playing with before. Fine in the REPL, not so for compilation: (defmacro get-var [f] `(let [~'out (var ~f)] ~'out))
05:45	PntBlnk	Hang on...
05:45	PntBlnk	That appears to work..
05:46	PntBlnk	I must apologize as it seems that I have wasted your time!
05:52	amalloy	happy hunting
05:53	PntBlnk	Yeah, thanks. Still not getting a final that's done and dusted feeling, but I appreciate the time you spent. Thanks!
06:19	kaeluka	what's the best way to learn clojure? book/webpage
06:25	kaeluka	hi, btw ;-)
06:26	Blkt	good day everyone
06:38	kaeluka	..seems pretty empty here, bikt :)
06:56	augustl	kaeluka: practicing :)
06:57	augustl	books and writing programs are my favorite ways of practicing
07:00	kaeluka	but sometimes, a certain web page is better than any book
07:00	kaeluka	i'm working with stuart halloway's programming clojure for now and maaaan, does this language look nice
07:01	kaeluka	have been dabbling around a bit a year ago or so but then got distracted
07:19	augustl	kaeluka: http://www.infoq.com/presentations/Simple-Made-Easy i
07:19	augustl	err
07:19	augustl	kaeluka: http://www.infoq.com/presentations/Simple-Made-Easy will also teach you alot about clojure
07:19	augustl	even though there's no clojure in it :)
07:20	kaeluka	awesome, great :)
07:22	kaeluka	btw: what's the editor/IDE most people use for clojure? i would very much like to stick with emacs, does that make sense?
07:24	augustl	I use emacs at least, but I'm not an authority on clojure ;)
07:24	raek	kaeluka: a lot of people use emacs for clojure
07:25	augustl	emacs is good for parens
07:26	raek	kaeluka: check out the stats :-) http://cemerick.com/2011/07/11/results-of-the-2011-state-of-clojure-survey/
07:26	kaeluka	yeah, i'm guessing that paredit mode has to be awesome for clj
07:27	raek	it is
07:27	raek	kaeluka: to get started follow the official readmes of clojure-mode (elisp lib) and swank-clojure (clojure lib)
07:27	raek	https://github.com/technomancy/clojure-mode
07:27	raek	https://github.com/technomancy/swank-clojure
07:29	raek	there are a lot outdated tutorials out there, so I just wanted show the correct ones early...
07:30	kaeluka	yeah, that's why i'm asking
07:30	kaeluka	thanks a lot!
07:32	raek	the no 1 cause of problems is having an incompatible version of slime installed in emacs (so beware of that if you come from Common Lisp)
07:33	raek	anyway, most of the things you need to look out for are explained in the "Troubleshooting" section of the swank-clojure readme
08:15	kaeluka	raek: no, got no slime installed currently; thanks
08:35	bpr	is flatten lazy?
09:44	TimMc	&(take 5 (flatten (range)))
09:44	lazybot	⇒ (0 1 2 3 4)
09:44	TimMc	bpr: ^
09:55	raek	~flatten
09:55	clojurebot	flatten is rarely the right answer. Suppose you need to use a list as your "base type", for example. Usually you only want to flatten a single level, and in that case you're better off with concat. Or, better still, use mapcat to produce a sequence that's shaped right to begin with.
11:05	bpr	TimMc, raek: Thanks guys. I actually did wind up using mapcat
11:24	zzach1	Compiling clj files with java works using a command line "java -cp clojure.jar:.:classes -Dclojure.compile.path=classes clojure.lang.Compile app.hello ". The resulting class has no debug / line number information. Which option can be used to compile including such information?
11:38	yoklov	wow. I'm surprised cljs has ``read''
11:44	rindolf	Hi all. What's the best way to learn Clojure?
11:45	pandeiro	rindolf: what's your task?
11:46	rindolf	pandeiro: I want to learn Clojure in order to learn a new language.
11:46	rindolf	pandeiro: no direct task, except maybe implementing Golf challenges in Clojure or whatever.
11:46	pandeiro	but what are you trying to do, specifically?
11:47	pandeiro	i would once you know the syntax, find code that does things that are somehow related to the problems that interest you
12:08	mindbender	please how can I get cljs.repl on my project classpath so that I can (require '[cljs.repl :as repl]) without errors
12:09	mindbender	is there a jar of some sort for clojurescript?
12:11	ldopa	mindbender: i haven't had any problems using the one from lein-clojurescript
12:11	ldopa	mindbender: http://clojars.org/org.clojars.bartonj/clojurescript
12:13	mindbender	ldopa: thanks, I think that's just what I needed
13:51	yoklov	>:/ code, you and i both know you were running at twice this speed yesterday.
13:59	yoklov	hrm. how do people profile clojure code?
15:48	zyfo	Have some noir code that works great on two servers, but when I try to run it locally (lein deps && lein run) I get the following exception: "resource-response does not exist (route.clj)". Any ideas?
16:00	Guest88848	Hello, I am new to clojure and would like help using something like (reduce (map reduce +)) on a list of lists of numbers to obtain the sum.
16:00	Guest88848	Is this an okay place to ask such questions?
