#clojure – 2011-06-23

2011-06-23

00:37	rhdoenges	I am proficient with vim and learning clojure. Is vimclojure/slimv effective or should I just learn emacs?
00:38	Scriptor	rhdoenges: I've heard some complaints about it, and learning the basics of emacs isn't too hard
00:40	rhdoenges	Scriptor: I guess I'll go watch the PeepCode screencast about emacs.
00:42	rhdoenges	I might as well learn emacs if I'm going to be learning clojure; I can kill two birds with one stone
00:55	scgilardi	inc
00:55	amac	dec
00:55	amalloy	(juxt inc dec)
00:58	amac	that is very zen.
01:01	rhdoenges	#(dec (inc %))
01:02	amalloy	amac: juxt is always the right answer
01:06	rhdoenges	but what is the question
01:09	amac	juxt is also the question
01:13	rhdoenges	,(map #(% 1) ((juxt juxt juxt) inc dec))
01:13	clojurebot	([2 0] [2 0])
01:13	rhdoenges	(juxt juxt juxt)
01:16	Scriptor	,(doc juxt)
01:16	clojurebot	"([f] [f g] [f g h] [f g h & fs]); Alpha - name subject to change. 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)]"
01:18	rhdoenges	,(map #(flatten (% 1)) ((juxt juxt juxt) inc dec))
01:18	clojurebot	((2 0) (2 0))
01:18	rhdoenges	,(flatten (map #(% 1) ((juxt juxt juxt) inc dec)))
01:18	clojurebot	(2 0 2 0)
01:19	rhdoenges	[I like juxt. I dislike juxt.]
01:34	halfprogrammer	[I Ꞁıʞǝ ɾnxʇ˙ I pıslıʞǝ ɾnxʇ˙]
01:36	rhdoenges	halfprogrammer: shit is deep, man.
01:36	halfprogrammer	rhdoenges: ɥǝ ɥǝ ɥǝ
01:37	rhdoenges	halfprogammer: what are you using to generate that text?
01:37	Scriptor	rhdoenges: it's somewhere online, just google for 'upside down text'
01:37	halfprogrammer	rhdoenges: I am using weechat for irc. There is a 'flip' plugin available for that
01:38	rhdoenges	I shall try to write it in clojure now.
01:59	amalloy	rhdoenges: write it in clojure? isn't it basically: (1) look up unicode for upside-down letters, (2) type in a hashmap literal?
02:14	zakwilson	Your new password is clojureql.core.RTable@bb998e4d <-- I don't think that's quite right.
02:19	hiredman	I bet those password strength widgets would give it a pass
02:34	amalloy	hiredman: two pairs in a row of repeated characters? laughably weak
02:43	rhdoenges	amalloy: I'm learning clojure. it will be good for me
02:44	rhdoenges	amalloy: also I'm curious about stdout/stdin
02:44	amalloy	rhdoenges: cool beans. every problem is a good candidate for a learning project
02:46	amalloy	i love clojure (functional programming in general, i guess?). if my code compiles, nine times out of ten it yields the right answer
03:00	sunnibo	I'm making a twitter client with clojure-twitter. (twitter/favorites) works well, but (twitter/destroy-favorite) does'nt work. it generates a 400 error and tells me that I have to use GET method. it's weird.
03:01	sunnibo	http://dev.twitter.com/doc/post/favorites/destroy/:id says the supported request method is POST and DELETE
07:42	void_	how do I turn {:a 1 :b 2} into (list :a 1 :b 2) ?
07:42	opqdonut	,(apply concat (seq {:a 1 :b 2}))
07:42	clojurebot	(:a 1 :b 2)
07:43	opqdonut	(the order of the key-value pairs will be undeterministic)
07:43	opqdonut	(unless you use sorted-map)
07:46	void_	doesn't matter, thanks!
08:00	jaju	,(+ 1 2)
08:00	clojurebot	3
08:00	jaju	(+ 1 2)
08:00	clojurebot	3
08:03	solussd	clojure is deterministic!
08:04	solussd	(* 1 2)
08:04	clojurebot	2
08:04	solussd	hm
08:05	jaju	(println "hello")
08:05	opqdonut	,(println "hello")
08:05	clojurebot	hello
08:05	opqdonut	slightly different
08:06	jaju	oh thanks!
08:06	jaju	interesting
08:06	jaju	clojurebot accepts function definitions too?
08:06	jaju	trying
08:06	jaju	(defn mul [x y] (* x y))
08:06	jaju	,(defn mul [x y] (* x y))
08:06	clojurebot	DENIED
08:07	jaju	,(defn mul [x y] (* x y))
08:07	clojurebot	DENIED
08:07	jaju	oh cool!
08:07	Fossi	nope
08:07	Fossi	it's sandboxed
08:07	Fossi	and def is evil ;)
08:07	jaju	my first time here - so, exploring a bit :)
08:07	manutter	I think you can use letfn to make temporary functions though
08:07	Fossi	you can also message it in private
08:07	jaju	cool stuff!
08:07	Fossi	if you wanna toy around a whole lot :)
08:08	manutter	,(letfn [(plus [a b] (+ a b))] (plus 1 2))
08:08	clojurebot	3
08:08	manutter	heh :)
08:08	jaju	oh no - just exploring. I'd rather have clojure discussions! :)
08:08	jaju	cool - thanks for the tip
08:08	Fossi	there's also sexpbot and another one
08:08	Fossi	#()
08:08	manutter	The letfn is more useful on 4clojure.org
08:08	Fossi	#(= 1 1 )
08:08	jaju	maybe - when I do not have a clojure repl around, this might be handy!
08:08	Fossi	#(+ 1 1)
08:09	manutter	##(= 1 1)
08:09	sexpbot	⟹ true
08:09	Fossi	ah, that was it
08:09	manutter	I've seen some cool bot commands but I don't remember may of them
08:09	jaju	##(+ 1 1)
08:09	sexpbot	⟹ 2
08:09	manutter	~botsnack
08:09	clojurebot	Thanks, but I prefer chocolate
08:09	jaju	nice
08:09	manutter	~chocolate
08:09	clojurebot	No entiendo
08:10	jaju	haha
08:10	manutter	clojurebot: you lie.
08:10	clojurebot	Gabh mo leithscéal?
08:10	manutter	clojurebot: help
08:10	clojurebot	http://www.khanacademy.org/
08:10	manutter	lol
08:10	jaju	:)
08:10	jaju	clojurebot: what's your name?
08:10	clojurebot	amespaces are (more or less, Chouser) java packages. they look like foo.bar; and corresponde to a directory foo/ containg a file bar.clj in your classpath. the namespace declaration in bar.clj would like like (ns foo.bar). Do not try to use single segment namespaces. a single segment namespace is a namespace without a period in it
08:11	manutter	not the answer I was looking for, but a good link nonetheless
08:13	jaju	~corn
08:13	clojurebot	Huh?
08:13	jaju	~corn
08:13	clojurebot	excusez-moi
08:13	jaju	~coffee
08:13	clojurebot	Pardon?
08:14	jaju	(doc +)
08:14	clojurebot	"([] [x] [x y] [x y & more]); Returns the sum of nums. (+) returns 0."
08:14	jaju	(source +)
08:14	jaju	,(source +)
08:14	clojurebot	java.lang.Exception: Unable to resolve symbol: source in this context
09:13	manutter	I shiver a little when I have to document code that uses variables like $magic_function_name
09:16	manutter	then again, it IS php...
09:40	kryft	Hmm, what's the difference between PersistentArrayMap and PersistentHashMap? JoC apparently refers to the former as a hash map.
09:44	manutter	I'm getting the impression that PersistentArrayMaps maintain order but can't hold as many values, whereas PersistentHashMap can be bigger but does not maintain order
09:45	manutter	That's googling for both terms and skimming a few blogs, so don't quote me
09:46	TobiasRaeder	yeah just based on the names the array map probably holds all data in an array the hash map in a hash based tree/trie
09:46	TobiasRaeder	-> lookups in long/large arrays might get expensive but should keep insertion order
09:46	TobiasRaeder	since the hashmap will distribute the data based on the hashes there is no way to reconstruct the insertion order
09:47	TobiasRaeder	but faster lookup on larger maps
09:47	TobiasRaeder	same here - dont quote me on it ;) just a quick look at the implementation and thinking about the names
09:50	kryft	Sounds logical. I was just confused by JoC page 98: ".. a new hash map can be created idiomatically ..: (into {} [[:a 1] [:b 2]])
09:50	kryft	"
09:52	TobiasRaeder	yeah
09:52	TobiasRaeder	{} creates an array map and not a hash-map
09:52	TobiasRaeder	seems like a typoe
09:52	pjstadig	The ordering of keys by PersistentArrayMap is just coincidental, and not a guarantee, the array map is just faster for smaller number of keys, and automatically promotes to a PersistentHashMap at some threshold (8 keys I think)
09:52	TobiasRaeder	mhm
09:52	TobiasRaeder	&(class (hash-map))
09:52	sexpbot	⟹ clojure.lang.PersistentArrayMap
09:53	TobiasRaeder	thats really interesting
09:53	manutter	&(class {})
09:53	sexpbot	⟹ clojure.lang.PersistentArrayMap
09:53	kryft	&(class (hash-map :a 1))
09:53	sexpbot	⟹ clojure.lang.PersistentHashMap
09:53	pjstadig	i think the term "hash map" can be applied to both since they both implement the same set of methods
09:53	pjstadig	i wouldn't get hung up on the names
09:54	manutter	&(class {:a 1 :b 2 :c 3 :d 4 :e 5 :f 6 :g 7 :h 8 :i 9})
09:54	sexpbot	⟹ clojure.lang.PersistentHashMap
09:55	manutter	Yeah, I think "hash map" is a non-implementation-specific term meaning "access values by key"
09:55	pjstadig	,(class {:a 1 :b 2 :c 3 :d 4 :e 5 :f 6 :g 7 :h 8})
09:55	clojurebot	clojure.lang.PersistentArrayMap
09:55	pjstadig	,(class {:a 1 :b 2 :c 3 :d 4 :e 5 :f 6 :g 7 :h 8 :i 9})
09:55	clojurebot	clojure.lang.PersistentHashMap
09:55	manutter	PHashMap vs PArrayMap are just implementation details that should be ignored in most cases.
