#clojure – 2011-11-03

2011-11-03

00:07	zakwilson	I want to start a repl in my namespace from my -main function. What's the easiest way to do this? Looking at clojure.main/repl, it doesn't seem to be able to do that.
00:14	hiredman	zakwilson: why not?
00:15	zakwilson	hiredman: because I'm bad at reading documentation. Looks like :init #(in-ns 'my-ns) does the trick.
00:28	brehaut	enlive nerds: thoughts on (defn enlive-render-wrap [f] (fn [req] (update-in (f req) [:body] #(if (:tag %) (enlive/emit* %) %)))) ? (removing template, and just using snippets, relying on the middleware to render)
01:27	callen	korma looks nice.
01:33	ibdknox	callen: hopefully ;)
03:33	callen	https://github.com/getwoven/clj-time/blob/master/src/clj_time/core.clj trying to use this code, having a weird issue specific to inside a defn.
03:33	callen	(defn unix-now [] (use 'clj-time.core) (in-secs (interval (epoch) (now))))
03:33	callen	that fails
03:34	callen	simply using the namespace inside a repl works.
03:34	hiredman	yeah, don't do that
03:34	callen	hiredman: feel free to explain the idiomatic approach.
03:34	hiredman	use is a global
03:35	callen	what should I use? in-ns?
03:35	hiredman	why do you want to use either?
03:35	hiredman	don't use any of them like that
03:35	callen	I need to use clj-time.core
03:35	callen	it has a namespace defined
03:35	callen	how do I make my helper function work?
03:35	hiredman	so put a (:use clj-time.core) in your (ns ...) form
03:35	Raynes	You don't. You want to use the namespace in your ns form.
03:36	ibdknox	woah
03:36	Raynes	What's up?
03:36	ibdknox	they used Noir in the 2nd edition of programmin clojure?
03:36	ibdknox	programming*
03:37	callen	Raynes hiredman my naive attempt at doing what you just said failed.
03:37	Raynes	If you makes you happy, I plan to use it in my book as well.
03:37	hiredman	well, learn how to do and do it better
03:37	callen	as you said, it's a global and appears to break other things.
03:37	callen	hiredman: I asked for the idiomatic approach.
03:37	hiredman	read the docs, read some code
03:37	callen	hiredman: I have the pragprog book in front of me.
03:37	hiredman	I gave it to you
03:37	callen	hiredman: and the code.
03:38	ibdknox	callen: the idiomatic approach is to do one of two things
03:38	ibdknox	callen: in your ns declaration (ns my-things.foo ..)
03:38	hiredman	if you failed to make the idiomatic approach work, you need to read more docs, look at more examples
03:38	ibdknox	callen: you can either do :use or :require
03:38	callen	I don't know the idiomatic approach, I just spent half an hour ctrl-f'ing through the book
03:38	callen	why :use instead of use?
03:39	Raynes	Because that's what ns takes.
03:39	ibdknox	callen: in general, if you do use, you want to do it like so (:use [some-thing :only [the-thing-I-need]])
03:39	callen	a new and interesting error that I don't understand.
03:39	clojurebot	error-kit is http://pragprog.com/magazines/2009-07/when-things-go-wrong
03:39	ibdknox	callen: if you do require, you then need to prefix your usages with the alias you give the required namespace: (:require [clj-time.core :as time])
03:40	callen	(:use 'clj-time.core [in-secs interval epoch now]) like so?
03:40	Raynes	callen: We don't even know what your doing. We don't know what your code is, how you're running it, what error you're getting, etc.
03:40	callen	Raynes: I tried to isolate the question so that it wouldn't be necessary.
03:41	Raynes	Isolate the code and show us a small example that is breaking. It'll be easier to help you like that.
03:41	callen	I did print my function.
03:41	ibdknox	callen: (:use [clj-time.core :only [in-sexs interval epoch now]])
03:41	callen	(defn unix-now [] (in-secs (interval (epoch) (now))))
03:41	callen	that's what's breaking
03:41	callen	says in-secs doesn't exist.
03:41	Raynes	Show us the entire file.
03:41	ibdknox	in a gist
03:41	Raynes	Yes.
03:41	Raynes	gist.github.com
03:41	callen	this is partially mutilated
03:42	callen	in my attempt to follow your advice.
03:42	ibdknox	that's ok
03:42	Raynes	Well, mutilate it less partially and show us the file.
03:42	Raynes	Or just show us the unmutilated file.
03:42	Raynes	I'm sure we can take it. :p
03:42	ibdknox	totally can't. I might explode
03:42	Raynes	We need to see how you're using clj-time. Your ns declaration and stuff.
03:42	callen	https://gist.github.com/1335982
03:43	callen	in-secs, interval, epoch, now are all clj-time
03:43	ibdknox	yep
03:43	callen	and are all representative of how the library's documentation instructs itself to be used.
03:43	ibdknox	so you just need what I wrote before
03:43	callen	I tried that.
03:43	ibdknox	(:use [clj-time.core :only [in-sexs interval epoch now]])
03:43	ibdknox	but not sexs
03:43	callen	ibdknox: look dude, it's in the comments
03:43	ibdknox	lol
03:43	ibdknox	no
03:43	ibdknox	you did something entirely different
03:44	ibdknox	look at my form
03:44	ibdknox	and your form
03:44	callen	see it
03:44	callen	trying it
03:44	callen	java.lang.IllegalStateException: Var irc/unix-now is unbound.
03:44	callen	?_?
03:45	callen	Unable to resolve symbol: in-secs in this context
03:45	callen	still doesn't work.
03:45	callen	updating gist.
03:46	callen	ibdknox: https://gist.github.com/1335982
03:46	ibdknox	you still don't have the same form
03:46	ibdknox	(:use [clj-time.core :only [in-secs interval epoch now]])
03:46	ibdknox	note that there's no apostrophe and that it's all wrapped in a vector
03:47	ibdknox	yours has a quote
03:47	ibdknox	and is missing the vector
03:47	ibdknox	I'm surprised that compiles
03:47	callen	clj-time.core [Thrown class java.lang.ClassNotFoundException]
03:48	callen	(:use [clj-time.core :only [in-secs interval epoch now]])
03:48	callen	it's identical now.
03:48	callen	and it fails with that exception.
03:48	hiredman	ibdknox: please make a note of the location of the closing paren of the ns form
03:48	ibdknox	it needs to be part of your ns declaration
03:48	ibdknox	hiredman: yeah, just noticed
03:49	ibdknox	I will gist exactly what that should look like
03:49	ibdknox	one sec
03:49	ibdknox	https://gist.github.com/1335992
03:50	ibdknox	callen: ^
03:50	callen	I already got it
03:50	Raynes	callen: Unrelated entirely to your project, but if you're doing IRC-related stuff, check out http://github.com/Raynes/irclj sometime. It wont help you if you're just doing this as a learning experiment, but it might be interesting to you in the future. lazybot runs off of it.
03:50	callen	the datetime object is fucked up
03:50	callen	Raynes: love it, thank you.
03:51	callen	ibdknox: thank you for the help, why does the namespace stop before all the func defs?
03:51	callen	ibdknox: I'm taking my cue from code that I'm building upon, but didn't write de novo.
03:51	callen	ibdknox: shouldn't all the defs/defns be part of the ns?
03:51	callen	ibdknox: at least if I'm following the example of other code I've seen.
03:51	Raynes	You can use and require namespaces dynamically, but you should almost never want to.
03:51	ibdknox	callen: think of it as setting a marker
03:52	ibdknox	callen: everything after here, until you find another ns decl, or something that resets the marker (like in-ns) is part of this one
03:52	ibdknox	we can't expand macros with the bots can we?
03:52	callen	ibdknox: then how does the reference in the defn which isn't part of the ns have a valid reference to clj-time.core's functions?
03:52	brehaut	Raynes: if ypu are pimping irclj does that mean you are working on it again?
03:53	Raynes	ibdknox: Sure, as long as it isn't bad.
03:53	Raynes	brehaut: Not at the moment, but I fully intend to partially rewrite it once I've got some time.
03:53	Raynes	Like, after the conj and I get caught back up with my book.
03:53	ibdknox	callen: because use loads all of the symbols from that other namespace into the current one
03:53	callen	sweet. all my code works now.
03:53	ibdknox	,(doc refer)
03:53	clojurebot	"([ns-sym & filters]); refers to all public vars of ns, subject to filters. filters can include at most one each of: :exclude list-of-symbols :only list-of-symbols :rename map-of-fromsymbol-tosymbol For each public interned var in the namespace named by the symbol, adds a mapping from the name of the var to the var to the current namespace. Throws an exception if name is already mapped to somethin...
03:54	brehaut	Raynes: fair enough :) if im not busy at the time, i'd be happy to help
03:54	callen	ibdknox: that doesn't make a huge amount of sense to me.
03:54	ibdknox	yeah that's not very clear
03:54	callen	ibdknox: I'm used to python which has a very flat and easy to understand scoping/namespace mechanism.
03:54	Raynes	ibdknox: Not trying to pimp lazybot here, but if you do that kind of thing in lazybot, he gists the truncated output for you.
03:54	ibdknox	oh?
03:54	Raynes	brehaut: That's be great.
03:54	Raynes	&(range 10000)
03:54	callen	ibdknox: I have no idea why references loaded via :use would be pushed out of an inline namespace into the one above it.
03:54	lazybot	⇒ (0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 ... failed to gist: Connection reset
03:54	Raynes	Well, that's too long.
03:55	Raynes	But you get the point.
03:55	ibdknox	yeah
03:55	ibdknox	callen: sorry inline namespace?
03:55	callen	the ns ends before the rest of the code
03:55	callen	that, structurally to my limited understanding, looks like it has ceased to be relevant at that point.
03:55	callen	the fact that it can mutate what is accessible in what is outside of it makes little sense to me.
03:55	ibdknox	##(macroexpand ns)
03:55	lazybot	java.lang.RuntimeException: Can't take value of a macro: #'clojure.core/ns
03:56	ibdknox	##(macroexpand '(ns))
03:56	lazybot	clojure.lang.ArityException: Wrong number of args (0) passed to: core$ns
03:56	ibdknox	##(macroexpand '(ns blah))
03:56	lazybot	⇒ (do (clojure.core/in-ns (quote blah)) (clojure.core/with-loading-context (clojure.core/refer (quote clojure.core))))
03:56	ibdknox	lol
03:56	Raynes	Well, your ns doesn't 'end' until the file is done.
03:56	ibdknox	callen: like I said, it sets a marker
03:56	callen	then why declare the ns?
03:56	Raynes	ns sets the ns for the code.
03:56	Raynes	Because it sets...
03:56	ibdknox	because you need to know what ns you're in
03:56	ibdknox	lol
03:56	callen	then why do people put stuff inside the ns block if you don't need to?
03:57	ibdknox	you do need to
03:57	callen	my code works
03:57	ibdknox	yes it does
03:57	ibdknox	lol
03:57	Raynes	Not that's necessarily your fault for not understanding. Could easily be our fault for failing to explain things properly.
03:57	callen	what changes if I move my defns and defs inside the ns?
03:57	Raynes	The problem is that I'm writing a book, and if I can't explain it...
03:57	ibdknox	I'm not sure, probably something terrible
03:58	callen	after herbert, I don't trust writing a book as a mark of expertise.
03:58	Raynes	The problem is that you think ns does something that it does not do.
03:58	callen	ibdknox: if you can't explain the difference, then how do you know you should do it one way or the other?
03:58	callen	Raynes: I have no idea at all what it does.
03:58	Raynes	The parentheses do not indicate the scope of the namespace.
03:58	ibdknox	callen: sorry, are you trying to be jackass now?
03:58	Raynes	ns sets the namespace for the entire file.
03:58	callen	I've established that.
03:58	callen	no, I'm not
03:58	callen	I'm actually quite sad and tired right now
03:58	callen	and just trying to understand code that already works.
03:58	ibdknox	perhaps you should sleep
03:59	callen	I won't until I understand this properly.
03:59	callen	you won't be rid of me that easily.
03:59	Raynes	Things might be clearer when the sun rises.
03:59	ibdknox	lol
03:59	ejackson	sleep is my ultimate debugging tool
03:59	callen	Chapter 2.4 "Vars, Bindings, and Namespaces" in the pragprog book doesn't explain it very well, at least from my POV
04:00	ibdknox	as Raynes said, the only thing you really need to understand is that you set the namespace for the entire file
04:00	ibdknox	it's idiomatic to place your uses and requires there
04:00	callen	and as I said earlier, I understood that much.
04:00	callen	I don't understand the point of it yet.
04:00	Raynes	No, I don't think so.
04:01	callen	is it to group everything you're doing in that file into that var?
04:01	ibdknox	no.
04:01	ibdknox	it's to group it in the namespace
04:01	Raynes	Everything you're questioning relies on Clojure setting the scope of the ns to the ns macro's scope itself, which is not the case and thus invalid.
04:01	ibdknox	this isn't actually different from python...
04:01	Raynes	You actually might be looking into it too much.
04:01	callen	I already sdaid I understood that.
04:01	callen	I keep trying to tell you
04:01	ibdknox	haha
04:01	callen	I understand that it sets it well past the end-paren.
04:01	ibdknox	alright man
04:01	callen	I am trying to understand the rest of its purpose.
04:02	ibdknox	You've got it all.
04:02	ibdknox	there isn't one
04:02	Raynes	I'm tired and quite sad right now.
04:02	ibdknox	don't use it :)
04:02	ibdknox	hahaha
04:02	callen	I am trying to explain that I understand a particular iota of the question at hand
04:02	callen	and you keep re-raising the single iota I do understand
04:02	callen	and aren't fucking listening to me
04:02	callen	when I say I understand that one bit
04:02	callen	and that we established that earlier
04:02	callen	than you patronize me
04:02	callen	and laugh at me when I try to get you to understand that
04:02	callen	this channel was better a year ago.
04:02	ibdknox	callen: you're antagonizing the people trying to help you
04:03	Raynes	You're raising questions that don't actually exist, I think. I can't tell. Do you actually have a question to ask?
04:03	ibdknox	I'm not sure how much better I can react
04:03	callen	Raynes: I asked earlier and was told they didn't know the answer.
04:03	callen	Raynes: I'll ask again.
04:03	Raynes	The last thing you asked was "what happens if I put the defs in the ns declaration" which makes utterly no sense given what you claim to already understand.
04:03	hiredman	ibdknox: /ignore
04:03	Raynes	We're not trying to patronize you or be mean to you or anything like that.
04:03	Raynes	We love to help.
04:03	callen	Raynes: currently my ns closes after a :use and an :import. (end-paren)
04:03	callen	hiredman: that doesn't help either.
04:04	callen	hiredman: you're borderline trolling when you publicly declare you're ignoring people.
04:04	callen	Raynes: now, I have my defs and defns after that end-paren, that is, it's not part of the syntactic portion of the ns.
04:04	Raynes	Right.
04:04	callen	Raynes: I understand FULLY that they are going to be in the ns.
04:04	callen	Raynes: What I do not understand is that if I shift it to the syntactic portion of the ns, what it changes, if anything.