16:01	_carlos_	hi!
16:02	_carlos_	,(defn factorial [n] (-> n (+ 1) (range 1) (apply *)))
16:02	clojurebot	#<Exception java.lang.Exception: SANBOX DENIED>
16:02	_carlos_	#(defn factorial [n] (-> n (+ 1) (range 1) (apply *)))
16:02	_carlos_	(defn factorial [n] (-> n (+ 1) (range 1) (apply *)));
16:02	_carlos_	._.
16:03	_carlos_	do I need to be registered to run code?
16:03	William_Sharkey	(reduce + 0 (map reduce + 0 '( '(1 2) '(3 4)) ))
16:04	_carlos_	anyway, this code doesn't work because ClassCastException clojure.lang.LazySeq cannot be cast to clojure.lang.IFn clojure.core/apply (core.clj:600)
16:04	William_Sharkey	my code does not work because #<IllegalArgumentException java.lang.IllegalArgumentException: Don't know how to create ISeq from: clojure.core$_PLUS_>
16:04	_carlos_	I know I should use ->> but I don't understand why the lazy seq doesn't work the oppositte way
16:05	_carlos_	William_Sharkey: hahaha.. I thought you were some parrot bot for a minute
16:06	_carlos_	William_Sharkey: why do you have 0 as an argument?
16:07	_carlos_	William_Sharkey: what do you want to do?
16:07	raek	William_Sharkey: you probably want (reduce + (apply concat [[1 2] [3 4]]))
16:08	raek	apply concat turns it into [1 2 3 4] and then you can simply sum that collection
16:09	_carlos_	raek: do you know why I can't use first-thread?
16:10	William_Sharkey	Raek:thank you.
16:10	William_Sharkey	Carlos: I wanted to sum a list of lists of numbers.
16:11	raek	William_Sharkey: (map reduce + 0 ...) doesn't make much sense. maybe you wew thinking about something like (map (fn [x] (reduce + 0 x) ...)?
16:12	raek	_carlos_: the bots don't allow you to define global vars. you can use anonymous functions though
16:12	raek	_carlos_: ##(let [factorial (fn [n] (-> n (+ 1) (range 1) (apply *)))] (factorial 5))
16:12	lazybot	java.lang.ClassCastException: clojure.lang.LazySeq cannot be cast to clojure.lang.IFn
16:13	raek	so, the (-> ...) macro expands to this: (apply (range (+ n 1) 1) *)
16:13	bobhope	Are there any clojure templating libraries? I want to general source code for a different language from a clojure program
16:14	_carlos_	bobhope: I heard about fleet
16:14	raek	_carlos_: can you write the function body without ->?
16:15	_carlos_	raek: I can write with ->> last-thread macro, but I wanted to understand why it doesn't work with first-thread ->. for me both make sense
16:16	raek	well, they don't mean the same thing
16:16	raek	_carlos_: do you know about the difference?
16:16	_carlos_	raek: yes, I know, but both directions make sense from substitution point of view
16:16	raek	what does (apply (range (+ n 1) 1) *) mean?
16:17	raek	lets assume n = 5
16:17	raek	(apply (range (+ 5 1) 1) *)
16:17	raek	(apply (range 6 1) *)
16:17	raek	(apply () *)
16:17	raek	BOOM
16:17	bobhope	_carlos_, fleet seems interesting, but it seems like they only designed for xml-esque languages
16:17	raek	() is not a function
16:18	raek	so to me the (-> ...) version does not make sence, since the functions end up with the arguments in the wrong order
16:19	_carlos_	raek: n is applied as last argument of (+ 1), making it (+ 1 n), not (+ n 1)
16:20	raek	_carlos_: yes, since (+ a b) is equal to (+ b a) this is not a problem.
16:20	raek	_carlos_: but what about the call to 'range' and 'apply'?
16:20	_carlos_	raek: ha, sorry, just a sec
16:23	_carlos_	raek: why did you apply the result of (+ 5 1) to the first argument of range? range also takes the result in its last argument
16:23	raek	_carlos_: I'm innocent! -> did it :-)
16:24	raek	_carlos_: the -> macro rewrites the code in a fixed pattern. it knowns nothing about what the different argument means
16:25	_carlos_	raek: my bias was that is was putting it in the end of the expression. my bad! thank you very much
16:26	raek	think of -> as a transformation that turns (-> a (b0 b1 b2 ... bn) ...more...) into (-> (b0 a b1 b2 ... bn) ...more...)
16:26	raek	and (-> a) into a
16:27	_carlos_	raek: I understand your previous example step by step eval.. my bad :-)
16:50	_carlos_	is it just me or most times clojure code is easier to read in a one liner than broken into several lines like most books do?
16:52	jodaro	clojure.main quit unexpectedly while using the jna8943583805303982763.tmp plug-in.
16:52	jodaro	woopsie
16:56	brehaut	_carlos_: just you?
16:57	Raynes	Spaces are stupid too! (foo,bar(baz),(foo,bar))
16:57	brehaut	Raynes: why have you used such long winded symbols?
16:57	Raynes	Bah!
16:58	raek	Raynes: you use commas for indentation too? :P
16:58	Raynes	raek: Of course.
