#clojure – 2012-08-15

2012-08-15

00:34	tolstoy	Is there a way to shut down Emacs (when slime is attached to swank) without killing the JVM and/or causing an exception in Swank?
00:35	tolstoy	,sayonara and ,quit cause swank to shut down the JVM, so that's not it….
00:35	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: sayonara in this context, compiling:(NO_SOURCE_PATH:0)>
00:35	tolstoy	Oops. Sorry clojurebot.
00:41	amalloy	tolstoy: the only way i know is to start a swank server independent of emacs (ie, lein-swank+slime-connect, rather than clojure-jack-in)
00:41	amalloy	then you can just kill the swank buffer to disconnect
00:42	tolstoy	amalloy: That's what I've done. "lein trampoline run" and in the start up sequence I start up swank (just for fun).
00:42	shawnlewis	any opinions on clojure (sql) database migration and query libraries?
00:43	tolstoy	When I kill emacs and hit "yes" when it asks me to terminate that connection, the app throws a stack trace, but still works. So there's that.
00:44	amalloy	tolstoy: you're doing something weird, then. i do the same thing and it works fine
00:44	tolstoy	Okay. Good to know.
00:44	shawnlewis	looks like lobos and korma were good options, but no longer work together due to https://github.com/budu/lobos/issues/43
00:45	amalloy	the startup code looks like (do (stuff) (swank/start-server :host host :port port) (stuff))
00:46	tolstoy	Yeah, that's what I have, except no :host param.
00:47	tolstoy	(defn- start-swank() (log/info "starting swank") (swank/start-server :port 4005))
00:48	tolstoy	exception in control loop java.lang.InterruptedException java.lang.Exception: Error reading swank message
00:49	tolstoy	Oops, that's a [] not (). Anyway, probably just some mismatch somewhere along the line.
00:51	tolstoy	Hm. Maybe I need slime-clj.
00:51	amalloy	NOOO
00:51	amalloy	so old and evil
00:51	tolstoy	Ah, ok. Phew!
00:54	xeqi	tolstoy: I would get a stack trace when I closed emacs as well, can't remember if M-x slime-disconnect would cause it
00:55	amalloy	you're talking about closing all of emacs, not just disconnecting? i dunno then; i never close emacs
00:56	tolstoy	xeqi: Yeah, slime-disconnect gives me the same stack trace in the JVM/swank process.
00:56	tolstoy	The st happens regardless. But "slime-disconnect" was what I was looking for.
00:57	amalloy	https://gist.github.com/3356076 to fix the bug that allows emacs to stop running
00:58	xeqi	shawnlewis: theres also https://github.com/weavejester/ragtime for migrations
00:58	tutysra_	can i tell whether a project should use lein2 or can be used with lein1 by looking at the project.clj file
00:58	xeqi	though in clojars technomancy just made his own script for us https://github.com/ato/clojars-web/blob/master/src/clojars/db/migrate.clj
00:59	xeqi	tutysra_: some have a :min-lein-version, :source-path vs :source-paths is a way, :profiles are lein2 only
01:00	xeqi	there are other keys, but those seem to be the main ones
01:01	amalloy	tutysra_: a strong indicator is whether the dependency is listed as [clojure "1.4.0"] or [org.clojure/clojure "1.4.0"] -- lein2 no longer allows the shorthand
01:01	shawnlewis	xeqi: cool
01:01	uvtc	amalloy: interesting. Thanks.
01:02	xeqi	amalloy: ohh, thats a good one
01:02	uvtc	Oh, also, I think a ":profiles" keyword would indicate lein2.
01:03	uvtc	whoops --- xeqi said that already.
01:05	tutysra_	amalloy, I could see all dependencies having full path - org.clojure/clojure "1.4.0" etc, there is no :min-lein-version or : profiles does it still indicate it is lein2?
01:05	amalloy	shrug
01:06	amalloy	most of the time lein2 works. if it doesn't, try lein1
01:06	uvtc	tutysra_: this might be helpful to you: https://github.com/technomancy/leiningen/wiki/Upgrading
01:11	tutysra_	amalloy, uvtc - issue is resolved, I was using a library which uses clojure 1.4.0 with my project which uses clojure 1.3.0, changed my project version and I could compile the files
01:13	uvtc	tutysra_: nice ☺
01:17	samrat	how can I do this: https://www.refheap.com/paste/4359 without using :when twice?
01:19	uvtc	Maybe :when (or (> v 1) (< v 20)) ?
01:20	xeqi	I was thinking `and`, but I never use :when
01:20	uvtc	Oh, whoops. :)
01:21	samrat	uvtc: thanks, putting `and` worked
01:21	uvtc	xeqi: what do you use instead?
01:21	uvtc	samrat: y/w :)
01:22	xeqi	filter
01:23	xeqi	not for any good reason
01:23	xeqi	just the habits I brought with me
01:24	uvtc	Ah. Seems like a `for` with `:when` might be better for perf; instead of first creating a big seq then filtering it afterwards.
01:25	uvtc	But I guess for most mortal-sized lists it doesn't matter too much. :)
01:25	xeqi	yeah, lack of stream fusion would hurt there
01:27	uvtc	I don't know what "stream fusion" is. Perhaps a new style of music? :)
01:31	xeqi	think smart compiler that turns multiple map/filter/reduce calls into a single walk down the seq
11:41	rplevy	jcromartie: https://github.com/technomancy/swank-clojure/ , M-x clojure-jack-in
11:42	jcromartie	thanks
11:46	dnolen	SWEET, http://www.50ply.com/cljs-bench/
11:46	dnolen	http://www.50ply.com/blog/2012/08/14/cljs-bench-facelift/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+50plyBlog+%2850Ply+Blog%29
11:47	mmitchell	technomancy: hey there, still hacking on my app trying to get s3wagon private to work. You mentioned hooking into leiningen via repl to do some debugging, how would I do that?
11:49	mmitchell	does my repo key name (in :dependencies) need to be "private", literally? Or is it by the mere presence of "s3p://" that s3 wagon kicks in?
11:53	jcromartie	wow, Emacs 24, latest clojure-mode and paredit
11:53	jcromartie	and "{" still doesn't work!
11:55	clgv	humm repl-y in lein2 is a backstep compared to lein1.x + rlwrap...
11:57	nDuff	jcromartie: I'm using emacs 24 with the emacs live package, and {} works fine for me.
11:58	jcromartie	emacs live package?
11:58	nDuff	jcromartie: https://github.com/overtone/emacs-live/
11:58	nDuff	...watch the video linked from the README; it's pretty awesome.
11:59	mmitchell	anyone here using s3wagon private for their dependencies?
11:59	jcromartie	hah, does it actually add that glow to Emacs?
11:59	jcromartie	for some reason I thin that's an embellishment
12:00	nDuff	jcromartie: It's from a video of a presentation
12:00	nDuff	jcromartie: ...so not an intentional post-processing as much as a camera-in-a-dark-room effect.
12:00	jcromartie	no
12:00	jcromartie	"@Pat Sadly it's post-recording FX."
12:00	jcromartie	:)
12:00	nDuff	Huh.
12:01	rplevy	clgv: I've been wondering if repl-y plans to clean up its act. All I ever see are spews of terminal escape codes
12:01	rplevy	but it's fine
12:01	borkdude	rplevy what version of lein2/reply are you using (just wondering)
12:02	rplevy	borkdude: let me check...
12:02	rplevy	preview 7
12:03	clgv	rplevy: yeah thats pretty annoying. I use Pos1 or End often...
12:04	borkdude	rplevy maybe you should file it as a bug on github
12:05	rplevy	good idea, I will make it a point to do that. I just assumed it was known.
12:16	clgv	I hope repl-y improves vastly until lein2 goes stable...
12:16	S11001001	nDuff: the lines in the screenshots are perfectly straight
12:17	clgv	I guess there is no rlwrap-fallback built-in anymore?
12:17	S11001001	implying the camera was very, very far away
12:18	nDuff	S11001001: Conferences tend to have good video recording setups in terms of ideal positioning..
