#clojure – 2014-08-27

2014-08-27

00:40	danielcompton	On my gravestone when I die: This tombstone is not part of GNU Emacs.
01:22	amalloy	celwell: yes, that's exactly what you can/should do
03:41	andrewchabmers	\join #clojurescript
03:45	andrewchambers	how hard is it to call c++ code from clojure?
03:47	beamso	you would be calling out from clojure through java (jni or jna) to call the c++ code
03:47	andrewchambers	hmm
03:47	andrewchambers	im trying to work out the best way to read in llvm bitcode files into clojure
03:48	andrewchambers	and convert them to native clojure data structures
03:50	beamso	http://stackoverflow.com/a/17401949 ?
03:52	cYmen	Morning #clojure!
04:08	sm0ke	,(instance? java.util.Map {}_)
04:08	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: _ in this context, compiling:(NO_SOURCE_PATH:0:0)>
04:09	sm0ke	hurm this seems to work on my repl
04:09	sm0ke	,(clojure-version)
04:09	clojurebot	"1.7.0-master-SNAPSHOT"
04:09	sm0ke	##(instance? java.util.Map {}_)
04:09	lazybot	⇒ true
04:10	sm0ke	ha
04:10	sm0ke	##(clojure-version)
04:10	lazybot	⇒ "1.4.0"
04:10	engblom	,(/ 1 0)
04:10	clojurebot	#<ArithmeticException java.lang.ArithmeticException: Divide by zero>
04:11	sm0ke	seems like a bug in compiler to me ##(doc instance?) ;as doc has only 2 arity
04:11	lazybot	⇒ "([c x]); Evaluates x and tests if it is an instance of the class c. Returns true or false"
04:12	sm0ke	##(assoc {} 1 {}_)
04:12	lazybot	java.lang.RuntimeException: Unable to resolve symbol: _ in this context
04:12	sm0ke	very peculiar
04:17	amalloy	uhhhh...peculiar in that you typed a _ in the middle of nowhere?
04:18	sm0ke	##(instance? java.util.Map {}_)
04:18	sm0ke	yes
04:18	lazybot	⇒ true
04:20	sm0ke	looks funny right?
04:21	amalloy	it was fixed in 1.6, i think. maybe 1.5
04:21	amalloy	instance? used to be totally awful if you called it with any number of args other than 2
04:21	amalloy	&(instance? String)
04:21	lazybot	⇒ false
04:21	amalloy	&(instance?)
04:21	lazybot	clojure.lang.ArityException: Wrong number of args (0) passed to: core$instance-QMARK-
04:21	amalloy	0 was fine too, i guess
04:22	sm0ke	hurm how is this possible, is instance? a macro?
04:23	sm0ke	only possible explaination
04:23	amalloy	it's...well, no
04:23	amalloy	there are stranger things in this world than what you consider possible
04:23	sm0ke	weel then how can it take more than 2 arity
04:23	sm0ke	yea right, that explaination is awesome
04:23	wjlroe	is it a special form?
04:24	amalloy	no, it's actually a regular function
04:24	amalloy	but there's a compiler intrinsic to turn literal calls to instance? into a single jvm bytecode op instead
04:24	sm0ke	##(inc 1 2)
04:24	lazybot	clojure.lang.ArityException: Wrong number of args (2) passed to: core$inc
04:24	amalloy	and in previous versions of clojure that intrinsic was written wrong
04:24	sm0ke	wow it says wrong arity, thank god
04:24	amalloy	such that it didn't complain about the wrong number of args
04:25	amalloy	or something like that. i think the intrinsic system is actually smarter than that, and it was actually a special case in the compiler
04:25	wjlroe	ah right - as it's a convenience function that calls java methods
04:25	andrewchambers	belated goodmorning cYmen
04:27	andrewchambers	beamso: Those seem pretty out of date
04:39	cYmen	I'd like to work on a medium/large clojure project. Is there a list of clojure open source projects available somewhere?
04:40	cataska	2014-08-27 16:31 *** andrei_ QUIT Remote host closed the connection
04:40	cataska	Oops
05:31	visof	,(sorted-map {:a 1, :d 2, :b 3})
05:31	clojurebot	#<IllegalArgumentException java.lang.IllegalArgumentException: No value supplied for key: {:b 3, :d 2, :a 1}>
05:31	visof	why this don't work?
05:32	visof	,(sorted-map :a 1, :d 2, :b 3)
05:32	TEttinger	,(apply sorted-map {:a 1, :d 2, :b 3})
05:32	clojurebot	{:a 1, :b 3, :d 2}
05:32	clojurebot	#<IllegalArgumentException java.lang.IllegalArgumentException: No value supplied for key: [:a 1]>
05:32	TEttinger	that was mine
05:32	TEttinger	that errored I mean
05:33	TEttinger	visof, it seems like it should. you may be looking for sort
05:33	TEttinger	,(class (sort {:a 1, :d 2, :b 3}))
05:33	clojurebot	clojure.lang.ArraySeq
05:33	TEttinger	,(class (sorted-map :a 1, :d 2, :b 3))
05:33	clojurebot	clojure.lang.PersistentTreeMap
05:33	TEttinger	hm
05:39	blorg	any1 doing clj gamedev?
05:39	justin_smith	,,(into (sorted-map) {:a 0 :b 1 :c 2})
05:39	clojurebot	{:a 0, :b 1, :c 2}
05:41	blorg	((fn [xs n] (flatten (partition (dec n) n nil xs))) [1 2 3 4 5 6 7 8] 3)
05:42	justin_smith	~flatten
05:42	clojurebot	flatten is rarely the right answer. Suppose you need to use a list as your "base type", for example. Usually you only want to flatten a single level, and in that case you're better off with concat. Or, better still, use mapcat to produce a sequence that's shaped right to begin with.
05:44	blorg	,,(filter (complement nil?) (map #(if (= %1 1) %2) (cycle (conj (into [] (repeat (dec 3) 1)) 0)) [1 2 3 4 5 6 7 8] ) )
05:44	clojurebot	(1 2 4 5 7 ...)
05:44	justin_smith	,((fn [xs n] (apply concat (partition (dec n) n nil xs))) [1 2 3 4 5 6 7 8] 3)
05:44	clojurebot	(1 2 4 5 7 ...)
05:44	blorg	whats up with the commas?
05:45	blorg	,( whats (up (with (the (commas huh?)))))
05:45	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: whats in this context, compiling:(NO_SOURCE_PATH:0:0)>
05:45	justin_smith	,"it triggers the bot"
05:45	clojurebot	"it triggers the bot"
05:46	blorg	," blorgon is my master!!!"
05:46	clojurebot	" blorgon is my master!!!"
05:46	blorg	," I fear the daeleks!!"
05:46	clojurebot	" I fear the daeleks!!"
05:47	blorg	nice gimmick...
05:48	blorg	,((comp drop-last #(interleave %2 (repeat (count %2) %1))) 0 [1 2 3])
05:48	clojurebot	(1 0 2 0 3)
05:50	blorg	,"OBEY THE BLORGONS justin_smith!"
05:50	clojurebot	"OBEY THE BLORGONS justin_smith!"
05:50	justin_smith	,(drop-last (interleave [1 2 3] (repeat 0)))
05:50	clojurebot	(1 0 2 0 3)
05:53	SagiCZ1	what other ~ tips does the bot know?
05:53	SagiCZ1	~atom
05:53	clojurebot	Titim gan éirí ort.
05:54	SagiCZ1	is that irish?
05:54	blorg	~~
05:54	justin_smith	~factoids
05:54	clojurebot	Excuse me?
05:54	clojurebot	Excuse me?
05:54	justin_smith	SagiCZ1: I believe so
05:54	blorg	~'~
05:54	clojurebot	Pardon?
05:54	blorg	~`~
05:54	clojurebot	No entiendo
05:54	SagiCZ1	~blorg
05:54	clojurebot	Huh?
05:54	blorg	ok ill leave u be
05:55	blorg	a whole REPL with one purpose in life: listining to an IRC channel...
05:55	justin_smith	(inc clojurebot)
05:55	lazybot	⇒ 43
05:55	blorg	funny
05:55	blorg	~clojurebot
05:55	clojurebot	clojurebot ignores several people at hiredman's whim
05:56	justin_smith	3d6
05:56	clojurebot	17
05:56	blorg	who built him?