16:59	raek	does someone know how to turn off the "fn" to "ƒ" transformation that emacs starter kit has by default?
17:00	raek	maybe I could just comment out these lines in my checkout: https://github.com/scgilardi/emacs-starter-kit/blob/5937360cc91d352380d95801d1a0cebecfb47d96/modules/starter-kit-lisp.el#L85
17:01	_carlos_	if computer languages supported more expressive symbols, variables could be, say chinese characters. code would be much more compact
17:02	raek	clojure does not forbid them
17:02	_carlos_	raek: really? didn't know
17:03	brehaut	,'ß
17:03	clojurebot	ß
17:04	brehaut	,((fn [ß] (inc ß)) 1)
17:04	clojurebot	2
17:04	raek	it isn't really set in stone which characters are allowed, but alphanumeric characters (even non-ascii ones) are ok
17:05	brehaut	raek: although clojure.org says "Symbols begin with a non-numeric character and can contain alphanumeric characters and *, +, !, -, _, and ? (other characters will be allowed eventually, but not all macro characters have been determined)"
17:05	_carlos_	,((fn [世界] (inc 世界)) 1)
17:05	clojurebot	2
17:05	_carlos_	:o
17:06	brehaut	_carlos_ please dont write apljure now!
17:06	raek	.clj-files are always in UTF-8, so all unicode characters can be used in clojure source code
17:06	_carlos_	haha
17:09	brehaut	raek: yeah sorry. i understand that, i was just mentioning that if you read clojure.org, you might get a different idea
17:10	brehaut	raek: rather than trying to contradict you
17:13	amalloy	raek: .clj files are required to be UTF-8?
17:15	raek	amalloy: I'm pretty sure that clojure will always decode them using UTF-8, but maybe I should try to back up my claim
17:17	raek	https://github.com/clojure/clojure/blob/master/src/jvm/clojure/lang/Compiler.java#L6870
17:18	raek	there
17:19	amalloy	interesting. that means someone whose OS uses a different encoding by default will be unable to compile their files
17:19	raek	amalloy: how?
17:19	brehaut	amalloy: you mean e.g. windows and UCS16 ?
17:20	raek	OS'es do not transcode dat just like that
17:20	raek	*data
17:21	raek	OS'es often have some notions of default encodings, but that doesn't affect much
17:21	_carlos_	from what I see, if my file is encoded with a different encoding, it will blow
17:21	amalloy	well. not an area i'm an expert in, but it seems like if i open up a new text file using some japanese character-encoding, then any characters which aren't encoded the same as UTF8 will be read incorrectly
17:22	raek	yes. this is an issue the text editor deals with rather than the OS
17:22	brehaut	oh. UCS-2 sorry
17:22	bobhope	What is the difference between extends? and satisfies? for protocols?
17:22	_carlos_	raek: so a language plugin for the editor, detecting the target language would always try to force that encoding
17:22	raek	and since the encoding is not actually stored anywhere, a text editor can only guess the encoding
17:23	brehaut	,(doc satisfies?)
17:23	clojurebot	"([protocol x]); Returns true if x satisfies the protocol"
17:23	raek	bobhope: check on a instance vs a class I think
17:24	_carlos_	raek: based on the extension would by concept be enough to guess de decoder
17:24	_carlos_	*the decoder
17:24	bobhope	raek, thanks
17:24	bobhope	that seems like a reason to have both
17:24	raek	_carlos_: yes, for some text-based formats (e.g. Clojure source files and JSON files) that is true
17:25	raek	but most text-based formats are broken and do not specify an encoding or lets the user choose but does not provite any way of soting that choice
17:27	raek	provide storing
17:29	bobhope	If I want to export a bunch of vars from another namespace when my namespace is imported, should I use ns-utils/immigrate?
17:31	raek	bobhope: there is no perfect way of doing that in clojure
17:31	bobhope	Here's my scenario
17:32	bobhope	I want to provide a different numeric tower, so I want to use clojure.algos.generic +-*/
17:32	bobhope	but make it so that anyone who includes my library will see those exposed to them as well
17:32	bobhope	and they could even add their own extensions to those multimethods
17:32	bobhope	how should I do this?
17:32	raek	if you use immigrate and redefine the vars in the source namespace, they won't be updated in the namespace into which they were immigrated
17:33	bobhope	I'd be happy to cut & paste the code from clojure.algos.generic into my project and call it a day
17:33	bobhope	is immigrate that functionality?
17:34	bobhope	and is immigrate in clojure 1.3? It looked like it was only in the monolithic contrib
17:34	raek	bobhope: it is in monolithic contrib, i.e. not in 1.3
17:35	raek	http://dev.clojure.org/display/doc/FAQ
17:35	bobhope	raek, can I use monolithic contrib and some clojure contrib 1.3 stuff?
17:35	bobhope	I've never tried
17:35	raek	"How can I make a namespace A that "exports" all the public symbols from namespaces B and C?"