12:18	S11001001	not that perfect
12:20	S11001001	the CCD pixels would have to perfectly line up with those displaying the different color of the statusbar
12:22	technomancy	clgv: you can trampoline to running clojure.main/main with rlwrap
12:22	technomancy	clgv: but if you're having problems with reply, please report them; for most people it works way better than the horrid stuff I put in 1.x
12:23	clgv	technomancy: fuctional keys seem not to work, e.g. Pos1, End
12:23	rplevy	I just opened an issue on repl-y for the escape code noise in emacs shell buffer
12:23	technomancy	the only problems I've heard are from having .inputrc interfere and not honoring TERM=dumb
12:23	technomancy	rplevy: that's the TERM=dumb issue, I think it's fixed if you run from master
12:24	technomancy	I don't know of a good solution around .inputrc other than just deleting it unfortunately
12:24	clgv	oh ok. so it might improve next preview/stable of leiningen?
12:24	rplevy	ok great, repl-y master or lein master?
12:24	technomancy	clgv: if it's TERM=dumb, yes. I don't know about the inputrc issues or end key.
12:25	technomancy	rplevy: lein master. if you have a checkout try removing .lein-classpath just to be sure
12:25	rplevy	great, thanks
12:25	technomancy	but honestly there's not much point to running it in an emacs buffer unless nrepl.el isn't working for you
12:25	clgv	rplevy: your name is too similar to "repl-y". I always read you like this ;)
12:27	rplevy	clgv: haha
12:29	rplevy	clgv: if I were to write a repl (as if there aren't enough) I would call it replevy, which is an actual word. my nick is based on first, middle initials and last name,
12:29	rplevy	.
12:31	clgv	write a customizable + extendable swing repl ;)
12:32	rplevy	I haven't tried out nrepl, because I mostly use swank-clojure. But sometimes it is easiest to use lein repl within a project, and I live in emacs.
12:33	trptcolin	clgv: we just found out about the home/end thing - see https://github.com/jline/jline2/issues/54
12:33	rplevy	repl-y is a lot nicer than the lein 1.x repl
12:34	technomancy	as the author of 1.x repl, I feel qualified to assert that it sucks
12:34	trptcolin	on ubuntu, anyway, it seems there's something outside of any INPUTRC that sets up keybindings in bash/zsh
12:37	trptcolin	… which we don't [yet?] have access to in jline
12:38	technomancy	you say in the issue it'll be in the next preview of lein; does that mean it's in beta10 of reply or are you shooting for getting beta11 into leiningen preview8?
12:39	otfrom	right. I'm off to the Clojure dojo. Catch you all tomorrow.
12:39	technomancy	trptcolin: ^
12:39	trptcolin	in beta10. i don't have any commits on top of beta10 yet
12:39	technomancy	ok, cool
12:40	trptcolin	this home/end thing is jline, but [at least so far] appears to be fixable via adding bindings to ~/.inputrc
12:40	clgv	technomancy: remaining question: when will preview8 be released?
12:40	technomancy	trptcolin: is there a wiki page or something that covers inputrc?
12:40	hyPiRion	clgv: When it's done, I suspect.
12:40	technomancy	I seem to see a fair number of questions about it
12:40	trptcolin	that's a good idea
12:40	technomancy	clgv: by the end of the week
12:41	clgv	technomancy: oh that's sooner than I expected for "when it's done" ;)
12:43	trptcolin	clgv: if you have a chance, i'd be interested to know whether this workaround fixes your Home/End issue: https://github.com/jline/jline2/issues/54#issuecomment-7747262
12:44	technomancy	if anyone has other considerations for things to go into preview8, please speak up
12:46	clgv	trptcolin: no it did not. I directly uncommented those two lines in /etc/inputrc
12:48	clgv	trptcolin: oh right. it has to be in the local .inputrc -- there it works!
12:50	cemerick	technomancy: this is happening today?
12:50	cemerick	preview8, I mean?
12:50	technomancy	cemerick: possibly. at least before the weekend.
12:51	cemerick	OK. Hopefully I can get a beta9 of nREPL out tomorrowish
12:51	cemerick	Though maybe that's cutting it too close.
12:51	technomancy	what kind of changes does it have?
12:51	technomancy	in terms of potential for instability
12:52	cemerick	No changes, just additions; though I should probably read the changelog before guaranteeing that.
12:52	cemerick	technomancy: nevermind, just run with what's been in there.
12:53	technomancy	oh yeah?
12:53	eggsby	M-. in nrepl is magic
12:53	eggsby	nrepl.el *
12:53	cemerick	Yeah. beta9 will happen this week, and then I can get to grinding down all the rough edges for a final release.
12:54	cemerick	But whatever goes out should percolate in master for at least a little while before being flung at innocents.
12:54	technomancy	heh; sure.
12:54	technomancy	leiningen actually doesn't even pull in nrepl directly, just via reply
13:07	spurrow	hi... i have a question about a general programming concept, but i'm not sure what the name is... the idea is you define different method bodies for particular values of the methods arguments
13:08	spurrow	is it just multi methods?
13:08	ivan	variable-arity functions
13:08	ivan	er, maybe multiple arity
13:09	ivan	er, pattern matching?
13:09	spurrow	so i might do something like: fib(0) = 1; fib(1) = 1; fib(n) = fib(n-1) + fib(n-2);
13:09	spurrow	(not clojure, sorry!)
13:10	ivan	looks like pattern matching
13:10	ivan	https://github.com/clojure/core.match/
13:10	technomancy	yeah, that's pattern matching
13:10	technomancy	you could sorta do it with multimethods, but they're lower-level
13:11	spurrow	oh great, core.match is exactly what i'm looking for
13:12	S11001001	compiling code that uses core.match brings some of the scalac experience to clojure.core/load :)
13:12	spurrow	S11001001: what do you mean by that?
13:12	spurrow	is scala famed for its pattern matching?
13:12	S11001001	scalac is famed for taking ages to compile things
13:12	spurrow	lol
13:14	dnolen	S11001001: heh, you must have lots of crazy patterns. haven't put much thought yet on possible optimizations ...
13:15	dnolen	spurrow: it's an idea that popular outside of mainstream languages - you should read up on pattern matching, CLOS generic methods, predicate dispatch
13:16	S11001001	dnolen: maybe; I'll break it down for you if you're interested, but I think only https://bazaar.launchpad.net/~scompall/+junk/clojure-stuff/view/head:/src/com/nocandysw/cloj_dummy/scala/adt.clj#L80
13:18	dnolen	S11001001: i don't see any relevant to core.match there.
13:18	dnolen	s/any/anything
13:18	S11001001	fnmatch is a thin wrapper for match
13:20	dnolen	S11001001: ah, w/o seeing what kind of code that typically generates and w/o seeing how often you rely on it - can't say much.
13:20	S11001001	I figured as much
13:25	spurrow	in clj files, why is code not nested within (ns) ?
13:29	nbeloglazov	spurrow: I think it's like syntactic sugar: to avoid indent before each top level command like defn or def.
13:30	spurrow	bummer...why negotiate with terrorists?
13:34	dnolen	S11001001: I poked around is the only place you're using those macros in scala.clj? If so then long compile times seem strange. running all the tests w/ a warmed up JVM takes 1s
13:40	l1x	o hai!
13:40	l1x	i am trying to get java.jmx working with my project but no luck yet
13:40	uvtc	hi, l1x.
13:40	l1x	i am following this doc
13:41	l1x	https://github.com/clojure/java.jmx/
13:42	S11001001	dnolen: yes
13:42	l1x	https://github.com/l1x/cljmx/blob/master/src/cljmx/core.clj#L10
13:42	l1x	this errors out that \* is not supported
13:43	l1x	https://gist.github.com/3361874
13:43	l1x	if somebody have a minute for a clojure noob, i would appreciate
13:43	dnolen	S11001001: how long does it take for you? and how are you compiling? interactively or at the command line w/ lein?
13:45	S11001001	alternately with C-c C-l and C-c C-k depending on mood
13:45	uvtc	l1x: if that thing on line 10 is a string, you need to put double-quotes around it "\":\""
13:46	uvtc	s/is/is supposed to be/
13:46	l1x	thx i try that
13:49	l1x	uvtc: actually (jmx/mbean-names ":") works
13:49	l1x	now i need to find a way to print out what it returns
13:51	uvtc	l1x: maybe `(doseq [i (jmx/mbean-names ":")] (println i))`?
13:52	treehug	l1x: that line 10 looks like it was copied out of the clojure doc string which would have required the quotes
13:52	l1x	you are awesome dude!
13:52	l1x	thx!
13:52	uvtc	l1x: sure, y/w :)
13:59	dnolen	S11001001: so does it take 1s to compile? 10s? 1 minute?