05:56	justin_smith	https://github.com/hiredman/clojurebot
05:57	justin_smith	note the ~clojurebot factoid also mentions hiredman
05:57	blorg	'(def clojurebot dead)
05:57	blorg	,(def clojurebot dead)
05:57	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: dead in this context, compiling:(NO_SOURCE_PATH:0:0)>
05:58	blorg	,(def clojurebot nil)
05:58	clojurebot	#'sandbox/clojurebot
05:59	SagiCZ1	i love the bot though
05:59	blorg	he is super cute
05:59	SagiCZ1	yup\|
05:59	TEttinger	I can answer questions about lazybot if anyone needs em
06:00	SagiCZ1	TEttinger: Nice try
06:00	blorg	lazybot?
06:00	clojurebot	lazybot is forget lazybot
06:00	SagiCZ1	but not its clojurebot-time!
06:00	SagiCZ1	(doc println)
06:00	clojurebot	"([& more]); Same as print followed by (newline)"
06:00	blorg	any1 read the post about how clojure/lisp coders are happier?
06:00	SagiCZ1	no?
06:00	TEttinger	##(clojure.string/join" "(repeatedly 2000(fn [](apply str(concat[(rand-nth ["rh""s""z""t""k""ch""n""th""m""p""b""l""g""phth"])](take(+ 2(* 2(rand-int 2)))(interleave(repeatedly #(rand-nth ["a""a""o""e""i""o""au""oi""ou""eo"]))(repeatedly #(rand-nth ["s""p""t""ch""n""m""b""g""st""rst""rt""sp""rk""f""x""sh""ng"]))))[(rand-nth ["os""is""us""um""eum""ium""iam""us""um""es""anes""eros""or""ophon""on""otron"])])))))
06:00	clojurebot	no is tufflax: there was a question somewhere in there, the answer
06:00	lazybot	⇒ "thategon pounotus phthortor phthomertus maspanes taspaustis choisterkeros choifophon kortor mospamium pausheros pangospos boifum rhoseros netor choufanes mauchus kananes maumon rhobus thages thautespis laspongeum pouchor mertobium zetamophon postachotron taunopum no... https://www.refheap.com/89505
06:01	TEttinger	the refheap link is handy
06:01	SagiCZ1	TEttinger what just hapenned?
06:01	TEttinger	it's a random greek-like word generator
06:01	TEttinger	in one IRC line
06:01	TEttinger	411 chars
06:02	TEttinger	413 if you include the prepend for lazybot
06:02	SagiCZ1	sweet! does lazybot link the log automatically if the output is too long?
06:02	blorg	wt is diff lazy vs clj?
06:02	TEttinger	yep!
06:02	TEttinger	blorg, different authors, different purposes
06:02	SagiCZ1	,(take 456 (range 1000))
06:02	clojurebot	(0 1 2 3 4 ...)
06:03	blorg	k
06:03	TEttinger	##(take 456 (range 1000))
06:03	lazybot	⇒ (0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 ... https://www.refheap.com/89506
06:03	SagiCZ1	ok so its double hashmark for lazybot
06:03	TEttinger	only lazybot links to refheap
06:03	TEttinger	there's othrs than ##
06:03	SagiCZ1	(inc lazybot)
06:03	lazybot	⇒ 30
06:03	TEttinger	I think $ or & works, one of them
06:03	SagiCZ1	ok.. good work btw.. !
06:04	justin_smith	double hashmarks are for ##(print "inline" "calls")
06:04	lazybot	⇒ inline callsnil
06:04	TEttinger	but ##(+ 1 1) can be inside another message
06:04	lazybot	⇒ 2
06:04	justin_smith	&"only works at the beginning of the line"
06:04	lazybot	⇒ "only works at the beginning of the line"
06:04	blorg	this irc shud be renamed to shoe-your-clojurebot-day
06:04	blorg	*show
06:04	TEttinger	SagiCZ1, heh, it's Raynes and amalloy_'s work, I just extended the plugins for my fork that wouldn't be a good fit here
06:05	blorg	bring-your-clojure-bot-to-scholl-day
06:05	SagiCZ1	TEttinger: I see.. :)
06:06	SagiCZ1	i better start doing some real coding in clojure.. its so hard to learn any new languages.. at first i dont have enough experience to really code but i dont really learn if i am just doing artifical exercises :/
06:06	SagiCZ1	i had the same problem with java 5 years ago but ive gotten through taht somehow
06:07	blorg	u know 4clojure? (sagiCZ1)
06:07	SagiCZ1	blorg: yep
06:07	SagiCZ1	almost done with that
06:07	blorg	hmm.. im only at like problem 50 or so..
06:07	SagiCZ1	mm.. me too
06:08	SagiCZ1	there is about 150 so i guess it doesnt really qualify as being "almost done" sorry
06:08	blorg	but i need clj for some project so i know ill pull thru
06:08	SagiCZ1	i need clj too
06:08	SagiCZ1	but its too early to start on that project..
06:08	SagiCZ1	i would be rewriting all of it
06:08	SagiCZ1	need some small toy projects before that
06:08	blorg	ye..
06:09	blorg	genetic algo is small enuf..
06:10	blorg	a small MS-paint clone.. in clj...
06:10	blorg	IamDrowsy too
06:10	SagiCZ1	cool ideas.. ive done some evolutionary algos too.. clojure just seemed so very natural with that
06:12	cYmen	IamDrowsy?
06:12	blorg	cYmen?
06:12	SagiCZ1	blorg?
06:12	blorg	clojurebot?
06:12	clojurebot	clojurebot is amazing
06:13	SagiCZ1	lazybot?
06:13	clojurebot	lazybot is forget lazybot
06:13	SagiCZ1	...but i still dont know how to replace java's objects in clojure.. i know i talk about it all the time here.. objects just make sense to me
06:13	cYmen	Interesting...
06:13	blorg	objects are bad
06:14	SagiCZ1	are they?
06:14	cYmen	SagiCZ1: Which part do you miss? The type-checking? Inheritance?
06:14	blorg	yes...
06:14	SagiCZ1	none of that.. just the data storage and methods that mutate their state.. i just miss mutability i guess
06:14	ivan	I miss the state
06:14	cYmen	I mean an object is just a bunch of data and a bunch of related methods.
06:15	SagiCZ1	cYmen: data that can be mutated
06:15	cYmen	But you can just use a map or tuple or whatever and still store data i.e. state.
06:15	blorg	u can hack objects together in any lang
06:15	blorg	almost..
06:15	SagiCZ1	map is ok.. but its immutable
06:15	cYmen	But why does it matter if you do object.mutate() instead of object = mutate(object)?
06:15	ivan	SagiCZ1: the common thing is to use an atom
06:15	SagiCZ1	atom is for special cases though isnt it
06:16	blorg	im trying to do OOP rehab
06:17	SagiCZ1	so what i would do.. is i would have a global variable atom.. and functions with sidefect that mutate the atom.. it just seems like a horrible misuse of clojure
06:17	cYmen	SagiCZ1: I think this is perfectly appropriate in some cases.
06:17	blorg	try exploring haskell mbe.
06:17	justin_smith	cYmen: it matters because the affects none of the data structures embedding object
06:18	blorg	its less usefull IMO than clj.. but it forces u to think diff
06:18	justin_smith	*the latter
06:18	SagiCZ1	blorg: i love the JVM.. can't live without it.. the apache math library is a lifesaver.
06:18	cYmen	justin_smith: I still can't parse that sentence. :)
06:18	SagiCZ1	justin_smith: could you elaborate please?
06:18	blorg	ye i agree that haskell isnt very usefull
06:21	justin_smith	if you have (def foo [o]) (def o (bar o)) does not change foo at all - (.mutate o) does change the contents of foo
06:22	SagiCZ1	yeah.. so it has no state
06:23	justin_smith	SagiCZ1: check out some of the alioth shootout haskell code
06:23	justin_smith	if you want it to, it can definitely work in terms of pure mutation
06:23	justin_smith	(that just happens to lead to ugly terrible code - but it performs well)
06:24	SagiCZ1	i see.. but i dont want ugly code :)
06:24	justin_smith	I find it informative to compare the ugly ass unidiomatic haskell code, to the more idiomatic ocaml that performs nearly as well
06:25	blorg	my end goal is make a very "modular"/programable game. since i can do the OOD for games in my sleep its very hard for me to switch to a functional mindset.. but thats the only way to go...