04:04	callen	Raynes: in idiomatic Clojure code, it's usually part of the ns body.
04:04	callen	Raynes: such as in clj-time.core
04:04	callen	from what I can tell.
04:04	Raynes	Uh, really?
04:04	callen	Raynes: I am asking why and what it changes if anything.
04:05	Raynes	Can you show me an example?
04:05	callen	Raynes: I could be wrong, let me double-check.
04:05	ibdknox	callen: there's never code aside from :use :require, and a couple others inside of an ns decl
04:05	callen	fuck, no, nevermind.
04:05	callen	Raynes: nope. the way my code is written is idiomatic.
04:05	Raynes	Anyways, I'm not going to ignore people when I still have at least a partial chance at helping them.
04:05	callen	Raynes: nevermind.
04:05	callen	Raynes: I was mistaken.
04:05	callen	Raynes: hence confusion.
04:06	callen	ibdknox: that explains a lot, thank you.
04:06	ibdknox	we tried to clear that up, sorry it took so long.
04:06	Raynes	Nobody here but maybe hiredman actually wants to see people dissatisfied with the help they receive in this channel.
04:06	Raynes	;)
04:07	callen	Raynes ibdknox thank you both for tolerating my long enough to eliminate the source of my confusion.
04:07	callen	s/my/me/g\
04:07	Raynes	callen: Anyways, you can probably understand now why it sounded silly to us and like you weren't listening. What you were asking really didn't make much sense at all! If we got snarky, it's because we were probably getting a little frustrated at trying to make sense of it all. If we came off that way, I think I speak for both of us when I apologize for it.
04:08	ibdknox	that's an official record for me: over 12 hours on the HN front page
04:08	ibdknox	/win
04:08	Raynes	Good work./
04:08	ejackson	ibdknox: nice one !
04:09	callen	ibdknox: what'd you post?
04:09	ibdknox	callen: korma
04:09	callen	ibdknox: oh god I am excited about that one!
04:09	Raynes	http://news.ycombinator.com/item?id=3188609
04:09	ibdknox	though in terms of points, http://iwbyp.chris-granger.com got more
04:09	ibdknox	amusingly noir was lower than both
04:10	callen	Raynes: part of the problem is that I am a bit of a lone-wolf and can end up uh... waldeinsamkeit.
04:10	Raynes	God, Noir is so great.
04:10	ibdknox	lol forrest loneliness?
04:10	callen	Raynes: so I come to weird conclusions and misunderstandings that can require a lot of stack-unrolling to get to the bottom of.
04:10	Raynes	Hah. I found a new word.
04:10	callen	ibdknox: lost in the woods.
04:11	Raynes	$trans gr en waldeinsamkeit
04:11	lazybot	Languages not recognized.
04:11	Raynes	Oh yeah, that's broken anyway.
04:11	ibdknox	callen: I know, I speak German :)
04:11	callen	ibdknox: it expresses how I've approach programming, to my chagrin, very well.
04:11	callen	approached*
04:12	callen	I'm quite envious of people that can speak of this or that mentor they had in uni.
04:12	ibdknox	now I kind of feel like I should go buy the programming clojure book
04:12	Raynes	I've never had a mentor.
04:12	ibdknox	neither have I
04:12	ibdknox	I never had a formal CS education
04:12	callen	ibdknox: you mean the book that failed to save me from my own bizarre conceptions?
04:12	Raynes	As a matter of fact, the only time I've ever met another programmer was at a conference.
04:12	Raynes	Last year, no less.
04:12	callen	ibdknox: ditto. most of my CS background is pirated bibliographies and OCW.
04:13	ibdknox	callen: apparently noir is in the 2nd edition
04:13	callen	ibdknox: been working as a programmer for 5, nearly 6 years.
04:13	callen	ibdknox: whoa @ noir
04:13	callen	Raynes: conj?
04:13	Raynes	Yes
04:13	callen	well I have the log-structured data I wanted now.
04:13	callen	now I can start loading this into a database.
04:14	callen	I might have to write a parser first though.
04:14	callen	yeah, parser first. balls.
04:14	ibdknox	it dawned on me the other day that I'd been programming professionally for just shy of 8 years
04:14	callen	Raynes ibdknox thanks again, I can rest well now. Goodnight.
04:15	ibdknox	which is a bit ridiculous
04:15	Raynes	Night.
04:15	callen	ibdknox: tell that to the junior below me who is one year younger than me and has been working professionally for a few months.
04:15	callen	ibdknox: he finds me bewildering
04:15	callen	ibdknox: I was explaining b-trees and database indexes to him the other day
04:15	callen	ibdknox: said junior has two masters degrees.
04:15	Raynes	I'd hate to see him meet me.
04:16	ibdknox	haha Raynes would eat him for breakfast
04:16	callen	essprain?
04:16	ibdknox	I'm 24.. and I'm old ;)
04:16	callen	I'm 23, and old, and tired.
04:16	Raynes	I'm 17, and mostly just hungry.
04:16	callen	I'm positive the motorcycle is making me older.
04:17	ibdknox	hah and I'm apparently a beowulf cluster lol
04:17	ibdknox	I'm excited to actually meet fogus
04:18	ibdknox	he strikes me as a fun guy
04:18	Raynes	Met him. He's short.
04:18	Raynes	Fun, but short.
04:18	ibdknox	haha
04:18	callen	lol whut
04:18	Raynes	Shorter than I expected in any case.
04:18	ejackson	old ? you kids.
04:18	callen	<--- audibly cracking up
04:18	ibdknox	ejackson: I meant it sarcastically
04:18	callen	ejackson: I know plenty of 50 year old unabomber types that love Forth.
04:18	ibdknox	ejackson: though compared to Raynes...
04:18	ejackson	well, yes.
04:19	Raynes	I expected some big guy with sunglasses until I nearly stepped on him on my way to my room.
04:19	ibdknox	lol
04:19	ibdknox	he's that short?
04:19	ibdknox	I'm only 5'7"
04:19	Raynes	Well, he is about as short as I am.
04:19	callen	5' 9" here.
04:19	callen	which is?
04:19	Raynes	Probably around the same height, I think.
04:19	Raynes	5'5 or so.
04:19	Raynes	Maybe 5'4.
04:19	callen	that's short
04:19	ejackson	I thought he was a giant green Minotaur...
04:19	ibdknox	huh, you're right, I expected him to be much taller
04:20	callen	ditto.
04:20	ibdknox	alright.. I've been up way too damn long
04:20	callen	me too.
04:20	callen	it's pacific time, but still bad enough.
04:20	callen	I have work tomorrow as always.
04:20	callen	lkgnmdsklfhmdfgh
04:20	ejackson	ibdknox: hehe, I'm just waking up here England time.
04:20	callen	cheers all. It'll be a two-gallon coffee mug day.
04:20	callen	g'night
04:21	Raynes	It is precisely 3:17AM right now.
04:21	ibdknox	ejackson: haha well then, all day to play with Korma ;)
04:21	Raynes	Night boys and girls.
04:21	ejackson	cheers kids :P
04:21	callen	Raynes: That's 1320308288 to you, heathen!
04:22	Raynes	ejackson: Night gramps.
04:22	ibdknox	lol
04:23	ejackson	get off my damn lawn !
04:23	ibdknox	in case anyone is interested: HN/twitter brought in about 5000 viewers to sqlkorma.com
04:23	ejackson	nice stat indee.
04:23	ibdknox	I like looking at where they're from, it always ends up surprising me some
04:24	ibdknox	more than half were outside the US
04:24	ejackson	that's promising
04:25	ibdknox	the UK was the top of the european countries actually
04:25	ibdknox	ejackson: well done
04:25	ejackson	oh, that was me with some perl
04:25	ibdknox	haha user-agent: curl
04:26	ibdknox	damn :p
04:26	ibdknox	alright. I'm outta here. Catch you folks later
04:26	ejackson	later
05:03	Blkt	good morning everyone
08:10	jscott1989	hey, if i want a linked list, so i can maintain an arbitrary order (specifically where i can insert and element at any position) how would i go about that?
08:13	jscott1989	ignore my stupid question
08:13	jscott1989	:)
08:13	jscott1989	solved it
09:11	bpr	is there a way to explore java packages? I'm thinking something like a cross between clojure.contrib.repl-utils/show and clojure.core/ns-*
09:16	scottj	well you can inspect packages in Slime but it doesn't give much that's useful
09:50	clgv	Is there any documentation support for functions with options in clojure yet?
09:50	clgv	This is often annoying when you have several layers where the options are passed through.
09:55	dpritchett	Raynes: Really dug the tryclojure on heroku post. Well done.
09:55	cemerick	clgv: beyond enumerating the options in a docstring, you mean?
09:56	clgv	cemerick: that is one good thing. I would go further and try some transitive analysis on the options for function calls that have options as well
09:57	cemerick	clgv: If you enumerate the options in a keyword destructuring form, they'll also show up in (:arglists (meta #'foo))
09:58	cemerick	FYI: The official unofficial homes of 'Clojure Programming', coming soon from O'Reilly: http://twitter.com/ClojureBook and http://clojurebook.com
09:58	clgv	cemerick: nice url you captured there^^
09:59	cemerick	I also got clojure-book.com and clojureprogramming.com, but I decided to go simple. :-)
10:02	davidrupp	cemerick: Looking forward to buying the ebook. Any idea why it's not available through the normal (non-Safari) Rough Cuts?
10:03	clgv	oh, I didnt know that you shortcut it like that (defn f[x & {:keys [bla blubb]}] ...) to not have to create a hashmap yourself
10:04	clgv	but it seems to work on 1.3
10:04	cemerick	davidrupp: No idea, sorry. Are you talking about their direct early access thing? I heard something about it, but have no control/visibility into those sorts of things.
10:05	cemerick	I can try to beat the bushes on that…
10:05	davidrupp	Yeah, most of their ebooks are available for direct download, no Safari required, even in Rough Cuts.
10:05	davidrupp	Not a huge concern.
10:06	cemerick	You can download it through safari AFAIK…
10:19	clgv	I think about this syntax: (defn+opts f [x] [opt1 [opt2 10]] ...) is this idiomatic or is there a better way to specify the optional params with default values?
10:22	cgray	hi, where do macros go so that the clojurescript compiler can find them?
10:23	kephale	clgv: i'm a little confused by opt1 not having something after it but what about (defn f [x & {:keys [opt1 opt2] :or {opt1 blah opt2 10}}] …)
10:23	clgv	kephale: no, I try to improve that defn statement to something more useful
10:24	clgv	kephale: I aim for specifying default values but also being able to omit them
10:25	kephale	clgv: aha cool… one of my complaints about the current approach (although it isn't exactly relevant to this situation) is that the value of :keys is technically implied by the value of :or since everything must have a default
10:26	clgv	hmm but maybe I have to separate it as well
10:29	kephale	or have a single map where you explicitly specify nil for things that don't have defaults, which wouldn't be pretty
10:30	neotyk	Good morning everyone!
10:30	licenser	has anyone managed to get clojure-jack-in work with cake?
10:31	kephale	i think i like the syntax you suggest, but maybe making a bit more uniform with the existing approach: (defn+opts f [x & [opt1 [opt2 10] …] body)
10:31	kephale	oops, missing a ]
10:34	clgv	kephale: I dont like the & ^^
10:34	clgv	it suggest the other semantic I think
10:34	clgv	maybe I introduce a sign for it. should be possible. lets check
10:35	kephale	maybe a pipe
10:35	clgv	or a question mark since they are optional
10:36	kephale	hrm perhaps, though in the case where x is some function predicate that could get confusing
10:36	clgv	kephale: example?
10:36	TimMc	,(let [\| 5] (* \| 2))
10:36	clojurebot	10
10:37	clgv	(defn+opts f [x ? {opt1 9, opt2 10}] ... ) and (defn+opts f [x ? [opt1, [opt2 10]] ... ) looks well
10:38	clgv	ok I'll note question mark and pipe
10:38	kephale	(defn+opts f [float? ? options] …)
10:39	clgv	hmm the pipe has more visual impact. thats true
10:39	clgv	is it used for anything yet?
10:40	kephale	well apparently TimMc wants to use it as a variable
10:40	clgv	lol, that's no harm - he can continue to do that^^
10:41	clgv	my implementation can ignore that^^
10:41	clgv	like defrecord uses '& as variable name ;)
10:45	cark	clgv: so you're going to make your own destructuring too ?
10:45	cark	throwing away the good work of our language developers for estetics only ?
10:47	clgv	cark: no I do not. I improve optional parameter handling for my project
10:51	cark	that's a whole cnan of worm you're opening here
10:51	cark	consider this :
10:51	cark	,(let [f (fn [& {[a1 a2] :a}] [a1 a2])] (f :a [1 2]))
10:51	clojurebot	[1 2]
10:52	cark	there is destructuring going on
10:52	clgv	cark: I do not have to consider this. I guess you did not get my intentions ;)
10:52	cark	recursive destructuring should i say
10:53	cark	ohwell ok =)
10:56	clgv	you recheck when it's ready and I found it stable enough to put it on github ;)
10:58	cgray	has anyone successfully used macros with clojurescript? i'm having trouble getting the compiler to find them...
10:59	PPaul	is there a function that is like project, but returns a set of hash-maps with with all the keys but the ones project was given?
11:00	PPaul	or do i have to do something like project, and difference?
11:05	Chousuke	PPaul: sounds like you could just use dissoc to remove the keys
11:06	PPaul	yeah, but project works over a set of hashes
11:06	PPaul	sticking to set stuff while working with sets seems like the right thing to do
11:10	PPaul	i think i don't need it anyway, oh well
11:10	PPaul	^_^
11:20	PPaul	how do i map over a hash-map... (map first/second {}) works, but when i do (map (fn [key val] val) {}) i get errors related to the num of args passed to the function
11:20	PPaul	is it because i'm not destructuring?
11:20	PPaul	(fn [[key val]] val) is what i should use?
11:21	PPaul	woo, answered my own question
11:28	TimMc	PPaul: Why not use val?
11:29	TimMc	,(map val {:a 1, :b 2})
11:29	clojurebot	(1 2)
11:29	PPaul	i want to use both
11:29	TimMc	OK.
11:30	PPaul	i projected, and i want to create a new hash with the projection (it's a compression alg)
11:30	cgray	is there a way to get cljsc to tell me what its classpath is?
11:30	cgray	,(vals {:a 1, :b 2})
11:30	clojurebot	(1 2)
11:31	PPaul	i have a static method that takes in 2 Str args... is this the right sig for it?
11:31	PPaul	(:gen-class
11:31	PPaul	:methods [#^{:static true}[statsTransform [String String] void]]))
11:32	PPaul	also, can i use void return even if it does return something (i'm guessing i can because all clojure functions return)
11:33	licenser	Raynes: you're there mate?
11:47	ljos	What is best practice for writing to file in Cljoure? I
11:48	licenser	ljos: depends on what you want, spit is a very easy way :)
11:48	ljos	I only need easy. Nothing fancy :)
11:49	licenser	that spit is your friend
11:49	licenser	,(doc spit)
11:49	clojurebot	"([f content & options]); Opposite of slurp. Opens f with writer, writes content, then closes f. Options passed to clojure.java.io/writer."