17:35	brehaut	bobhope: i think its possible yes, if you dont mind a screed of warnings
17:36	raek	bobhope: no. there won't be any monolithic contrib for clojure 1.3. (and the one from 1.2 won't work in 1.3)
17:36	bobhope	Ok, I'll probably just reimplement the numeric tower the way I want it then
17:36	bobhope	back to basics :)
17:37	brehaut	ah. its possible to use monolithic contrib with new contrib in 1.2
17:38	raek	yes. the new contrib libs are usually version independent. (some parts of monolithic aren't therefore monolithic as a whole isn't)
17:39	raek	bobhope: immigrate is pretty small, so if you really need to use it it's probably easier to just copy and paste it into your own code
17:42	amro	hi
17:42	amro	I seem to have broken lein
17:43	amro	Exception in thread "main" java.lang.NoClassDefFoundError: clojure.core.protocols$fn__5393
17:43	amro	this after I installed some java stuff
17:43	amro	I'm not sure exactly what, I installed a few while trying to get jni.h
17:45	mrevil	i there a way to set different environment variables when I running tests as opposed to development a environment? I need to control which database my noir tests connect to. How do I determine that code is being executed in the context of a test or swank? Ruby on Rails has 3 environments with different initialization scripts, is there something similar pre-baked into noir, clojure.test or midje?
17:48	clj_newb	are variables that start with _ reserved by clojure?
17:48	clj_newb	like _foo ?
17:49	raek	amro: are there files in your classes/ directory? if so, have you tried deleting them?
17:49	raek	sounds like you have old ahead of time compiled code lying around
17:49	brehaut	clj_newb: ive never seen that in clojure code. do you have an example?
17:49	clj_newb	I want to define something like _Animal
17:49	clj_newb	which is a wrapper around Animal.
17:49	clj_newb	where Animal is a defrecord.
17:50	Raynes	There is nothing special about underscores. That said, what you want to do is nuts.
17:50	clj_newb	Raynes: why?
17:51	brehaut	to start with, why do you want a wrapper?
17:51	clj_newb	styptype checking
17:51	Raynes	Because Clojure isn't Perl. We like to use words instead of symbols.
17:51	brehaut	styptype?
17:51	clj_newb	type checking
17:52	brehaut	clojure is a dynamic language; type checking is not idiomatic
17:53	jodaro	so i'm looking at clojure.tools.logging
17:53	clj_newb	brehaut: I have never been able to write more than 10K LOC w/o typing.
17:53	clj_newb	brehaut: Is there a list of good practices of "writing largescale clojure apps" ?
17:53	bobhope	test your code
17:53	bobhope	with unit tests
17:53	bobhope	and automated tests
17:53	jodaro	in (defn log*, there is a line about using agents
17:53	amro	raek: I broke lein in general- it doesn't work at all
17:54	bobhope	if your testing consists of running examples in a repl, or anything done by hand, it's called "experimentation", not testing
17:54	bobhope	;)
17:54	jodaro	and it says "for performance reasons, an agent will only be used when invoked within a running transaction ..."
17:54	brehaut	clj_newb: write small, pure functions that can be easily reasoned about
17:54	jodaro	is it particularly expensive to use send-off?
17:54	brehaut	clj_newb: and leverage HoFs to not write as much code
17:56	clj_newb	what is HoFs ?
17:56	brehaut	clj_newb: higher order functions
17:56	brehaut	map being a classic example
17:56	clj_newb	yeah; I think the problem is that I'm writing to write haskell in clojure
17:59	amalloy	hiredman: i have a terrible suspicion that Keyword was made non-final because it now contains non-final fields. those two things obviously shouldn't be related, but it's quite a coincidence otherwise
18:03	blakesmith	Is there an idiom to do the following more concisely?
18:03	blakesmith	&(map (fn [x y] [x y]) (range 10) (range 10))
18:03	lazybot	⇒ ([0 0] [1 1] [2 2] [3 3] [4 4] [5 5] [6 6] [7 7] [8 8] [9 9])
18:03	brehaut	(map vector …
18:04	blakesmith	Oh, neato.
18:04	blakesmith	Since vector takes variable arguments...
18:05	blakesmith	Sweet, thanks!
18:05	blakesmith	Er, variable argument length.
18:33	Raynes	brehaut: ohai
18:33	brehaut	Raynes: hello
18:33	Raynes	brehaut: Feel like testing my Vim plugin for me?
18:33	brehaut	Raynes: haha
18:34	Raynes	>_>
18:34	brehaut	Raynes: i dont know how to use vim
18:34	Raynes	Oh, I thought you were a vimmer.
18:34	brehaut	Raynes: nope. just a pot stirrer
18:34	Raynes	Heh
18:34	brehaut	i dislike all text editors :P
18:34	Raynes	:)
18:34	brehaut	just happen to dislike some less than others
18:34	Raynes	It's unfortunate that I know so few vimmers.
18:35	Raynes	But then again, I couldn't even get Emacsers to test my elisp when I wrote it.
18:35	Raynes	cough technomancy cough
18:38	guns	Raynes: I'm in the process of streamlining my vim/clojure workflow. I wouldn't mind beta testing your software
18:39	Raynes	guns: It's at https://github.com/Raynes/refheap.vim if you want to give it a go.