09:56	djpowell	Hmm, is there a bug with aset on 2-d arrays?
09:56	djpowell	,(aset (make-array Integer/TYPE 3 3) 0 0 1)
09:56	clojurebot	1
09:56	djpowell	grr
09:56	djpowell	#(aset (make-array Integer/TYPE 3 3) 0 0 1)
09:56	manutter	,(doc make-array)
09:56	clojurebot	"([type len] [type dim & more-dims]); Creates and returns an array of instances of the specified class of the specified dimension(s). Note that a class object is required. Class objects can be obtained by using their imported or fully-qualified name. Class objects for the primitive types can be obtained using, e.g., Integer/TYPE."
09:56	djpowell	what ver is sexpbot, and how do i talk to it?
09:57	manutter	&(doc make-array)
09:57	sexpbot	⟹ "([type len] [type dim & more-dims]); Creates and returns an array of instances of the specified class of the specified dimension(s). Note that a class object is required. Class objects can be obtained by using their imported or fully-qualified name. Class objects f... http://gist.github.com/1042576
09:57	mrBliss	&(clojure-version)
09:57	sexpbot	⟹ "1.2.0"
09:57	djpowell	&(aset (make-array Integer/TYPE 3 3) 0 0 1)
09:57	sexpbot	⟹ 1
09:57	mrBliss	,(clojure-version)
09:57	clojurebot	"1.2.0"
09:57	djpowell	&(clojure-version)
09:57	sexpbot	⟹ "1.2.0"
09:57	djpowell	ah
09:57	manutter	(doc aset)
09:57	clojurebot	"([array idx val] [array idx idx2 & idxv]); Sets the value at the index/indices. Works on Java arrays of reference types. Returns val."
09:57	djpowell	ok - it fails in clojure 1.3
09:57	pjstadig	neither of the bots is running 1.2.1...tsk tsk
09:58	manutter	(insert famous obiwan quote here...)
10:01	djpowell	it seems to be a bug anyway, i'll file a report
10:09	djpowell	ah, it is the new default behaviour of longs that is doing it - the aset value is getting converted to a Long rather than an Integer
10:10	djpowell	oh, hang on; the aset-int method works. everyone normally seems to suggest avoiding that
10:10	gfrlog	I am getting very strange behavior from "lein uberjar" -- it's throwing a java.util.zip.ZipException for various files in src/resources/public
10:11	gfrlog	doesn't seem to be a particular file, because renaming or deleting the offending file doesn't help
10:11	djpowell	on windows?
10:11	gfrlog	I should say the exception message is "duplicate entry: <filename>"
10:11	gfrlog	I'm on linux
10:12	djpowell	ah k. I seemed to have an issue with odd file permissions recently, but it doesn't sound like that
10:12	gfrlog	guess I should see if "lein jar" fails as well
10:12	djpowell	do you have a file of the same name in src and resources?
10:12	gfrlog	nope
10:12	gfrlog	nothing weird like that
10:13	gfrlog	same exception for "lein jar"
10:14	gfrlog	"lein ring server-headless" succeeds though, and I believe it properly serves the files in src/resources/public
11:01	pyr	clojure 1.3 has a deref with timeout and timeout-val possible,nice!
11:16	technomancy	gfrlog: there's probably a duplicate in your dependencies. maybe uberjar needs debug logging to track that kind of thing down
11:21	fliebel	Life is still a vector.
11:26	chouser	##(= == =)##
11:26	sexpbot	⟹ false
11:26	chouser	good, just checking.
11:26	Fossi	#(- _ -)
11:26	Fossi	##(- _ -)
11:26	sexpbot	java.lang.Exception: Unable to resolve symbol: _ in this context
11:26	TobiasRaeder	what is the most idiomatic way to convert a string to an int?
11:27	pjstadig	Integer/parseInt
11:27	Fossi	i thought there was a wrapper somewhere
11:27	TobiasRaeder	mhm
11:27	TobiasRaeder	kk
11:27	TobiasRaeder	if i find it ill use it otherwise just write something
11:28	jcromartie	what do y'all think of DDD? http://domaindrivendesign.org/resources/ddd_terms
11:28	jcromartie	seems like a good match
11:28	chouser	parseInt is good. You can use (read), but you may be surprised by the results.
11:28	manutter	a good match for clojure?
11:28	manutter	jcromartie: ^^
11:29	jcromartie	yes manutter
11:29	tomoj	the DDD book I have has like three good chapters then a bunch of OOP BS
11:29	jcromartie	hah
11:30	tomoj	guess I ought to slog through it to try to get the point anyway :/
11:34	jcromartie	I think that glossary maps really nicely to lots of clojure concepts
11:34	jcromartie	especially the difference between value and identity
12:14	timvisher	hi all
12:14	arohner	timvisher: hi
12:14	timvisher	is there a function I can easily combine with some that will return the thing matched from a seq?
12:14	timvisher	hoo boy that was a bad explanation.
12:14	timvisher	let me gist something
12:15	timvisher	https://gist.github.com/1042882
12:15	timvisher	basically, i find myself someing over a seq a lot and then returning the thing that I was searching for by duplicating it
12:15	timvisher	because the only thing I know to do is to use =
12:16	timvisher	which is a predicate
12:16	arohner	the return value of some is the thing it matched
12:16	timvisher	i know
12:16	timvisher	not quite though
12:16	timvisher	it's the return value of the function that returned non-nil
12:17	timvisher	so in other words, ,(some #(= 1 %) [1 2 3 4 5])
12:17	timvisher	,(some #(= 1 %) [1 2 3 4 5])
12:17	clojurebot	true
12:17	timvisher	whereas
12:17	timvisher	,(some identity [nil nil 1])
12:17	clojurebot	1
12:17	arohner	(some #(if (= "png" %) % nil) extensions)
12:17	mrBliss	,(or (first (filter #(= 1 %) [1 2 3 4 5])) :else)
12:17	clojurebot	1
12:18	arohner	oh yeah. I do first filter a lot
12:18	timvisher	arohner: wow, not sure why i didn't think of that
12:18	timvisher	goes to show how blind you can get when you're stairing at your own code
12:18	timvisher	thanks for the tips, arohner and mrBliss
12:28	gfrlog	deffing things that depend on IO is really messy I'm finding out :(
12:28	tomoj	isn't #(if (= "png" %) % nil) #{"png"}
12:29	no_mind	any plans for proposing clojure miniconf at linux.conf.au ?
12:29	tomoj	,(doc some)
12:29	clojurebot	"([pred coll]); Returns the first logical true value of (pred x) for any x in coll, else nil. One common idiom is to use a set as pred, for example this will return :fred if :fred is in the sequence, otherwise nil: (some #{:fred} coll)"
12:29	arohner	tomoj: doh. yeah, that's better
12:50	redline6561_	Hi. Sorry if this question is dumb but are there plans to extend leiningen to handle a use-case more like cljr? My background is CL and I miss being able to fire up SLIME and have access to all my globally installed libs. :-P
12:53	manutter	redline6561_: Have you looked at jark?
12:53	redline6561_	manutter: Nope. Haven't heard of it.
12:54	davekong	Is using a map or doing the equivalent with a list comprehension more efficient? Is one way considered more idiomatic than the other?
12:54	manutter	ah, you might like it, http://icylisper.in/jark
12:54	technomancy	redline6561_: why not create a project to contain your libs?
12:54	manutter	(I think that's the right url...)
12:54	manutter	(yeah, that works)
12:55	redline6561_	technomancy: Mostly to be able to experiment with new libs and try things that I might not necessarily have a specific project in mind for.
12:55	redline6561_	Admittedly, I could have (or leiningen could ship with) a default/experiments project for screwing around.
12:55	redline6561_	It's just extra work vs my old start emacs, M-x slime process...so I was curious. :)
12:56	redline6561_	manutter: Interesting, thanks!
12:56	technomancy	redline6561_: well, the JVM prevents you from adding new deps without restarting anyway
12:56	technomancy	so it's never going to be like CL
12:56	cemerick	technomancy: maybe have `lein repl` add the most recent versions of all artifacts in ~/.m2 to the classpath if there's no project.clj? :-P
12:56	redline6561_	technomancy: :( Yeah, that's the impression I got.