11:50	ljos	Thank you :)
11:50	licenser	you're more then welcome :)
11:51	ljos	Another question. I know this is not excactly best practice, and it sounds like I'm doing things I shouldn't, but is there a way I can capture global variables?
11:52	ljos	So if I have (def a 1) (defn retra [] a) (let [a 3] (retra)) I want that to return 3, but let doesn't capture globals.
11:53	dnolen	ljos: you want binding.
11:53	dnolen	ljsos: and you need to declare a to be dynamic
11:53	dnolen	(def ^:dynamic a 1)
11:54	dnolen	ljos: I'm assuming you're using 1.3.0
11:54	ljos	ok. Sweet! And yes I am.
11:55	ljos	Thank you.
11:57	PPaul	(:gen-class
11:57	PPaul	:methods [#^{:static true}[statsTransform [String String] bool]]))
11:57	PPaul	bool is throwing a compile error
11:57	ljos	binding will only work inside it's own lexical scope right? It won't do stupids things like actually changing the variable for functions outside of the scope that might be running at the same time under other scopes?
11:57	PPaul	what should i use instead (i can't find what options i have in docs or gen-class help sites)
11:58	ljos	its*
11:58	dnolen	ljos: it's dynamic binding. It will change that vars value until it exits.
11:58	dnolen	any code that is code called within it will see the new value.
11:59	PPaul	nm i have to use boolean (really not familiar with java anymore)
11:59	ljos	dnolen: yes, but if I have code running in parallel, will it see it?
11:59	ljos	That is not within the binding.
12:00	dnolen	ljos: dynamic binding is thread local. however threads spawned by the current will also see the new value.
12:01	hugod	assuming those threads are spawned by a clojure core function…
12:01	ljos	dnolen: That was what I asked. So the change won't be global, only within the lexical scope of the binding, right?
12:01	dnolen	ljos: dynamic scope
12:03	ljos	dnolen: ah. I read on what dynamic is vs lexical, I had some missconception of what lexical meant. Thank you for the help. That answered everything really..
12:04	dnolen	ljos: np
12:14	ThreeCups	I'd like to take this map {:a [[0 0] [0 1]], :b [[1 0] [2 0]]} and transform it to this vector [[[[0 0] :a] [[0 1] :a]] [[[1 0] :b] [[2 0] :b]]]. I'm sure there's an elegant way to do it, but I'm stumped
12:20	halfprogrammer	,((apply map vector ((juxt vals keys) {:key1 1 :key2 2})))
12:20	clojurebot	#<ClassCastException java.lang.ClassCastException: clojure.lang.LazySeq cannot be cast to clojure.lang.IFn>
12:21	halfprogrammer	ThreeCups: think something like this should work
12:21	halfprogrammer	,((apply map vector ((juxt vals keys) {:key1 1 :key2 2}))
12:21	clojurebot	#<ExecutionException java.util.concurrent.ExecutionException: java.lang.RuntimeException: EOF while reading>
12:21	halfprogrammer	,(apply map vector ((juxt vals keys) {:key1 1 :key2 2}))
12:21	clojurebot	([2 :key2] [1 :key1])
12:22	halfprogrammer	got it as expected this time :)
12:22	halfprogrammer	does this serve your purpose?
12:22	ThreeCups	halfprogrammer: Thanks! I think this'll work. Let me give it a try
12:26	ThreeCups	Never used juxt before... reading the docs to understand it
12:31	amalloy	~juxt
12:31	clojurebot	juxt is usually the right answer
12:34	jodaro	i love juxt
12:34	jodaro	used it for the first time the other day
12:37	technomancy	the biggest problem with juxt is that it's too awesome
12:37	technomancy	you're always tempted to use it where you shouldn't
12:37	ejackson	a
12:37	ejackson	spiderman function
12:37	jodaro	you mean like when i go
12:37	jodaro	(ns ...
12:37	jodaro	(juxt ....
12:38	scottj	anyone know other lisp dialects that have juxt builtin?
12:38	jodaro	500 lines
12:38	jodaro	)
12:38	technomancy	scottj: liskell? =)
12:38	halfprogrammer	,(mapcat (fn [m] (let [k (key m) vx (val m)] (zipmap vx (repeat k)))) {:a [1 2 3] :b [4 5 6]})
12:38	clojurebot	([3 :a] [2 :a] [1 :a] [6 :b] [5 :b] ...)
12:39	halfprogrammer	ThreeCups: even this seems to work :)
12:39	ThreeCups	halfprogrammer: thanks again! still trying to wrap my mind around juxt :)
12:40	ThreeCups	All functions passed to juxt have to have the same arity, correct?
12:41	halfprogrammer	think the arity is also fixed to 1
12:42	ThreeCups	,(doc juxt)
12:42	clojurebot	"([f] [f g] [f g h] [f g h & fs]); Takes a set of functions and returns a fn that is the juxtaposition of those fns. The returned fn takes a variable number of args, and returns a vector containing the result of applying each fn to the args (left-to-right). ((juxt a b c) x) => [(a x) (b x) (c x)]"
12:42	halfprogrammer	,((juxt + - *) 1 2 3)
12:42	clojurebot	[6 -4 6]
12:42	halfprogrammer	i was wrong :)
12:44	ljos	Is there a way to use relative paths in spit? I want to do this: (spit "~/newfile" "content of new file") .
12:45	kephale	hurrah for the triumvirate: juxt, comp, partial
12:45	ThreeCups	halfprogrammer: in your first solution, you has this: (juxt vals keys), is that gauranteed to return the vals and keys in the same order?
12:45	technomancy	ThreeCups: it's guaranteed, yes
12:46	pdk	zipmap yo
12:46	halfprogrammer	technomancy: even if the map does not give the guarantee?
12:46	technomancy	halfprogrammer: right, it's deterministic for a single map
12:48	amalloy	ljos: that's not relative, it's absolute with $HOME expansion
12:49	ljos	amalloy: oh.. yeah ofc.
12:49	ljos	I'm getting tired.
12:50	amalloy	anyway, you can do it by hand with (System/getProperty "user.home"), or i'm sure there are some libraries out there to make java pretend it's posix
12:50	amalloy	but normally your shell expands ~, not your filesystem, so java doesn't do it
12:51	TimMc	,(doc sort-by)
12:51	clojurebot	"([keyfn coll] [keyfn comp coll]); Returns a sorted sequence of the items in coll, where the sort order is determined by comparing (keyfn item). If no comparator is supplied, uses compare. comparator must implement java.util.Comparator."
12:51	TimMc	,(sort-by #(do (pr %) %) '[5 4 3 2 1])
12:51	clojurebot	54534352423251413121
12:51	clojurebot	(1 2 3 4 5)
12:52	TimMc	That's disappointing.
12:53	TimMc	In Ruby you can apparently define shuffle as array.sort_by { rand } because the keyfn (rand) is only called once per index.
12:53	hiredman	clojurebot: what?
12:53	clojurebot	what is moustache
12:53	amalloy	TimMc: it's a tradeoff
12:53	ljos	amalloy: that worked. Thank you.
12:53	TimMc	hrm
12:54	amalloy	suppose the predicate is super-cheap, such as identity. you want to double your storage costs and increase complexity in order to save calling this function?
12:54	amalloy	whereas clojure makes it trivial to get ruby's behavior
12:54	bobnodes	oh hi
12:54	bobnodes	anyone know of a sandbar replacement?
12:55	amalloy	&(let [f #(do (pr %) %)] (map second (sort-by first (map f ['5 4 3 2 1]))))
12:55	lazybot	java.lang.IllegalArgumentException: Don't know how to create ISeq from: java.lang.Long
12:55	scottj	bobnodes: noir
12:55	technomancy	,(sort-by (memoize #(do (pr %) %)) '[5 4 3 2 1])
12:55	clojurebot	54321
12:55	amalloy	&(let [f #(do (pr %) %)] (map second (sort-by first (map (juxt f identity) ['5 4 3 2 1]))))
12:55	clojurebot	(1 2 3 4 5)
12:55	lazybot	⇒ 54321(1 2 3 4 5)
12:55	bobnodes	I've spent the last 8 hours trying to make it authorize a simple login form. But this seems completely broken in any recent version
12:55	technomancy	amalloy: beat ya =P
12:56	TimMc	technomancy: Beat me, too. :-/
12:57	amalloy	technomancy: yours doesn't call f at all for duplicate items, which doesn't seem to actually align with the ruby behavior he wanted
12:57	technomancy	hm; yeah I could go either way on that
12:59	amalloy	i guess i could go super-efficient by using pop instead of second :P
13:03	TimMc	amalloy: annotate!
13:03	amalloy	give the man a prize
13:04	TimMc	TimMc: botsnack
13:04	TimMc	TimMc: Thanks! Om nom nom!!
13:09	TimMc	pandeiro: ^ amalloy made a much nicer version of my shuffle-by-sort.
13:10	pandeiro	TimMc: i already modified your version to work with cljs but i'd like to take a look, link?
13:10	amalloy	TimMc: i read somewhere that using sort-by rand to shuffle leads to not-very-good shuffling
13:10	amalloy	so i'm not sure you'd want to do this
13:10	TimMc	amalloy: We were discussing that yesterday.
13:10	TimMc	amalloy: pandeiro is working in cljs, where shuffle isn't defined.
13:11	TimMc	pandeiro: (map second (sort-by first (map (juxt rand identity) [5 4 3 2 1])))
13:11	amalloy	TimMc, pandeiro: https://gist.github.com/805546
13:11	TimMc	But use Fisher Yates or Knuth instead.
13:11	amalloy	a functional impl of Fisher-Yates
13:12	amalloy	with the nice side benefit that it's lazy - you can take only N items out of the front and the rest won't be shuffled
13:14	amalloy	looking at it again it seems like it would be nicer to implement using lazy-seq directly instead of mapping first over an iterate
13:25	amalloy	TimMc, pandeiro: updated gist with an impl that is probably more performant
13:28	TimMc	I see a shuffle in there.
13:28	amalloy	TimMc: (fn shuffle ...)
13:28	TimMc	Oh!
13:28	TimMc	shadowed, got it
13:28	amalloy	$mail fliebel remember back when we were looking at various lazy-shuffle implementations? i finally realized i should be using a transient vector and it's about four times faster than my last impl
13:28	lazybot	Message saved.
13:32	amalloy	$mail fliebel forgot the link! https://gist.github.com/805546
13:32	lazybot	Message saved.
13:35	ThreeCups	,(into [] {:a 1, :b 2})
13:35	clojurebot	[[:a 1] [:b 2]]
13:35	ThreeCups	Is there a better way to do this?
13:36	duck1123	Is there anything easier than set/difference for removing a value from a set, or better yet, something like a dissoc-in that will work on maps containing sets?
13:36	duck1123	ThreeCups: ,(vec {:a 1 :b 2})
13:37	technomancy	duck1123: update-in plus disj?
13:37	duck1123	technomancy: disj is what I was missing
13:37	technomancy	yeah. I was a little confused as to why dissoc didn't work on sets at first.
13:38	technomancy	I think it makes sense since sets aren't associative, but it's a bit surprising at first.
13:39	duck1123	much nicer, thanks
13:45	TimMc	amalloy: Having trouble understanding lazy-shuffle. The innermost "coll" is the remaining elements, right?
13:45	amalloy	yeah
13:45	TimMc	The last line of shuffle... are you doing an exchange?
13:45	amalloy	right
13:46	TimMc	Oh, I get it! You add a random element to the lazy seq, and put the top element in its place.
13:46	amalloy	yes
13:46	TimMc	then pop the duplicate
13:46	TimMc	Cute.
13:47	ThreeCups	How do I flatten [[1 2] [3 4]] to [1 2 3 4]? (reduce concat [[1 2] [3 4]])?
13:47	chouser	yes, or apply concat or mapcat seq
13:47	amalloy	ThreeCups: strongly prefer not producing [[1 2] [3 4]] in the first place
13:48	gfredericks	amalloy: :D
13:48	ThreeCups	amalloy: one thing at a time :). Thanks for the advice
13:48	amalloy	eg, much nicer to ##(mapcat (fn [[x y]] [(inc x) y]) {1 2 3 4 5 6}) than to ##(apply concat (map (fn [[x y]] [(inc x) y]) {1 2 3 4 5 6})
13:48	lazybot	⇒ (2 2 4 4 6 6)
13:48	TimMc	for for for
13:48	amalloy	&(apply concat (map (fn [[x y]] [(inc x) y]) {1 2 3 4 5 6}))
13:48	lazybot	⇒ (2 2 4 4 6 6)
13:49	amalloy	indeed. i always find the for version of the latter somewhat unpalatable though, and it's slower as well
13:50	amalloy	&(for [[x y] {1 2 3 4 5 6}, item [(inc x) y]] item) ; yuck
13:50	lazybot	⇒ (2 2 4 4 6 6)
13:54	TimMc	I thought you were all for & juxt these days
13:56	amalloy	it's true
13:59	chouser	'for' rocks for golfing
14:03	hiredman	I just wish there was something like mapcat for for, you can kind of hack it in, but it's not pretty
14:04	fdaoud	,(flatten [[1 2] [3 4]])
14:04	clojurebot	(1 2 3 4)
14:05	amalloy	~flatten
14:05	clojurebot	flatten is clojure.contrib.seq-utils/flatten
14:05	amalloy	ugh
14:05	amalloy	clojurebot: forget flatten \|is\| clojure.contrib.seq-utils/flatten
14:05	clojurebot	I forgot that flatten is clojure.contrib.seq-utils/flatten
14:05	raek	,(flatten [[[1 2] [3 4]] [[5 6] [7 8]]])
14:05	clojurebot	(1 2 3 4 5 ...)
14:06	ibdknox	flatten is one of those scary functions
14:06	raek	,(apply concat [[[1 2] [3 4]] [[5 6] [7 8]]])
14:06	clojurebot	([1 2] [3 4] [5 6] [7 8])
14:06	amalloy	clojurebot: flatten \|is\| rarely the right answer. What if your "base type" is a list? 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.
14:06	clojurebot	In Ordnung
14:06	ibdknox	you'll use it, not think about needing to retain a nested collection and then you're screwed
14:07	amalloy	~flatten
14:07	clojurebot	No entiendo
14:07	ibdknox	lol
14:07	amalloy	hiredman: haallllp
14:08	hiredman	:(
14:10	chouser	amalloy, ibdknox: correct.
14:10	chouser	aren't you happy you have my approval? :-P
14:10	amalloy	hiredman: seriously though what did i get wrong there? i know he sometimes gets confused when the factoid has an "is" in there, so i wrapped the first one with pipes to get his attention
14:10	amalloy	but i'd like to make sure he knows this fact so i can stop crusading
14:11	hiredman	yeah
14:11	ibdknox	chouser: I'm checking that box on my list of things I need before I die :p
14:11	chouser	heh, I bet.