14:02	S11001001	dnolen: 6s
14:04	dnolen	S11001001: slow enough to be annoying. I'll look into possible optimizations when I pick up core.match again.
14:04	l1x	treehug: yep
14:04	l1x	treehug: using ":" works
14:07	S11001001	not too bad, considering: https://bazaar.launchpad.net/~scompall/+junk/interp-mixed--mainline/view/head:/src/main/scala/com/nocandysw/interpmixed/Tree.scala takes 14s to compile :)
14:20	l1x	how can i iterate through a hashmap?
14:20	l1x	(doseq [[k v] (jmx/mbean-names ":")] (prn k v))
14:20	l1x	this does not work :(
14:20	l1x	Exception in thread "main" java.lang.UnsupportedOperationException: nth not supported on this type: ObjectName
14:21	hiredman	l1x: I don't thing mbeen-names returns a hashmap
14:22	l1x	sorry i meant hashset
14:22	hiredman	do you understand what a hashset is?
14:22	hiredman	and do you understand what the [k v] destructuring in the doseq above is doing?
14:24	l1x	yep, my bad
14:27	l1x	hiredman: is there something like in ruby the puts variable.class which tells me what is what
14:29	dnolen	craziness, core.logic querying for Eclipse project from a Smalltalk SOUL dev
14:34	`fogus	I have no idea what that sentence means
14:35	dnolen	`fogus: http://soft.vub.ac.be/SOUL/
14:36	dnolen	`fogus: SOUL embeds Prolog into Smalltalk, they did some incredible pattern directed metaprogramming work w/ it.
14:36	`fogus	ooooo! Now it's starting to make sense
14:36	`fogus	(for some value of sense)
14:36	`fogus	This just gave me an idea for my EL talk
14:50	nDuff	...so, I'm looking at korma, and it's not clear to me what the expected workflow is for injecting runtime configuration values.
14:51	acheng	hello. destructuring fn args question: can defaults specified by :or be computed from other key/values that were supplied in the call? i.e. :or {key2 (inc key1)}
14:52	l1x	if prn print this [:HeapMemoryUsage {:committed 85000192, :init 0, :max 129957888, :used 17351208}], how can i access the individual elements in the hash?
14:52	nDuff	...since defdb assigns directly to a var, and expects all the settings for a database to be already present when it's run, I'm not clear on what the sane approach is to wire up connections after namespace load time.
14:53	dnolen	ibdknox: my last bit about SOUL might interest you as well ...
14:55	haspaker	Why won't Clojure read numbers beginning with zero?
14:55	hyPiRion	,012
14:56	clojurebot	10
14:56	hyPiRion	octal.
14:56	Chouser1	acheng: that's a great question
14:56	haspaker	Ah
14:56	Chouser1	apparently it depends
14:56	haspaker	Thanks
14:56	uvtc	,0000
14:56	clojurebot	0
14:56	`fogus	,08
14:56	clojurebot	#<ExecutionException java.util.concurrent.ExecutionException: java.lang.NumberFormatException: Invalid number: 08>
14:56	`fogus	That's where octal gets ya!
14:57	hyPiRion	It's kind of scary when you work with floats though.
14:57	hyPiRion	,065.4321
14:57	clojurebot	65.4321
14:58	hyPiRion	...well, in Java at least.
14:58	@chouser	,((fn [{a :a, b :b, :or {b (inc a)}}] [a b]) {:a 5})
14:58	clojurebot	[5 6]
14:59	@chouser	((fn [{a :a, :keys [b], :or {b (inc a)}}] [a b]) {:a 5})
14:59	@chouser	,((fn [{a :a, :keys [b], :or {b (inc a)}}] [a b]) {:a 5})
14:59	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: a in this context, compiling:(NO_SOURCE_PATH:0)>
15:01	acheng	chouser: i wonder why that would be the case when you use :keys
15:02	acheng	chouser: seems like one can choose between [1] using :keys and adding a let binding to the body of fn to handle those cases and [2] not using :keys and having a more verbose arg list
15:04	@chouser	,(destructure '[{a :a :keys [b] :or {b (inc a)}} {:a 5}])
15:04	clojurebot	[map__187 {:a 5} map__187 (if (clojure.core/seq? map__187) (clojure.core/apply clojure.core/hash-map map__187) map__187) b ...]
15:04	technomancy	nDuff: indeed; that's a major design flaw in korma
15:05	@chouser	hmph. cut off.
15:05	@chouser	,(prn (destructure '[{a :a :keys [b] :or {b (inc a)}} {:a 5}]))
15:05	clojurebot	[map__214 {:a 5} map__214 (if (clojure.core/seq? map__214) (clojure.core/apply clojure.core/hash-map map__214) map__214) b ...]
15:05	@chouser	,(pr-str (destructure '[{a :a :keys [b] :or {b (inc a)}} {:a 5}]))
15:05	clojurebot	"[map__241 {:a 5} map__241 (if (clojure.core/seq? map__241) (clojure.core/apply clojure.core/hash-map map__241) map__241) b ...]"
15:05	@chouser	heh
15:05	acheng	chouser: running low on quarters?
15:06	acheng	chouser: (didn't mean for that to sound snarky in any way)
15:06	@chouser	heh, I'm not sure what you mean
15:06	@chouser	anyway, here is the order the bindings are done in: && (take-nth 2 (destructure '[{a :a :keys [b] :or {b (inc a)}} {:a 5}]))
15:07	@chouser	,(take-nth 2 (destructure '[{a :a :keys [b] :or {b (inc a)}} {:a 5}]))
15:07	clojurebot	(map__29 map__29 b a)
15:07	hyPiRion	Seems like order is important here.
15:07	@chouser	,(take-nth 2 (destructure '[{a :a b :b :or {b (inc a)}} {:a 5}]))
15:07	clojurebot	(map__56 map__56 a b)
15:08	@chouser	yeah, so the "or" part is done at the moment the target (b in this case) is bound
15:08	hyPiRion	,((fn [{a :a, b :b :or {b 2, a (inc b)}}] [a b]) {})
15:08	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: b in this context, compiling:(NO_SOURCE_PATH:0)>
15:08	hyPiRion	But when reversed, it's okay.
15:08	hyPiRion	,((fn [{b :b, a :a :or {b 2, a (inc b)}}] [a b]) {})
15:08	clojurebot	[3 2]
15:09	@chouser	,(take-nth 2 (destructure '[{:keys [a b c d e]} {}]))
15:09	clojurebot	(map__145 map__145 e d c ...)
15:09	@chouser	:keys are bound in reverse order (!)
15:10	acheng	chouser: is that for a good reason?
15:10	@chouser	I doubt it was intentional
15:10	hyPiRion	I suspect it's conj on a list in action.
15:11	acheng	chouser: hyPiRion: would it be bad to rely on this behavior of :keys then?... and so one should just skip the use of :keys for this situation?
15:11	@chouser	seems like relying on :or expressions having access to other keys in the same map would be to rely on an undocumented implementation detail. Risky in the long term.
15:11	acheng	yeah
15:12	nbeloglazov	&(doc destructure)
15:12	lazybot	⇒ "([bindings]); "
15:12	acheng	thanks both! followup question: how does one become awesome?
15:12	@chouser	You're already awesome, you just have to find people who agree.
15:13	@chouser	Wait, scratch that. I don't think that's a good philosophy.
15:13	trptcolin	lol
15:13	nbeloglazov	I like it :)
15:13	hiredman	chouser: too late
15:13	hiredman	you said it, someone slap it on a tshirt
15:13	@chouser	acheng: Commenting on Yegge posts makes one awesome.
15:14	@chouser	no, this is getting worse.
15:14	hiredman	chouser: why don't you stick to "reading is good", seems pretty safe
15:14	@chouser	:-)
15:15	hyPiRion	We can do it the Hickey way:
15:16	hyPiRion	Awesome - Causing awe or terror; inspiring wonder or excitement.
15:16	mattmoss	Faster? Stronger? More functional?
15:16	hyPiRion	(Look it up in a dictionary)
15:17	uvtc	hammock-driven awesomeness
15:17	acheng	he also said not everything is awesome. so there's a limited number of spots
15:18	hiredman	https://www.youtube.com/watch?v=UPq1_9DsLKk
15:20	@chouser	hiredman: now that is awesome.
15:21	hiredman	and it describes software developement pretty well
15:21	mattmoss	hiredma: I actually had six billion dollar man in mind, but that works too.