12:56	technomancy	cemerick: can't tell if you're serious. =)
12:57	redline6561_	cemerick +1000
12:57	cemerick	technomancy: sure, why not? Print a warning, maybe with a link that explains what's going on, and let 'em have at it.
12:57	pjstadig	ugh
12:57	manutter	what if your .m2 contains incompatible libs?
12:58	pjstadig	terrible idea
12:58	technomancy	doesn't perf degrade with a ginormous classpath?
12:58	sot	Hi, Does anyone know in what new module str-join is in?
12:58	cemerick	only when loading resources
12:58	manutter	but while we're on this topic, does lein support like a user.clj global file?
12:58	cemerick	Clearly not intended for "real work" — but it'd be hard to beat for experimentation and newbies.
12:58	technomancy	maybe a separate "mega-repl" task; probably not suitable for the default
12:59	pjstadig	manutter: clojure supports it,if it's on the classpath
12:59	manutter	maybe lein isn't the right place for that, though
12:59	redline6561_	mega-repl sounds acceptable to me.
12:59	technomancy	manutter: there's ~/.lein/init.clj which runs in lein's process
13:00	manutter	yeah, I'm trying to figure out how to have some util routines that are available whenever I fire up a repl
13:00	redline6561_	I like clojure a lot and have been glad to see leiningen gain traction...but the remaining extra functionality found in cljr/jark/cake is functionality I really like.
13:00	redline6561_	Would prefer not to have several build/lib tools.
13:00	redline6561_	Mega-repl is certainly the main thing in my mind. Less concerned with scripting + persistent jvm.
13:00	technomancy	I've yet to see a convincing explanation for how that's different from just creating a throwaway project
13:01	technomancy	but mega-repl would be easy-peasy to implemnet as a plugin
13:01	cemerick	redline6561_: I think we're going to see continued proliferation of such tools for a while.
13:01	manutter	but if I just do clojure-jack-in then I haven't necessarily launched my repl from lein, so lein isn't the right place for it
13:01	technomancy	manutter: maybe you want user-level plugins?
13:01	redline6561_	That's totally fair, not trying to be pushy. It was just a late evening and I went, "Ugh. Create a pretend project? No." and went to bed.
13:01	manutter	redline6561_: it seems to me that the focus of lein and jark are somewhat different, build tool vs repl tool
13:02	redline6561_	^^ technomancy. It's certainly a workable solution, if not an ideal one.
13:02	redline6561_	manutter: True. cemerick: Duly noted. Thanks.
13:02	technomancy	redline6561_: I think psychologically it's different. I'm just not sure if technically it is
13:02	redline6561_	UX is all about psychology. :)
13:02	technomancy	maybe if there were just a project auto-generated in ~/.lein/scratch or something
13:02	redline6561_	Technically, yeah, it's pretty much identical.
13:03	manutter	technomancy: if I install a plugin with "lein plugin install", does that make it available to all my projects automatically?
13:03	technomancy	manutter: yeah
13:05	manutter	Ok, so I create a project with lein new, but then I just fire up emacs and do clojure-jack-in, will it stlll be there?
13:05	redline6561_	technomancy: Thanks for leiningen. Excited for search in 1.6. Will be quite nice for a Quicklisp guy like myself. :D
13:06	technomancy	manutter: sure
13:06	arohner	davekong: no, one isn't more idiomatic than the other. it's user preference
13:06	manutter	technomancy: cool! I'd have thought you needed a lein swank or something, but this suits my purposes
13:07	technomancy	jack-in just delegates to swank
13:08	mattmitchell	how can i easily convert a hash map into a vector: {:names ["one" "two"] :id 100} would become [:names ["one" "two"] :id 100] ???
13:08	pjstadig	starting a repl is not the same thing as a build tool, and perhaps having a standard CLI for clojure would be a better solution
13:08	mattmitchell	(map vec {}) almost does it, but I want the top level flattened
13:09	gfrlog	mattmitchell: ##(apply concat (seq {:names ["one" "two"] :id 100}))
13:09	sexpbot	⟹ (:names ["one" "two"] :id 100)
13:09	mattmitchell	gravity: awesome thank you
13:09	gfrlog	I guess pass that to vec if you really do need a vector
13:09	technomancy	pjstadig: a novel idea!
13:09	gfrlog	mattmitchell: no problem :)
13:10	gfrlog	##(mapcat {:names ["one" "two"] :id 100})
13:10	sexpbot	java.lang.IllegalArgumentException: Wrong number of args (1) passed to: core$map
13:10	gfrlog	##(mapcat identity {:names ["one" "two"] :id 100})
13:10	sexpbot	⟹ (:names ["one" "two"] :id 100)
13:10	gfrlog	maybe that's a little simpler
13:12	mattmitchell	gfrlog: cool, yeah i like that
13:14	gfrlog	mattmitchell: note that the order of the pairs is not defined
13:15	mattmitchell	gfrlog: you mean ordering of the keys could possibly change?
13:26	gfrlog	mattmitchell: yes
13:26	gfrlog	,(map seq [{:foo 1 :bar 2} {:bar 2 :foo 1}])
13:26	clojurebot	(([:foo 1] [:bar 2]) ([:bar 2] [:foo 1]))
13:27	gfrlog	that is apparently a bad example
13:27	gfrlog	probably small maps use arrays
13:27	gfrlog	but I don't think that's something you're supposed to depend on
13:28	mattmitchell	gfrlog: ok, so is it possible a key/value could get messed up in the resulting list?
13:28	gfrlog	mattmitchell: I mean that you should assume {:foo 1 :bar 2} could come out either as [:foo 1 :bar 2] or [:bar 2 :foo 1]
13:28	gfrlog	but NOT anything like [:foo 2 :bar 1] or [1 2 :foo :bar]
13:28	gfrlog	at least not with the code we used above
13:28	mattmitchell	gfrlog: ahh ok yes, that's fine
13:29	mattmitchell	thank you for clarifying
13:29	gfrlog	no problem
13:31	gfrlog	technomancy: removing my src/resources directory causes the [uber]jar tasks to succeed -- does that go along with or contradict your "duplicate dependencies" suggestion?
13:32	technomancy	gfrlog: must be a dupe between your project and a dep then
13:32	technomancy	if the error message doesn't make it clear what file is conflicting, that should be addressed
13:33	gfrlog	technomancy: it does, but each time I delete a file a new file is the culprit. Some of the files have rather unique names too
13:33	technomancy	do you have your own project as a dependency? =)
13:33	gfrlog	I have 4 files in src/resources total, and it blamed all 4 of them as I deleted the others
13:34	gfrlog	technomancy: I hope not :)
13:34	gfrlog	technomancy: the project's not on clojars or anything like that
13:35	gfrlog	technomancy: I guess I'll start hacking away at the project until I get a minimal example of the error
13:37	technomancy	gfrlog: actually uberjar keeps a running set of all files already included and automatically omits dupes
13:37	technomancy	so somehow your project is working around that
13:37	gfrlog	technomancy: did I mention than "lein jar" fails the same way? so it's not an uberjar thing
13:37	technomancy	oh, I didn't catch that
13:39	gfrlog	technomancy: and "lein ring server" succeeds, and the app correctly serves the static files from src/resources
13:39	gfrlog	I've never had an issue with that, and also never tried building a jar until now
13:52	gfrlog	I bet this is me not using :resources-path correctly
13:52	gfrlog	cuz I have the project down to about nothing and it's still failing
13:54	cemerick	dnolen: See, it looks like you have quite the cheering section, at the very least. :-)
13:55	dnolen	cemerick: heh
14:01	gfrlog	technomancy: if this isn't a minimal example I don't know what is :) https://github.com/fredericksgary/isuckatlein
14:02	gfrlog	I'm sure it'll end up being some missing key in the project.clj
14:04	dnolen	though it's a bit disappointing that people think types for Clojure are "advanced", "hard". A logic engine gives you most of a type system for "free". The HM type checker for Prolog is 890 LOC. I don't expect a HM type checker on top of core.logic for Clojure to be significantly more than that.
14:05	dnolen	in Typed Racket it sounds like a lot of the actual complexity comes from integrating untyped code.
14:09	technomancy	gfrlog: mv src/resources .
14:09	technomancy	it's odd that it would fail that way though
14:10	amalloy	technomancy, gfrlog: foo-whatever.bar is on the classpath twice, with different paths. that sounds to me like a bad plan
14:10	technomancy	I see what's going on; the mechanism for trimming the filename is the same for files in src or in resources
14:10	technomancy	so it gets confused if there is overlap
14:10	technomancy	however, there shouldn't bever be overlap
14:11	pyr	hey
14:11	gfrlog	so I'm misunderstanding how to use :resources-path?
14:11	pyr	i'm having a clojure problem that's related to my understanding of java i think
14:11	technomancy	gfrlog: you only set :resources-path if for some reason you don't want to keep your resources in the resources/ dir in the project root
14:12	technomancy	if resources/ exists, it will use it without any configuration in project.clj
14:12	pyr	i have an exception occurring nested within futures and when I catch it and print the full trace
14:12	gfrlog	technomancy: okay. I guess I don't mind moving it.