14:11	hiredman	my guess would be you need to escape the following is'es (\is) but the real answer is clojurebot's parser needs an overhaul
14:11	amalloy	ibdknox: chouser will never give you approval to die. you'll be kept alive in a vat, inventing useful new tools
14:12	amalloy	clojurebot: flatten is rarely the right answer. What if you need to use a list as your "base type"? 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.
14:12	clojurebot	c'est bon!
14:13	amalloy	in the meantime, i'll just rephrase without "is" :P
14:13	TimMc	flatten?
14:13	clojurebot	flatten is rarely the right answer. What if you need to use a list as your "base type"
14:13	ibdknox	hahaha
14:13	amalloy	~botsmack
14:13	clojurebot	clojurebot evades successfully!
14:13	ibdknox	~botsmack
14:13	clojurebot	Owww!
14:14	amalloy	~flatten
14:14	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.
14:14	chouser	~for
14:14	clojurebot	for is a loop...in Java
14:15	chouser	~for
14:15	clojurebot	for is not a loop
14:15	ibdknox	...
14:15	ibdknox	lol
14:15	chouser	clojurebot: for is complected
14:15	clojurebot	You don't have to tell me twice.
14:15	ibdknox	that seems to be the new word these days :p
14:15	amalloy	chouser: oh boy, tell us about LOOP
14:16	dnolen	clojurebot: LOOP is complect complected
14:16	clojurebot	You don't have to tell me twice.
14:17	chouser	though interestingly, 'for' creates a lexical space in which some things can be simpler (not to mention easier) than outside the 'for'
14:18	cgray	amalloy: heh, i was just about to guiltily use flatten before i thought "i'll just have a look at #clojure for a second"
14:18	ibdknox	chouser: hm?
14:19	amalloy	cgray: have you repented?
14:19	amalloy	accept mapcat into your heart
14:19	cgray	amalloy: indeed, mapcat is really what i wanted
14:20	ibdknox	I've wanted flatten exactly once
14:20	ibdknox	and it was actually the right thing to use
14:20	technomancy	it's pretty trivial to flatten with tree-seq if that's what you really want
14:21	technomancy	kind of surprised when that got promoted
14:21	ibdknox	it was for my version of that music as data thing: https://github.com/ibdknox/music-as-data
14:21	amalloy	agreed. my guess was that flatten is something that's taught sometimes as an interesting problem and someone wanted to show off that it's easy in clojure
14:22	technomancy	I dunno; for showing off purposes implementing it with tree-seq gets more wows
14:23	lobotomy	wrote another blog post on clojure: http://lobotomys.blogspot.com/2011/11/chess-problem-solver.html
14:23	lobotomy	this one's a bit long. executive summary: straightforward dfs is straightforward
14:24	technomancy	flatten is literally (partial tree-seq coll? seq), right?
14:24	Raynes	technomancy: Heh, we actually realized we needed to do that in clojail last night. Flatten wasn't flattening maps, so you could bad stuff.
14:24	technomancy	oh, sequential? instead of coll?
14:24	Raynes	&{:foo (eval '(+ 3 3))}
14:24	ibdknox	,(doc tree-seq)
14:24	lazybot	⇒ {:foo 6}
14:24	tolstoy	I saw in the Joy of Clojure free chapter something like: (defn [& (keys :foo 1 :bar 2)] …). Is that documented anywhere?
14:24	clojurebot	"([branch? children root]); Returns a lazy sequence of the nodes in a tree, via a depth-first walk. branch? must be a fn of one arg that returns true if passed a node that can have children (but may not). children must be a fn of one arg that returns a sequence of the children. Will only be called on nodes for which branch? returns true. Root is the root node of the tree."
14:25	Raynes	could do bad stuff*
14:25	amalloy	technomancy: right
14:25	pjstadig	technomancy: wow, you're right that is more readable than flatten :-p
14:25	technomancy	wait, filter complement of sequential? ? dude.
14:25	technomancy	as pjstadig would say: remove, man.
14:25	technomancy	=)
14:25	Raynes	tolstoy: http://stackoverflow.com/questions/3337888/clojure-named-arguments
14:27	tolstoy	Raynes: Thanks. I guess it's not a canonical technique?
14:27	Raynes	Sure it is. Why wouldn't it be?
14:27	tolstoy	There's not a page on clojure.org. ;)
14:27	ibdknox	hah
14:28	Raynes	I guess it just isn't officially documented. It has been possible for a couple of Clojure versions now.
14:28	tolstoy	I see there's a contrib.
14:28	Raynes	But there are plenty of stuff like that SO post to learn about it.
14:28	tolstoy	Yep.
14:28	ibdknox	huh, I'm surprised there isn't a heading for destructuring
14:28	amalloy	tolstoy: i don't think there's a page on clojure.org documenting that if you want to multiply a number by three you should use (* 3 x), but that's got to be the canonical technique
14:29	ibdknox	it's probably one of most useful, yet least used by new people, things in clojure
14:29	mdeboard	I want destructuring in Python.
14:29	mdeboard	So bad.
14:29	tolstoy	Sometimes I see techniques here and there, and I wonder how I could have found out about them if not randomly searching for something completely different. So, I was hoping there was some secret page I had missed….
14:30	ibdknox	tolstoy: at some level it's a matter of building things on top of the concepts that you know already exist
14:30	amalloy	that's how programming works though. you read source and notice people doing neat stuff that you want to steal
14:30	mdeboard	tolstoy: Reading a couple of good books on the topic is the best way to learn things about new technologies
14:30	tolstoy	amalloy: Ah, come on. I'll not bite. I was just curious if there was a page somewhere that kind of had interesting idioms for typical kinds of things. That's all. I didn't mean to complain or criticize.
14:30	ibdknox	mdeboard: eh, I didn't read any books. I read code.
14:30	mdeboard	amalloy: Eh some techniques are sufficiently advanced that you dont' even know where to begin.
14:30	mdeboard	ibdknox: How long have you been programming with lisps?
14:30	amalloy	mdeboard: for sure, books are good
14:30	ibdknox	mdeboard: this is my first
14:31	amalloy	mdeboard: mine too. started last summer
14:31	ibdknox	soooo, 8ish months?
14:31	ibdknox	no
14:31	technomancy	oh, wait did you say "lisps"?
14:31	ibdknox	10 months
14:31	mdeboard	lithpth
14:31	ibdknox	lol
14:31	technomancy	it's like in the Three Amigos. "We could go for a walk and you could kiss me on the veranda." / "Oh, uh... Lips would be fine."
14:32	mdeboard	ibdknox: It was hard for me to find good projects written with Clojure that provided grokkable examples.
14:32	mdeboard	but I'm a newb to programming in general
14:32	ibdknox	mdeboard: I've been working on that ;)
14:32	mdeboard	technomancy: I program with my face. 1. slam face into keyboard 2. roll 3. ??? 4. Push to github
14:32	mdeboard	ibdknox: I've noticed.
14:32	ibdknox	I have a fun idea that I might try to convince Raynes to help me with
14:33	ibdknox	I want to use Clojail to create a "trynoir"
14:33	amalloy	ibdknox: you should try perl, eh?
14:33	ibdknox	oh god
14:33	ibdknox	I hate perl... even more than I hate PHP
14:33	ibdknox	which is saying something
14:33	Raynes	ibdknox: Sounds like fun.
14:33	Raynes	Perl 6 is interesting.
14:34	mdeboard	What was the Clojure project that was just on HN a couple days ago? I think cemerick authored?
14:34	mdeboard	Not the SQL DSL
14:34	callen	as a fan of Indian food, I find the nomenclature of that SQL DSL to be pleasing.
14:35	mdeboard	Oh, wait, Python DOES have destructuring, of a sort.
14:36	mdeboard	args, *kwargs
14:36	mdeboard	Anyway, sorry.
14:37	amalloy	mdeboard: it has tuples and splatting, which does get you a fair chunk of the way there
14:38	mdeboard	amalloy: I'm only familiar with "splatting" in the context of references in C
14:39	mdeboard	Never heard it applied to (not '(C, C++))
14:39	ThreeCups	I have a fn that takes a map and a keyword (fn [m k] ...). I want to either create a new set containing k or conj k into a set that already exists as a value in the map. How could I do this? Or is my old nature imperative style still creeping up?
14:39	amalloy	mdeboard: as i understand it that's what ruby calls the *args construct, and possibly also python? i've never heard of it in C
14:39	sdbo	Does anyone know a particularly good Resque-like backgrounding library for Clojure?
14:39	amalloy	http://stackoverflow.com/questions/2322355/proper-name-for-python-operator
14:39	mdeboard	amalloy: May haev just been a former CS professor's word for it
14:40	technomancy	sdbo: there's this: https://github.com/technomancy/die-roboter
14:40	technomancy	hasn't seen much real-world usage yet, but it's really simple; only about 100 LOC
14:40	mdeboard	die-bart-die
14:40	technomancy	mdeboard: http://www.cul.de/images/autotoolscg.jpg
14:41	fdaoud	mdeboard: "no one german could be harmful!" ?
14:41	tolstoy	Raynes: Thanks for your pointer. I guess I'm after "restructuring", which led me to: http://blog.jayfields.com/2010/07/clojure-destructuring.html
14:41	mdeboard	fdaoud: nailed it
14:41	fdaoud	:)
14:41	tolstoy	Raynes: I guess I should dip in to Joy of Clojure soonish. ;)
14:42	mdeboard	tolstoy: it's really a great book, only tech book I've ever read straight through like a novel. I mean, until the agents/refs/etc part.
14:42	sdbo	technomancy: that's pretty close to what I was hacking around on building— you just set up the execution context in the worker before calling roboter/work?
14:43	tolstoy	mdeboard: The free chapter pretty much backs up what you say. I've got some pressing problems I'm using Clojure for, but it'll be nice to go back and see how to improve.
14:43	technomancy	sdbo: the execution context meaning database connections, etc?
14:43	fdaoud	mdeboard: +1. It really is tremendously well-written. I wonder who used the word "propensity".. chouser, fogus, or the editor?
14:44	sdbo	technomancy: database connections, namespaces, etc
14:44	technomancy	sdbo: right, pretty much
14:44	mdeboard	tolstoy: I wish I had some problems that were of a nature to warrant solving them with Clojure.
14:45	mdeboard	fdaoud: I don't follow
14:45	technomancy	sdbo: I'm actually working on adding a -main function that you can pass the number of workers you want to set up and things like that
14:46	fdaoud	mdeboard: I was just impressed by the vocabulary. Like in chapter 9, "Newcomers to Clojure have a propensity to hack away at namespace declarations until they appear to work." I admit having to look up "propensity".
14:46	tolstoy	mdeboard: Well, I don't have that. I just do stuff like query mongo, shove data into file, upload, etc, or web apps, or web services.
14:46	mdeboard	tolstoy: Oh, i think I'd get a shoe shoved into inappropriate places if I wrote production code in anything but Python. :P
14:46	tolstoy	mdeboard: But I would really like to see if I can see these tasks in a completely new way.
14:47	mdeboard	fdaoud: Ah
14:47	sdbo	technomancy: this is basically perfect, I think, haha
14:47	tolstoy	mdeboard: Like incoming rabbitmq messages as an infinite sequence, etc.
14:47	mdeboard	tolstoy: Even utilities. Plus it wouldn't be fair. If I got hit by a truck, maintainability of the code base shouldnt' be impacted
14:48	tolstoy	mdeboard: I'm doing our production code in Scala, and all the side-issues (monitoring, etc) in Clojure.
14:49	fdaoud	clojurebot: scala
14:49	clojurebot	{((x: Any, y: Any) => (f: Function2[Any, Any, Any]) => f(x, y))(1, 2)((x: Any, y: Any) => x)}
14:49	technomancy	sdbo: the error handling still needs a bit of help, but I think the rest of it is pretty solid
14:49	tolstoy	Hah!
14:49	mdeboard	tolstoy: Ah that's a little more reasonable.
14:50	mdeboard	I think eyes would pop at anythign but Python here.
14:50	tolstoy	I find I really like the type system stuff in Scala (esp. what I can use of Algebraic types). But when I work in Clojure, I don't miss it.
14:50	fdaoud	mdeboard: if you get hit by a truck, don't worry about the maintainability of your code..
14:50	tolstoy	I do miss "linked" actors, though, as a way to manage resources (not necessarily concurrency).
14:50	mdeboard	fdaoud: :P
14:51	sdbo	technomancy: ohhh, does wabbitmq connect to other AMPQ servers? I can't seem to get RabbitMQ running on my cluster
14:51	tolstoy	maintainability: I figure if you can document the interfaces, and keep your apps to single-purpose functionality, someone could just rewrite the thing if they need to.
14:51	technomancy	sdbo: probably, but I haven't tried it
14:52	tolstoy	(But then I'm not writing trading software.)
14:54	sdbo	technomancy: need to run atm, but I'll definitely try this out—perfecto
14:54	technomancy	sdbo: definitely interested in feedback
14:54	kzar	Is there a way to split a decimal number into it's component parts? So for 1.5 I need to get 1 and 5
14:56	amalloy	kzar: uhm. what result do you want from 1.49?
14:56	amalloy	(my point here is that 1.5 is the same as 1.50000000000)
14:56	kzar	amalloy: Hmm I'm not sure, thing is I need to handle a timezone in decimal. The function I use expects two parameters, hours offset and minutes offset
14:57	mdeboard	kzar: http://clojure.github.com/clojure/clojure.core-api.html#clojure.core/mod
14:57	mdeboard	oh.
14:59	kzar	Oh hang on, how does this sound (int (* 60 (rem offset 1)))
14:59	TimMc	~XY
14:59	clojurebot	xy is http://mywiki.wooledge.org/XyProblem
14:59	TimMc	hah
15:00	TimMc	kzar: What are you trying to accomplish?
15:00	jodaro	shot in the dark here, but is anyone using zeromq and clojure?
15:00	kzar	TimMc: Pretty sure I solved it, just tried it out
15:00	amalloy	~anyone
15:00	clojurebot	Just a heads up, you're more likely to get some help if you ask the question you really want the answer to, instead of "does anyone ..."
15:00	technomancy	jodaro: IIRC it's doable but annoying since you have to use some native libs that aren't packaged up nicely
15:01	jodaro	yeah
15:01	jodaro	i'm through all of that pain, fortunately
15:01	jodaro	but now i've got new pain
15:01	Raynes	amalloy: Don't be mean. jodaro is mah buddy. <3
15:01	kzar	TimMc: Say 1.5 hours is 1 hour and 30 minutes, so if I do (int 1.5) I get 1 hour and (int (* 60 (rem 1.5 1))) gives me minutes
15:01	TimMc	kzar: Right, but what is this in service of?
15:02	endform	Can anyone tell me the differences between anonymous functions defined with #() and (fn [] ())?
15:02	kzar	I explained above, that's what a function I use needs
15:02	jodaro	sorry, anyone currently active and willing to share
15:02	amalloy	endform: those are the same
15:02	TimMc	kzar: OK. I won't try to dig any deeper.