06:26	blorg	justin_smith: hmm.. what?
06:26	lvh_	I'm having some difficulty understanding when to use futures, promises and core.async channels. I have an api for getting something out of a key value store. futures don't seem appropriate because it's not cpu-blocking (although it could be that, because you're wrapping a synchronous API, blocking a thread is really what it has to do)
06:27	lvh_	promises feel most appropriate, but everything else in my code already has channels
06:27	lvh_	does a channel that will only ever return one thing (an error or a success) and then close make any sense?
06:27	justin_smith	blorg: what, what?
06:28	blorg	"ugly ass unidiomatic haskell code, to the more idiomatic ocaml that performs nearly as well"
06:29	justin_smith	right, on alioth
06:29	justin_smith	http://benchmarksgame.alioth.debian.org/ pick a benchmark, compare haskell code to ocaml code, and their performance
06:29	SagiCZ1	oh.. now i know what alioth is
06:29	justin_smith	http://benchmarksgame.alioth.debian.org/u64q/haskell.php
06:30	justin_smith	sorry, this is the link I meant to share http://benchmarksgame.alioth.debian.org/u64q/benchmark.php?test=all&lang=ghc&lang2=ocaml&data=u64q
06:32	SagiCZ1	most of clojure programs are slower than java and some take more lines of code.. thats surprising
06:33	blorg	justin_smith: both seem ugly to me...
06:37	blorg	this http://benchmarksgame.alioth.debian.org/u32/program.php?test=binarytrees&lang=racket&id=2 however seems so nicer. and only twice slower
06:49	blorg	hello glorbon
06:50	glorbon	hello blorgon
06:54	SagiCZ1	are you two, by any chance, each others evil twin?
06:57	cYmen	justin_smith:Very often you can replace mutation with getting a new object except in cases very there are multiple references to the same thing which is exactly the case when you generally don't want mutation. :)
06:57	justin_smith	yeah, that's kind of what I was trying to say
06:58	glorbon	YES WE ARE !!
06:58	glorbon	(which makes us both evil!)
06:58	blorg	SO EVIL
07:00	glorbon	cYmen: u r sayin one can do good OOP?
07:01	mindbender1	Is there a clojure library wrapping java.nio?
07:02	Bronsa	mindbender1: https://github.com/pjstadig/nio
07:02	cYmen	glorbon: I don't think I said anything about OOP
07:03	glorbon	ook.. sry then
07:04	vijaykiran	clear
07:04	vijaykiran	meha
07:04	mindbender1	Bronsa: thanks.
07:04	Bronsa	np
07:09	justin_smith	mindbender1: there is also an nio.file wrapper
07:10	mindbender1	justin_smith: how can I access it?
07:10	justin_smith	https://github.com/ToBeReplaced/nio.file I am a contributor, prs welcom
07:11	mindbender1	justin_smith: cool, thanks. Checking it out.
07:55	glorbon	so quiet
07:56	SagiCZ1	im listening!
07:58	glorbon	are you? are . . . you . . .
08:00	glorbon	,"lets party"
08:00	clojurebot	"lets party"
08:00	glorbon	we agree then
08:03	glorbon	,((fn do-u-love-me-clojurebot? [_] (str "YES!"))0)
08:03	clojurebot	"YES!"
08:03	SagiCZ1	~party
08:03	clojurebot	Pardon?
08:04	SagiCZ1	why are u passing zero to it?
08:04	glorbon	y not
08:05	glorbon	idk actually..
08:05	glorbon	,((fn do-u-love-me-clojurebot? [_] (str "YES!")))
08:05	clojurebot	#<ArityException clojure.lang.ArityException: Wrong number of args (0) passed to: sandbox/eval73/do-u-love-me-clojurebot?--74>
08:05	glorbon	o.. thts y..
08:05	glorbon	,((fn do-u-love-me-clojurebot? (str "YES!")))
08:05	clojurebot	#<CompilerException java.lang.IllegalArgumentException: Parameter declaration str should be a vector, compiling:(NO_SOURCE_PATH:0:0)>
08:05	SagiCZ1	,((fn do-u-love-me-clojurebot? [] (str "YES!")))
08:05	clojurebot	"YES!"
08:06	SagiCZ1	this way
08:06	glorbon	k
08:06	glorbon	coolio
08:07	SagiCZ1	and btw..
08:07	SagiCZ1	,(= (str "YES!") "YES!))
08:07	clojurebot	#<RuntimeException java.lang.RuntimeException: EOF while reading string>
08:08	glorbon	(#(do % "YES"))
08:08	SagiCZ1	,(= (str "YES!") "YES!"))
08:08	clojurebot	true
08:08	glorbon	((#(do % "YES"))0)
08:08	glorbon	(#(do % "YES") 0)
08:08	glorbon	?
08:08	glorbon	oh forgot commas
08:08	glorbon	heh
08:09	glorbon	,(#(do % "YES") 0)
08:09	clojurebot	"YES"
08:11	hyPiRion	constantly yes.
08:11	samflores	,(#(constantly "YES!"))
08:11	clojurebot	#<core$constantly$fn__4178 clojure.core$constantly$fn__4178@12b700f>
08:12	glorbon	,(#(constantly "YES!") 0)
08:12	clojurebot	#<ArityException clojure.lang.ArityException: Wrong number of args (1) passed to: sandbox/eval51/fn--52>
08:12	samflores	,(doc constantly)
08:12	clojurebot	"([x]); Returns a function that takes any number of arguments and returns x."
08:12	glorbon	hmm
08:13	glorbon	,((constantly "YES!") 0)
08:13	clojurebot	"YES!"
08:13	nathan7	remember, #() makes a function
08:14	nathan7	,(#(do %& 42) 0)
08:14	clojurebot	42
08:14	samflores	I always forgot constantly makes a fn :D
08:15	glorbon	,(reduce (constantly "YES") ["does" "clj" "love" "me?"])
08:15	clojurebot	"YES"
08:15	glorbon	aww shucks
08:16	samflores	*forget
08:16	glorbon	huh?
08:17	glorbon	(can I add stuff to the bot or is it protected via clojail and such?)
08:17	hyPiRion	it's semiprotected
08:18	glorbon	elaborate?
08:18	andrewchambers	is there a tool for pretty printing edn?
08:22	hyPiRion	clojure.pprint/pprint ?
08:25	lvh	What happened to clojure.contrib again? Especially clojure.contrib.base64
08:25	lvh	(I want base64 urlencoding and decoding)
08:25	lvh	Sorry, base64url*
08:26	lvh	nothing to do with urlencoding; just base64 with a url-safe alphabet :)
08:42	lvh	argh
08:42	spradnyesh	if there are multiple expressions (w/ side-effects) in a let body, will they be executed serially or in parallel?
08:42	lvh	there was a library on github that did this right for clojure (base64url enc/dec)
08:42	spradnyesh	i'm seeing parallel execution, but want serial
08:45	spradnyesh	for example, in (let [x 1] (do-a) (do-b)) i'm seeing errors which should happen only if do-b is executed before do-a has finished
08:45	dnolen_	spradnyesh: serial
08:45	spradnyesh	i haven't put either in a future/Thread
08:45	dnolen_	spradnyesh: then present a minimal case that people can look at - but it sounds implausible
08:45	spradnyesh	dnolen_: thanks. that is what i expected too; but that's not what i'm seeing :(
08:47	spradnyesh	do-a cleans-up something from mongodb, and do-b reads from mongodb and processes and writes back. i'm seeing a duplicate key error
08:47	spradnyesh	also, i get the error back almost instantaneously, whereas do-a should have taken atleast 1-2 sec to complete
08:48	dnolen_	spradnyesh: sounds like a problem with a) mongodb b) your mongodb driver
08:48	spradnyesh	i'm thinking of putting do-a inside a future, and then referencing it from do-b. that should work, right?
08:48	spradnyesh	dnolen_: i'm using monger-1.7.0
08:49	dnolen_	spradnyesh: I don't know anything about the mongodb but I can say w/ confidence Clojure execution is serial. Look for the problem in more likely sources.