15:22	uvtc	mattmoss: six *million?
15:22	acheng	uvtc: inflation
15:24	qmx	cemerick: fwiw, bought your book :)
15:24	pandeiro	cemerick: you mentioned an http lib to me last week, specializing in API routing, can you remind me what it's called please?
15:25	cemerick	pandeiro: bishop, perhaps? That's cmiles74's.
15:25	cemerick	qmx: Hey, thanks :-D
15:25	cemerick	I hope you like it.
15:26	pandeiro	cemerick: i think that's it; cheers
15:26	clojurebot	cemerick: Yeah, MacKay's book is great too.
15:38	georgek	hi, I just found twhile using hiccup hat there's a bug in java.net.uri such that it chokes on the pipe character; is this something I can file a bug report with Java for? I'm Googling around but not seeing where to go. Also what might be a fix?
15:39	emezeske	georgek: http://stackoverflow.com/questions/1828641/how-to-parse-a-uri-like-this-in-java
15:39	emezeske	georgek: I don't think it's a bug in java.net.uri -- you just need to encode the special character
15:40	georgek	ah cool, thanks emezeske
15:40	hiredman	georgek: http://www.ietf.org/rfc/rfc2396.txt says pipes are "unwise" in uris and must be escaped
15:40	hiredman	oh, bleh, I guess stackoverflow works too
15:41	hyPiRion	Use "%7C" instead of "\|".
15:41	hiredman	use a a urlencoder
15:42	hiredman	I like how the rfc defines a character class called "unwise"
15:42	hiredman	those guys are a hoot and a half
15:43	emezeske	hiredman: Heh, stackoverflow was just the first hit on google :)
15:43	hiredman	!!!!
15:43	hiredman	new loper-os.org post
15:44	hiredman	fuh, not about lips though
15:44	hiredman	lisp
15:47	zerokarmaleft	hiredman: but still proclaiming himself a prophet, more or less?
15:48	hiredman	zerokarmaleft: with an extensive self quotation
16:05	TimMc	Well, he's spot-on here.
16:12	zeromodulus	I tried getting my compojure project to recognize my main namespace, i.e. :main hello-world.core, but it says it can't be found. what am I doing wrong?
16:13	zeromodulus	the code is there, under src/hello-world/core.clj
16:13	georgek	does anyone know if I can use pomegranate with nrepl.el? When I did (require '[cemerick.pomegranate :as pomegranate]) it wasn't found on classpath
16:14	georgek	I mean as you can with lein2
16:17	solussd	does assoc create a new record when used with records? I thought it returned a hash-map…. is that a recent change?
16:17	zeromodulus	welp, I give up, I'll try again when things are more mature and documented. see ya.
16:17	gfredericks	solussd: I believe it will only switch to a hash-map if you dissoc one of the declared keys
16:18	solussd	so under the hood it's newing up a new record
16:18	gfredericks	solussd: and afaik it's been the same behavior since the beginning
16:18	solussd	cool
16:18	cemerick	I guess 4 minutes is the half-life of a question on irc. Verified experientially. :-P
16:19	S11001001	solussd: try dissocing a vital key
16:20	solussd	yup, that turns it into a hash-map
16:20	solussd	cool, that's exactly how I expect/want it to work
16:21	S11001001	oh gfredericks said that already
16:21	S11001001	gfredericks: indeed it's always been like this
16:27	djanatyn	quick question. I have a list of cities, and a function that takes two cities and finds the distance between them.
16:27	djanatyn	how I can I find the totaly distance in an idiomatic way?
16:27	gfredericks	between all pairs?
16:28	S11001001	shortest route that visits all cities :)
16:28	gfredericks	:P
16:28	@Chouser	Heh, I was about to suggest that.
16:28	djanatyn	Well, no, not yet. I just want to find out the length of a route given a list of cities
16:29	djanatyn	that's next!
16:29	gfredericks	djanatyn: partition 2
16:29	S11001001	gfredericks: needs less skipping
16:29	TimMc	djanatyn: Is this a closed loop?
16:29	aperiodic	partition 2 1
16:29	S11001001	,(doc partition-by)
16:29	metellus	why not reduce?
16:29	clojurebot	"([f coll]); Applies f to each value in coll, splitting it each time f returns a new value. Returns a lazy seq of partitions."
16:29	@Chouser	and reduce
16:29	gfredericks	(apply + (for [[a b] (partition 2 1 cities)] (distance a b)))
16:29	metellus	wait, I know why not.
16:29	S11001001	gfredericks: ah, there it is
16:29	gfredericks	man that was like 8 people answering at once
16:29	djanatyn	mmm. I wanted to use reduce but my function wasn't set up in a way that would work.
16:30	@Chouser	djanatyn: it's good for your function to just take two cities. And wanting reduce was also right
16:30	gfredericks	(->> cities (partition 2 1) (map (partial apply distance)) (reduce +))
16:30	@Chouser	you just need gfredericks's use of 'for' or similar to glue them together
16:31	djanatyn	I knew to find the sum of the cities I could just #(reduce + %)
16:31	djanatyn	errm, the sum of the distances. I bet there's even a sum function already implemented in contrib or something too
16:31	TimMc	I believe (apply + ...) is better since it allows + to decide on a computation order.
16:31	acheng	zeromodulus: you're gone but i wonder if you have the - vs _ issue.
16:32	@Chouser	TimMc: for reasons I don't deeply understand, Rich has recommend 'reduce' in general and for + in particular, though of course your argument is exactly why 'apply' is better for 'str'
16:33	gfredericks	when I run a ring/jetty server, what would cause it to return a gzip'd response?
16:33	TimMc	reduce makes me feel like more of a badass, but apply seems more correct.
16:34	gfredericks	does the response have to be above some threshold size?
16:34	dnolen	@Chouser: I've always suspected that something like reducers was the reason.
16:35	gfredericks	but reduce has order as part of its API
16:35	@Chouser	dnolen: yeah I wondered that as I was typing just now, but of course multi-arg + could choose to use reducers if it wants to.
16:35	gfredericks	so it couldn't really do anything else without breaking backwards compat
16:35	dnolen	Chouser: which wouldn't compose right?
16:35	@Chouser	gfredericks: good point, where + could know that order doesn't matter and use reducers.
16:37	@Chouser	hm, of course apply calling a variatic fn promises to give you a lazy seq, which is itself ordered and so can't use reducers, right?
16:37	gfredericks	well that's prior to adding
16:38	hiredman	Chouser: I don't recall seeing rich recommend reduce in general and for + in particular
16:38	hiredman	I seem to recall apply + being faster
16:38	gfredericks	so assymptotically it can still be more efficient but for something fast like + probably not
16:38	@Chouser	while if you were using reduce yourself you might have an easier time realizing you could use a reducer to good effect for the particular collection at hand.
16:39	hiredman	,(time (reduce + (range 1e4)))
16:39	clojurebot	"Elapsed time: 1.349744 msecs"
16:39	clojurebot	49995000
16:39	hiredman	,(time (apply + (range 1e4)))
16:39	clojurebot	"Elapsed time: 5.568422 msecs"
16:39	clojurebot	49995000
16:39	hiredman	huh
16:39	hiredman	~clojurebot
16:39	clojurebot	clojurebot is a multimap
16:39	hiredman	~benchmarking
16:39	clojurebot	Cool story bro.
16:39	Bronsa	i remembered apply + being faster too
16:39	hiredman	~clojurebot
16:39	clojurebot	clojurebot is amazing
16:39	hiredman	clojurebot: no you are not
16:39	clojurebot	Excuse me?
16:40	Bronsa	*remember
16:40	@Chouser	hiredman: you're going to make me grep the IRC logs, huh?
16:41	metellus	,(time (reduce + (range 1e4)))
16:41	clojurebot	"Elapsed time: 154.705691 msecs"
16:41	clojurebot	49995000
16:41	metellus	there's more than just apply vs. reduce going on here
16:42	hiredman	Chouser: I think I found it
16:42	hiredman	b
16:43	@Chouser	http://clojure-log.n01se.net/date/2009-01-12.html#13:48
16:43	hiredman	apply lets the function choose how to handle var args, reduce (with the protocol stuff) lets the collection drive how reduction is done
16:45	hyPiRion	Oh, the old common lisp loop.
16:46	TimMc	Function vs. collection: FIGHT!
16:46	TimMc	What we need here is a protocol.