15:02	amalloy	&'#()
15:02	lazybot	⇒ (fn* [] ())
15:03	endform	amalloy: I have heard that before, but I get different behavior from them, also when I run `(#()) and `((fn [] ())) the output differs
15:03	amalloy	&(#())
15:03	lazybot	⇒ ()
15:03	hiredman	() is the empty list
15:03	amalloy	&((fn [] ()))
15:03	lazybot	⇒ ()
15:03	kzar	TimMc: Javascript offset is given as decimal clj-time offset functions need minutes and hours
15:03	endform	#((Thread/sleep 1000) (+ 1 %)) returns a null pointer exception for me, but the same function written with fn, does not
15:04	amalloy	endform: that's not the same function
15:05	amalloy	&'#((Thread/sleep 1000) (+ 1 %))
15:05	lazybot	⇒ (fn* [p1__8621#] ((Thread/sleep 1000) (+ 1 p1__8621#)))
15:05	TimMc	endform: You're trying to call the result of Thread/sleep
15:05	TimMc	endform: Count your parens.
15:06	endform	TimMc: right, okay,
15:06	endform	TimMc: thanks!
15:06	TimMc	Throw a do in there.
15:06	endform	they differ in that the do is implied then in using the fn form
15:07	TimMc	Sort of...
15:07	endform	i'm not sure if that is the right way of saying that
15:07	TimMc	It's more that a fn accepts multiple forms (and wraps them in a do.)
15:07	amalloy	endform: the () is implied by the #() form
15:07	amalloy	is how i would describe it
15:07	hugod	jodaro: storm uses zeromq - not sure whether from clojure or from java
15:07	TimMc	&'#(fn [x] x)
15:07	lazybot	⇒ (fn* [] (fn [x] x))
15:08	TimMc	&'#(fn [x] x x)
15:08	lazybot	⇒ (fn* [] (fn [x] x x))
15:08	drewr	hugod: from clojure (zilch lib)
15:08	TimMc	hmmph
15:08	licenser	Hi, I have some odd problems with swank/slime lately they refuse to connect and jack-in only gives strange errors like this: http://pastie.org/2805988
15:08	TimMc	Urgh, I see what I was doing.
15:08	TimMc	&'(fn [x] x x)
15:08	lazybot	⇒ (fn [x] x x)
15:09	TimMc	OK, right. Quote only reveals reader expansion.
15:09	jodaro	hugod: thanks, looked there (and nathanmarz gave me some tips yesterday)
15:11	tolstoy	licenser: Maybe try 'lein plugin install swank-clojure 1.3.3' on the command line?
15:11	endform	TimMc: do you know where #() is documented? I cannot find it for the life of me on clojure.org
15:12	licenser	tolstoy: so I usually use cake :P that does not make things better I know
15:12	licenser	oh at least a longer struck trace :D
15:13	gregh	endform: http://clojure.org/reader under Macro Characters
15:13	jodaro	endform: http://clojure.org/reader#The Reader--Reader forms
15:13	jodaro	woops
15:13	jodaro	repost
15:13	endform	radical
15:13	devn_	cowabunga dude.
15:14	licenser	tolstoy: thanks that really got me a step ahead :)
15:14	tolstoy	licenser: The other day I had the problem of starting Emacs via the OSX Doc, and it couldn't find lein in /usr/local, so had to add some stuff to my .emacs file. It all seems simple after technomancy helps you. ;)
15:15	devn	Anyone going to the conj smart enough to reserve a hotel for the 9th want to split the cost?
15:15	licenser	teehee
15:15	devn	I just reviewed my reservation and for some reason I thought it'd be a great idea to get a room for the 10th and 11th.
15:15	devn	facepalm
15:17	licenser	java.lang.NoSuchMethodError: clojure.lang.RestFn.<init>(I)V (internal.clj:1) isn't exactly good right?
15:17	callen	devn: I dunno man, that kind of stupid might be infectious.
15:17	callen	devn: what if I catch something? :\
15:18	amalloy	licenser: mismatched binary versions usually
15:18	licenser	oi mismatched binary version? How does that happen o.O?
15:19	amalloy	compile against 1.2 and run against 1.3?
15:19	licenser	ahha sneakuy
15:23	devn	callen: haha
15:24	devn	callen: if the worst kind of stupid I can pass on is sucky calendaring I think I'm in pretty good shape
15:25	gfredericks	okay, I'm sitting down and trying to tackle this 'unbound fn' problem now
15:27	licenser	I start to think clojure hates me because I betrayed it with erlang :(*
15:27	gfredericks	erlang: finally a syntax everyone can agree on.
15:27	Bronsa	is it normal for lein uberjar to take likely a minute to complete?
15:28	Bronsa	*like
15:28	gfredericks	Bronsa: sounds typical to me. Something about cramming all the jars together.
15:28	Bronsa	ok
15:28	gfredericks	lein jar is usually much faster
15:28	Bronsa	let's try
15:28	Bronsa	indeed you're right
15:29	gfredericks	I've just learned to not want to make uberjars every 5 minutes
15:30	gfredericks	maven usually has better ways to do things
15:30	amalloy	gfredericks, Bronsa: uberjar AOT-compiles all your dependencies
15:30	Bronsa	woah,.
15:31	Bronsa	that explains a lot
15:31	licenser	oh it really does, lein swank crashes, cake swank hangs when emacs connects
15:31	amalloy	because your main ns needs to be aot-compiled, which means its dependencies do too
15:32	gfredericks	amalloy: now I know!
15:33	zerokarmaleft	gfredericks: wrt erlang => http://i.imgur.com/eXSQf.png
15:34	gfredericks	zerokarmaleft: :) classic
15:34	jodaro	heh
15:35	licenser	aha found the crulpit
15:36	licenser	clojure-contrib is version 1.1.0 and clojrue 1.2.0
15:38	Raynes	licenser: If you're using cake from the develop branch, it also supports clojure-jack-in now.
15:38	devn	Anyone here play with protoge?
15:38	devn	protege*
15:39	gfredericks	devn: yes.
15:40	gfredericks	s/~anyone~/~anyone"
15:40	licenser	Raynes: that is good to know :)
15:40	gfredericks	(dvorak-type)
15:40	gfredericks	s/type/typo
15:45	cemerick	mdeboard: did you find what you were looking for?
15:46	mdeboard	cemerick: I was just trying to remember what the project was (of yours?) that was on HN a couple/few days ago. Not the SQL DSL
15:47	cemerick	mdeboard: all my public open source stuff is at https://github.com/cemerick FWIW
15:47	mdeboard	Maybe I'm misremembering
15:49	licenser	is there a way to do a soft dependence with maven? like saying "If you have clojure contrib, please take 1.2.0"?
15:54	cemerick	licenser: that's an optional dependency
15:55	licenser	cemerick: then clojure should have clojure contrib in the right version as a optional dependencie
15:56	cemerick	hrm, that's a fair point
15:56	hiredman	meh
15:56	licenser	yup
15:56	hiredman	clojure doesn't depend on contrib in anyway
15:57	cemerick	it's water under the bridge at this point anyway
15:57	licenser	hiredman: but it crashes if something else does :P
15:57	cemerick	hiredman: but there's nothing wrong with libs limiting minimum versions of known downstream dependents
15:57	cemerick	licenser: there is no "clojure-contrib" anymore, anyway
15:58	hiredman	licenser: right, if you do the wrong thing, it crashes
15:58	hiredman	so just stop doing that
15:58	licenser	hiredman: not me, someone else did
15:58	hiredman	whoever
15:59	licenser	what happened is that one dependencey required clojure-contrib and meven/lein figured to get 1.1.0 while the project uses 1.2.0 of clojure
15:59	hiredman	I understand
15:59	licenser	so if there are optional dependencies and it is known that clojure 1.2.0 does not work with contrib < 1.2.0 why not put it in
15:59	licenser	they are from the same 'distributor' in the end
16:00	TimMc	I don't see how it could hurt.
16:00	cemerick	it doesn't, but it also isn't relevant anymore
16:00	licenser	I mean I understand that you don't put a random library like that
16:00	licenser	cemerick: yes it is a lot of libraries require old versions
16:00	TimMc	cemerick: Wishful thinking.
16:01	licenser	it isn't if you have SOTA clojure and all your libs do the same ;)
16:01	cemerick	Well, we can't go re-release clojure 1.2.0 with different dependency info.
16:01	TimMc	Why not?
16:01	licenser	of cause but perhaps 1.2.1 ^^
16:01	TimMc	Oh, I see -- not that exact version.
16:01	licenser	or 1.3.1 then
16:01	hiredman	SOTA?
16:01	cemerick	1.3.x has nothing to do with clojure-contrib at all
16:01	licenser	State Of The Art :)
16:03	TimMc	cemerick: ...and everything that uses 1.2 probably explicitly asks for 1.2.0 or 1.2.1, instead of the awkward [1.2.0,1.3.0)
16:03	licenser	cemerick: still it cranes if a old version of c.c is present
16:04	licenser	well the point of it is c.c 1.1.0 seems to be incompatible with 1.2.0+ since a lot of old libs require 1.1.0 it would be nice to have clojure be aware of this and block it :)
16:04	cemerick	licenser: except that would require changing your clojure version to, what, 1.2.2? Might as well require c.c [1.2.0] and be done with it?
16:05	TimMc	technomancy: Can we have leiningen turn 1.2.* into [1.2.0,1.3.0) in deps? :-)
16:05	cemerick	:-O
16:05	licenser	cemerick: for me yes I did that already and am fine I just expect that I am not the only one that runs into this problem and learns the hard way what the cryptic error message java throws on you means
16:05	TimMc	I'm guessing the latter syntax is a Maven-ism.
16:06	gfredericks	are there any common gotchas with maven and SNAPSHOTs?
16:06	licenser	also c.c 1.2 does the same with clojure 1.3
16:06	licenser	as in crash
16:06	hiredman	snapdon't
16:06	cemerick	TimMc: also a lein-ism as long as no one attempts to reinvent dependency resolution and repositories
16:06	licenser	so I think binding the contrib version within clojure is a bad idea
16:07	licenser	at least I don't see any reason why it would do any harm
16:07	gfredericks	hiredman: so when you work on several maven projects simultaneously, you just version-bump every time?
16:07	licenser	I mean about 10 second of work on the pom of clojure, saves generations of clojuruans the trouble to learn that including libraries means better double check that they don't depend on a old version of contrib
16:07	hiredman	gfredericks: I think I would just cry a lot
16:08	TimMc	hiredman: That's reassuring.
16:08	hiredman	move to lein, use checkout dependencies
16:08	cemerick	bleh
16:08	cemerick	only helpful if you own the world
16:08	gfredericks	I'm not sure that'd go very well with this java project
16:08	licenser	hiredman: I use lein, it downloaded the wrong c.c happily and crashed in my face with 200 lines of java excpetions
16:08	gfredericks	which depends on my clojure project
16:08	cemerick	gfredericks: what problem are you having in particular?
16:09	hiredman	licenser: wah wah wah
16:09	TimMc	licenser: He was responding to gfredericks re: lein
16:09	hiredman	cemerick: how so?
16:09	gfredericks	cemerick: I keep doing "lein deploy" with my snapshotted project, but my java project keeps building with the same old jar
16:09	licenser	hiredman: You failed to produce any argument just play nay-sayer, mind to explain why?
16:09	gfredericks	cemerick: I can even see in my ~/.m2 that the new jars are comming in with their timestamped names, but the generic jar is still the old version
16:09	licenser	TimMc: ah sneaky :) sorry hiredman
16:10	cemerick	gfredericks: is the Java project using maven?
16:10	gfredericks	cemerick: yeah
16:10	cemerick	It'd try running `mvn -U …` (which will update all snapshot versions).
16:10	gfredericks	cemerick: it's building a war, so I can easily crack it open to see what it included
16:10	gfredericks	cemerick: will try that now
16:10	cemerick	if that doesn't work, nuke the relevant subdirectory in ~/.m2/repository/whatever, and try again
16:11	cemerick	hiredman: checkout dependencies require…a checkout
16:11	hiredman	licenser: because clojure doesn't depend on contrib? so why would marking it as a dependency make any kind of sense? it is semantically incorrect, you are trying to hack around a problem without solving it by smashing a square peg into a round hole
16:11	hiredman	cemerick: right, which if you are hacking on several projects at once you have
16:11	cemerick	if I want to use head of some library I'm not working on, then I need to go clone its repo. Not fun.
16:11	gfredericks	cemerick: yeah, that's definitely my backup plan. I'd like it to work properly though, not hackily.
16:12	licenser	hiredman: I see that point that is why I suggested a (how I now learned it is called) optional dependencie
16:12	cemerick	hiredman: right, if I'm hacking on the projects in question. Doesn't help if someone else is hacking, and I want to benefit from the hacking without futzing with having their repo.
16:12	hiredman	cemerick: right, but you can then commit and push your changes
16:13	cemerick	gfredericks: it is unfortunately a fact of life that .m2/repository needs flushing every now and then. That's gotten quite a bit better now that aether is maturing, FWIW.
16:13	hiredman	licenser: "dependency" clojure does not depend on clojure-contrib, so even if the result is the effect you want, a solution that ends up marking it as so is incorrect
16:14	cemerick	hiredman: commit to what? If I'm following foo's HEAD, and I'm not involved in the development of foo, then using snapshots I don't have to have a clone of foo (and update my clone of foo, and bar, and baz, whenever they have changes).
16:14	gfredericks	cemerick: the -U seems to have made a difference. I'll add that in, thanks.
16:15	cemerick	checkout deps are a great solution for a single developer, but hardly a general solution
16:15	hiredman	cemerick: hacking on several projects at once implies making changes to several at once
16:15	licenser	hiredman: sorry that is just words that you argue, having named it 'things-that-are-relavant-for-this' just would have not sounded so good as dependencies
16:15	cemerick	hiredman: that's not the only use case of snapshots
16:15	hiredman	cemerick: correct, but is the case gfredericks asked about
16:15	hugod	gfredericks: zi can do checkout dependencies for maven projects, but unfortunately relies on fixes in maven post 3.0.3
16:15	cemerick	gfredericks: should only be necessary when you see things not updating properly; using it all the time will not do friendly things to your build time :-)
16:16	gfredericks	cemerick: :/ I'd hate to have to always check
16:16	gfredericks	my goal here is to automate everything
16:16	gfredericks	I prefer slow to inconsistent
16:17	licenser	anyway since it won't happy anyway I just give it up
16:19	srid	can (fn [a [o b]] (o a b)) be shortened any further?
16:20	mdeboard	.
16:21	chouser	srid: in some contexts you could almost replace that with ->
16:21	srid	a and b are numbers, and o is one of the math operators
16:22	srid	this is http://www.4clojure.com/problem/135 btw
16:22	hiredman	bleh
16:22	chouser	Is this new?
16:22	srid	yea
16:22	hiredman	isn't there a #clojure-golf channel or something?
16:22	chouser	heh
16:23	TimMc	it's called #4clojure
16:24	amalloy	~rimshot
16:24	TimMc	SANBOX DENIED
16:25	TimMc	licenser: I suppose it should really be that clojure-contrib 1.2 declares a dependency on clojure 1.2.
16:25	TimMc	What was I thinking...