14:12	pyr	it seems it doesn't go "deep" enough
14:12	technomancy	gfrlog: in fact, resources is explicitly for things that don't belong in src anyway
14:12	technomancy	so src/resources is kind of contradictory
14:12	amalloy	pyr: each thread has its own stack
14:13	pyr	http://pastie.org/2112344
14:13	gfrlog	technomancy: how about this: :dev-resources-path "test/resources"
14:13	pyr	amalloy: the pastie is the function I use to catch the trace
14:13	gfrlog	technomancy: i.e., is there a better place for dev-resources?
14:14	technomancy	gfrlog: same thing; just mkdir test-resources
14:14	technomancy	(will be moved to dev-resources in 2.0)
14:14	gfrlog	technomancy: okay. Cool, thanks.
14:14	amalloy	pyr: you're explicitly catching e and e.getCause
14:14	amalloy	what about the cause of the cause of e?
14:14	technomancy	no need for all this nesting
14:14	amalloy	and so on down the line
14:15	hiredman	,(doc root-cause)
14:15	clojurebot	I don't understand.
14:15	hiredman	clojurebot: jerk
14:15	clojurebot	Gabh mo leithscéal?
14:15	hiredman	,(require 'clojure.stacktrace)
14:15	clojurebot	nil
14:15	hiredman	,(doc root-cause)
14:15	clojurebot	Excuse me?
14:15	pyr	ah ok
14:15	amalloy	when java prints a stacktrace, it explicitly runs through all the nested .getCause's and prints them all. you only do two
14:15	hiredman	clojurebot: for real?
14:15	clojurebot	Gabh mo leithscéal?
14:16	pyr	ok, alright, I didn't get what getCause was, didn't read thoroughly enough
14:16	pyr	so yeah, I should instead loop while getCause returns something
14:16	pyr	and print my trace
14:17	pjstadig	or just use clojure.stacktrace/root-cause
14:17	pjstadig	i think hiredman is trying to tell you that root-cause already does the looping for you
14:17	pjstadig	'(doc clojure.stacktrace/root-cause)
14:17	pjstadig	,(doc clojure.stacktrace/root-cause)
14:18	clojurebot	"([tr]); Returns the last 'cause' Throwable in a chain of Throwables."
14:18	amalloy	pjstadig: but it doesn't
14:18	amalloy	well. it does. but it only returns the last exception, not all the ones on the way
14:18	pjstadig	sure
14:18	pjstadig	but usually the last one is the interesting one
14:19	pyr	thanks for the clarifications anyhow :)
14:19	pyr	i still get lost in java details, my burden for coming from a non jvm background i guess
14:22	pyr	easier than I had feared, thanks to leiningen mostly
14:22	gfrlog	I'm a fan of leiningen
14:33	timvisher	anyone know off the top of their heads how to get clj-time (or joda time) to honor DST?
14:33	choffstein	Hey all. I have developed a jetty app that runs perfectly on my local machine when I run it as a jar, but I would like to deploy it as a war. I've heard this can be done with a few leiningen plugins, but am having some serious trouble figuring out how to package it all up and how to construct my servlet. Anyone here familiar with building and deploying jars that wouldn't mind me asking them a few questions in private?
14:34	choffstein	(only to not clutter this channel with stupid questions)
14:34	timvisher	i thought it did it automatically but i'm seeing incorrect results that are an hour slow
14:34	timvisher	choffstein: go for it, bro. i do nothing but ask stupid questions here and i haven't been banned yet!
14:34	timvisher	'fraid i won't be much use to you, though
14:34	Scriptor	choffstein: don't worrying about cluttering the channel, it's not all that busy
14:34	timvisher	i'm still pre-deployment with my web app
14:35	Scriptor	and other might be interested in the discussion!
14:37	choffstein	Well, I guess the first question would be is: does anyone know of a very, very barebones example of a lein project set up to be deployed as a war using ring?
14:38	choffstein	Every tutorial online seems to do it a different way
14:38	timvisher	choffstein: what about this? https://github.com/weavejester/compojure-example
14:38	choffstein	The second question would be: I use a "public" directory for my static files (css / javascript) in the root directory of my lein project -- any idea how to get this wrapped in to my war file?
14:39	ibdknox	you just need to put it in a resources folder
14:39	timvisher	in fact, is this what you're looking for? https://github.com/weavejester/lein-ring
14:41	choffstein	...wow, that was simple
14:41	choffstein	I need to stop spending 2 hours trying to solve things myself and just come ask you guys
14:41	timvisher	,drinks
14:41	clojurebot	java.lang.Exception: Unable to resolve symbol: drinks in this context
14:41	timvisher	boo
14:41	timvisher	used to #emacs
14:41	timvisher	anyway,
14:41	timvisher	drinks all around!
14:42	timvisher	can anyone help with my clj-time question?
14:42	timvisher	why might it not appear to be honoring DST?
14:48	choffstein	Any idea why ring's warp-file might not be playing nice with the resources folder?
14:48	choffstein	e.g. saying files not found?
14:50	timvisher	not off hand but you could verify through an http watcher that the paths are actually correct
14:51	timvisher	I had issues with utf-8 in my app
14:53	timvisher	unfortunately, i only know ring as it's exposed by compojure
14:53	choffstein	holy shit. it works...
14:53	choffstein	This isn't supposed to work...
14:55	timvisher	heh
14:55	timvisher	that could be a good thing
15:03	timvisher	can anyone help with my clj-time question?
15:03	timvisher	why might it not appear to be honoring DST?
15:04	amalloy	timvisher: not really enough information there for anyone to help, i think? i don't know anything about clj-time, but if i did i don't think i'd know what, specifically, your problem is
15:05	timvisher	let me explain then
15:05	timvisher	i'm playing with clj-time at the repl
15:06	timvisher	I do (now) and it returns, as expected, this instant in GWT
15:06	timvisher	GMT*
15:06	Somelauw	(- \h \a)
15:06	timvisher	except that instead of 20 being returned for (hour (now))
15:06	timvisher	I get 19
15:07	Somelauw	= 7 or something in java. But can I get the same result in clojure without translating to int?
15:07	timvisher	as if DST wasn't in effect at the moment
15:07	jcromartie	Somelauw: no
15:07	timvisher	make sense?
15:07	timvisher	everything I've read about joda time is that it handles DST automagically
15:07	jcromartie	Somelauw: (defn char- [& chars] (apply - (map int chars)))
15:08	jcromartie	and if you want, (defn char- [& chars] (char (apply - (map int chars))))
15:08	jcromartie	to keep it in the char type
15:08	jcromartie	so, (char- \z 1) => \y
15:09	Somelauw	okay, thansk
15:11	ibdknox	timvisher: are you sure you have the locale set correctly?
15:11	timvisher	don't even know enough to tell you
15:12	timvisher	is there a ! function I have to call in clj-time to set that?
15:13	ibdknox	try this:
15:13	ibdknox	(def time-format (tform/formatter "h:mma" (ctime/default-time-zone)))
15:13	ibdknox	whoops
15:13	ibdknox	you'll need to require:
15:13	ibdknox	[clj-time.core :as ctime]
15:13	ibdknox	[clj-time.format :as tform]
15:14	ibdknox	(tform/unparse time-format (ctime/now))
15:14	ibdknox	see if that gives you the correct time
15:16	timvisher	that indeed seems to do it
15:16	timvisher	though it only returns it for GMT
15:16	timvisher	why wouldn't the time-zone-for-offset -5 work, though?
15:16	ibdknox	weird, that returns PST for me
15:16	timvisher	are they returning different things?
15:17	timvisher	actually
15:17	timvisher	i'm calling it wrong
15:19	ibdknox	(tform/show-formatters)
15:19	ibdknox	that will show you every format
15:19	ibdknox	one of which is :local-time
15:20	ibdknox	so you could also do this:
15:21	ibdknox	(tform/unparse (:local-time tform/formatters) (ctime/now))
15:21	ibdknox	I believe
15:24	timvisher	that's strange, my show-formatters doesn't include local-time
15:25	timvisher	so the issue seems to be that the DateTime returned by (time/to-time-zone (time/now) (time/default-time-zone)) is in fact correct
15:25	timvisher	but the call (time/to-time-zone (time/now) (time/time-zone-for-offset -5)) is off by an hour
15:26	timvisher	which doesn't make sense to me because they should be returning the same TimeZone
15:27	ibdknox	hm :local-time is in the source
15:27	ibdknox	what version are you using?
15:28	timvisher	it's strange because it in fact works, but it's not listed when I call the show function
15:29	ibdknox	I see, does it get you what you want?
15:29	timvisher	it actually returns 19
15:29	timvisher	which is very odd
15:29	timvisher	version is 0.3.0
15:30	timvisher	https://gist.github.com/1043415
15:34	ibdknox	I'm not sure, using my custom formatter and setting it to the default-time-zone gives me PST as I would expect
15:35	timvisher	what if you call it with (time-zone-for-offset -8) ?
15:35	timvisher	or whatever the offset is for PST
15:36	ibdknox	it's off by one then
15:36	ibdknox	11:35 instead of 12:35
15:36	timvisher	interesting
15:36	timvisher	maybe i'm misunderstanding the semantics of that call
15:37	ibdknox	well given that DST is a national thing
15:37	ibdknox	not a time-zone thing
15:37	ibdknox	that's what I would expect
15:37	ibdknox	other countries will be in that time zone, so it's giving me the mathematically correct time
15:37	choffstein	Anyone have any experience deploying war files to amazon's elastic beanstalk?