08:49	spradnyesh	dnolen_: what do you think about my using "future" as a bypass/solution as suggested above?
08:50	dnolen_	spradnyesh: sounds ok to me.
08:51	cYmen	Is it possible that the mongodb call returns immediately because there is no return value it needs to wait for?
08:51	cYmen	So even though it isn't done it returns already?
08:53	spradnyesh	cYmen: i don't think so; http://docs.mongodb.org/manual/reference/method/db.collection.update/ says that update returns result of update. but thanks for pointer :)
08:53	spradnyesh	dnolen_: i'll try that out
08:58	cYmen	dnolen_: Is it possible that clojure simply runs the commands serially withou waiting for them to return?
09:00	spradnyesh	found out the issue: do-a has a (map (some-side-effect) some-seq) as it's last expr. this made the some-side-effect lazy. trying to do a dorun on it now
09:03	cYmen	ah well :)
09:11	ebaxt	I've been using Stuart Sierra's "Clojure Workflow, Reloaded" based on tools.namespace for over a year now. I've recently started to experience PermGen exceptions when reloading my project. Could this have anything to do with http://dev.clojure.org/jira/browse/CLJ-1152? I haven't looked at the mechanics of tools.namespace reloading so I don't know if it might be related...
09:13	stuartsierra	ebaxt: Don't know if it's related, but I've certainly experienced PermGen problems when (re)loading large projects (hundreds of source files).
09:14	stuartsierra	tools.namespace removes namespaces before reloading them, so things like protocols & multimethods should be GC'd.
09:15	ebaxt	stuartsierra: about 10K loc in 138 files
09:17	stuartsierra	ebaxt: That might be enough to fill the PermGen, depending on your JVM settings.
09:17	stuartsierra	See if increasing the PermGen size fixes the problem.
09:17	ebaxt	stuartsierra: Any PermSize suggestions? Lein repl use :jvm-opts from user profiles right?
09:18	stuartsierra	ebaxt: Look up the default PermGen size for your JVM, make it bigger, see if that fixes the problem :)
09:20	ebaxt	stuartsierra: hehe, thx, I'll do that. BTW, we're using some libs that keep state in refs, could that be part of the problem?
09:20	stuartsierra	ebaxt: don't know
09:21	ebaxt	k
09:21	stuartsierra	but I doubt it
09:21	hyPiRion	ebaxt: not sure how that would affect permgen
09:21	stuartsierra	it might break in other ways when you try to reload though
09:29	ebaxt	:jvm-opts ^:replace [] in project.clj will ignore :jvm-opts in .lein/profiles.clj?
09:37	stuartsierra	ebaxt: not sure; test or try asking later when the U.S. West Coast wakes up :)
09:38	SagiCZ1	clojure is trending https://www.google.cz/trends/explore#q=clojure
09:40	TimMc	Finland, interesting.
09:41	hyPiRion	Norway whut
09:41	hyPiRion	I feel like there's like 5 Clojure devs over here
09:41	llasram	Add one more and it's up 20%
09:42	Foxboron	hyPiRion: Norwegian?
09:42	hyPiRion	Foxboron: yes
09:43	Foxboron	is there actually anyone doing professional CLojure development in Norway?
09:44	hyPiRion	Foxboron: yeah, pretty sure ebaxt is
09:44	Foxboron	oh, thats fun. Got a friend who wrote his master thesis in Clojure
09:44	Foxboron	been dabbling with it since autumn 2012
09:45	Foxboron	<- Norwegain aswell (to get that straight)
09:45	hyPiRion	I am working in Clojure at work.
09:45	Foxboron	hyPiRion: there was also a Clojure talk at Boosterconf IIRC
09:46	hyPiRion	Foxboron: oh nice. I know augustl is working with Clojure as well, but not sure if that's professionally or on a hobby basis
09:46	ebaxt	Foxboron: We are, (small startup ardoq.com) I know a couple of other small ones and I think schibsted (https://github.com/schibsted)
09:47	Foxboron	ebaxt: thats fun. I had no idea there where places in Norway doing that
09:48	hyPiRion	Oh, and I think Bekk has a single project in Clojure. But they seem to prefer Scala -- I think they have 3 projects with Scala.
09:48	ebaxt	Foxboron: There are :) Kodemaker has at least one project, maybe more magnars / augustl ?
09:49	Foxboron	i know magnars had a Clojure online thingie developing a zombie game
09:49	Foxboron	hyPiRion: got a friend working at Bekk
09:50	Foxboron	he is mostly working in Javascript tho
09:51	hyPiRion	Foxboron: What I feel like is sort of a problem is that we don't have a Norwegian user group for Clojurists. That's probably why people don't see so many of these Clojure projects
09:52	Foxboron	hyPiRion: yeah, but there is a user group for functional langs in Oslo?
09:53	ebaxt	hyPiRion: You're based in Trondheim right?
09:53	hyPiRion	ebaxt: I was until June. Finished my master's then went down to Oslo
09:55	hyPiRion	Foxboron: yeah, true. But I still don't feel it's easy to figure out who to ask if I were to join a Clojure company.
09:55	hyPiRion	/s/were/wanted/
09:57	ebaxt	hyPiRion: agree, most of the people I know are using Clojure at work I met at http://www.meetup.com/Oslo-Socially-Functional/
09:59	hyPiRion	ebaxt: How often are there events with OsloSF? Seems like roughly monthly or so?
10:00	TimMc	Oslo, San Francisco, California, USA
10:01	ebaxt	hyPiRion: Every two months aprox. Trying to increase the number of meetings last i heard
10:04	hyPiRion	ebaxt: cool. I'll try to join and see if I can help out with anything
10:05	ebaxt	hyPiRion: Good idea :)
11:11	acagle	buffer 1
11:15	mdeboard	false
12:02	virmundi	Hello. I’ve got a question about naming standards. I’ve got a method that will take a map argument and execute an http call from it. The client gets to pick the HTTP method. So it could be a get or a post. Should the method’s name end with !?
12:04	bbloom	virmundi: no not really
12:04	bbloom	virmundi: ! does not mean side effectful
12:04	bbloom	otherwise entire subsections of your library would have bang on everything
12:04	justin_smith	do you literally mean a method? Clojure functions are objects.
12:05	virmundi	justin_smith: I mean function. I try to avoid the OO side of the underlying host. Old terms for a different pardigm and all.
12:05	virmundi	so the api has create-database!
12:05	CookedGryphon	the basic rule of thumb is, could you put it in a swap! statement
12:05	CookedGryphon	i.e. would it matter if it got executed 3 times over instead of one?
12:06	CookedGryphon	if it would matter, put a bang on it to warn other people
12:06	TimMc	meh
12:06	TimMc	I appreciate the thoroughness, but I think that's impractical.
12:06	bbloom	i probably wouldn't put a ! on create-database
12:06	bbloom	especially if it's idempotent
12:07	virmundi	bbloom: since it’s a rest call, the results will change. So POST first time, returns 201. POST a second time on the resource returns a 409 (I think, I just remember not 201).
12:08	bbloom	virmundi: just use ! to highlight a rare non-transaction-safe method among an otherwise transaction-safe api
12:08	bbloom	virmundi: if everything is effectful, then nothing is special enough to justify a !
12:09	teslanick	What is the actual standard for using ! anyway? I usually use it as a "be careful" or a "this is really impure" signifier
12:09	teslanick	i.e. non-idempotent.
12:09	justin_smith	&(filter #(re-matches #".*!$" %) (map (comp name first) (ns-publics 'clojure.core)))
12:09	justin_smith	no lazybot?
12:09	justin_smith	,(filter #(re-matches #".*!$" %) (map (comp name first) (ns-publics 'clojure.core)))
12:09	virmundi	bbloom: I thought that. Things get even weirder if the client uses the async feature of the driver, because that creates a resource for later lookup.
12:09	clojurebot	("set-error-mode!" "set-agent-send-executor!" "disj!" "conj!" "pop!" ...)
12:09	lazybot	java.lang.SecurityException: You tripped the alarm! ns-publics is bad!
12:10	mpenet	,(doc pop!)
12:10	clojurebot	"([coll]); Removes the last item from a transient vector. If the collection is empty, throws an exception. Returns coll"
12:10	mpenet	ah transient
12:11	bbloom	(doc persistent)
12:11	clojurebot	Huh?