16:26	licenser	TimMc: so that would mean that when releasing c.c 1.2 you already know that 1.3 will be incompatible but yes that would be the better way
16:27	cemerick	If only we had a time machine?
16:27	hiredman	I do
16:27	hiredman	but I get messages about corrupt backups at least once a month and have to start over, drives me nuts
16:28	licenser	on a bigger scale libraries requiering outdated things can be pretty scary especially if they are not explicit about it what would be the best way to handle this?
16:28	TimMc	cemerick: When you were saying something about "lein-ism", were you saying that that syntax is native to leiningen, or that notion of version ranges is?
16:28	hiredman	avoid those libraries
16:28	TimMc	ugh
16:28	cemerick	TimMc: lein inherits nearly all of the dependency resolution semantics of maven (necessarily so), including version ranges.
16:28	brehaut	hiredman: my time machine did the same for about 3 months, and then the disk died completely
16:29	hiredman	:(
16:29	licenser	that is a rather harsh way but yes it would solve the problem
16:29	TimMc	cemerick: Semantics, OK. But not syntax, surely!
16:29	cemerick	TimMc: insofar as version numbers are strings, yes
16:29	TimMc	Hrm.
16:29	cemerick	The syntax is not really problematic. The semantics are more questionable.
16:30	cemerick	[5,12) isn't exactly unprecedented, etc
16:30	srid	,(-> 3 (- 1))
16:30	srid	,(#(-> % %2) 3 '(- 1)) ;; fail
16:30	TimMc	cemerick: I was sure that Maven would represent it as <version><lower include="true">1.2.0</lower><upper include="false">1.3.0</upper></version>
16:30	cemerick	heh
16:30	TimMc	but I gues even the XMLers have a limit.
16:31	hiredman	licenser: you are saying "I want to use library x, but it's dependecies are outdated" and I am saying "well then you cannot use it" and you are responding with "but I really want to"
16:31	licenser	hiredman: nah I say is there a way to make it better
16:31	brehaut	TimMc: youd think so would you…
16:31	srid	chouser: how would you wrap -> that in anonymous function?
16:32	licenser	hiredman: I mean it might be that there is a super simple way I just don't know about :P
16:32	TimMc	srid: -> is a macro, it is a syntax transformer
16:32	TimMc	he comes, etc.
16:33	licenser	but since none went "oh that again, just do lein show-me-where-the-outdated-contrib-comes-from I guess something else is wrong
16:33	hiredman	the only real correct answer is the library needs to be brought up to date in it's dependencies, or you need to log yourself to its dependencies
16:33	hiredman	either of which can have large repercussions
16:34	TimMc	,(println "You OK, clojurebot?")
16:35	licenser	I really want the result to be "No."
16:35	chouser	srid: yeah, I don't see how.
16:35	srid	in general, ` (fn [a [o b]] (o a b)) ` can't be shortened using anonymous functions?
16:36	TimMc	srid: Which 4clojure problem?
16:36	srid	135
16:36	licenser	hiredman: okay perhaps my approach was wrong. How about making it easier to find those problems or prevent them from crashing in a strange java stack trace that no normal person understand (save for those who had the problem themselves or wrote clojure ;)
16:36	TimMc	licenser: The exceptions are a problem for sure.
16:37	TimMc	Or maybe there should be a lein troubleshoot plugin!
16:37	hiredman	licenser: a. don't aot unless you are absolutely required b. educate your self, java stack traces are very informative, learn to read them
16:37	TimMc	"It looks like you have incompatible versions of those libs."
16:37	TimMc	hiredman: Nonsense, there's some crazy shit in those traces, you're just used to it.
16:37	hiredman	b. is really a must for anything
16:38	licenser	even so I figured out how to get me a dependence tree I don't see where contrib 1.1 comes from
16:38	TimMc	They may have gotten better, but they're still mystifying on many occasions. Especially certain syntax errors.
16:38	hiredman	languages are just another abstraction, they can an do break, and you depend on them, so treating them as a black box is dumb
16:38	amalloy	chouser: hah, what a funny way to solve that problem
16:39	srid	? you stalking people's solutions?
16:39	hiredman	licenser: I don't recall the intermediate debugging steps you went through, but the first cause you mentioned does not indicate that something is pulling in contrib 1.1
16:39	dakrone	b
16:39	amalloy	srid: http://www.4clojure.com/problem/solutions/135
16:39	srid	hmm. not looking, until i compress my solution
16:39	hiredman	it means that some library was aot compiled against some other version of clojure, that was not abi compatible
16:40	arohner	licenser: lein pom && mvn dependency:tree \| less
16:40	licenser	hiredman: I know but someone here explained that, the stack trace: http://pastie.org/2806452 indicates that I use clojure 1.3 with something that is compiled for clojure 1.2
16:40	licenser	which was pretty close, I use clojure 1.2 for something - that for some reason is compiled with clojure 1.1 namely clojure-contrib 1.1.0
16:41	amalloy	srid: want a hint?
16:41	licenser	now I don't require contrib at all but something does so I went to put 1.2.0 in fixed - solved the crash
16:42	licenser	now I did more or les what arohner said to trace down which library is the one that requires 1.1.0 - answer is, none - my core project does
16:42	arohner	is there a function for 'or'ing a bunch of predicates together? #(or (foo? %s) (bar? %s)) -> (magic-fn foo? bar?)
16:43	hiredman	some
16:43	licenser	http://pastie.org/2806467 would be the dependence tree
16:44	amalloy	clojure 1.3 adds every-pred or something like that
16:45	ThreeCups	I'm trying to write a function that take a 2-d matrix of maps and updates a particular map. The function takes a matrix, coords (two ints in a vector), a fn that takes a map and other args, and the other args that are passed to fn. The definition is like this:
16:45	ThreeCups	(defn update-cell [matrix coords f & args] (update-in matrix coords f args))
16:45	amalloy	&(#(list % %&) 1 2 3 4) ; sric
16:45	lazybot	⇒ (1 (2 3 4))
16:45	ThreeCups	But I'm having trouble with args and getting them through to the fn call. I get this error: java.lang.IllegalArgumentException: Wrong number of args (2) passed to: user$eval137$fn (NO_SOURCE_FILE:0).
16:45	ThreeCups	If I take out args, then everything works fine, but I want to be able to pass in optional args. Any tips?
16:45	amalloy	ThreeCups: (apply update-in ...)
16:45	TimMc	chouser: Nice! (looking at solution)
16:46	brehaut	i was going to be productive today, and now you all have me trying to solve 4clj problems.
16:46	licenser	aha found the bugger
16:46	ThreeCups	amalloy: Thanks! it worked great
16:49	srid	amalloy: btw, is that (% %&) that hint? i didn't know of this implicit destructuring before. (don't need hint yet)
16:49	amalloy	srid: yes, that was all
16:49	chouser	TimMc: heh, thanks.
16:49	srid	oh ok. and recur is allowed in anonmy functions?
16:49	amalloy	srid: try it and see?
16:50	cgray	in clojurescript, it appears that you can't have a function with the same name as the namespace. is this a bug?
16:50	srid	&#(if % (recur (dec %)) %) 5)
16:50	lazybot	⇒ #<sandbox6471$eval8674$fn__8675 sandbox6471$eval8674$fn__8675@17019f7>
16:50	srid	that hung my repl; restartring ..
16:50	ibdknox	cgray: it is an issue that has come up before, but I thought someone fixed it
16:51	srid	,(#(if (pos? %) (recur (dec %)) %) 5)
16:51	cgray	ibdknox: ok
16:51	amalloy	srid: clojurebot is down for the count atm, i think
16:51	devn	:(
16:51	devn	I can play clojurebot...
16:51	TimMc	haha
16:51	TimMc	devn: ~botsmack
16:52	TimMc	slowbot
16:52	devn	=> 0
16:52	devn	srid: ^
16:52	TimMc	srid: ##(#(if (pos? %) (recur (dec %)) %) 5)
16:52	lazybot	⇒ 0
16:53	srid	devn: ,(loop [x 1] (recur x)
16:53	TimMc	lazybot is alive
16:54	TimMc	amalloy: Did you change something about lazybot's eval responses? I get a "?" now instead of an arrow.
16:54	ibdknox	TimMc: utf?
16:55	hiredman	amalloy: did someone do something that locked up clojurebot? down for the coutn how?
16:55	cgray	ibdknox: it looks like the issue is still there... i just updated to the latest github version
16:55	TimMc	ibdknox: Shouldn't have changed since like last week, at least not on this computer.
16:55	Raynes	TimMc: Nothing has changed.
16:55	Raynes	&(+ 3 3)
16:55	lazybot	⇒ 6
16:56	TimMc	hrmf
16:57	Raynes	,(+ 3 3)
16:57	clojurebot	6
16:57	Raynes	~botsnack
16:57	clojurebot	Thanks! Can I have chocolate next time
16:57	Raynes	Sure.
16:58	TimMc	The last sign of life had been 14:57 < amalloy> ~anyone
16:58	devn	srid: Still working on your query...
16:58	srid	buggy bot. that's a syntax error!
16:58	TimMc	srid: Not for lazybot.
16:59	TimMc	&(+ 1 2
16:59	lazybot	java.lang.RuntimeException: EOF while reading, starting at line 1
16:59	TimMc	o\__/o
16:59	Raynes	TimMc: We got rid of that.
16:59	TimMc	OK.
16:59	TimMc	Why?
16:59	clojurebot	TimMc: because you can't handle the truth!
16:59	Raynes	It used fnparse which is a bit of an abandoned project and we needed to update to 1.3.
16:59	Raynes	Or, I wanted to update to 1.3 in any case.
17:00	TimMc	Aha.
17:00	Raynes	But I absolutely want it back if someone wants to do it in some other way.
17:00	TimMc	Are there any cljs eval bots around?
17:00	TimMc	Raynes: (while EOF (add-parens))
17:00	Raynes	http://github.com/flatland/lazybot <-- src/lazybot/plugins/clojure.clj if anybody wants to give it a shot.
17:01	Raynes	lazybot can do Haskell, Clojure, and JRuby.
17:01	Raynes	No cljs.
17:01	amalloy	TimMc: well, it used to handle (let [x (inc y] x)
17:01	srid	(eval (-> a b)) --- is it possible to resolve `a` and `b` before the the form is passed to `eval`?
17:01	amalloy	which your strategy is not so good on
17:01	Raynes	And the jruby isn't allowed at the moment because I haven't bothered adding JVM sandboxing (which I can do fairly easily with clojail).
17:01	Raynes	But Haskell certainly works.
17:02	Raynes	$he 2 + 2
17:02	lazybot	⇒ 4
17:02	TimMc	srid: For-kludge-er won't take eval.
17:02	srid	hmm
17:04	TimMc	unless you're clever about it
17:04	TimMc	Not golf-friendly-though.
17:05	TimMc	s/y-/y /
17:07	devn	bahaha, srid: You are a jerk. I forgot about that REPL and my computer just started to get awfully loud.
17:10	cemerick	FYI: http://mostlylazy.wordpress.com/2011/11/03/coming-soon/
17:11	licenser	Raynes: :D long time no see
17:11	Raynes	Indeed.
17:11	Raynes	How have you been?
17:12	licenser	good good :) got a new server, lots of fun and experience in setting stuff up, a actual work related clojure project which is darn awesome to have and my gf is spending a month with me :D
17:12	licenser	how about you?
17:12	TimMc	whelp
17:12	Raynes	I'm fantastic. :)
17:12	TimMc	devn: It turns out that (loop [] (future (loop [] (recur))) (recur)) is a terrible idea.
17:13	licenser	glad to hear ^^
17:13	devn	haha
17:13	TimMc	I'm on another computer now...
17:13	devn	licenser: yeah, where have you been?
17:13	licenser	I read your blog and noticed that you moved try-clojure to heroku, should I changed to try-clojure.org domain?
17:13	Raynes	Naw, I've got a redirect up.
17:13	licenser	devn: hiding under a pile of Hardware related work and some erlang coding
17:13	Raynes	It's all good.
17:14	licenser	:)
17:14	Raynes	The primary domain is tryclj.com now anyway.
17:14	licenser	kk
17:14	TimMc	My computer is now a space heater.
17:14	licenser	sweet TimMc :)
17:15	Raynes	TimMc: I take of my pants, fold my legs, and sit with my laptop in my lap to keep warm in the winter.
17:15	Raynes	off*
17:15	TimMc	Raynes: And run a linear fork bomb?
17:15	Raynes	Right, yes.
17:15	licenser	heh
17:15	TimMc	'cause that's what I just did.
17:15	TimMc	And I have 8GB of RAM, so fuck me. Stupid Java process won't die for a while.
17:16	licenser	TimMc: neat, I tried that once to see if OSX fixed the "bug" in bash :P
17:16	licenser	for the record, they didn't
17:16	devn	licenser: cool, I just started working on the Manning Erlang and OTP book
17:16	licenser	yea that got me hooked, it is really good
17:16	licenser	if you ever are interested to look at some code, I've a darn interesting project in erlang
17:16	devn	the guy was like "that is not a beginner book, but it is a great book"
17:17	devn	so i immediately purchased it. beginner books are a bummer.
17:17	devn	90 pages of what an array is.
17:17	devn	licenser: I would be delighted to. :)
17:17	devn	brb
17:17	licenser	cool cool
17:20	ibdknox	technomancy: ping?
17:21	technomancy	ibdknox: what are the haps
17:22	ibdknox	technomancy: would you be interested in version of lein new that used a lib designed to spit out project structures?
17:22	technomancy	ibdknox: yeah, there's a "lein spawn" plugin that does that
17:23	callen	anybody going to the clojure meetup in SF tonight?
17:23	technomancy	I'd be interested in mainlining it or (something like it; I haven't looked at it much) for 2.0
17:23	technomancy	oops, paredit slurp that or in
17:23	ibdknox	technomancy: excellent. There's a good chance Raynes or I might build such a lib.
17:24	technomancy	have you evaluated spawn?
17:24	ibdknox	if that one doesn't look good
17:24	ibdknox	was supposed to be on that sentence
17:24	Raynes	I'm looking at it now.
17:25	Raynes	I think we'll have to put some work into taking the Leiningen stuff out.
17:25	Raynes	I need this to work for both cake and leiningen.
17:25	Raynes	But it looks good so far.
17:25	cgray	is there an equivalent to the int function in cljs?
17:26	ibdknox	cgray: (js/parseInt)
17:26	licenser	Raynes: out of curiosity how does the cake dev version handle jack-in?
17:26	Raynes	licenser: The same way leiningen does, for the most part.
17:28	licenser	what I mean is if I do clojure-jack-in how does it decide weather to use cake or lein?
17:35	callen	whether
17:36	Raynes	licenser: Ah, you have to set the command it uses.
17:36	Raynes	(setq clojure-swank-command "cake jack-in %s")
17:36	Raynes	That ought to do it.
17:38	technomancy	any seattle folk are welcome to swing by early =)
17:38	licenser	Raynes: sneaky :) thanks.