16:48	gfredericks	amalloy: ping
17:06	djanatyn	wow, I didn't know there was a clojure meetup group in pittsburgh :O
17:07	TimMc	Is it active?
17:07	djanatyn	Yes, they're having a meetup on September 11th. I plan on attending.
17:07	djanatyn	The only other usergroup I've been to was the ruby usergroup, and to be honest, I'm not really into Ruby.
17:09	haspaker	Is there an equivalent of "reduce" that returns all values instead of just the very last one?
17:10	brehaut	reductions
17:10	brehaut	,(reductions conj [] [1 2 3])
17:10	clojurebot	([] [1] [1 2] [1 2 3])
17:10	haspaker	Brilliant
17:12	leandro	how do i get the current function's name?
17:13	nDuff	leandro: The current function often won't _have_ a name at all. What's the context?
17:13	leandro	log, i'd line to log this information.
17:13	leandro	line=like
17:14	nDuff	nod; ahh.
17:14	nDuff	The normal JVM-centric methods work, though they're pretty slow.
17:14	nDuff	IIRC, log4j has that support built in
17:14	nDuff	so there's something you can use in your format strings which will indicate where your code is
17:15	nDuff	having it be user-configured that way is probably best -- that way they can choose whether to eat the performance penalty
17:15	leandro	it helped soo much. ty.
17:17	leandro	nDuff: i'll continue to use a hardcoded string. =)
17:40	solussd	why does this fail? &(let [b 5] (for [item '[b]] (eval item)))
17:40	solussd	,(let [b 5] (for [item '[b]] (eval item)))
17:40	clojurebot	#<Exception java.lang.Exception: SANBOX DENIED>
17:40	solussd	well, it fails with a b doesn't exist in my repl
17:40	solussd	if I (def b 5), it works
17:41	emezeske	solussd: Why do you have the quote on '[b] ?
17:41	dnolen	solussd: eval doesn't work with locals.
17:41	solussd	emezeske: bc I want the name of the symbol in the body of the for
17:41	solussd	dnolen: thanks
17:41	bosie	is there a documentation that is better than the official one?
17:42	acagle_	/flush
17:42	solussd	dnolen: is there a way to get the local value?
17:42	dnolen	bosie: Clojure official docs are nice IMO - but there are some truly excellent books now as well if you're just starting.
17:42	dnolen	solussd: not w/ eval.
17:43	bosie	dnolen: just starting. java/ruby background
17:43	solussd	clojuredocs.org is nice, sometimes.
17:43	dnolen	bosie: O'Reilly Clojure Programming is pretty darn good
17:43	bosie	dnolen: i find them undigestable, like examples would be helpful ;)
17:43	bosie	solussd: nice
17:44	bosie	dnolen: bought it
17:44	Raynes	$examples eval
17:44	lazybot	You must pass a name like clojure.core/foo or, as two arguments, clojure.core foo.
17:44	dnolen	bosie: covers language, idioms & tooling. clojuredocs.org is nice for examples. accessible from lein2 repl as well.
17:44	Raynes	$examples clojure.core/eval
17:44	lazybot	https://www.refheap.com/paste/4400
17:44	Raynes	These are extracted from clojuredocs.
17:44	bosie	how would i go about figuring out what line 1 does: http://biomunky.wordpress.com/2010/03/17/clojure-read-a-file/
17:44	Raynes	Well you could ask here like you just did.
17:44	Raynes	:p
17:45	bosie	Raynes: hah, sure but i couldn't figure out from the docs how to read it
17:45	Raynes	Line one of the first example?
17:45	bosie	Raynes: yes
17:45	Raynes	The namespace stuff is particular hard to understand.
17:45	bosie	Raynes: the whole '[] business
17:45	Raynes	The docs for them are designed to be thorough and not necessarily easy.
17:46	Raynes	In this case, it's "Mr. Clojure sir, could you please let me use the 'reader' function/macro from the 'clojure.java.io' namespace? I don't want to be able to access everything, just 'reader'."
17:47	bosie	the equivalent of import clojure.java.io.reader
17:47	bosie	i take it?
17:47	solussd	dnolen, I can accomplish what I want with macrolet!!!! :D
17:47	dnolen	bosie: one thing you'll encounter is a lot of things that require special syntax in other languages tend to be first class features in Clojure.
17:47	dnolen	solussd: nice
17:47	Raynes	Well, no, because you just said has no meaning in Clojure.
17:47	Raynes	:p
17:47	bosie	Raynes: sure
17:47	solussd	i've wanted an excuse to use it.
17:48	bosie	Raynes: so what is '[] ?
17:48	bosie	a function in itself?
17:48	dnolen	bosie: so yes, that dynamically loads the file according. '[...] will contain the usual syntax for importing.
17:48	Raynes	It's a quoted vector.
17:48	bosie	hm
17:49	Raynes	It means "Hey, please don't evaluate the things inside this vector." Since 'use' is a function and not a macro, it can't control evaluation and thus if you did (use [foo :only [bar]]), Clojure would try to resolve the symbols 'foo' and 'bar' and it would fail.
17:49	bosie	oh
17:50	bosie	hm
17:50	bosie	so in what "state" is [] in?
17:50	nDuff	"state"?
17:50	clojurebot	mutable state is bad
17:50	dnolen	bosie: it's just a piece of data.
17:51	bosie	nDuff: in the sense that it kindad is a vector but apparently not evaluated, so it just gets 'held' somehow without being executed?
17:51	bosie	like a lazy sequence?
17:51	dnolen	,(let [a b] [[a b] '[a b]])
17:51	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: b in this context, compiling:(NO_SOURCE_PATH:0)>
17:51	metellus	,(let [a 1, b 2] [[a b] '[a b]])
17:51	clojurebot	[[1 2] [a b]]
17:51	dnolen	oops, thx metellus
17:52	dnolen	bosie: ^
17:52	Raynes	bosie: It isn't anything spectacular. It just isn't evaluated.
17:52	bosie	k
17:52	Raynes	bosie: In Clojure, things are read and then they are evaluated. A quote prevents that evaluation part of things.
17:52	bosie	Raynes: so its simply held as a string until needed?
17:53	nDuff	bosie: The vector is evaluated as a vector.
17:53	nDuff	bosie: ...the symbols within it aren't resolved.
17:53	nDuff	bosie: ...but there's no kind of lazy / just-in-time behavior involved.
17:53	bosie	k
17:53	Raynes	FWIW, Clojure behavior is often a lot more simpler than you think.
17:54	bosie	Raynes: so it isn't like haskell? ;)
17:54	Raynes	Haskell is super easy.
17:54	Raynes	:D
17:54	bosie	Raynes: i didn't find it easy at all :(
17:54	bosie	makes me wonder if i am not cut out for FP
17:54	nDuff	bosie: ...you just get clojure.lang.Symbol objects rather than, well, whatever foo or bar would point to if they were defined.
17:54	Raynes	bosie: I was kidding.
17:54	Raynes	Haskell can be overwhelming sometimes.
17:55	Raynes	Things like what we're talking about here can be difficult to understand at a distance. Once you've used things a bit, it'll start to make sense.
17:56	bosie	Raynes: k
17:56	bosie	Raynes: i honestly hope so ;)
17:57	metellus	,(let [a 1] [(class a) (class 'a)])
17:57	clojurebot	[java.lang.Long clojure.lang.Symbol]
17:57	haspaker	Is there a function that returns the first item of a lazy sequence that satisfy a certain condition?
17:58	joegallo	(first (filter ... ...))
17:58	haspaker	Of course
17:58	haspaker	Feel a bit stupid now :>
17:59	Raynes	haspaker: It isn't immediately obvious to most people, but it works because filter is lazy and thus just getting the first element doesn't cause (much) unnecessary computation.
17:59	joegallo	sometimes (some .......) is the right thing, though, too. But the semantics are different.
18:00	haspaker	Yeah
18:00	joegallo	(first (filter x y)) returns the first value from y for which x returns truthy
18:00	joegallo	(some x y) returns the first truthy return value of x, which always tricks me.
18:00	haspaker	I mostly felt stupid because I have used (first (filter …)) on other occasions but did not realize I could do it here as well
18:01	joegallo	no worries. we shall strike this conversation from the chat logs, it'll be like it never happened.