15:37	timvisher	but why then would the default-time-zone return something different than the your time-zone calculated via an offset
15:38	ibdknox	it's the locale
15:38	timvisher	choffstein: have you seen https://github.com/weavejester/lein-beanstalk
15:38	choffstein	jesus, weavejester just has everything
15:38	ibdknox	#<CachedDateTimeZone America/Los_Angeles>
15:38	ibdknox	is what (ctime/default-time-zone)
15:38	timvisher	the man is prolific
15:39	ibdknox	returns
15:39	timvisher	ah
15:39	timvisher	so they're returning different types
15:39	timvisher	that's interesting
15:39	ibdknox	yep
15:40	timvisher	so what do you think the usefulness of time-zone-for-offset would be then?
15:40	timvisher	given that you would have to manually keep track of DST.
15:40	timvisher	i can't think of any reasonable uses for it off the top of my head
15:40	ibdknox	No one else has DST ;)
15:41	timvisher	actually, London is in DST at the moment
15:41	gfrlog	why wouldn't you arbitrarily change your clocks?
15:41	timvisher	there are a number of locations that do this awful thing to us
15:41	timvisher	:)
15:41	ibdknox	lol
15:41	gfrlog	we're demonstrating that the time is whatever we say it is
15:41	ibdknox	it's better to stick to a standard, I think
15:41	gfrlog	if we can't do that it's a slippery slope
15:42	ibdknox	but I'm sure there's a magical way to get all of the DST's to work out
15:42	ibdknox	you'll likely need to look more into Joda to figure that out
15:44	VT_entity	hey, all. Anyone have any experience with "seque"
15:44	choffstein	okay ... very specific question. When using "lein ring war", I keep getting a "ZipException: duplicate entry" error with my servlet.clj file. Anyone know why that may be happening? Is it a name clash?
15:44	choffstein	"lein war" works fine...
15:44	gfrlog	that's weird...
15:45	VT_entity	I'm trying to peek and pop off of a seque, but I think I'm doing it wrong, and the documentation is really slim.
15:45	davekong	I spent some time trying to figure out how to quickly load a 120MB binary file (unsigned ints) into an array of longs and eventually came up with code that works but it still pretty slow (90s on my machine). I am wondering if there is an obvious way to speed this up by a factor of ten or so. http://www.pastie.org/2112719
15:45	timvisher	gfrlog: i'm already convinced that you should store only UTC times
15:45	gfrlog	choffstein: do you have a resources directory?
15:45	timvisher	i'm now concerned with display
15:45	gfrlog	timvisher: amen
15:45	timvisher	that's where i started running into the issue we've been discussing
15:45	choffstein	gfrlog: yep
15:46	gfrlog	choffstein: it's just "/resources"?
15:46	gfrlog	not anywhere weird?
15:46	choffstein	gfrlog: Exact error: "Exception in thread "main" java.util.zip.ZipException: duplicate entry: /WEB-INF/classes/portal/servlet.clj (NO_SOURCE_FILE:0)"
15:46	davekong	to-byte-array from contrib.io is very fast, but it is unclear to me how to write a to-long-array equivalent
15:46	timvisher	it looks like the way to set it up will be to store the long "America/New_York" time-zone in the customer account rather than the offset, which is what i was hoping for
15:46	choffstein	yep, nothing weird. "lein ring server" works fine.
15:47	gfrlog	choffstein: Okay. I had a strangely similar problem earlier today that technomancy helped me figure out. Sounds like it's probably unrelated though.
15:48	choffstein	ah, gotcha. thanks for trying anyway :)
15:49	timvisher	choffstein: i'm sure you've checked that you don't actually have 2 servlet.clj files?
15:52	choffstein	as far as I can tell, I don't :)
15:52	timvisher	just making sure
15:52	timvisher	how familiar are you with jarring from the commandline?
15:52	timvisher	could you try to do it outside of lein?
15:52	timvisher	and see if you still get the same problem
15:53	choffstein	not hugely familiar. i've always been a c++ / ruby guy
15:53	timvisher	or maybe see if any of the various and sundry IDEs out there support a lein project and try exporting?
15:54	timvisher	hmm
15:54	timvisher	fraid I'm entirely out of my depth at this point
15:54	timvisher	i only ever built wars 'by hand' using the ant task
15:54	timvisher	and i believe lein is backed by maven
15:54	timvisher	hope someone can help
15:54	amalloy	timvisher: lein is backed by maven and ant
15:54	choffstein	it's alright. i'll figure it out
15:55	amalloy	maven for dependency resolution, ant for building, iiuc
15:57	choffstein	hmmm ... looks like "lein ring uberjar" is expecting everything to be in WEB-INF/ ... that could be a problem
15:57	timvisher	ah good ol `wtf`
15:57	timvisher	shame it doesn't work well with M-\|
15:58	choffstein	hmmm ... or maybe not. gotta dig deeper into the source.
15:58	choffstein	"use the source, choffstein..."
16:06	choffstein	hmm ... just changing my servlet.clj to another file name and changing all my servlet references to that new name in the project and web.xml files made it work. That is very ... very odd.
16:09	timvisher	choffstein: just had a though
16:09	timvisher	thought*
16:10	timvisher	it's possible that ring uberjar is generating a servlet.clj file for you
16:10	timvisher	auto'magically' as it were
16:10	timvisher	and then yours is conflicting
16:10	timvisher	are you sure there isn't an alternative to get things into the servlet.clj file?
16:11	choffstein	yeah, that is what I am guessing (hence why I tried changing it) -- but i'm not seeing that in the source
16:11	timvisher	maybe passing a map in project.clj or something?
16:11	choffstein	well, it works fine now. I just had to rename and remap. it's just strange that it seemed to be a restricted name -- but i'm not seeing that in the lein-ring source code
16:12	timvisher	you could always ask weavejester on github
16:13	timvisher	he's ordinarily quite responsive in my experience
16:13	timvisher	in fact, i just realized i've been assuming he's a guy this whole time
16:13	timvisher	hmm
16:14	amalloy	timvis, her
16:15	timvisher	:)
16:15	choffstein	on the internet, we're all just dudes pretending to be women
16:16	timvisher	ah the days of aol chat rooms, they were different times
16:16	timvisher	better yet, compuserve chat rooms
16:16	timvisher	but then again, here i am in irc
16:16	timvisher	wondering whether weavejester is a guy or a girl
16:16	timvisher	i hate clearcase
16:16	timvisher	...
16:16	Somelauw	timvishster
16:18	gfrlog	timvi-she-r
16:19	gfrlog	I think I got lucky. I had no idea before making the comment that his named had both "she" and "her" as substrings.
16:19	gfrlog	aw crap I just assumed timvisher was a dude
16:20	timvisher	gfrlog: there you go assuming!
16:20	timvisher	you know when you assume you make an ass out of all of us.
16:20	gfrlog	no that's what happens when you assallus
16:24	OscarZ	hi.. what's the relation between Clojure and Scala ? Both seem to be FP languages on top of JVM.. are there some specific issues with Scala where Clojure wanted to go other way?
16:25	OscarZ	I'm just trying to get an overview of the FP languages that are available..
16:26	technomancy	OscarZ: the syntaxes and type systems are the most obvious differences. also scala is hybrid OO/FP while Clojure is "just" FP.
16:27	Somelauw	They are both functional but differ a lot I think.
16:27	__name__	Clojure is lispy, Scala is not at all.
16:28	timvisher	OscarZ: I'd recommend googling. Reams have been written on this subject
16:28	timvisher	i think __name__ has the most fundamental difference though
16:29	gfrlog	(inc __name__)
16:29	sexpbot	⟹ 1
16:29	__name__	Yay, I am 1 now.
16:29	Somelauw	I think clojure is easy to compile but might not always work the first time. Scala is really hard to compile, but often works the first time.
16:29	gfrlog	Somelauw: that sounds like static vs. dynamic
16:29	technomancy	gfrlog: (swap! ~name inc)
16:30	OscarZ	Ok.. I'm a Java programmer trying to learn some FP.. I've been looking at Haskell a bit before
16:30	gfrlog	(swap! technomancy inc)
16:30	gfrlog	technomancy: what I meant is I agree :)
16:30	technomancy	though I've always been partial to dinc
16:30	gfrlog	d for what?
16:30	technomancy	double
16:30	gfrlog	phonetics?
16:30	technomancy	(def dinc (comp inc inc))
16:31	gfrlog	ah ha
16:31	timvisher	OscarZ: fwiw, i chose Clojure over Scala in your situation because I knew Clojure would push me pretty strongly towards FP where Scala seemed much more like Groovy with FP tacked on
16:31	technomancy	very extreme situations may call for oinc
16:31	__name__	oinc?