18:40	Raynes	Haha
18:41	technomancy	also, I wish the word "infidel" could be used more without bringing religious connotations into it
18:41	technomancy	it's a pretty good word.
18:41	Raynes	technomancy: ARE YOU SAYING MY GOD IS FAKE?
18:42	guns	Raynes: You may want to leverage the `autoload` plugin method to avoid load errors at startup
18:42	brehaut	Raynes: not fake, just old fashioned and not self hosting
18:43	technomancy	what; no, I'm not saying anything... let's just meet on Mt. Caramel with some altars and see whose gets lit.
18:43	Raynes	guns: I have no clue what autoload is.
18:43	guns	Raynes: it's vim's lazy plugin loading system
18:44	guns	Gundo.vim is a pretty good example of a python plugin that employs autoload
18:45	Raynes	guns: It looks like I'd basically just put my stuff that checks if Python is loaded and my command! statement in the plugin stuff and the rest in autoload.
18:45	guns	Raynes: yep. also, how to install pyperclip.py ?
18:46	Raynes	You don't need pyperclip. Why did you think you needed that? :o
18:46	Raynes	Oh man, did I include that?
18:46	guns	line 33:
18:46	guns	Traceback (most recent call last):
18:46	guns	File "<string>", line 1, in <module>
18:46	guns	IOError: [Errno 2] No such file or directory: 'pyperclip.py'
18:46	clojurebot	Titim gan éirí ort.
18:47	Raynes	guns: That's an accidental commit. Hold on.
18:49	Raynes	guns: Pull again. Sorry, I forgot to commit some stuff (and accidentally committed other stuff).
18:51	guns	Got a 413 (request too large) when trying to post ~24k lines (unicode table). Also, is refheap hosted on your server?
18:51	guns	I suppose that's reasonable (too many lineS)
18:52	technomancy	as long as you don't get 418 when trying to upload coffeescript
18:52	guns	haha. had to look that one up
18:53	Raynes	guns: Yes, refheap is hosted on my server.
18:54	Raynes	guns: Also, I guess the server doesn't like large requests. I have the maximum paste size capped at like 60KB, but this error wasn't because of that.
18:55	guns	Raynes: it's not really your problem, but pbcopy/pbpaste is broken from within Tmux sessions. I get around that myself by using a clipboard plugin for rxvt-unicode
18:56	guns	I'm not sure of a good way to solve it from your perspective though
18:56	Raynes	It's a pain in the ass to copy stuff programmatically and have it work everywhere.
18:56	Raynes	You can still copy the link though, at least.
18:56	emezeske	guns: what is pbcopy?
18:57	guns	emezeske: xclip for OS X
18:57	guns	basically
18:57	emezeske	ahh, gotcha
18:57	emezeske	I was gonna suggest xclip :P
18:57	Raynes	guns: How large is your unicode table? In bytes.
18:58	guns	966360 bytes
18:59	technomancy	emezeske: hey, is there any specific reason you're publishing lein-cljsbuild using the emezeske group-id?
18:59	Raynes	guns: Yeah, that'd be way over what refheap is designed to accept. nginx is probably rejecting it before it even gets to me though.
18:59	guns	Raynes: :) I was just stress testing
19:00	Raynes	:)
19:01	guns	Otherwise, it all works quite well, visual selections and all. I would autoload, though, for people that move their .vim setup to different machines and don't want to easy_install * for every unix account
19:01	Raynes	Yeah, I agree. I just didn't know about it until now. I'll definitely do that in a bit.
19:01	Raynes	Thanks a lot for testing it.
19:01	guns	np
19:01	Raynes	Clojurians vimmers are a rare breed.
19:08	oakwise	Raynes: I'm one, but I just started down the dark road to emacs today
19:08	oakwise	the propaganda got to me
19:08	Raynes	Heh.
19:08	Raynes	I'm actually a big Emacs fan. I'm just currently experimenting with other editors.
19:08	Raynes	Which is good, since somebody had to write plugins for them for refheap.
19:09	technomancy	and pillage them for their good ideas?
19:10	emezeske	technomancy: no specific reason other than... I don't know what I'm doing? :)
19:10	Raynes	technomancy: Vim has a lot of good ideas. VimL isn't one of them.
19:10	emezeske	technomancy: is there a preferred way to do that?
19:11	guns	Haha. Vimscript is a real drop kick. It does have a map() fn, fwiw. no reduce though.
19:11	oakwise	Raynes: "It's a pain in the ass to copy stuff programmatically and have it work everywhere" <-- does the * register not work?
19:11	guns	oakwise: OS X doesn't use X SELECTION or CLIPBOARD
19:12	technomancy	emezeske: heh; sure, it's conventional to just specify lein-cljsbuild, which ends up setting the group-id to the same as the artifact-id
19:12	guns	pbcopy is the workaround for OS X, but not for people who use X on OS X (like me)
19:12	oakwise	"*y works for me on osx
19:12	emezeske	technomancy: ahh, I did not know that
19:13	technomancy	emezeske: actually I just brushed up the plugin guide for Leiningen 2.0; I would love to hear your feedback on it since you're coming at it from a fresh perspective
19:13	guns	oakwise: what terminal and what vim, I'm curious
19:13	guns	?