12:11	bbloom	(doc persistent!)
12:11	clojurebot	"([coll]); Returns a new, persistent version of the transient collection, in constant time. The transient collection cannot be used after this call, any such use will throw an exception."
12:11	bbloom	the bang doesn't have a firm meaning
12:12	virmundi	the thing that made me wonder about the ! is that clojure.java.jdbc has insert! and execute! where as query is normal.
12:12	bbloom	it can be effectful, or non-idempotent, or dangerous, or whatever
12:12	bbloom	it's just to make it stand out when such an effect may not be expected
12:12	bbloom	virmundi: use your discretion
12:12	virmundi	Sounds fair. I appreciate the advice from all involved.
12:13	stuartsierra	There's no universal rule for `!` but it sometimes means "modifies mutable state", e.g. `swap!`
12:13	bbloom	(doc alter) ; stuartsierra
12:13	clojurebot	"([ref fun & args]); Must be called in a transaction. Sets the in-transaction-value of ref to: (apply fun in-transaction-value-of-ref args) and returns the in-transaction-value of ref."
12:13	bbloom	really, it means "pay attention, dummy" :-)
12:13	virmundi	:)
12:14	justin_smith	sorry guys, I got a massive lag suddenly
12:14	justin_smith	virmundi: in that case, (regarding avoiding the OO side of the underlying host), a function is self contained, a method belongs to some object
12:15	virmundi	justin_smith: I’m know. I’m jst not using precise language. Method is a hold over from the Java days. I first developed in VB so it took me a while to say method rather that function or sub.
12:15	stuartsierra	bbloom: like I said, not universal, also 'alter-var-root'
12:16	virmundi	justin_smith:I’ll try to be more precise.
12:17	justin_smith	np - not trying to be pedantic, when I first saw your question I really didn't know if you were talking about a normal function or an object made via perhaps gen-class with a method on it
12:18	justin_smith	and of course I use the relatively imprecise term "function" when really I mean "object implementing IFn"
12:18	noncom	are bitwise operations like (bit-shift-left), (bit-and) and the likes on ints in Clojure as fast as in Java ? or do we pay for some boxing and stuff ?
12:18	noncom	do i better do bitwise stuff in java ?
12:18	noncom	if i care for speed..
12:18	justin_smith	noncom: depends on how well things are hinted and / or hotspot compiled
12:19	noncom	uh..
12:19	justin_smith	noncom: there is a library that warns about numeric boxing and reflection... one moment
12:23	justin_smith	noncom: ztellman/primative-math provides reflection and boxing warnings (normal warn-on-reflection won't actually show this) https://github.com/ztellman/primitive-math
12:24	noncom	thanks :)
12:24	justin_smith	if you aren't boxing or reflecting, and you turn on unchecked-math, the numeric part of your code should be competitive with java
12:25	noncom	i still have to master all this..
12:26	stuartsierra	Just write Java.
12:26	noncom	:)
12:28	technomancy	shell out to bc
12:31	rhg135	inlime assembly 2.0
12:31	rhg135	inline*
12:32	justin_smith	create a REST service to do arithmetic, implemented in fortran
12:32	rhg135	whoa
13:04	bridgethillyer	It’s an interesting aspect of the Clojure ecosystem that it lives in the Java world
13:04	bridgethillyer	And I often wonder about the experience of people who are coming to Clojure with no Java experience
13:04	justin_smith	bridgethillyer: that was me - I had lisp experience but no java experience
13:05	bridgethillyer	Some things must be Totally Mysterious to them, and my guess is that is not a very comfortable experience
13:05	technomancy	same
13:05	bridgethillyer	And I just had an ah-hah! lightbulb moment about this
13:05	justin_smith	java is tedious, but not that surprising (most of the time)
13:05	noprompt	why doesn't metadata for clojurescript namespaces get added to cljs-ns? https://github.com/clojure/clojurescript/blob/master/src/clj/cljs/analyzer.clj#L1134-L1135
13:05	bridgethillyer	I am a {retired? reformed? defeated?} java programmer
13:05	technomancy	the main thing that confused me early on was the JDK's IO stuff
13:05	technomancy	but that was before clojure.java.io existed
13:06	bridgethillyer	So I have a LOT of experience with Maven
13:06	justin_smith	similar to man pages, javadoc is terrible to read, but once you learn how you can get the info you need quickly
13:06	bridgethillyer	But if you are not a Java programmer, likely you have little experience with Maven
13:06	mdrogalis	Maven will steal your lunch money. :/
13:06	noprompt	i probably should have aimed that at dnolen_ or bbloom
13:06	bridgethillyer	So here’s a tidbit of Maven know-how that comes in handy:
13:07	bridgethillyer	Releases you just get with leiningen
13:07	bridgethillyer	Voila they magically appear
13:07	aperiodic	technomancy: do you know when the stack overflow on dependency resolution issue started?
13:08	bridgethillyer	But SNAPSHOTS you actually have to go get yourself or add Sonatype repository
13:08	technomancy	aperiodic: AFAIK it's been around forever, but :pedantic? just got turned on by default in 2.4.3
13:09	bridgethillyer	So if your library is unreleased, someone is going to be totally mystified why it’s not there when they try to require it
13:09	aperiodic	technomancy: ok, so can it be worked around by turning pedantic off?
13:09	technomancy	aperiodic: I think so
13:09	technomancy	aperiodic: if not, try master and let me know
13:10	mdrogalis	bridgethillyer: Yeah, it's mildly infuriating. D:
13:10	technomancy	bridgethillyer: that's not any different from ruby
13:10	bridgethillyer	Anyhow, just found it interesting to actually uncover one of those weirdnesses about the non-Java experience. Generally I wouldn’t see them since I don’t have the perspective
13:10	aperiodic	technomancy: it's definitely fixed in master, I just misunderstood when you said it was a 2.4.3 issue yesterday (since my colleague just reproduced on 2.3.4)
13:11	bridgethillyer	technomancy: Well, it’s helpful to be aware of it
13:11	technomancy	aperiodic: gotcha, cool
13:11	technomancy	bridgethillyer: actually now I'm curious... in what language do you get access to unreleased libs without asking for them specifically?
13:11	technomancy	I guess you mean in systems that work from git directly?
13:11	bridgethillyer	technomancy: If you’re aware of it, you can do things to help orient non-Java people (assuming they are a big enough population to serve that way… I assume they are)
13:12	technomancy	bridgethillyer: I'm skeptical this has anything to do with java/non-java
13:13	bridgethillyer	technomancy: fair point; I guess it’s more that you don’t know how it happens because the machinery is specific to Java
13:13	justin_smith	my big surprise when I was new to clojure: the fact that you didn't install things via some analog of git clone, configure, make, make install
13:13	bridgethillyer	technomancy: So it is mysterious and difficult to diagnose/see
13:13	rhg135	juarez, thats just c(++)
13:14	rhg135	damn tab
13:14	bridgethillyer	technomancy: I’m not suggesting that anything that is happening is the wrong way to do it
13:14	rhg135	i can never get used to a chan with so many ppl
13:14	rhg135	justin_smith***
13:14	bridgethillyer	technomancy: I’m pointing out the user experience of it
13:15	technomancy	bridgethillyer: sure, it's always difficult to see things from the perspective of someone new.
13:16	bridgethillyer	technomancy: exactly and that is what I was able to experience
13:16	technomancy	bridgethillyer: that said, most snapshots are on clojars; contrib libs seem to like going out of their way to make things more complicated.
13:16	justin_smith	it's not just c - many languages have clone / build / install for libs (at least for the core language - this includes all the versions of common lisp and scheme I know of, ocaml, haskell...)
13:16	bbloom	noprompt: namespaces are not reified at runtime, so there's no way to get to the metadata from cljs
13:16	bbloom	noprompt: do you need it form a macro?
13:16	bridgethillyer	technomancy: yes, exactly… which makes for an even more mysterious experience
13:16	bbloom	s/form/from
13:16	noprompt	bbloom: yep. it's the macro use case.
13:16	technomancy	justin_smith: that's just legacy ocaml stuff; these days everything is done from opam. IIUC the same is true of haskell/cabal.