16:31	gfrlog	(apply comp (take 8 (repeat inc)))
16:31	timvisher	it's not impossible or even terribly difficult to write non-functional code in Clojure, but the whole language pushes you that way
16:31	timvisher	whereas Scala is far more agnostic
16:31	hiredman	e.g. if you have an off by one error and are missing a long
16:31	technomancy	gfrlog: exactly
16:32	timvisher	it seems like the 'best' Scala programmers write mostly FP, but it takes awhile to get there
16:32	technomancy	clojurebot: oinc is octo-inc: (apply comp (take 8 (repeat inc)))
16:32	clojurebot	Ik begrijp
16:32	amalloy	Somelauw: heathen. clojure code works the first time, almost always. it's so functional that you don't have to be careful with bookkeeping
16:32	OscarZ	timvisher: Ok I see, I'd really like to learn the FP way of thinking and I guess its better to go with more "pure" FP .. that's why I looked at Haskell first
16:32	gfrlog	timvisher: I've heard some java programmers describe doing similiar things in java
16:32	gfrlog	just with less closures I guess
16:32	timvisher	gfrlog: true
16:32	OscarZ	but find it too big leap to actually start using Haskell in some real work where the existing code base is in Java
16:32	timvisher	I've heard that if you really want to push/punish yourself, you go Haskell
16:33	timvisher	But it's also been a life long dream of mine to learn lisp so
16:33	timvisher	..
16:33	timvisher	ah
16:33	gfrlog	clojurebot: oinc?
16:33	clojurebot	oinc is octo-inc: (apply comp (take 8 (repeat inc)))
16:33	timvisher	well yeah, if you need to integrate on the JVM you basically should choose between Scala and Clojure
16:33	amalloy	anyone have a link to fogus's 140-character obfuscated-oinc tweet?
16:34	OscarZ	But I understand Haskell is really elegant language in terms of being "pure FP", so I guess thats a good place to learn at least some basic concepts
16:34	timvisher	yeah
16:34	timvisher	I liked Channel9s Intro to FP by Meijer
16:34	gfrlog	OscarZ: as somebody with almost no experince in haskell, I agree
16:34	timvisher	,google Channel9 Intro to Functional Programming Meijer
16:34	clojurebot	java.lang.Exception: Unable to resolve symbol: google in this context
16:34	timvisher	,,google Channel9 Intro to Functional Programming Meijer
16:34	clojurebot	java.lang.Exception: Unable to resolve symbol: google in this context
16:34	timvisher	bah
16:34	amalloy	(defn oinc "fourteen" [n] (reduce + n (map #(%) (repeat (.length ((keyword (apply str (map char [100 111 99]))) (meta (var oinc)))) #(*)))))
16:34	timvisher	how do I do that
16:34	gfrlog	use $google with sexpbot
16:34	amalloy	$google hello
16:34	sexpbot	First out of 118000000 results is: Hello
16:34	sexpbot	http://www.hellomagazine.com/
16:34	gfrlog	maybe ~google works for clojurebot
16:34	timvisher	$google Channel9 Intro to Functional Programming Meijer
16:34	sexpbot	First out of 978 results is: Erik Meijer's MSDN Channel 9 lecture series on functional ...
16:34	sexpbot	http://lambda-the-ultimate.org/node/3642
16:35	amalloy	gfrlog: yes, i think so
16:35	gfrlog	~google sexpbot
16:35	clojurebot	First, out of 149 results is:
16:35	clojurebot	Introduction to sexpbot » Bathroom Reading Material
16:35	clojurebot	http://blog.raynes.me/?p=82
16:35	timvisher	looks about right
16:35	OscarZ	thanks, I'll have a look on that video
16:37	timvisher	but honestly, I learned more about FP by watching through the Clojure blip.tv channel and reading through The Joy of Clojure than anywhere else, probably
16:37	timvisher	at least to start off with
16:37	timvisher	not to mention getting ad-hoc code reviews here in this fine channel
16:37	timvisher	^_^
16:37	OscarZ	ok :)
16:40	choffstein	anyone have any experience using lein-beanstalk?
16:41	timvisher	god i am a nerd...
16:41	__name__	gfrlog: That must have been hard.
16:42	gfrlog	__name__: I was just thinking how weird that was in hindsight. Learning about the actor stuff at the same time.
16:42	__name__	Erlang is painful I guess.
16:42	__name__	Its syntax is just, well.
16:42	nishant	can anyone here help me with clj-http?
16:42	gfrlog	it definitely is. I found it fun in that kind of painful way that writing assembly code is also fun.
16:42	__name__	I find C fun in that way.
16:43	gfrlog	yeah. I still think it is a strange mix of high-level and low-level
16:43	gfrlog	kind of exotic I guess
16:43	amalloy	~anyone
16:43	clojurebot	Please do not ask if anyone uses, knows, is good with, can help you with <some program or library>. Instead, ask your real question and someone will answer if they can help.
16:43	drewr	b
16:44	gfrlog	drewr: c
16:45	drewr	sry little slip of the finger there
16:46	gfrlog	:)
16:50	amac	Clojurebot seems cranky
16:50	amac	~botsnack
16:50	clojurebot	Thanks! Can I have chocolate next time
16:52	amalloy	that's new
16:55	gfrlog	clojurebot: botsnack?
16:55	clojurebot	Thanks! Can I have chocolate next time
16:55	gfrlog	it was a question.
17:01	ibdknox	does anyone know of an alternative to clojure.contrib.find-namespaces? I'm trying to remove my dependency on contrib in Noir
17:03	ibdknox	nm, tools.namespace appears to be the replacement
17:09	Raynes	ibdknox: Where are you using it in noir?
17:10	ibdknox	to load a dir's worth of views
17:10	ibdknox	as opposed to individually requiring them
17:10	ibdknox	noir.server/load-views
17:18	ibdknox	Raynes: thoughts?
17:21	Raynes	ibdknox: I was just wondering how you were using it. I thought I was going to need it for something myself recently, but found all-ns. Looks like you actually need it though.
17:22	ibdknox	Raynes: ah, yes. I think in this case it's the way to go.
17:23	bdesham	off-topic, but does anyone know of a C++ equivalent of marginalia?
17:24	ibdknox	bdesham: not off the top of my head, but it's based on Docco, so searching for a C++ docco might get you somewhere
17:24	bdesham	ibdknox: great, thanks
17:25	stuartsierra	noweb is a language-agnostic version of Knuth's original literate programming thing.
17:26	bdesham	stuartsierra: will look into that too, thanks
17:33	sot	Hi. Is using future the same as creating a thread in java? I have a few background processes that I need to spawn, and am using future, but am seeing some erratic behaviour.
17:35	stuartsierra	Futures use the same thread pool as send-off.
17:37	stuartsierra	So each future gets its own thread, but the thread pool is allow to cache and reuse the thread after the future completes.
17:37	stuartsierra	gotta go
17:38	sot	I'm using zeromq and have a number of event loops that listen and send on sockets, and wondering if using future to start these loops is correct.
17:47	hugod	anyone know if something like contrib.condition will be in 1.3?
17:48	technomancy	hugod: it won't be in 1.3, but http://github.com/scgilardi/slingshot is the successor of c.c.condition and is awesome.
17:49	hugod	technomancy: thanks, missed that
17:49	technomancy	hugod: it hasn't been widely publicized
17:55	bsteuber	where can I put init stuff for the user-namespace like (use 'clojure.repl) with leiningen/swank setup?
17:55	technomancy	put :repl-init foo.repl in project.clj where foo.repl is the namespace
17:56	bsteuber	yeah but I have like 10 projects
17:56	bsteuber	and would prefer a global solution
17:57	technomancy	hm... you could do it with a hook in ~/.lein/init.clj, but I guess that should be standardized. open an issue?
17:57	bsteuber	I expected .clojure/user.clj or swank-clojure-init-files to work, but neither did with this setup
17:57	bsteuber	ok I'll do that :)
17:58	hiredman	you can create a project where you keep your stuff
17:58	hiredman	and add it as a checkout to all your other projects
17:59	bsteuber	checkouts are nice in general, just stumpled across them
17:59	bsteuber	technomancy: could you elaborate on the exact hook I'd need?
18:01	technomancy	bsteuber: off the top of my head: (add-hook #'leiningen.compile/get-readable-form (fn [orig j project form init] (orig j project form `(do ~init (load-file "/home/me/.lein/user.clj")))))
18:02	technomancy	basically expanding the init arg to every eval-in-project invocation
18:02	bsteuber	great, I'll try that (and still file a ticket) :)
18:16	choffstein	anyone here have any experience distributing clojure web projects as .war files?
18:18	bsteuber	choffstein: got it working once or twice - what's your problem?
18:19	hiredman	choffstein: lein ring is nice
18:20	choffstein	"lein ring server" seems to work fine, but when I create a .war using "lein ring uberwar" and upload it to amazon's aws beanstalk, it doesn't seem to work at all. I was wondering if there is a good way to locally test the war file created by lein ring without having to install apache and tomcat.
18:21	bsteuber	I think I used jetty for that
18:21	hiredman	choffstein: you've seen lein-beanstalk?
18:22	choffstein	yeah -- lein-beanstalk doesn't seem to work for me
18:23	choffstein	ah, I'll try jetty
18:23	bsteuber	technomancy: great, just added needed use/require clauses and it works!
18:24	bsteuber	so thanks a lot
18:27	timvisher	choffstein: installing tomcat or jetty and running a war file is dead simple
18:27	timvisher	it might be worthwhile deploying it locally first just to test it.
18:27	choffstein	I'm about to try jetty
18:27	timvisher	nice
18:29	technomancy	bsteuber: if you want it nicer in leiningen 1.6, you could submit a patch =)
18:30	bsteuber	ok I'll try :)
18:30	technomancy	maybe a leiningen.core/add-init macro that could swap a form into an atom that leiningen.compile/get-readable-form could use?