19:13	technomancy	https://github.com/technomancy/leiningen/blob/master/doc/PLUGINS.md if you get the chance
19:13	oakwise	guns: I used to use pbcopy/xsel scripty workarounds, but * seems to work on my platforms now
19:13	emezeske	technomancy: I'll definitely take a look, probably sometime today or tomorrow
19:13	oakwise	iterm and...
19:14	emezeske	technomancy: so, lein-cljsbuild depends on emezeske/clojurescript and emezeske/closure-library. should those be self-group-id'ed as well?
19:14	oakwise	guns: vim 7.3
19:14	Raynes	oakwise, guns: I couldn't find a way to use Vim stuff to copy text that was obtained from an http response. Black magic.
19:14	emezeske	technomancy: or does that matter?
19:14	guns	Oh, I see. The iTerm devs are pretty hardworking
19:14	technomancy	emezeske: no, if it's a library you didn't write then it's important not to have the group-id match the artifact-id
19:14	guns	they just added a bunch of tmux support I heard
19:15	emezeske	technomancy: interesting, okay
19:15	technomancy	org.clojars.emezeske is more common, but plain emezeske is fine too.
19:15	emezeske	technomancy: cool, thanks
19:15	technomancy	emezeske: if you read TUTORIAL.md in the same directory as PLUGINS.md, it covers those questions too =)
19:15	oakwise	guns: I just tried it in Terminal and it works fine
19:15	emezeske	technomancy: so, having the group-id match the artifact-id is used to indicate the "canonical" source?
19:15	emezeske	technomancy: awesome!
19:15	technomancy	emezeske: zactly
19:16	guns	oakwise: /usr/bin/vim, and my copy of vim don't work that way
19:16	guns	oakwise: do you compile with --with-darwin?
19:16	technomancy	emezeske: so there are a handful of things in those docs that only apply to leiningen 2, but for the most part it will apply to 1.x
19:16	guns	I use --without-darwin, so that's probably my problem
19:17	emezeske	technomancy: okay
19:17	oakwise	guns: I think I may have compiled from source to get ruby support. Don't remember the flags.
19:19	guns	oakwise: Yeah, they must have added OS X pasteboard support in the past couple of years. I switched to rxvt-unicode and started compiling vim without mac-specific features
19:19	guns	But then in my case, Raynes could still use the "* or "+ register to copy into my system clipboard
19:20	guns	so perhaps you are right
19:20	Raynes	Hrm.
19:21	technomancy	emezeske: I spent a few hours playing with cljs-one on Friday and most of it was spent wishing it used something like lein-cljsbuild =)
19:22	emezeske	technomancy: haha, I wonder if I should ping those people to see if there's some way to coordinate
19:23	technomancy	I don't expect they would be particularly receptive.
19:23	technomancy	https://github.com/brentonashworth/one/issues/64
19:25	technomancy	when I talked to Stu H at the conj he mentioned that he still doesn't use Leiningen in some cases because he thought it was lacking some basic feature that had been implemented since 1.1.0 or so.
19:25	emezeske	technomancy: I did not know about checkout dependencies
19:25	emezeske	yikes!
19:25	emezeske	knowing about checkout deps is going to make my life easier
19:26	ibdknox	emezeske: I'm with technomancy, I'm not sure you'll get very far with them :(
19:26	oakwise	Raynes: vim.command("let @* = %r" % res)
19:26	oakwise	that should replace the need for xerox I think
19:26	oakwise	although you may have to do some escaping (?)
19:27	ibdknox	they seem to like to reimplement things I've built :p
19:27	emezeske	heh
19:27	oakwise	guns: what does without-darwin get you?
19:27	emezeske	ibdknox: I kinda did that too with cljs-watch vs lein-cljsbuild :P
19:27	guns	oakwise: better X integration
19:27	emezeske	ibdknox: don't hate me!
19:28	ibdknox	emezeske: quite the opposite, I was being lazy and waiting for lein 2.0 before I did anything, now I don't have to :)
19:28	guns	Raynes: remember, too that /usr/bin/vim on Lion is not compiled with "* or "+ registers, so you'll still need the pbcopy stuff
19:28	emezeske	ibdknox: haha, nice
19:29	Raynes	guns: Yeah, it seems like using xerox is the more general way to go here.
19:29	ibdknox	I'll probably resurrect noir-cljs as soon as an official clojurescript goes up somewhere
19:29	oakwise	ah my bad
19:29	emezeske	yeah, official cljs would be nice
19:49	seancorfield	emezeske: i installed cljsbuild 0.0.4 as a global plugin and started to see AOT version compatibility errors - i didn't spend time tracking it down, just uninstalled the plugin (and will try again with a dev-dependencies entry) - just mentioning it as an FYI
19:50	technomancy	0.0.4? that's old news, man
19:50	emezeske	seancorfield: yeah, cljsbuild (not lein-cljsbuild) is super broken
19:51	emezeske	there's a bug in the clojure compiler that does not save the :macro metadata in class files
19:51	seancorfield	i meant lein-cljsbuild - i was just being lazy typing
19:51	emezeske	ah, gotcha, there was actually a non-lein version for a while (that I wish I could eradicate)
19:51	seancorfield	0.0.4 was emezeske/lein-cljsbuild - nice to see you've switch to just lein-cljsbuild
19:52	emezeske	I'll mess around with the global-plugin way and see if I can repro
19:52	emezeske	I'd guess that the problem is the same as I mentioned (:macro metadata)
19:52	emezeske	Unfortunately, it means that the clojurescript compiler cannot be AOTed. Well, it can, but it will not work correctly
19:53	notsonerdysunny	Is there a way to have stacktrace at the time error display the whole thing instead of it cutting it at some point because it is too large..