17:39	Raynes	That isn't documented yet because the next version of clojure-mode will define clojure-swank-command as a defcustom so that it can be edited using the Emacs customization interface. Also, cake jack-in isn't in a release yet, so I didn't want to confuse people.
17:40	replaca	callen: I'll be there. You?
17:40	licenser	Raynes: I see I see :)
17:40	callen	replaca: it was between clojure and django, I'm going to the clojure one instead. so yeah, I'll be there. I work nearby weatherbill.
17:41	licenser	also reading your blog I found it curiose that you dived into the topic of media servers as well :)
17:41	technomancy	is weatherbill hosting?
17:41	callen	da.
17:41	technomancy	cool
17:41	ThreeCups	Is there a way to get around reduce's need for it's fn argument to take 2 args? I'd like to pass in more than 2.
17:41	technomancy	one of their Seattle employees is a regular up here
17:42	callen	they seem to be prolific in general.
17:42	cgray	ThreeCups: just put them in a vector and destructure it in your fn
17:43	ThreeCups	cgray: Thanks, was hoping for a different solution, but that will work. I don't really like it because the optional arg is required.
17:44	ThreeCups	That means I have to write my function like (defn [a b & c] ...) and c is required by the fn
17:44	cgray	ThreeCups: no, i meant something like (defn foo [[a b] c] ...)
17:45	cgray	and it should return a vector of two values
17:45	replaca	callen: cool. I work close too, so it's a great location.
17:45	ThreeCups	cgray: okay, that makes sense. let me give it a try...
17:46	replaca	callen: see you there
17:46	callen	replaca: indeed. My name's Chris btw.
17:46	replaca	callen: And I'm Tom
17:46	ThreeCups	cgray: I don't understand the the "and it should return a vector of two values"
17:47	ThreeCups	I'll have a look at the reduce source and that will prolly answer my confustion
17:47	cgray	ThreeCups: reduce calls the fn repeatedly with the last thing you return and the next thing in the list...
17:50	technomancy	so, any ideas why I can access a postgres db just fine on the command line but not via jdbc? the DB is owned by a postgres user with the same name as my unix user, which means I shouldn't need a password, but c.j.jdbc freaks out claiming the server requires a password.
17:50	technomancy	(side note: this is why people are nuts about nosql)
17:50	clojurebot	'Sea, mhuise.
17:50	technomancy	~botsmack
17:50	clojurebot	clojurebot evades successfully!
17:50	technomancy	...
17:50	technomancy	clojurebot: forget (side note: this \|is\| why people are nuts about nosql)
17:50	clojurebot	I forgot that (side note: this is why people are nuts about nosql)
17:51	hiredman	postgres somtimes decides to be whacky and do auth via identd
17:52	gfredericks	okay, I have now ceased getting unbound fn errors and am now getting an error about one of the generated fn classes not being found
17:52	gfredericks	the exception relates to a class ending in "$iter__94__98$fn__99", while when I examine the jar the relevant class has slightly different numbers
17:52	hiredman	sounds like you are mixing aot and classloader wonkyness
17:53	gfredericks	I bet so :/
17:53	hiredman	oh, no
17:53	technomancy	oh ffs
17:53	technomancy	ident sameuser only works using unix domain sockets
17:53	hiredman	thats not it, you should clean out your classfiles
17:53	technomancy	that's it. mongodb here I come.
17:53	hiredman	technomancy: well, it makes sense, other wise how can you tell it is the same user?
17:55	technomancy	hiredman: sure, but the point is that this is a waste of time.
17:55	amalloy	ThreeCups: what would it mean for reduce to take three args? what args would it pass?
17:56	technomancy	setting up a postgres DB that you can actually access takes in excess of seventeen steps if you count all the times you have to create the DB, realize you've done something subtly wrong, dropping it, and recreating it.
17:56	hiredman	technomancy: you could just allow anyone to use the db without a password, just like most nosql dbs
17:58	gfredericks	hiredman: does "clean out your classfiles" imply anything more than `lein clean`? :/
17:58	seancorfield	i must admit, on the (two) occasions i've setup postgres, i've found it to be crazy fussy to install / configure / use and i can't understand why anyone raves over it :(
17:59	hiredman	gfredericks: rm -rf classes
17:59	ThreeCups	amalloy: it would take an accumulator, the next value to add to the accumulator, and any other args needed to calculate the addition of value to the accumulator
17:59	duck1123	There used to be a library that had a fn for shelling out that I can't seem to find anymore. Does anyone know of any good ones?
17:59	hiredman	and make sure the jar you end up using was generated after you did the rm -rf
17:59	duck1123	I've got pallet.stevedore, but that only seems to generate the scripts
18:00	technomancy	seancorfield: "GRANT ALL ON db.* TO $USER" is 90% of the reason mysql is widely used
18:00	hugod	duck1123: clojure.java.shell
18:01	amalloy	ThreeCups: but how would that work? either the "other args" are part of the next value (and you need only two args), or they're part of the accumulator (same thing), or they're the same every time. if they're the same every time, you can just make them part of the reducer function
18:01	duck1123	hugod: Well there it is. thanks
18:02	seancorfield	technomancy: and the fact that for dev work you don't need to create an o/s level user for mysql etc etc etc
18:02	gfredericks	hiredman: did everything over quite carefully. It looks like the jar for my project contains in itself a reference to the class, even though the class it contains has the slightly different numbers
18:03	gfredericks	(i.e., when I `grep <classname> <jarfile>`, grep says it matches)
18:04	gfredericks	I'm using a macro that emits a deftype...would this make sense if it was somehow running twice during compilation?
18:04	hiredman	possibly
18:05	hiredman	the differing numebrs tends to mean something got compiled twice
18:05	hiredman	also, if you are emitting a deftype inside a let or something
18:05	gfredericks	they all differ by one, which is odd, as it doesn't make sense if there's a global incremented variable for the unique IDs
18:07	gfredericks	the deftype is at the top level. The weirdest thing going on is that I'm doing (require ...) a couple times in the body of the only function of my deftype, just to make sure the ns's are loaded.
18:07	hiredman	ugh, yeah, no
18:07	hiredman	don't do that
18:07	ThreeCups	amalloy: I think I understand what you're saying. I'm thinking about it a little differently. I'm updating a 2-d vector using reduce. I have a sequence of coordinate, value pairs e.g. ([[0 0] :value]...) and I have a fn that I use to update a matrix
18:07	ThreeCups	(defn update-matrix-cell [matrix coords f & args] (apply update-in matrix coords f args))
18:07	gfredericks	okay, then I don't know how to make sure my class works
18:07	ThreeCups	I also have an intermediate fn between reduce and update-matrix-cell that takes 3 args
18:07	ThreeCups	(defn update-matrix-cell-for-specific-data [matrix coords data] ... (update-matrix-cell ...))
18:08	ThreeCups	It's this update-matrix-cell-for-specific-data that I'd like to call from reduce. I've "solved" this in the meantime by doing destructuring.
18:08	gfredericks	hiredman: without that I get unbound fn errors
18:08	hiredman	gfredericks: ok, add it, but use resolve
18:09	hiredman	or, have the java side load the required namespaces
18:09	amalloy	ThreeCups: (reduce (fn [acc x] (update-matrix-cell acc (do-stuff-with-x))) cells) or something?
18:09	gfredericks	hiredman: the latter would be clunky, as I'd rather not edit the java (different project). My macro is emitting symbols for the require statements
18:09	gfredericks	e.g., (require '~(-> ns str symbol))
18:10	hiredman	use ns-name
18:10	hiredman	and use resolve
18:10	gfredericks	so (require (resolve '~(ns-name ns)))?
18:10	hiredman	no
18:11	hiredman	when you go to use something from the required ns, use resolve
18:11	hiredman	(reuire x) ((deref (resolve 'x/y)) 1)
18:11	gfredericks	oh wow
18:11	gfredericks	the only thing I have after the require is ~@body
18:11	gfredericks	so I guess anytime I call the macro I need to resolve everything?
18:13	gfredericks	I had no idea interop from clojure was so hairy
18:14	gfredericks	am I doing something wrong? Is it really not possible to supply a class to a java proj via a jar without adding java code to bootstrap it?
18:16	srid	,((first '(+ x x)) 2 3) ;; strange result
18:16	clojurebot	3
18:16	srid	shouldn't that be 5?
18:16	amalloy	no
18:16	gfredericks	,('+ 2 3)
18:16	clojurebot	3
18:16	amalloy	&(= + '+)
18:16	lazybot	⇒ false
18:16	amalloy	&(:+ 1 20) ; same as this
18:16	lazybot	⇒ 20
18:16	srid	%('whatever 1 2)
18:17	srid	membership lookup? returns 2nd arg as default.
18:17	gfredericks	have symbols always done that?
18:17	amalloy	&('+ '{* 1} 4)
18:17	lazybot	⇒ 4
18:17	srid	,((resolve (first '(+ x x))) 2 3) ;; works n ow
18:17	clojurebot	5
18:18	amalloy	&('+ '{* 1 + 8} 4)
18:18	lazybot	⇒ 8
18:18	amalloy	gfredericks: for as long as anyone can remember, anyway. probably for as long as keywords have done it
18:18	srid	4clojure refuses resolve.
18:19	amalloy	&(resolve (symbol "eval"))
18:19	lazybot	java.lang.SecurityException: You tripped the alarm! resolve is bad!
18:20	amalloy	that's why - with resolve you can get eval, from which you can get anything you want
18:20	srid	problem 121 - close to lowest code golf ...
18:20	brehaut	amalloy: does it really ;)
18:20	gfredericks	,(list (symbol "eval"))
18:20	clojurebot	(eval)
18:24	gfredericks	is require idempotent?
18:24	ThreeCups	amalloy: that will work, and I think it's what I've done but removing the "middle man". I want to think about updating my matrix with the fn
18:24	ThreeCups	(defn update-matrix-cell [matrix coords f & args]
18:24	ThreeCups	This seems to be a pretty clean/clear fn. And it makes sense that I'd have another function that update the matrix for a specific "data type" at a specific coordinate, hence the fn
18:24	ThreeCups	(defn update-matrix-cell-for-specific-data [matrix coords data] ... (update-matrix-cell ...))
18:24	ThreeCups	Now I have a matrix and a sequence of coordinate/data pairs and want to call my handy function to update the matrix for this type of data. But I can't because reduce takes a fn that takes 2 args.
18:24	ThreeCups	I've got around this by creating an intermediate fn
18:24	ThreeCups	(defn intermediate [matrix [coords data]] (update-matrix-cell-for-specific-data matrix coords data)
18:24	ThreeCups	It'd be nice if I could hide this fn somewhere, i.e. make it anonymous or something (I'm too new to clojure so I don't know how to do this)
18:25	brehaut	ThreeCups: let + fn or letfn ?
18:26	ThreeCups	brehaut: will give it a try. I've used it once and forgot about it :)
18:26	technomancy	gfredericks: yeah
18:32	ThreeCups	brehaut: Thanks! worked like a charm
18:32	gfredericks	oh, I was wrong about my jar containing a reference to the non-existing class. apparently I used grep poorly.
18:36	gfredericks	oh man, I think it's a serialized object
18:37	gfredericks	are clojure datatypes not safely serializable in the java-sense? or maybe just lazy seqs?
18:44	gfredericks	I can imagine if java tries to serialize a lazy seq it couldn't do anything other than serialize the internal fn itself
19:02	hiredman	gfredericks: the other thing is java serializing requires the classes to be present at both ends, dones't play well with dynamic byte code generation
19:04	gfredericks	hiredman: are non-lazy datatypes safer? or do they have hidden references to fns?
19:05	hiredman	they are safer, you really should be able to serialize a lazy-seq
19:08	gfredericks	I think I finally got the example-of-a-use-case working, though I'm not sure if serialization will come up later. Will try it out.
19:08	gfredericks	hiredman: thanks for the help.
19:09	hiredman	np
19:42	scottj	"There are several upcoming books on Clojure from multiple publishers. I think there's been more books on Clojure in 2 yrs than Lisp in 20 ;)"
19:42	scottj	no way, no matter how you pick the two years (since programming clojure is already over two years old).
19:43	scottj	let over lambda, land of lisp, practical common lisp, lisp small pieces, ansi common lisp, on lisp, paip, plai, little schemer, seasoned schemer, reasoned schemer. and if you go 22 years or so you get horn and cltl2
19:43	callen	PC is the one I have.
19:43	callen	I can't wait for v2.0
19:43	callen	I want some Noir.
19:44	scottj	of course the tweet suffers from the same flaw as people who complain that the US isn't building as many bridges as China.
19:50	technomancy	don't forget the art of the metaobject protocol
19:51	scottj	oh wow I thought that was older so I didn't bother checking
19:52	technomancy	not really fair to count both CL and Scheme though
19:53	scottj	actually I bet if you just do 89-91 CL beats any three year pear of clojure
19:56	scottj	oopcl, amop, cltl2, paip, horn3
19:59	tolstoy	How long until "Clojure & Spring"?
20:00	tolstoy	Heh. ;)
20:00	TimMc	$kill
20:00	lazybot	KILL IT WITH FIRE!
20:01	tolstoy	Movie: The Spring. A group of hackers go down to Antartica to uncover an old architecture, when...
20:03	ThreeCups	How would I get the x/y coordinate pairs from a 2-d matrix? E.g. (coord-pairs (->> (repeat 2 {}) vec (repeat 2) vec)) would return [[0 0] [0 1] [1 0] [1 1]]
20:05	gfredericks	ThreeCups: I'd get the dimensions and do something like ##(for [x (range 3) y (range 2)] [x y])
20:05	lazybot	⇒ ([0 0] [0 1] [1 0] [1 1] [2 0] [2 1])
20:12	technomancy	dakrone: how do you feel about accepting a string for :basic-auth?
20:12	technomancy	it would make for easier round-tripping via java.net.URI since that class exposes user/password only via .getUserInfo as a single colon-separated string.
20:12	technomancy	(as well as continuing to accept a seq of course)
20:13	technomancy	also: why do I always come up with these questions for dakrone _after_ 6pm mountain time?
20:14	hiredman	:/
20:14	mdeboard	So, question. I was just reading http://blog.factual.com/clojure-on-hadoop-a-new-hope and he has several variables preceded by `?`. Is this some convention or does it have a syntactic meaning I'm missing
20:15	mdeboard	or is it just something he likes doing
20:15	hiredman	it's a cascalog/datalog/logic programming thing/convention
20:16	nathanmarz	it distinguishes variables from regular variables
20:16	hiredman	doesn't have a meaning in clojure outside of those domains
20:16	nathanmarz	errr
20:16	nathanmarz	from regular values
20:16	nathanmarz	e.g., if you did this: (<- [?a ?b] (source ?a) (+ val ?a :> ?b))
20:16	nathanmarz	then "val" is just a normal Clojure variable, and its value is substituted into the query
20:16	nathanmarz	whereas ?a and ?b are part of the query
20:17	mdeboard	oh hey nathan
20:17	mdeboard	marz
20:17	mdeboard	I was just looking at Cascalog
20:17	mdeboard	I se
20:17	mdeboard	e
20:17	mdeboard	I have a need for such functionality at work.