18:31	bsteuber	why not just support .lein/user.clj?
18:31	bsteuber	and do what you did with your hook when it exists
18:32	bsteuber	that's what I'd do without knowing better :)
18:32	technomancy	yeah, no reason to have an atom really. just be sure to honor leiningen.util.paths/leiningen-home instead of hard-coding ~/.lein
18:33	bsteuber	good
18:40	sjl	So for my first non-toy project in Clojure I'm working on a simple Minecraft bot. If anyone has a few minutes to look at my code so far and tell me how to clean it up and make it more idiomatic, I'd love it.
18:40	sjl	https://bitbucket.org/sjl/clojurecraft/src/c6c71945637c/src/clojurecraft/core.clj
18:40	sjl	There's a lot of repetition that I think could probably be removed.
18:41	the-kenny	First thought on write-packet: use case, this does a constant-time lookup
18:42	the-kenny	Or implement this with multimethods, this would make it extensible (probably unnecessary)
18:43	the-kenny	Also, you can pass assoc more than one key value pair: (assoc m :foo 42 :bar 23) will assoc :foo and :bar
18:43	sjl	the-kenny: wow, I completely missed case. Thanks!
18:43	sjl	Yeah, I knew that, but now that you mention it I can clean up those ugly assoc chains with it too
18:43	the-kenny	Oh, and when you just want to populate a map (like in all the read-stuff), just use the map-syntax
18:44	the-kenny	{:foo 23, :bar 42} (the , is optional)
18:44	the-kenny	This removes the need for most assoc and ->
18:45	sjl	the-kenny: well, about that, is it guaranteed to parse the vals in the order I list them?
18:45	the-kenny	Oh, I haven't thought of that. Not sure.
18:45	sjl	the-kenny: the -read-whatever's need to be executed sequentially because the bytes are sent in a specific order
18:45	the-kenny	I understand
18:45	sjl	That's why I did the -> thing, I wasn't sure and just wanted to get a player on the screen, hah
18:46	sjl	the-kenny: same deal with the multiple-arg assocs
18:46	the-kenny	I'm pretty sure the order of multi-arg assocs is guaranteed
18:47	sjl	hmm
18:47	the-kenny	If order is important and if you don't want the whole assoc stuff to be gone, you could also do something like (zipmap [(-read-int stream) ...] [:eid ...])
18:48	bsteuber	or write it like (into {} [[:first (do-somehting)] [:second ...
18:48	sjl	mm, I think I'd rather keep the pairs together for readability
18:48	the-kenny	bsteuber's suggestion is more readable :)
18:49	sjl	but is the evaluation order of the vector of pairs guaranteed?
18:49	sjl	(same question as the eval-order of map literals, really)
18:49	bsteuber	yes for vectors it is
18:49	bsteuber	but I'd still just use one big assoc :)
18:51	sjl	ok, according to http://clojure.org/evaluation it looks like function arguments are guaranteed to be eval'ed left to right
18:51	sjl	so one big assoc should work
18:52	choffstein	egads. I thought this would be easy. jetty doesn't seem to like my war file. can't find "web-app" description ... and yet, there it is in WEB-INF/web.xml...
18:52	sjl	it still feels... dangerous, though.
18:53	amalloy	sjl: it's guaranteed. doesn't get any safer than that :P
18:53	the-kenny	sjl: You can always take a look at the implementation of assoc :)
18:53	sjl	Heh, I'm not worried about now -- I'm worried about a few months from now when the implementation of assoc changes :)
18:54	rkbuchmann	That would mean that evaluation in general would change
18:54	rkbuchmann	and that is very unlikely
18:55	amalloy	sjl: can you point me at a place where you're worried about order of evaluation? there's a lot of code there and i don't want to scan through all of it
18:56	amalloy	oh, read-packet-login
18:56	the-kenny	every read/write operation :)
18:56	sjl	amalloy: Yeah, any of the read-packet-whatever ops
18:57	sjl	rkbuchmann: Yeah, seems like a pretty significant change, so I guess it's safe to assume it won't happen any time soon.
18:59	choffstein	okay, can anyone with some jetty experience give me a bit of help? I am having some trouble figuring out if there is something wrong with my .war, or if there is something wrong with my jetty setup...
18:59	bsteuber	I'm sure order of function evaluation is guaranteed in the "spec"
19:00	amalloy	sjl: fwiw, the current compiler does eval map literals in order
19:00	amalloy	i don't know whether it's guaranteed
19:00	sjl	amalloy: Hmm, I think I'll stick with assoc. Not much more verbose, and the wiki says it's safe.
19:00	bsteuber	ah right that should be okay too :)
19:00	bsteuber	just after the map is constructed the order can change
19:01	amalloy	yes
19:02	choffstein	damn you jetty. you should not be this hard to figure out. what the crap.
19:03	bsteuber	probably it's just some simple thing you need to do different
19:04	bsteuber	did you check your web.xml for syntax errors?
19:08	choffstein	yep. Nothing that i can see.
19:08	choffstein	I tried turning the warnings off in jetty, but get a whole new slew of problems
19:08	choffstein	"sealing violation"
19:09	bsteuber	might help to start from some example project on github
19:09	choffstein	seems I am hunting an endangered species...
19:09	bsteuber	which of course I can't find right now :)
19:09	amalloy	(inc choffstein)
19:09	sexpbot	⟹ 1
19:09	choffstein	eh?
19:10	bsteuber	(inc choffstein)
19:10	sexpbot	⟹ 2
19:10	bsteuber	ok this is for polls
19:10	bsteuber	thought some nice way to insult people via sexpbot
19:11	bsteuber	at least "you're a null" is insulting in German
19:12	amalloy	choffstein: seal-hunting made me laugh. so you get a sexpbot karma credit
19:12	choffstein	ah, gotcha
19:12	bsteuber	(dec choffstein)
19:12	sexpbot	You want me to leave karma the same? Fine, I will.
19:12	amalloy	!
19:12	bsteuber	lol?
19:12	amalloy	that sounds broked
19:13	rbuchmann	(doc inc)
19:13	clojurebot	"([x]); Returns a number one greater than num."
19:13	rbuchmann	bah
19:13	amalloy	$karma choffstein
19:15	amalloy	i'm probably the one who broke that, but it's not clear to me how
19:15	amalloy	(dec Raynes)
19:15	sexpbot	⟹ 4
19:15	amalloy	interesting. dec works
19:15	Raynes	(inc amalloy)
19:15	sexpbot	⟹ 10
19:15	amalloy	funny funny. cause i'm the one who should be getting dec'd for breaking shit :P
19:16	rbuchmann	(inc amalloy)
19:16	sexpbot	⟹ 11
19:18	choffstein	war files are confusing
19:18	rbuchmann	war is hell my friend... :)
19:18	choffstein	it seems war files + static resources (not served) + ring = no no
19:19	choffstein	I am trying to use files that don't get served, but rather get read by my webserver. I am finding out that such a situation doesn't really work so well...
19:19	rbuchmann	hm, but surely that must be a pretty common scenario...?
19:19	iwillig	hi guys, is there a way i can look up a function in a namespace by a string, i am thinking of something like python's getattr
19:19	choffstein	one would think
19:20	choffstein	this is getting rather frustrating. why does it work in "lein ring server" but not as a war under jetty. What in the world am I doing wrong...
19:20	Raynes	iwillig: look at resolve and ns-resolve. Might be what you're after. I don't use Python though, so don't hold me to it. ;)
19:21	hiredman	strip your project down to a single handler function and if that works slowly build it back up and see where it breaks
19:23	choffstein	hiredman: yeah ... that sounds horrible.
19:23	choffstein	hiredman: but unfortunately is looking like the way I will have to go...
19:24	hiredman	choffstein: look, if you know how to read stack traces and are familiar with the jvm you most likely won't have to, but if you aren't what other option is there?
19:24	choffstein	...I can read stack traces, thanks. these aren't stack trace issues.
19:25	choffstein	This is trying to figure out why jetty isn't liking my war file
19:25	hiredman	shrug I don't know what more you want, you just complain about jetty over and over
19:26	hiredman	no pastebins with errors, no explantions of this or that, just "I have this project, it doesn't work" for the last few hours
19:27	Raynes	Man, there are better ways to go about requesting more information.
19:27	hiredman	people compulsively click on urls, so it is very easy to get people to look at errors, logs, etc
19:27	choffstein	hiredman: fair enough. I'd love to give a pastebin with an error and a stack-trace, but there isn't one.
19:27	hiredman	then how do you know it isn't working?
19:27	choffstein	hiredman: the stacktrace is in jetty and only goes down into the configuration levels of jetty (e.g. the XML parser)
19:28	hiredman	choffstein: that is a stacktrace
19:28	choffstein	hiredman: so the only relevant part is the top level exception, which I have already asked about
19:28	choffstein	Basically, I'm not an idiot. If I had more information to give, trust me, I would.
19:28	hiredman	well then, strip it down and rebuild it then
19:28	hiredman	no, you have tons of information
19:29	hiredman	you have namespace names, you have files, you have bytecode you have xml
19:29	hiredman	you have everything
19:29	choffstein	hiredman: And sometimes, I can't give everything because they are proprietary projects. But the issues I am running into are generic enough and are about setting up your application to be appropriately distributed as a war file, which I am unfamiliar with.