19:54	seancorfield	fwiw, it doesn't seem to happen with 0.0.6 as a global plugin
19:55	technomancy	emezeske: is it this bug? http://dev.clojure.org/jira/browse/CLJ-130
19:55	technomancy	that's been driving me crazy since forever.
19:55	seancorfield	i just installed that and fired up a new swank / slime repl and ran the test that failed before and it didn't fall over :)
19:55	technomancy	I think the official line is "it will be fixed in the clojure-in-clojure compiler" =\
19:57	emezeske	technomancy: very well could be
19:57	emezeske	seancorfield: awesome, thanks for trying it!
20:00	seancorfield	it'll be a while before i can really spend some time with it... i need to experiment with integrating with fw/1 at some point
20:01	seancorfield	but work's keeping me a bit too busy and we're not using clj for web front end stuff (yet)
20:01	emezeske	seancorfield: I created https://github.com/emezeske/lein-cljsbuild/issues/10
20:01	emezeske	seancorfield: if you come across any more AOT weirdness, please note it there
20:02	emezeske	hopefully, the problem will just never come back :)
20:04	ibdknox	I really don't understand the argument against distributing cljs libs through maven
20:05	hiredman	rich has some kind of objection to the forcing of a linear progression of versions on to a project (at least from what I hear)
20:06	hiredman	which is part of maven's model I guess
20:06	ibdknox	people can version things however they want?
20:06	ibdknox	it can be a tree
20:06	ibdknox	lol
20:06	hiredman	sure, but the way maven versions things doesn't account for a tree
20:07	ibdknox	yeah. Is there some versioning scheme that does?
20:07	hiredman	I am not sure really, I know he mentioned "maven tradeoffs" on the mailing list at one point, not sure if he ever elaborated on that
20:07	hiredman	(I think I responded asking for clarification, but I am still in the dark)
20:09	hiredman	https://github.com/clojure/clojure/commit/59d3c724684c212fbb5eafaaaac30761c2c75a37 speaking of things I am in the dark about
20:10	ibdknox	oh weird
20:16	ibdknox	hiredman: there used to be a big performance gain with final classes, do you know if that's still the case?
20:17	hiredman	I don't
21:26	lynaghk	Does anyone know a clean way to check the arity of anonymous functions?
21:26	lynaghk	Preferably that works in ClojureScript.
21:27	brehaut	this is a guess, but unless there is metadata that tells you, there is probably no way to check arity of a clojurescript function, because there is no way to check the arity of a javascript function
21:27	lynaghk	Yeah, I was thinking that too
21:27	ibdknox	and functions don't have metadata in CLJS :(
21:28	lynaghk	I don't actually need the arity, I just need the more JavaScripty behavior of passing lots of args to functions with impunity
21:28	ibdknox	hm?
21:28	lynaghk	Basically, let (#(vector %) 1 2 3 4 5) eval to [1] happily
21:29	ibdknox	(first %&)
21:29	lynaghk	Chris, this is for an implementation of D3, actually.
21:29	lynaghk	I'm hoping I can let people write {:x #(whatever %)}
21:30	lynaghk	instead of having to always catch both the datum and the index; {:x (fn [d idx] (whatever d))}
21:30	ibdknox	ah
21:30	ibdknox	cljs let's you do that for free
21:30	ibdknox	by accident
21:30	ibdknox	unless that changed recently
21:31	lynaghk	oh
21:31	lynaghk	I was testing in a CLJ repl
21:31	lynaghk	go figure.
21:31	ibdknox	by virtue of the JS function implementation
21:31	ibdknox	you can pass in as many as you want
21:31	ibdknox	and they just get ignored
21:32	lynaghk	Sounds like some JS permissiveness is leaking into our glorious semantics.
21:32	ibdknox	one of the many subtle differences between CLJ and CLJS lol
21:32	technomancy	lynaghk: it used to work in 1.2; IIRC I opened an issue on that
21:32	technomancy	I'll let you guess what happened to the issue
21:32	ibdknox	heh
21:32	brehaut	technomancy: it was promptly fixed and closed
21:32	brehaut	amirite?
21:32	ibdknox	~guards
21:32	clojurebot	SEIZE HIM!
21:32	lynaghk	I mean, I can see how arity checking of anon functions is a feature.
21:33	technomancy	it could be that it's something planned for the "dynamicity knobs"
21:33	lynaghk	It makes sense in most cases; just this particular one where I'm expecting users to pass 'em in.