18:01	joegallo	;)
18:01	haspaker	Brilliant :P
18:03	dnolen	bosie: lots of new stuff to learn. fortunately people here are nice and helpful :)
18:03	bosie	dnolen: yup that seems
18:03	bosie	dnolen: to be the case. seems to be the case in quite a few FP languages ;)
18:06	bosie	thanks Raynes / dnolen
18:11	eggsby	why can't you use == on keywords?
18:11	eggsby	,(== :foo :foo)
18:12	clojurebot	#<ClassCastException java.lang.ClassCastException: clojure.lang.Keyword cannot be cast to java.lang.Number>
18:12	hiredman	,(doc ==)
18:12	clojurebot	"([x] [x y] [x y & more]); Returns non-nil if nums all have the equivalent value (type-independent), otherwise false"
18:12	hiredman	== is for numbers only
18:12	eggsby	ah, and =
18:12	eggsby	I see
18:13	eggsby	thanks hiredman
18:15	joegallo	,(doc identical?)
18:15	clojurebot	"([x y]); Tests if 2 arguments are the same object"
18:15	joegallo	(== :foo :foo)
18:15	joegallo	,(== :foo :foo)
18:15	clojurebot	#<ClassCastException java.lang.ClassCastException: clojure.lang.Keyword cannot be cast to java.lang.Number>
18:15	joegallo	,(identical? :foo :foo)
18:15	clojurebot	true
18:15	eggsby	hm
18:15	joegallo	phew, just checkin'
18:15	nkoza	there is a more succint way to do this, without using update-in?: (update-in {:a 1} [:a] #(inc %))
18:16	eggsby	nkoza: the #(%) is unnecessary
18:16	eggsby	,(update-in {:a 1} :a inc)
18:16	clojurebot	#<UnsupportedOperationException java.lang.UnsupportedOperationException: nth not supported on this type: Keyword>
18:16	eggsby	oops
18:16	eggsby	,(update-in {:a 1} [:a] inc)
18:16	clojurebot	{:a 2}
18:16	nkoza	using update-in is the common idiom to update a key in a map?
18:17	hyPiRion	,(merge-with + {:a 1} {:a 1})
18:17	clojurebot	{:a 2}
18:17	nDuff	nkoza: Very.
18:17	nkoza	(without nested maps)
18:17	metellus	,(assoc {:a 1} :a 2)
18:17	clojurebot	{:a 2}
18:17	nkoza	I thought update-in was only for nested maps
18:17	eggsby	nkoza: it works with nested maps, you just define your 'path' to the data
18:17	eggsby	,(update-in {:a {:b 1} [:a :b] inc)
18:17	clojurebot	#<ExecutionException java.util.concurrent.ExecutionException: java.lang.RuntimeException: Unmatched delimiter: )>
18:17	eggsby	oops
18:17	eggsby	:3
18:17	eggsby	you get the idea
18:17	amalloy	nkoza: yes, use update-in whether you have nested maps or not
18:17	nkoza	ok, thanks
18:18	ToxicFrog	amalloy: why is it preferred to use update-in for non-nested maps rather than assoc?
18:18	amalloy	(assoc x :birthday (inc (get x :birthday))) ;; so gross
18:18	metellus	I was about to ask the same thing when I realized that eggsby wanted to modify the existing value
18:19	amalloy	(update-in x [:birthday] inc) ;; so clear
18:19	nkoza	ToxicFrog: i was searching how to update a map with a transformation of the existing value, assoc is for putting a new one or replacing the previous
18:19	ToxicFrog	nkoza: aah, ok
18:20	amalloy	and it just gets more extreme when you're computing x somewhere else: you have to let-bind it to make the assoc version work
18:20	ToxicFrog	Yeah, I missed that bit
18:23	hyPiRion	Hm, is there a specific reason to have update-in, but not update? It seems weird considering assoc exists, even though assoc-in would be enough.
18:23	amalloy	for the same reason we have dissoc but not dissoc-in
18:23	amalloy	it's not entirely obvious what update does
18:24	mabes	I have a ns as a symbol, how can I resolve the ns so that I can then use ns-resolve with it?
18:24	mabes	,ns
18:24	clojurebot	#<Namespace sandbox>
18:24	mabes	I'd like to say something like:
18:24	amalloy	(update m [:x :y] inc) ;; what does this do, hyPiRion?
18:24	mabes	,(resolve 'sandbox)
18:24	clojurebot	nil
18:24	hiredman	http://clojure.org/namespaces
18:25	hyPiRion	amalloy: I myself consider it the same as (update-in m [[:x :y]] inc)
18:26	hyPiRion	Which is like ##(update-in {[:x :y] 1} [[:x :y]] inc)
18:26	lazybot	⇒ {[:x :y] 2}
18:26	mabes	,(find-ns 'sandbox)
18:26	clojurebot	#<Namespace sandbox>
18:26	mabes	thanks hiredman :)
18:26	hyPiRion	Exactly the same relation assoc and assoc-in has.
18:26	amalloy	hyPiRion: not exactly the same
18:26	amalloy	assoc and dissoc can accept multiple keys
18:27	hyPiRion	,(assoc [0 0] [0 1] inc)
18:27	clojurebot	#<IllegalArgumentException java.lang.IllegalArgumentException: Key must be integer>
18:27	amalloy	what you said is how it would probably work if it existed. but wouldn't it be equally useful to have it act like (-> m (update-in [:x] inc) (update-in [:y] inc))?
18:27	hyPiRion	Oh, gotcha.
18:28	amalloy	&(assoc nil :a 1 :b 2)
18:28	lazybot	⇒ {:b 2, :a 1}
18:28	amalloy	anyway, that's unclear, which i think is why it doesn't exist
18:30	hyPiRion	so not immediately obvious what the function does.
18:33	hyPiRion	So what's the reason for not accepting multiple key-vectors for assoc-in then?
18:39	djanatyn	one more quick question ^_^; if I have a list of maps, and I want to find the maximum value for one of the keys, how would I go about doing that idiomatically?
18:40	djanatyn	I can't just (max (map #(:key %) list)) because then I lose the other values in the map
18:42	amalloy	&(doc max-key)
18:42	lazybot	⇒ "([k x] [k x y] [k x y & more]); Returns the x for which (k x), a number, is greatest."
18:42	djanatyn	thank you amalloy!
18:43	djanatyn	man, clojure sure does have a lot of convenient functions
18:45	madsy	djanatyn: Yeah. Discovering all of them can be a chore :)
18:45	madsy	Even clojure.core has tons
18:45	cpinera	Hi all. I'm attempting to create an uberjar with Leiningen, and both clojure-1.3.0 and clojure-1.4.0 jars are being included. I'm assuming one or several dependencies are causing it. Is there a way for me to have leiningen not to include the jar for clojure-1.3.0?
18:45	hyPiRion	What's the best way of finding these nuggets? Just dive into the source code?
18:46	aperiodic	http://clojure.org/cheatsheet
18:47	madsy	cpinera: There should be a project.clj setting for excluding jars. Check the example files on the Leiningen github page
18:48	cpinera	madsy: I'm using :uberjar-exclusions [#"clojure-1.3.0.jar"] but the jar is still being included...
18:48	technomancy	cpinera: what version of leiningen?
18:48	cpinera	technomancy: I was on version 2 preview 7 and then I upgraded to the latest in the trunk
18:49	madsy	cpinera: Hm, not sure then. Maybe ask on #leiningen ?
18:49	cpinera	madsy: You are correct. Thanks!
18:49	technomancy	cpinera: is it a fork of clojure with a separate group-id?
18:49	hyPiRion	aperiodic: That doesn't cover much.
18:50	technomancy	hyPiRion: you pretty much have to hang out on IRC =)
18:50	aperiodic	hyPiRion: what else are you looking for? i've found it to be pretty comprehensive
18:50	technomancy	there's no substitute
18:50	hyPiRion	For example, you'll never find ##(unchecked-byte 255) there.
18:50	lazybot	⇒ -1
18:50	cpinera	technomancy: not as far as I can see. I got it directly from https://github.com/technomancy/leiningen, branch: master
18:50	technomancy	cpinera: i mean in your project
18:50	technomancy	cpinera: does `lein deps :tree` list two versions of clojure?
18:51	cpinera	technomancy: ah yes, you are correct. I'm doing Android development, and I'm using [android/clojure "1.4.0"] as a dependency
18:51	technomancy	cpinera: ok, in that case you have to use :exclusions
18:52	hyPiRion	But okay, I'll just hang in here to find those diamonds.