19:30	Raynes	choffstein: gist that stacktrace.
19:30	choffstein	Raynes: the jetty one?
19:30	bsteuber	if need the help so badly, put a minimal non-working setup on github
19:30	Raynes	Sure. Whatever stacktrace you're getting.
19:30	bsteuber	there's you missing somewhere
19:30	hiredman	^-
19:30	bsteuber	"a you"
19:30	bsteuber	whatever :)
19:30	Raynes	He seems fairly certain that your stacktrace is the key to infinity and beyond, so gist it and we can see.
19:31	choffstein	https://gist.github.com/1043893
19:31	Raynes	Before this channel turns into ##c
19:33	Raynes	choffstein: Furthermore, that is an excellent idea. If you can get a minimal example of what isn't working for people here to see, it'll be easier to debug the problem.
19:34	choffstein	Raynes: I'm doing it as we speak. Hopefully I'll have a barebones up in a few minutes.
19:34	hiredman	choffstein: what version of jetty?
19:34	Raynes	Cool. :)
19:34	choffstein	5.1.10
19:34	hiredman	very old
19:34	choffstein	installed it from macports.
19:34	hiredman	ugh
19:34	hiredman	never install jars via system package managers
19:35	choffstein	didn't know it was going to come as a jar.
19:35	choffstein	didn't know anything about jetty until a couple hours ago
19:35	hiredman	http://blogs.webtide.com/janb/entry/jetty_runner jetty-runner is what you want
19:35	hiredman	get the latest version, you'll be fine
19:36	choffstein	alright, thanks.
19:36	bsteuber	install all java libraries with leiningen
19:37	bsteuber	jarvana.com is great for finding the maven names
19:37	hiredman	I guess you could jetty-runner via lein, but I would just download the jar as needed
19:37	hiredman	java -jar jetty-runner.jar some.war
19:38	bsteuber	sorry, raynes :)
19:38	Raynes	<3
19:38	bsteuber	you mean for confusion with lein install?
19:39	Raynes	Maven blows away any common understanding of what 'install' means to people. jars are copied around and such.
19:40	Raynes	It's just confusing is all.
19:40	bsteuber	:)
19:43	bsteuber	good night everyone
19:44	bsteuber	or day or whatever :)
19:44	amalloy	~ugt
19:44	clojurebot	ugt is Universal Greeting Time: http://www.total-knowledge.com/~ilya/mips/ugt.html
19:45	bsteuber	haha, thanks, I'll keep that in mind :)
19:47	pdk	holy shit
19:47	pdk	joy of clojure in mobi/epub
19:47	Raynes	Calm down.
19:47	pdk	but god descended man
19:48	technomancy	pdk: that was like two thousand years ago
19:49	choffstein	Ah, fuck. Okay ... so jetty-runner makes my war file run perfectly.
19:49	pdk	i think jesus only had lisp 1.5 then though
19:49	hiredman	technomancy: so not available in epub?
19:49	choffstein	that's ... awesome and horrible at the same time
19:49	choffstein	Thanks for thinking of the version issue hiredman
19:49	hiredman	choffstein: right the xml lein ring
19:49	hiredman	generates was too new
19:52	choffstein	ahhh, that makes sense
19:52	choffstein	hmm. now all of a sudden I have to wonder why my war isn't working with the amazon beanstalk :D but that is a whole new can of worms....
20:30	choffstein	Anyone try using lein-beanstalk yet?
22:27	justinko	what's the best way to install aquamacs on os x?
22:30	dnolen	justinko: I recommend Cocoa Emacs 23 unless you really, really, realy don't want to learn standard Emacs bindings.
22:32	justinko	dnolen: aquamacs seems so popular, even Hickey uses it
22:34	justinko	dnolen: okay I see your point http://stackoverflow.com/questions/1096009/carbon-vs-aqua-vs-cocoa-emacs-builds
22:36	dnolen	justinko: acquamacs is fine, but I've run into weird edge cases here and there which made me abandon it.
22:38	Scriptor	any word on what the theme for clojure/conj would be?
22:39	cemerick	I didn't think there was to be a theme.
22:40	Scriptor	it mentions on the website how they want to "weave a common theme", apparently there was one last year?
22:42	justinko	dnolen: this how you installed cocoa emacs? http://emacsformacosx.com/
22:42	dnolen	justinko: yes
23:07	jamiltron	What is the easiest way to read a single character from the console? I have tried monadic-io-streams' read-char but that seems to always give me back an instance of the function no matter how I use it.
23:18	technomancy	handy little VM setup courtesy of justinlilly: https://github.com/Seajure/emacs-clojure-vagrant
23:18	technomancy	just in case M-x clojure-jack-in isn't easy enough...
23:19	justinko	technomancy: I'm about to buy your peepcode screencast
23:20	justinko	quick question: you guys map control to caps-lock?
23:20	justinko	like VIM users do?
23:20	technomancy	justinko: absolutely
23:21	Scriptor	I mapped escape to caps lock :)
23:21	justinko	Scriptor: so then you use control...?
23:22	dnolen	justinko: I got sick of Emacs pinky, control barely helps. Control -> Command, Meta -> Option
23:22	justinko	my pinky would break off
23:22	dnolen	thumbs are your strongest fingers spare your pinkies.
23:22	justinko	yeah command is fine
23:23	justinko	option is as bad as pinky though...
23:23	Scriptor	justinko: one advantage of windows on a macbook, left thumb always on the command button
23:23	Scriptor	eh, I just move my thumb a bit
23:23	gko	Is/will there be a unified Clojure/Common Lisp SLIME version?
23:24	dnolen	justinko, Many interesting Meta commands involve chords, no need for pinky on option.
23:24	justinko	if the left thumb get's tired of command, you can use the other! (on mac)
23:24	justinko	dnolen: yeah u use the ring finger for option, right?
23:25	technomancy	gko: there could be in the future if 0) the slime devs are open to it and 1) someone from Clojure land cares enough to coordinate with them.
23:25	technomancy	but 1) is probably not going to be me
23:25	justinko	dnolen: ring finger isn't bad
23:25	dnolen	justinko: depends really. I have a Kinesis, so I think my chording patterns are influenced by that.
23:26	islon	why I cannot do this? (apply and [true true false])
23:26	dnolen	justinko: with Kinesis it's trivial to use both thumbs and a finger on another key at the same time which is hard on standard keyboards.
23:29	justinko	dnolen: damn you going to make me buy a kinesis? :)
23:29	dnolen	justinko: heh, no. I really only use that at work.
23:30	dnolen	it's just influenced my typing to be more thumb-centric and less obsessed about moving off home keys.
23:32	symbole	islon: I believe it's because and is a macro? I'm probably wrong.
23:32	dnolen	islon: symbole: yup, and is a macro.
23:33	dnolen	,(macroexpand '(and true false false))
23:33	clojurebot	(let* [and__3468__auto__ true] (if and__3468__auto__ (clojure.core/and false false) and__3468__auto__))
23:36	justinko	dnolen: wait a minute, you maintain textmate-clojure, yet you use emacs?
23:37	islon	yeah i know
23:38	dnolen	justinko: barely maintain. I was hoping more people would step up to help me since it's so simplistic (thanks to cake). Oh well, the TextMate people seem to use and like it well enough.
23:38	islon	but i want to apply and to a seq
23:39	dnolen	,(every? [true false false])
23:39	clojurebot	java.lang.IllegalArgumentException: Wrong number of args (1) passed to: core$every-QMARK-
23:39	cemerick	islon: use (every? boolean [true true false …])
23:39	dnolen	,(every? identity [true false false])
23:39	clojurebot	false
23:40	dnolen	,(every? identity [true true true])
23:40	clojurebot	true
23:41	gko	technomancy: OK... but how do devs using both CL and Clojure manage this? My solution is to define a function slime, which will ask me if I want to use CVS (for CL) or ELPA (for Clojure) version and load the appropriate version... But then, I can't use the other version without side effects (after reloading) in the same Emacs session...
23:41	islon	cemerick: thanks!
23:42	cemerick	islon: dnolen's is semantically the same, FWIW
23:42	dnolen	gko: I think the CL / Clojure crossover is small, thus the state of things.
23:42	islon	ok
23:43	dnolen	personally I just use LispWorks Personal Edition if I'm messin' w/ CL. Nicer than Slime anyhow and you still get all the Emacs commands.
23:47	gko	dnolen: I use SLIME with LW Pro with slime-connect, because I can't have the listener and editor in the same window.
23:50	dnolen	gko: I see. Yeah that stinks. The more CL people use Clojure, hopefully the more the SLIME folks care to keep things compatible. I remember I tried in the early Clojure days at running both, but it was painful and I gave up.
23:51	symbole	Wasn't somebody working on an editor for Clojure?
23:51	dnolen	symbole: I think there's been a couple of attempts but nothing with legs.
23:52	symbole	Must be a diffucult project, considering that it has to be quite extensible.
23:56	cemerick	symbole: as in, a reimagination of emacs but using Clojure instead of elisp?
23:57	dnolen	symbole: I think a full fledged extensible editor would be too ambitious. Something lightweight, like text editor + REPL would be welcome for newcomers. DrRacket's editor is a good reference point.