21:33	ibdknox	lynaghk: yeah I just checked, that's what I did for my D3 clone
21:33	ibdknox	and it seems to work
21:33	brehaut	what is D3?
21:34	lynaghk	For future-proof-ness and a clear indicator that it's a gross situation I might just do some try/catches.
21:34	lynaghk	It's a DOM manipulation library focused on mapping abstract data to DOM elements.
21:34	ibdknox	brehaut: http://mbostock.github.com/d3/
21:34	ibdknox	very cool stuff
21:34	ibdknox	you can make really neat visualizations in it
21:34	brehaut	oh! ive seen this. its very cool
21:35	ibdknox	there's a simple CLJS implementation of it in pinot
21:35	technomancy	I really wish we'd hear more about the plans for dynamicity knobs
21:35	technomancy	since it also seems to be blocking locals-clearing, which is needed for good debuggers
21:35	ibdknox	technomancy: I really wish we'd hear more about plans.
21:35	ibdknox	:)
21:36	ibdknox	brehaut: a mere 77 lines! https://github.com/ibdknox/pinot/blob/master/src/pinot/draw/visualization.cljs
21:37	lynaghk	The basic idea of D3 is pretty simple, yeah. Clojure's conciseness surely doesn't hurt either
21:37	brehaut	ibdknox: nice :)
21:37	lynaghk	I'm working on a D3-like library in Clojure(Script), and for a lot of visualizations the compiled JavaScript (including the cljs runtime) is half the size of minified d3.js
21:38	ibdknox	lol
21:38	lynaghk	Then again, Bostock's library has a TON of great layouts, scales, &c.
21:39	lynaghk	That's one of the nice things about Closure though; no need to worry about library bloat.
21:39	brehaut	D3 is the successor to a similar library right?
21:39	lynaghk	Protovis was Bostock's earlier project
21:39	brehaut	thats the one
21:39	lynaghk	it had a custom scenegraph implementation, which made it somewhat hard to debug.
21:40	lynaghk	ibdknox, do you have any tips for writing Clojure targeting both cljs and the jvm?
21:41	emezeske	lynaghk: you can't do exceptions AFAICT
21:41	lynaghk	If I do this try/catch hack around the arity, I'm not sure how to try to catch both kinds of errors without angering the runtimes.
21:41	emezeske	lynaghk: also, macros are very tricky to share
21:41	lynaghk	I think you can catch errors in cljs, they're all just js/Error.
21:42	ibdknox	my suggestion would be to write it for Clojure first, then worry about CLJS
21:42	ibdknox	in terms of error handling
21:42	ibdknox	the best thing to do is probably to create your own macro that you then use everywhere
21:42	ibdknox	and swap a single specific file out for each impl
21:42	lynaghk	mm, that makes sense
21:42	ibdknox	basically hide the differences in one file that changes and go from there
21:43	ibdknox	otherwise it will be painful :(
22:17	devn	a leaderboard in clojure -- ref or atom?
22:17	devn	sorted-map or sorted-set?
23:30	devn	is it possible to create a ref using sorted-set-by and have it continue to sort as items are conj'd onto the map?
23:30	devn	onto the set*
23:32	devn	it doesn't appear so.
23:35	ibdknox	why would it being in a ref make a difference?
23:36	devn	idk -- but it doesn't seem to retain its sorting
23:37	ibdknox	hm
23:37	ibdknox	it worked for me?
23:39	ibdknox	,(let [r (ref (sorted-set-by >))] (dosync (alter r conj 4)) (dosync (alter r conj 1)) (dosync alter r conj 2))
23:39	clojurebot	2
23:39	ibdknox	lol
23:39	ibdknox	,(let [r (ref (sorted-set-by >))] (dosync (alter r conj 4)) (dosync (alter r conj 1)) (dosync alter r conj 2) @r)
23:39	clojurebot	#{4 1}
23:40	ibdknox	I shouldn't type code in IRC
23:40	ibdknox	,(let [r (ref (sorted-set-by >))] (dosync (alter r conj 4)) (dosync (alter r conj 1)) (dosync (alter r conj 2)) @r)
23:40	clojurebot	#{4 2 1}
23:42	devn	hm -- wonder what I'm doing wrong here
23:43	devn	ibdknox: how about when conjing maps?
23:43	amalloy	devn: do it without refs, and then add refs in once you have it working. there is no way the refs are relevant
23:43	ibdknox	devn: ^
23:44	amalloy	then, if it works without refs but not with them, re-inspect how you added them, cause that'll be the part you did wrong
23:44	devn	fair enough -- im rewriting this anyway
23:44	devn	i made a mess
23:45	amalloy	ibdknox: i liked your broken examples, though. definitely an error it's easy to make and easy to miss proofreading
23:46	ibdknox	yeah that last one is kinda scary
23:47	ibdknox	hadn't really thought about the fact that that would just fail silently
23:50	amalloy	ibdknox: proposed awful solution: (defmacro dosync [first & more] (if (coll? first) `(dosync* (do ~first ~@more)) `(dosync* (~first ~@more))))
23:51	ibdknox	lol