13:16	bbloom	noprompt: in that case, it's probably just an oversight. i can't speak for dnolen_, but i'd say patch is likely welcome
13:16	technomancy	justin_smith: also CL has quicklisp now, racket has had packages forever, chicken has eggs...
13:17	bridgethillyer	So I think the lesson to be taken away for library (cough contrib) authors is: if your library is unreleased, just go ahead and put this in your README:
13:17	justin_smith	technomancy: but for getting ocaml itself
13:17	technomancy	justin_smith: opam manages compiler installs too
13:17	justin_smith	technomancy: what was new is the language itself being managed by the package manager
13:17	bridgethillyer	To use Clojure and contrib library SNAPSHOTS in your Leiningen project, add the following to your project.clj: :repositories {“sonatype…
13:17	bbloom	noprompt: just add the metadata as :meta here: https://github.com/clojure/clojurescript/blob/245de2e752f638f5d06527a1cafed26b0d8d0da8/src/clj/cljs/analyzer.clj#L1166
13:17	bridgethillyer	So easy, makes things much easier for a set of your users
13:17	justin_smith	technomancy: and somehow I managed to be ignorant of the various language package managers for the most part
13:18	rhg135	those exist??
13:18	lazybot	rhg135: Uh, no. Why would you even ask?
13:18	aperiodic	technomancy: FYI turning pedantic off does work around the stack overflow
13:18	rhg135	shut up lazybot lol
13:19	noprompt	bbloom: awesome. i'll work up a patch.
13:20	noprompt	thanks.
13:20	technomancy	man, rebar. =(
13:20	justin_smith	technomancy: regardless of my relative ignorance, I was used to a paradigm where to try a language I could find the source, configure make and install, and then run it, without also needing to figure out a language specific dep management tool or packaging system or construct a "project", and this was my biggest surprise as a newcomer to clojure
13:21	technomancy	I unknowningly pulled in a patch that changed the dependencies, and suddenly I got test failures due to needing a newer version that I had no idea about. no dep updates, no warning, nothing.
13:21	technomancy	justin_smith: ah, gotcha
13:23	justin_smith	when my friend told me I had to download a shell script that downloaded another shell script and executed it and that was the right way to do clojure I was like "WTF YO"
13:24	dnolen_	noprompt: missing the backlog what's up?
13:25	noprompt	dnolen_: i was asking about metadata for cljs-ns. apparently it never gets used. for macro use cases that metadata is useful.
13:25	dnolen_	noprompt: what kind of metadata?
13:25	noprompt	dnolen_: namespace metadata.
13:26	dnolen_	noprompt: yes I never use this - what goes there that's missing?
13:27	noprompt	dnolen_: i'm writing a macro and i was hoping that i could read that metadata from cljs-ns but the only information that's there is :line, :column, etc.
13:27	dnolen_	noprompt: but what else is supposed to be there? I'm asking you a question because I don't know, hint, hint :)
13:28	bbloom	dnolen_: you can put arbitrary metadata on namespaces
13:28	bbloom	,(meta (the-ns 'clojure.core))
13:28	clojurebot	{:doc "Fundamental library of the Clojure language"}
13:29	bbloom	,(ns ^:omg my-ns)
13:29	clojurebot	nil
13:29	bbloom	,(meta (the-ns 'my-ns))
13:29	clojurebot	{:omg true}
13:29	dnolen_	ah right, yeah patch welcome that's an easy one
13:30	noprompt	dnolen_: cool, bbloom mentioned placing it here: https://github.com/clojure/clojurescript/blob/245de2e752f638f5d06527a1cafed26b0d8d0da8/src/clj/cljs/analyzer.clj#L1166
13:30	dnolen_	noprompt: you probably want to put it here https://github.com/clojure/clojurescript/blob/245de2e752f638f5d06527a1cafed26b0d8d0da8/src/clj/cljs/analyzer.clj#L1160
13:30	dnolen_	and we can dupe it into the analysis map as well
13:31	noprompt	dnolen_: cool, initially that's where i thought it might go.
13:32	noprompt	dnolen_: alrighty, well i'll work up a patch. fortunately i work with someone who knows the whole jira process (which i've been neglecting to learn). this'll get my feet wet.
13:32	noprompt	dnolen_, bbloom: thanks for the help. :)
13:32	mindbender1	where the issue tracker for org.clojure-andriod/clojure?
13:33	mindbender1	is the same for org.clojure/clojure?
13:39	toxmeister	clojars currently gives HTTP 500 (not sure if that's the best place to mention here...)
13:40	loliveira	hi, congomongo or monger?
13:41	Bronsa	toxmeister: it's a known issue with the homepage: https://github.com/ato/clojars-web/issues/235
13:43	toxmeister	bronsa: thx, it seems deploys are broken too… getting checksum failures. was going to double check my uploaded keys, but that will have to wait until the UI is fixed again
13:43	arrdem	loliveira: whichever, they both work fine
13:45	loliveira	arrdem: thank you. I will try monger. I’ve been using congomongo since forever and I ‘m wondering if I’d like monger.
13:46	arrdem	loliveira: I've used congomongo a before, just looked at monger and it doesn't look fundamentally different.
13:47	justin_smith	~mongodb
13:47	clojurebot	mongodb is an ORM for mmap
13:47	schmee	can `^:const` be used with arrays, like this? (def ^:const foo [6 3 0 7 4 1 8 5 2])
13:47	loliveira	arrdem: ty
13:48	schmee	I read the it only works with "primitive values", but I'm unsure of what that means
13:48	justin_smith	schmee: primitives are things in the jvm that are not objects
13:48	schmee	ahh, so int bool and the likes. I guess that const does nothing for arrays then?
13:48	justin_smith	http://docs.oracle.com/javase/tutorial/java/nutsandbolts/datatypes.html
13:49	justin_smith	right, an array is an object (as are vectors, and your example was a vector, not an array)
13:50	schmee	yeah, I always mix up vectors and arrays, using many different programming languages certainly doesn't help :P
13:50	schmee	thanks for the info
13:50	justin_smith	schmee: though I think :const can actually help on vars for non-primitive values, but Bronsa or arrdem would be better people to ask about the ins and outs of that
13:51	arrdem	Bronsa will know better than I do. I remember playing around with ^:const a while back and there is a bytecode difference in evaluating const tagged symbols but I don't think it's significant.
13:52	arrdem	s/symbols/vars/g ;; arguably pedantic
13:53	schmee	ok, I think I'll leave them out in that case
13:56	Bronsa	schmee: ^:const can be used with anything that has a clojure literal representation, but there are some issues; the only thing I'd feel safe using ^:const with are: numbers, strings and keywords
13:58	alexbaranosky_	Bronsa: curious what the issues are
14:06	Bronsa	alexbaranosky metadata might or might not be preserved
14:07	alexbaranosky_	Bronsa: thx
14:09	toxmeister	bronsa: I actually just chased a bug in related to ^:const attached to deftype instances a few hours ago. it behaved all fine until i started implementing a hash cache mechanism using mutable fields. then suddenly got completely weird compiler errors about missing field names. removing ^:const from some of the "const" preset instances, removed the issue...
14:10	Bronsa	yeah, don't use ^:const for deftypes
14:11	hiredman	don't use :const
14:11	hiredman	toxmeister: why are you using :const at all?
14:11	Bronsa	hiredman: that's a bit too much :), it's perfectly fine for numbers/strings/keywords
14:12	hiredman	Bronsa: but doesn't throw an error or warning for anything else
14:12	alexbaranosky_	what spots does the perf benefit of const really add up to anything substantial enough to matter?
14:12	Bronsa	right, it just explodes in your face later
14:13	hiredman	will be better off having maintainable code that doesn't use it
14:13	Bronsa	alexbaranosky_ (def ^:const x 23), you can use x as a primitive
14:14	hiredman	alexbaranosky_: I am sure there are cases where it matters, I suspect the majority of (ab)use is people who sprinkle it on as magic pixy dust
14:16	toxmeister	hiredman: i read somewhere (ages ago) that it would provide some minor perf difference, but I learned since that this doesn't make much sense for non-primitives
14:17	hiredman	well, there is a data point
14:18	hiredman	(I am counting that as one for magic pixy dust)
14:19	alexbaranosky_	pixie dust driven development
14:22	chenglou	what's a good beginner clojure/clojurescript rss feed to follow?