18:52	cpinera	technomancy: Understood. How would it look like? e.g. [clj-time "0.4.3" :exclusions [org.clojure]] ?
18:53	technomancy	cpinera: normally you would add it inside :dependencies, but for something like clojure there are too many libs pulling it in, so you wouldn't want to add an entry for each one. so you can use a top-level :exclusions [org.clojure/clojure]
18:53	amalloy	hyPiRion: you'll never learn them all anyway. i think my knowledge of clojure.core is pretty encyclopedic, and yet just today i found out about ##(doc release-pending-sends)
18:53	lazybot	⇒ "([]); Normally, actions sent directly or indirectly during another action are held until the action completes (changes the agent's state). This function can be used to dispatch any pending sent actions immediately. This has no impact on actions sent during a tr... https://www.refheap.com/paste/4401
18:54	hyPiRion	amalloy: That's a pessimistic look on it :(
18:54	cpinera	technomancy: Perfect! Thank you very much indeed, I really appreciate your help.
18:54	technomancy	no problem
18:54	amalloy	only if you feel pessimistic about it
18:55	amalloy	i also recently found out about ##(doc sequence), which is like the opposite of seq. i'd been using lazy-seq for that purpose till last week
18:55	lazybot	⇒ "([coll]); Coerces coll to a (possibly empty) sequence, if it is not already one. Will not force a lazy seq. (sequence nil) yields ()"
19:01	djanatyn	>_<
19:01	djanatyn	this is so much harder than it should be
19:02	djanatyn	so, I have a list of maps. the maps are in the format of {:distance foo :route bar}
19:02	djanatyn	I'm trying to find the map in that list that is the biggest. so I need to use 'max-key and apply.
19:03	hyPiRion	Is the biggest by distance or route, or both?
19:03	djanatyn	so, the function I'm using for max-key to compare is #(:distance %).
19:03	gfredericks	:distance works too
19:04	hyPiRion	#(:distance %) is not a nice key to compare with.
19:04	hyPiRion	or nevermind.
19:04	djanatyn	oh. okay, distance works
19:05	djanatyn	but when I try (apply #(max-key :distance %) list-i-mentioned-earlier), I get an error! because the function I'm using with apply only accepts one argument >_>
19:05	amalloy	(apply max-key :distance ...)
19:05	amalloy	djanatyn: you seem to just be wrapping everything in lambdas just in case :P
19:06	kenneth	hey, is there an option to `lein uberjar` to specify which filename to create the uberjar at
19:06	djanatyn	but I thought that function would apply
19:06	djanatyn	...
19:06	djanatyn	oh
19:06	djanatyn	headdesk
19:06	djanatyn	I do want to pass :distance to max-key :)
19:06	djanatyn	thank you.
19:06	gfredericks	~headdesk
19:06	clojurebot	Excuse me?
19:07	gfredericks	clojurebot: headdesk is botsmack
19:07	clojurebot	c'est bon!
19:07	metellus	~headdesk
19:07	clojurebot	headdesk is botsmack
19:07	gfredericks	~headdesk
19:07	metellus	~botsmack
19:07	clojurebot	headdesk is botsmack
19:07	clojurebot	Owww!
19:07	gfredericks	~headdesk
19:07	clojurebot	headdesk is botsmack
19:07	gfredericks	~headdesk
19:07	clojurebot	headdesk is botsmack
19:07	gfredericks	what happened to inference?
19:07	hyPiRion	~botsnack
19:07	clojurebot	Thanks, but I prefer chocolate
19:07	metellus	~chocolate
19:07	clojurebot	Cool story bro.
19:07	amalloy	gfredericks: it's working, you just got unlucky
19:08	gfredericks	amalloy: getting unlucky 4 times in a row is inconceivable
19:08	gfredericks	I have inconceived of it
19:08	djanatyn	...woooooooo
19:08	djanatyn	okay
19:08	amalloy	~headdesk
19:08	clojurebot	clojurebot evades successfully!
19:09	djanatyn	after a few days I now have a working implementation that will bruteforce any number of cities in the travelling salesman problem :D :D :D
19:09	djanatyn	in clojure!
19:10	gfredericks	djanatyn: so what're you going to sell?
19:11	djanatyn	uhh
19:11	djanatyn	clojure consulting?
19:22	djanatyn	I have a feeling (->) and (->>) are really important and useful functions
19:24	emezeske	djanatyn: s/functions/macros :)
19:24	gfredericks	they're just sugar
19:24	gfredericks	important and useful for readability perhaps
19:26	l1x	hey guys noob question, how can i create a hash in clojure?
19:27	TimMc	l1x: Are you talking about cryptographic primitives or associative data structures?
19:27	l1x	second
19:28	TimMc	l1x: http://clojure.org/data_structures#Data%20Structures-Maps%20%28IPersistentMap%29
19:28	l1x	this returns a hash (jmx/mbean "java.lang:type=Memory") i would like to store it and refernce the nested part
19:29	TimMc	Oh, a Java bean? You may want to `bean` on it.
19:32	l1x	in ruby i would reference it the following way parent[:nested_hash] (where parent ={:hash =>value , :nested_hash => {:super_deep_key => value}}
19:32	l1x	i would like to do something similar
19:34	emezeske	,(get-in {:a {:b 42}} [:a :b])
19:34	clojurebot	42
19:35	l1x	emezeske: thx
19:35	emezeske	l1x: You can of course just do it the long way too: &&(:b (:a {:a {:b 42}}))
19:35	emezeske	err
19:35	emezeske	,(:b (:a {:a {:b 42}}))
19:35	clojurebot	42
19:36	l1x	k
19:36	l1x	thx
19:40	jlewis	hi
19:41	zer	(let [m (transient {})]
19:41	zer	(dotimes [n 12]
19:41	zer	(assoc! m (keyword (str n)) n))
19:41	zer	(persistent! m))
19:41	jlewis	can you determine whether a PersistentHashMap has an entry with a null key in it?
19:41	zer	(let [m (transient {})] (dotimes [n 12] (assoc! m (keyword (str n)) n)) (persistent! m))
19:42	zer	&(let [m (transient {})] (dotimes [n 12] (assoc! m (keyword (str n)) n)) (persistent! m))
19:42	lazybot	⇒ {:0 0, :1 1, :2 2, :3 3, :4 4, :5 5, :6 6, :7 7}
19:42	emezeske	jlewis: I think contains? does what you want maybe
19:42	emezeske	,(contains? {:a 1} :b)
19:42	clojurebot	false
19:42	emezeske	,(contains? {:a 1 :b nil} :b)
19:42	clojurebot	true
19:43	zer	&(clojure-version)
19:43	lazybot	⇒ "1.4.0"
19:43	zer	,(clojure-version)
19:43	clojurebot	"1.4.0-master-SNAPSHOT"
19:43	zer	,(let [m (transient {})] (dotimes [n 12] (assoc! m (keyword (str n)) n)) (persistent! m))
19:43	clojurebot	{:0 0, :1 1, :2 2, :3 3, :4 4, ...}
19:44	jlewis	emezeske: i'm specifically asking about null keys, not null values
19:44	zer	,(clojure.pprint/pprint (let [m (transient {})] (dotimes [n 12] (assoc! m (keyword (str n)) n)) (persistent! m)))
19:44	clojurebot	#<ExecutionException java.util.concurrent.ExecutionException: java.lang.RuntimeException: java.lang.ClassNotFoundException: clojure.pprint>
19:44	amalloy	zer: :((((((((((((((((((((( transients aren't mutable objects
19:44	jlewis	the implementation has a special case for this - it has a boolean hasNull that it keeps around
19:44	zer	hm, this is how i treat them
19:44	emezeske	jlewis: Oh, I misread you
19:44	jlewis	i'd like to determine whether the map has a null key or not, but hasNull unfortunately is package private... :) this is from java, by the way
19:45	jlewis	package protected*
19:45	amalloy	zer: you are treating them wrong, then, which is why you don't get the behavior you want
19:45	emezeske	jlewis: (contains? {nil 42} nil) does not work for your purpose?
19:45	zer	should i work with the return value of assoc! instead?