20:18	mdeboard	I've got this colossal "cities.txt" file that is a csv of every city in the world, with its country and lat/long
20:18	mdeboard	1.5 mil lines or something
20:18	mdeboard	cool story bro
20:19	nathanmarz	cool
20:19	nathanmarz	you could certainly query that with cascalog
20:19	gfredericks	the '.txt' makes me imagine opening it in notepad
20:20	mdeboard	gfredericks: Worse. Iterating over every line, creating a Django ORM object out of the line, then saving it to the database.
20:21	mdeboard	gfredericks: It is unbelievably slow.
20:21	hiredman	(didn't go over well)
20:21	gfredericks	mdeboard: I'm not sure what you just said but I'm pretty sure it was offensive
20:21	mdeboard	gfredericks: I threw up.
20:21	hiredman	at least he didn't call you a cfmljure
20:22	gfredericks	:)
20:23	mdeboard	I'm frustrated just thinking about it.
20:26	mdeboard	It's stovepiping part of a bigger project, actually. I wish I hadn't brought itup.
20:27	TimMc	stovepiping!
20:27	mdeboard	indeed
20:27	TimMc	I first heard that phrase in a Charlie Stross book, re: gov't projects.
20:28	TimMc	Or is that even the same term?
20:29	mdeboard	TimMc: I spent a decade+ in the USMC, common metaphor :P I'm sure it was explained in the book but literal stovepiping is when a spent cartridge is incompletely ejected from the chamber, and gets caught between the bolt and ejection port
20:29	mdeboard	So it's just sticking straight up, preventing the weapon from completing its firing cycle
20:29	mdeboard	it being the spent casing
20:29	TimMc	Why does it get that name?
20:29	mdeboard	It looks like a stovepipe
20:29	TimMc	That's like a 4th meaning...
20:30	mdeboard	http://images.google.com/imgres?q=stovepipe&hl=en&biw=1600&bih=746&gbv=2&tbm=isch&tbnid=O_5jgmm1b31ftM:&imgrefurl=http://4photos.net/en/image:142-47576-Glock_23_stovepipe_images&docid=BW6DcnkfZoGYgM&imgurl=http://4photos.net/photosv2/130428_glock_23_stovepipe.jpg&w=300&h=200&ei=QDGzTqLoNvG42QXsjMHMDQ&zoom=1
20:30	mdeboard	wow, nice link
20:30	mdeboard	you're right, I used"literally" and shouldn't have
20:30	mdeboard	Literally a stovepipe: http://www.daviddarling.info/images/stove_pipe.jpg
20:30	mdeboard	Metaphorical weapon-related stovepipe, whence the term originates: http://4photos.net/photosv2/130428_glock_23_stovepipe.jpg
20:30	TimMc	Now that's a stovepipe.
20:31	TimMc	Wikipedia has two definitions: https://secure.wikimedia.org/wikipedia/en/wiki/Stovepiping
20:31	mdeboard	So by stovepiping the project I mean something that is between me and project completion
20:32	mdeboard	wow
20:32	mdeboard	ANd knowing is half the battle
20:32	technomancy	(defn getenv [x] ({} x (System/getenv x)))
20:33	technomancy	^ because no System/setenv exists
20:33	TimMc	mdeboard: And here I thought you meant you had some spaghetti... stovepipes...
20:33	TimMc	piping data from one component to another, but haphazardly.
20:36	mdeboard	TimMc: Interesting, I've really never heard of that. Can tell I wasn't an intel bubba.
20:38	mdeboard	TimMc: Thanks, actually.
20:39	TimMc	I still don't really get the gov't/intel sense of the phrase.
20:39	TimMc	I mean, at a gut level for what it means organizationally.
20:43	mdeboard	TimMc: AFAICT it uses the meaning of stovepipe i.e. "isolated vertical conduit" that vents material outside the building, to mean an isolated vertical conduit that vents "material" (information) outside the organization
20:43	mdeboard	Like when a colonel has the ear of a senator
20:43	mdeboard	and passes his information to influence policy
20:43	mdeboard	without distributing it through the proper channels to reach the same destination.
20:44	TimMc	I get that one.
20:44	TimMc	technomancy: What's the context?
20:45	technomancy	TimMc: you want to change an environment variable without restarting the whole JVM
20:45	technomancy	so you just slap it in the map in getenv
20:48	TimMc	{} is a placeholder for a ref of a map?
20:48	technomancy	it's just something you can edit inside the function at dev time
20:48	cark	is there a one liner that transforms :a to ([:a] [:a :a] [:a :a :a] ....) ?
20:49	cark	vectors not required,
20:49	TimMc	technomancy: Oh, some defaults.
20:49	hiredman	,(for [i (range 10)] (repeat i :a))
20:50	clojurebot	(() (:a) (:a :a) (:a :a :a) (:a :a :a :a) ...)
20:50	TimMc	,((partial iterate conj []) :a)
20:50	clojurebot	#<ArityException clojure.lang.ArityException: Wrong number of args (3) passed to: core$iterate>
20:50	cark	hiredman: ah thanks, i was looking at reductions
20:50	TimMc	urg
20:50	cark	but good enough !
20:50	ibdknox	,(map #(repeat % :a) (range 1 10))
20:50	clojurebot	((:a) (:a :a) (:a :a :a) (:a :a :a :a) (:a :a :a :a :a) ...)
20:52	amalloy	&(rest (iterate #(conj % :a) []))
20:52	TimMc	,(take 3 (drop 1 (iterate #(conj % :a) []))) ; cark
20:52	clojurebot	([:a] [:a :a] [:a :a :a])
20:52	lazybot	java.lang.OutOfMemoryError: Java heap space
20:52	ibdknox	lol
20:52	amalloy	haha sorry lazybot
20:52	amalloy	(dec amalloy)
20:52	lazybot	You can't adjust your own karma.
20:53	gfredericks	(inc amalloy)
20:53	lazybot	⇒ 19
20:53	gfredericks	serves you right
20:53	gfredericks	for trying to dec yourself
20:53	ibdknox	lol
20:54	cark	thanks all !
20:54	gfredericks	when I try to create a future in the body of an aleph channel-receiving function, I get a java.util.concurrent.RejectedExecutionException
20:54	TimMc	technomancy: http://stackoverflow.com/questions/318239/how-do-i-set-environment-variables-from-java/496849#496849
20:55	TimMc	:-P
20:55	TimMc	cark: Pyramid scheme of some sort.
20:55	technomancy	TimMc: my eyes!
20:55	TimMc	haha
20:56	gfredericks	It's not a real function if you haven't typed "Map<String, String>" at least four times.
20:56	technomancy	TimMc: that is a serious hack though, wow
20:56	technomancy	i wonder how many lines of clojure that would be
20:57	ibdknox	that is all kinds of evil
20:57	hiredman	,(let [[Map<String, String>] [1 2]] [Map<String, String>])
20:57	clojurebot	[1 2]
20:57	TimMc	ahaha
20:58	hiredman	clojurebot: symbols \|are\| kind of fun
20:58	TimMc	gfredericks: Map<LanguageCode, Map<String, Map<String, Set<Integer>>>>
20:58	clojurebot	In Ordnung
20:58	gfredericks	clojurebot: symbols?
20:58	clojurebot	symbols are kind of fun
20:58	cark	ah for completeness here is the reductions version :
20:58	cark	,(take 3 (reductions conj [] (repeat :a)))
20:58	clojurebot	([] [:a] [:a :a])
20:58	cark	but i'll take hiredman's version which is the most readable
20:59	cark	(for me)
20:59	TimMc	cark: but but but it doesn't share structure!
20:59	cark	ah indeed !
21:00	amalloy	technomancy: not as much, anyway
21:00	amalloy	er, TimMc
21:01	gfredericks	to share the most structure wouldn't we want to use cons instead of conj?
21:01	TimMc	technomancy: Farther down the page is a crossplatform hack.
21:01	hiredman	depends
21:02	cark	curse you TimMc, now i have to think about structure sharing =P
21:02	hiredman	cons is really only for constructing seqs
21:02	hiredman	conj is fine for sharing structure among vectors
21:02	cark	one can conj on a list
21:03	technomancy	http://p.hagelb.org/setenv.clj.html <= 7 lines, could be golfed further
21:03	gfredericks	,(map type [(cons 7 ()) (conj 7 ())])
21:03	clojurebot	#<ClassCastException java.lang.ClassCastException: java.lang.Long cannot be cast to clojure.lang.IPersistentCollection>
21:04	gfredericks	,(map type [(cons 7 ()) (conj () 7)])
21:04	clojurebot	(clojure.lang.Cons clojure.lang.PersistentList)
21:04	gfredericks	okay so I meant conjing onto a list I guess
21:04	technomancy	I'll stick with my crappy defn getenv workaround for now
21:07	TimMc	technomancy: I was just writing that! `lein new hurk`, (defn ohgod [...] ...)
21:09	jcromartie	for some reason, the bozo that wrote this code thought that a singleton object which operates as a state machine was somehow thread safe
21:09	jcromartie	and the words "thread safe single object for foo repository" are all over
21:10	TimMc	jcromartie: It is threadsafe because he used the @threadsafe Javadoc annotation.
21:10	jcromartie	not even
21:10	jcromartie	but dear god
21:10	jcromartie	I just realized this crap is all over
21:10	jcromartie	it's a wonder it works at all
21:11	amalloy	jcromartie: but there's only one! how could that be unsafe!!!!
21:11	TimMc	jcromartie: Ah, the mystery of life.
21:12	gfredericks	at work I've been using the Activiti BPMN engine, and in one of their webapps they have a 500 line custom implementation of a prefix trie. AND it's buggy. I rewrote it in clojure and it worked fine.
21:13	gfredericks	I thought of that because I'm pretty sure it's not threadsafe either, but of course my clojure replacement was.
21:13	jcromartie	:)
21:13	TimMc	OK, evangelism question. I want to start converting bits of a large multi-component Java project at work into Clojure. How can I do it incrementally?
21:13	TimMc	Everything is managed with Maven.
21:14	jcromartie	this is in 14 classes
21:14	jcromartie	they are all race conditions
21:14	jcromartie	they are all used unsynchronized
21:14	jcromartie	by multiple users on the web site
21:14	jcromartie	and they all say "Thread Safe"
21:14	gfredericks	jcromartie: if something doesn't work they just click the button again?
21:14	TimMc	Maybe it is a brand name.
21:14	jcromartie	and even among those classes there are OTHER race conditions
21:15	jcromartie	like where they copied temporary files to a file using new Guid() in .NET
21:15	jcromartie	which creates an empty GUID
21:15	jcromartie	every time
21:15	TimMc	ahaha
21:15	jcromartie	the same file, 00000000-0000-0000-0000-000000000000
21:15	jcromartie	is used for every concurrent user
21:15	jcromartie	I mean dear god
21:16	jcromartie	I knew concurrency was not always easy
21:16	TimMc	"Well, when I tested it, it worked!"
21:16	jcromartie	but this...
21:16	jcromartie	this is inexcusable
21:16	gfredericks	jcromartie: it's okay if they're all the same as long as they're globally unique
21:16	cark	indubitably one unique guid you got there
21:17	jcromartie	thankfully our user activity is measured in hits per minute
21:18	gfredericks	IUID: Initially Unique Identifier
21:18	TimMc	I figured.
21:18	TimMc	jcromartie: Why does that constructor even exist? Attractive nuisance.
21:19	jcromartie	TimMc: it's a .net struct
21:19	jcromartie	meaning it defaults all of the fields to a default (0) values
21:20	jcromartie	and, of course, since the geniuses copied and pasted all of this crap, it's the same race condition duplicated separately 14 times
21:20	jcromartie	used in slightly different ways
21:20	TimMc	hahaha
21:20	jcromartie	I mean
21:20	jcromartie	can we sue?
21:20	TimMc	lazybot: Can we sue???
21:20	lazybot	TimMc: Yes, 100% for sure.
21:20	gfredericks	jcromartie: for a better-than-average job? :)
21:22	TimMc	,(= java.lang.String (.getClass ""))
21:22	clojurebot	true
21:22	TimMc	technomancy: Why not (= c java.util.Collections$UnmodifiableMap)
21:23	technomancy	TimMc: sure, why not?
21:23	TimMc	since you kind of asked for golfing
21:23	TimMc	Also, prettier.
21:25	amalloy	(some #{j.u.C$UM} all-the-classes)
21:25	TimMc	Why does it even bother iterating?
21:27	technomancy	probably because it's java =P
21:27	TimMc	(let [c java.util.Collections$UnmodifiableMap ... ) works
21:29	TimMc	Seriously, WTF.
21:31	TimMc	I wonder if the author didn't know you could directly reference stuff like that.
21:32	jlf	technomancy: what's the recommended package to get clojure-jack-in working in a vanilla emacs 24?
21:32	technomancy	jlf: clojure-mode and lein plugin install swank-clojure 1.3.3 should do it
21:32	technomancy	jlf: clojure-mode either from marmalade or github
21:32	jlf	thanks
21:40	technomancy	ah cool; say hi to everyone for me
21:41	darevay	hi 4clojurers. For problem 86 (http://www.4clojure.com/problem/86) I have a solution (http://www.4clojure.com/problem/86) that causes ClassNotFoundException:clojure.core. Works fine locally. What have I done?
21:41	gfredericks	how does 1.3 handle vars in (future)?
21:41	darevay	... oops that solution link should be https://gist.github.com/1338452
21:46	TimMc	darevay: Yeah, WORKSFORME locally.
21:47	amalloy	darevay: something seems to be wrong with the set literal. i don't know how that can be, really
21:48	amalloy	ie, if i replace your #{} with [] and use (some #{n} ns), it works. obviously that shouldn't matter
21:51	darevay	amalloy: thanks. weird.
22:00	darevay	... and now I see how I was working way too hard to get the digits of a number. geesh :)
22:07	gfredericks	oh man. an hour of debugging and it turns out ztellman implemented his own map type. o_o
22:17	gfredericks	is there not an interface or protocol for conj?
22:19	cark	,(conj {} [:a 1])
22:19	clojurebot	{:a 1}
22:20	amalloy	gfredericks: IPersistentCollection.cons
22:23	gfredericks	amalloy: ah ha. Indeed that method is left out. thanks.
22:45	mdeboard	export JAVA_OPTS=-Xmx768m
22:45	mdeboard	oop
22:55	brehaut	ibdknox: ping?
23:02	leo2007	can M-. in slime jump to definition of `def'?
23:21	tali713	leo2007: i can't, here.
23:21	leo2007	tali713: thanks for the confirmatino.
23:22	leo2007	confirmation*
23:28	technomancy	leo2007: it should actually
23:29	leo2007	a bug?
23:29	technomancy	could be
23:45	technomancy	seancorfield: fixed that bug in clojure-jack-in that was causing issues with warn-on-reflection
23:45	technomancy	released a new clojure-mode version with the fix
23:47	aperiodic	is there a way to exit from a repl inside a repl?
23:47	aperiodic	without also exiting from the enclosing repl, that is
23:55	technomancy	sending an EOF should do it