14:23	hiredman	/win 19
14:25	kqr	I don't know who I talked to before, but I've tried the java interop a little more now and I'm really starting to like it
14:26	kqr	it was just unfamiliarity talking before when I was unsure about how good it was
14:27	samflores	quick survey: which one is more readable (or idiomatic)? https://gist.github.com/samflores/2c68a9bcb6b4742d6760
14:28	kqr	samflores, for so few function calls, 1 works
14:28	kqr	samflores, otherwise 3
14:29	amalloy	samflores: i like 3. but i'd surround the chained calls with parens, like (display-board!). i don't like that those are optional
14:29	loliveira	samflores: 3, but i liked 4
14:29	amalloy	i definitely hate 2. 10 is a bad place to start the arrow
14:29	kqr	I agree with amalloy on that
14:29	alexbaranosky_	i do 3
14:30	Jaood	indent 1 :P
14:30	arrdem	M-x cljr-thread-first-all RET :P
14:30	amalloy	kqr: on which? hating 2, or using parens?
14:30	alexbaranosky_	or indented 1
14:30	kqr	amalloy, I was about to say "the former" but you made it tricky by exchanging the order of them
14:31	amalloy	haha
14:31	kqr	amalloy, hating 2
14:31	amalloy	i try to make your answers to my questions as ambiguous as possible so i can pick which stance i think you're taking
14:31	kqr	that's a good strategy
14:32	kqr	I don't mind the no parens... with parens it looks more like actual function calls
14:32	kqr	like, fully saturated with arguments and all
14:32	Jaood	this wins: https://www.refheap.com/89513
14:32	kqr	which is probably not the intention
14:32	kqr	Jaood, i think so too for so few levels of nesting
14:37	amalloy	kqr, Jaood: http://stackoverflow.com/a/12716708/625403 is a summary of my oft-linked-to stance on how to use ->
14:37	amalloy	obviously you don't have to agree, but it's a useful viewpoint to be aware of
14:39	samflores	good point
14:40	elben	is there an easy way of pasting in strings with double-quotes into the REPL? e.g. my copy buffer is ‘“hello”’ and I want (def s “\”hello\””)
14:40	SagiCZ1	and know i know what amalloy looks like
14:40	SagiCZ1	*now i know
14:41	justin_smith	elben: in a naked repl or in an editor?
14:41	elben	naked, though i also use vim-fireplace
14:41	amalloy	SagiCZ1: well, you know what i looked like when i took my gravatar picture like...four years ago
14:42	amalloy	elben: not really
14:42	justin_smith	elben: if you put your paste in a temporary file, slurp should get you the right result
14:42	SagiCZ1	amalloy: i bet u look completely different now
14:43	technomancy	elben: step 1: don't use smart quotes. =)
14:43	justin_smith	bonus points: make a small shell script that takes the paste buffer, puts it in a temp file, then puts the path to the temp file in the paste buffer
14:43	elben	technomancy: that’s just my IRC client :)
14:43	amalloy	technomancy: or do use smart quotes, and then pasting them mid-string won't be an issue
14:43	amalloy	next-level thinking here
14:44	elben	i really do like python’s triple-quote, though. wonder if http://dev.clojure.org/display/design/Alternate+string+quote+syntaxes is going anywhere
14:44	amalloy	elben: no, it's not
14:44	TimMc	I have an utterly terrible answer for you.
14:44	technomancy	amalloy: brilliant
14:44	TimMc	Are you ready?
14:44	technomancy	I would really love to have «guillemot» support work like triplequoting in python though
14:45	TimMc	(def input (javax.swing.JOptionPane/showInputDialog "paste it"))
14:45	kqr	hahaha
14:45	kqr	that's actually not a bad answer
14:45	TimMc	Terrible, but practical.
14:45	amalloy	TimMc: well, i was considering suggesting (read-line)
14:45	TimMc	I don't know if it handles multiline.
14:45	arrdem	(def foo (slurp (io/resource "docs/user/foo")) ...)
14:45	amalloy	rich's opinion, i believe, is that you shouldn't be embedding wacky strings in your source code - anything weird you need should be in a resource file anyway, and as a side effect it then doesn't matter if it has " in it
14:46	amalloy	arrdem: justin_smith suggested that first
14:46	rhg135	what's wrong with ctrl-v assuming your term is nice
14:46	elben	TimMc: amazing
14:46	justin_smith	rhg135: you don't get proper escaping
14:46	rhg135	ah
14:46	rhg135	justin_smith, im with rich/amalloy
14:46	elben	amalloy: the common use case i have everyday is working with JSON in the repl
14:46	rhg135	keep it elsewhere
14:47	amalloy	the problem with TimMc's suggestion is that anyone who uses it on macos will be like "argh there's a dang system tray icon for my java process now"
14:47	justin_smith	rhg135: my answer also involved putting the wacky string in a file first
14:47	technomancy	there should be an M-x clojure-escape-yank though in elisp
14:47	justin_smith	technomancy: string-edit-at-point is handy, and works recursively
14:47	technomancy	nice
14:47	TimMc	amalloy: read-line would interact badly with the REPL, yeah?
14:48	justin_smith	comes with the string-edit package
14:48	amalloy	TimMc: if you have a lame repl
14:48	TimMc	shots fired
14:48	arrdem	amalloy: oh that wasn't apparent to me.
14:48	amalloy	(as, for example, i do)
14:48	amalloy	but it works fine in lein repl, for example
14:48	amalloy	which is what he was asking about
14:48	TimMc	(defn lame? [repl] (not= repl dr-racket))
14:49	kqr	what's the difference between when-let, if-let and if-some?
14:49	amalloy	TimMc: (def lame? (complement #{dr-racket}))
14:50	justin_smith	kqr: when-let has no else clause
14:50	kqr	i see
14:50	justin_smith	and if-some doesn't exist in my currently open repl, dunno what it is
14:50	justin_smith	(doc if-some)
14:50	clojurebot	"([bindings then] [bindings then else & oldform]); bindings => binding-form test If test is not nil, evaluates then with binding-form bound to the value of test, if not, yields else"
14:50	kqr	new in 1.6 apparently
14:51	amalloy	TimMc: you know what's cool about my repl, though? (swank.core/read-from-emacs-minibuffer "What is your name? ")
14:51	justin_smith	kqr: looks like it replaces (if-let [foo bar] (let [...])) by letting you put other bindings after the conditional one
14:51	technomancy	amalloy: nrepl-discover has that
14:52	kqr	justin_smith, where do you gather that?
14:52	technomancy	except hypothetically-portable
14:52	kqr	it's not obvious from the documentation i've found
14:52	amalloy	technomancy: does nrepl-discover actually get used? i never hear anyone but you talk about it
14:52	kqr	though i'm still new so I might be missing something obvious
14:52	amalloy	but i don't know about the ecosystem
14:52	justin_smith	kqr: never mind, I was misled by the name "bindings"
14:52	kqr	ah
14:53	technomancy	amalloy: I keep mentioning it in hopes that someone who actually still uses clojure will realize how cool it is
14:53	kqr	justin_smith, it looks like when-some tests for (not (nil? p))
14:53	kqr	justin_smith, while let-if tests for true
14:53	kqr	justin_smith, possibly
14:54	TimMc	elben: (def input (clojure.string/join \newline (take-while (complement empty?) (repeatedly read-line)))) will do multiline input, stopping at the first empty line.
14:55	TimMc	From standard lein repl, that is.
14:57	hlprmnky	newbie question time
14:58	hlprmnky	I have required core.async like so in my ns decl: [clojure.core.async :as async]
14:58	elben	TimMC, interestingly that only allows 1024 chars to be pasted in at a time?
14:58	hlprmnky	yet when I run a repl in that namespace, things like merge are clojure.core.async/merge, not clojure.core.merge as I would expect
14:58	Jaood	technomancy: was that phrase a pun? re:"someone who actually still uses clojure"
14:59	technomancy	Jaood: not intentionally?
14:59	amalloy	hlprmnky: did you actually :require it, or did you :use it?