19:45	amalloy	yes
19:45	amalloy	they're allowed to mutate, but not required to
19:46	zer	still a bit confusing, that my hash-map gets filled with 8 values instead of 1
19:46	zer	err 12
19:46	jlewis	emezeske: that always returns false
19:46	emezeske	,(contains? {nil 42} nil)
19:46	jlewis	,(contains? {nil 1} nil)
19:46	clojurebot	true
19:46	clojurebot	true
19:46	jlewis	err
19:46	jlewis	wtf
19:46	amalloy	zer: well, the first 8 times it decides to mutate, and the ninth time it decides to return a new object
19:46	jlewis	,(contains? {3 1} nil)
19:46	clojurebot	false
19:47	jlewis	?!?! i'm confused. i observed contrary behavior to this before
19:47	zer	ah ok, ty
19:47	zer	i wonder why you would want to return a new object instead of mutating
19:47	jlewis	emezeske: thanks, my test was screwed up. :)
19:48	zer	treating transients as mutable would be quite handy
19:48	amalloy	&(class (transient {}))
19:48	lazybot	⇒ clojure.lang.PersistentArrayMap$TransientArrayMap
19:48	emezeske	jlewis: No prob :)
19:48	amalloy	&(class (transient {1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20))
19:48	lazybot	java.lang.RuntimeException: Unmatched delimiter: )
19:48	amalloy	&(class (transient {1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20}))
19:48	lazybot	⇒ clojure.lang.PersistentHashMap$TransientHashMap
19:49	zer	aah right
19:49	amalloy	zer: and no, treating them as mutable wouldn't be handy; you'd be back in the soup of mutable-state
19:50	zer	well, i like it, when constructing things, just applying sequentiel mutations and not caring about the return value
19:50	amalloy	just use ->
19:50	georgek	hi, can anyone tell me what might be a typical problem that causes the error "clojure.lang.Var$Unbound cannot be cast to clojure.lang.Symbol"
19:51	amalloy	&(-> (transient {}) (assoc! :x 1) (assoc! :y 2) persistent!)
19:51	lazybot	⇒ {:x 1, :y 2}
19:51	zer	yes, this is what i will do from now on ty again
19:52	hiredman	georgek: calling resolve incorrectly
19:54	Gnosis-	what's the equivalent of subvec for a seq?
19:55	emezeske	Gnosis-: I think I usually end up using take and drop for that
19:55	Gnosis-	hmm, okay
19:57	georgek	thanks hiredman
19:57	Gnosis-	also, it looks like BigInts don't support a lot of bit-* operations
20:00	georgek	I'm working with a library with two files which I require in my ns; file A has a call to a function X. There's a function in file B that I think depends on the call to function X. If I require both files into my ns, is the call to function X in file A 'automatic'
20:00	georgek	in other words, so file B can then rely on it?
20:01	georgek	this is the lib, https://github.com/ngrunwald/clj-rome/tree/master/src/clj_rome
20:01	georgek	I think the translate fn in fetcher.clj relies on register-converters in reader.clj
20:09	Gnosis-	how do I test if something is an array?
20:12	georgek	Gnosis, do you mean this, http://stackoverflow.com/questions/9090396/testing-whether-an-object-is-a-java-primitive-array-in-clojure
20:12	Gnosis-	yeah
20:13	Gnosis-	also, why the hell are bytes signed in Java?
20:15	Gnosis-	georgek: the Class/forName thing worked for what I needed, thanks!
20:20	zer	shame, i cant do
20:20	zer	(-> (transient {}) #(if pred (assoc! % :a 'a) %) persistent!)
20:20	zer	is there any way that does not use let for each conditional or an atom?
20:27	gfredericks	,(-> (transient {}) (#(if pred (assoc! % :a 'a) %)) persistent!)
20:27	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: pred in this context, compiling:(NO_SOURCE_PATH:0)>
20:27	gfredericks	,(-> (transient {}) (#(if true (assoc! % :a 'a) %)) persistent!)
20:27	clojurebot	{:a a}
20:27	gfredericks	zer: you just gotta respect how -> works is all
20:28	zer	sweet
20:40	gfredericks	so if I want to be able to use #bytes [1 2 3], do I just go define that myself or is there a standard library or something for these things?
20:41	amalloy	gfredericks: do it yourself. but you need to use a namespaced tag; others are reserved for clojure.core
20:42	gfredericks	how does this work with lein libraries? I put the file in /src?
20:42	amalloy	yeah
20:42	amalloy	at one of the classpath's roots
20:43	gfredericks	coothx.
20:44	gfredericks	incidentally how does clojure or anything handle situations with identically named files all at the root of the classpath?
20:48	pepijndevos	you probably get the one thats on the path first?
20:48	gfredericks	I guess there must be an API for getting multiple files with one name
20:49	gfredericks	"When Clojure starts, it searches for files named data_readers.clj at the root of the classpath."
20:50	hiredman	gfredericks: classloaders have a getResources method
20:51	gfredericks	hiredman: entiendo
20:51	gfredericks	also the 1.4 changelog doesn't mention how you go about writing reader tags
20:51	gfredericks	does this preclude pr-str?
20:52	hiredman	you muck with print-method or whatever
20:52	gfredericks	ah
21:36	Scriptor	overtone seems to crash whenever I try running it from the slime repl
21:36	Scriptor	otherwise works fine from the command-line
21:37	casion	Scriptor: I've not been able to get overtone working with slime myself
21:40	Scriptor	casion: does java crash every time you try (use 'overtone.live) ?
21:40	casion	Scriptor: yep
21:40	Scriptor	have you tried with the external server?
21:40	casion	there was another mysterious freeze at times that I never could figure out
21:40	casion	no I havent, I gave up on it
21:41	tetrad_	HiroProtagonist: www.siliconmicrodisplay.com/st1080.html
21:41	casion	I use supercollider directly, so getting overtone working is not a high priority for me
21:52	gfredericks	(inc tagged-reader-literals)
21:52	lazybot	⇒ 1
22:06	aniero	(take 3 (repeatedly (constantly "heh")))
22:09	xeqi	,(repeate 3 "heh")
22:09	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: repeate in this context, compiling:(NO_SOURCE_PATH:0)>
22:09	xeqi	,(repeat 3 "heh")
22:09	clojurebot	("heh" "heh" "heh")
22:13	aniero	:)
22:15	aniero	is it better to do (zipmap ks (repeatedly (constantly val))) or (zipmap ks (repeat (count ks) val)) ? style choice?
22:15	casion	,"hehhehheh"
22:15	clojurebot	"hehhehheh"
22:15	casion	you guys should write more concise code ;)
22:17	xeqi	&(zipmap [:a :b :c] (repeat "heh"))
22:17	lazybot	⇒ {:c "heh", :b "heh", :a "heh"}
22:17	aniero	ohh, length is optional in repeat, thanks!
23:03	Gnosis-	how do I get a helpful stacktrace at the REPL?
23:03	Raynes	(.printStackTrace *e)
23:06	Gnosis-	it looks like a big mishmash of .invoke(...) to me :(
23:16	jlewis	when is it possible that a PersistentHashMap would return an entry with key and value = null, and yet not have hasNull set?
23:16	jlewis	i'm trying to track down a bug in my program, and it's proving difficult - I can't find out how that null entry gets in the map
23:22	muhoo	well, i'm 50% of the way to convincing a customer to let me deliver the user interface for an android tablet app in cljs. this will be interesting.
23:35	lpvb	/close
23:38	amalloy	jlewis: i don't know why you would ever find out that it has hasNull set
23:38	aniero	aha, (take-while identity ...) instead of (take-while (complement nil?) ...)
23:39	jlewis	amalloy: i'm using PersistentHashMap from java. I'm trying to debug this issue, and I'm noticing the unexpected behavior that there's a null->null pair in the entry set of the map, yet hasNull is not set
23:40	jlewis	i don't assoc a null anywhere... i'm kind of at my wit's end :)
23:40	amalloy	and you're looking at its hasNull in a debugger then?
23:40	jlewis	yes
23:42	amalloy	i don't think that behavior is possible, and i don't think anyone can debug it without seeing what code you're building up the map with
23:43	jlewis	i was afraid someone would say that... i don't have anything near to a minimal test case
23:46	jlewis	amalloy: oh, i suppose it is very much worth mentioning that i am using https://github.com/krukow/clj-ds, not clojure itself. perhaps it has a bug.
23:46	lynaghk	muhoo: we're doing a cljs iPhone app right now. Might be fun to compare notes sometime.
23:47	amalloy	gosh. probably it has a bug, then
23:47	jlewis	that is a serious bummer!
23:57	muhoo	lynaghk: cool, yes it would