14:59	hlprmnky	(e.g., (doc merge) returns the docstring for async/merge, not core/merge, and (merge {} {:a “foo”}) returns a channel, or dies because it can’t merge a keyword
14:59	Jaood	technomancy: :)
14:59	hlprmnky	amalloy, I don’t have a :use block for any of my namespaces
15:00	amalloy	mmmm. paste your ns form then, i guess
15:00	amalloy	technomancy: king of puns so subtle even he doesn't get them
15:00	amalloy	(i don't either)
15:00	hiredman	hlprmnky: you are likely using :use insteaad of :require
15:01	hiredman	oh
15:01	hiredman	I see that was already said
15:01	hlprmnky	oh haha
15:01	hiredman	(old repl)
15:01	hlprmnky	no, but I am doing something equally dumb
15:01	amalloy	ah, or an old repl, sure
15:01	justin_smith	kqr: frankly, I can't make if-some do anything if-let would not do
15:01	amalloy	justin_smith: try (if-some [x false] true false)
15:01	amalloy	vs if-let
15:01	justin_smith	,(if-some [a false] :yes :no)
15:01	clojurebot	:yes
15:01	justin_smith	ahh
15:01	justin_smith	there we go
15:02	hlprmnky	no but I do still have (:refer-clojure :exclude [map reduce into …etc.]) from before I decided that I wanted the core available and async off to the side
15:02	hlprmnky	amalloy: thanks for asking me to paste the ns and therefore inducing me to actually read the whole thing again
15:02	hiredman	that in no way explains why you would get core.asyncs merge instead of clojure cores
15:03	kqr	is there anyone here who uses vim and can recommend something to edit expressions?
15:04	hlprmnky	if I have excluded merge in that statement, and then :require [clojure.core.async :refer :all :as async], what would the expected value of ‘merge’ be?
15:04	Jaood	kqr: evil-mode :P
15:05	arrdem	hlprmnky: :as async will create async/* bindings, :refer :all will refer all symbols from core.async with no prefixes
15:05	kqr	Jaood, that's just evil
15:05	hlprmnky	The whole reason I decided to move to having :as async in there was that I didn’t fully understand what :refer-clojure was doing and made the conscious choice to not muck with it now
15:05	hlprmnky	arrdem: Ah! (also: d’oh)
15:05	hiredman	hlprmnky: :refer :all is the same as :use basically
15:05	amalloy	hlprmnky: :refer :all! you left out the key point
15:06	hlprmnky	I did
15:06	rhg135	kqr, paredit.vim
15:06	hlprmnky	I need to go reread that 8th light blog post about how to know what on earth I am doing with use, require, etc.
15:06	rhg135	that's what i used
15:07	Ven	kqr: evil-mode and slime :P
15:07	amalloy	c'mon you guys, vim has fine clojure tooling. you don't have to switch to emacs
15:07	catern	yes you do
15:07	catern	switch to emacs
15:07	amalloy	ask tpope or Raynes or something, those being the two vimmers i can think of at this moment
15:08	technomancy	no, emacs is ridiculous
15:08	hlprmnky	I was doing it right in <project>.core (just [clojure.core.async :as async], calls to (async/*)), doing it wrong in <project>.system, which hilariously over time has lost all its uses of async as I refactored stuff into functions in core
15:08	catern	at least it's not vim, regards ex-vim-user
15:08	kqr	rhg135, hm I also heard about another one that was supposedly better, but I can't for the life of me remember the name
15:09	justin_smith	kqr: I assume you are using fireplace?
15:09	rhg135	amalloy, no lighttable
15:09	rhg135	i USED to use vim
15:10	virmundi	I use to use Light Table, now I use vim
15:10	kqr	justin_smith, not yet, but that's in the pipes as well
15:10	virmundi	LT takes 1 gb of ram.
15:10	aperiodic	kqr: http://www.vim.org/scripts/script.php?script_id=3998
15:10	amalloy	rhg135: i don't know what i've said that you're disagreeing with? i don't think i addressed you, or said anything about what editor you're using
15:10	rhg135	i have a 3tb hdd
15:10	rhg135	if i needed more i could swap
15:11	rhg135	im not disagreeing
15:11	rhg135	vim was nice and i use it when i don't have x for lt
15:12	kqr	ah
15:12	kqr	the other alternative is vim-sexp
15:13	kqr	which is apparently made with clojure in mind
15:13	kqr	I don't know if paredit is
15:14	schmee	kqr: vim-surround gets the job done
15:14	aperiodic	paredit.vim is aware of clojure but not made only for clojure
15:14	schmee	https://github.com/tpope/vim-surround
15:15	schmee	kqr: + a whole lot of {command}ab and %{command}
15:18	TimMc	catern: Don't tell other people to switch editors just because you didn't like it. That's absurd.
15:19	aperiodic	kqr: if vim-sexp treats quotes exactly like parens and square braces, and supports the majority of the operations that paredit.vim does, let me know!
15:20	catern	TimMc: huh? I just have their best interests at heart, I was a diehard vim user for a long time
15:20	aperiodic	kqr: my main gripe with paredit.vim is it'll let you wrap quotes but not strip them
15:20	kqr	aperiodic, isn't that what you have vim-surround for?
15:21	technomancy	switching to emacs at the same time you're learning clojure is a pretty good way to get frustrated and give up.
15:21	aperiodic	kqr: yeah, I should probably get that
15:21	alloyed	aperiodic: vim-sexp will wrap quotes as you'd expect, not sure about stripping because I always use vim-surround instead
15:21	brainproxy_	technomancy: it's not that bad
15:21	alloyed	using tpopes mappings for vim-sexp make them feel like two sides of the same coin
15:23	TimMc	brainproxy_: I agree with technomancy from personal experience on this.
15:23	TimMc	Learning a new language and a new, powerful editor at the same time is just asking for sadness.
15:23	kqr	haha
15:23	kqr	either way
15:23	kqr	I've tried emacs for a few months
15:23	amalloy	TimMc: i learned them both at the same time and had fun with it, but i wouldn't recommend it as a strategy. certainly not to someone who specifically asked for help with a different editor
15:23	kqr	and I'd really like to like it
15:23	TimMc	Especially if you try to use integration features like SLIME.
15:23	kqr	but it's just not for me
15:23	emaphis	notepad and clojure is sadness too.
15:24	danielszmulewicz	Emacs is for life. Clojure, who knows?
15:24	kqr	I didn't even know editor wars were still a thing
15:24	TimMc	kqr: Some people are still fighting that war. It's like those Japanese soldiers on the islands after WWII.
15:25	rpaulo	there's just one war
15:25	rpaulo	emacs vs vi{,m}
15:25	amalloy	don't forget bash vs everyone else, since you had to bring in {,m}
15:26	amalloy	(obviously not an editor, but pretty similar kinds of wars)
15:26	rhg135	just use vim if you know it, emacs if you know it, or lt if you know neither
15:26	justin_smith	does anyone seriously defend vi[^m]* ? I prever non-vim vi, but I am only a casual vi user
15:26	amalloy	but i agree the editor wars are silly. use what you know, and always learn more things
15:27	TimMc	justin_smith: Such as vixjohyvwaw?
15:27	amalloy	TimMc: visudo
15:27	TimMc	ah, ok
15:27	justin_smith	TimMc: nvi, elvis, plain vi
15:27	rhg135	visudo actually calls EDITOR
15:28	justin_smith	vim overrides things like ## and ~ which I acutally like to use when doing a quick config file edit
15:28	rhg135	gentoo sets it to nano by default xD
15:28	danielszmulewicz	What is the editor that hipsters want to be seen with?
15:28	justin_smith	rhg135: so does debian
15:28	rhg135	D:
15:29	rhg135	it's a virus
15:29	rhg135	or plague
15:29	rhg135	to be less ambigous
15:29	aztak	danielszmulewicz: it used to be Atom, right?
15:29	justin_smith	danielszmulewicz: vintage TECO
15:29	danielszmulewicz	haha.
15:29	justin_smith	it's the heavy vinyl pressing of the editor world
15:30	TimMc	rhg135: Confirmed: https://en.wikipedia.org/wiki/Nanoviridae
15:30	emaphis	hipsters use ed
15:30	rhg135	i was joking...
15:30	rhg135	but wow ik it!
15:30	aztak	hehe... about editors: https://www.kickstarter.com/projects/1203633826/notepad-conf
15:31	kqr	haha
15:32	kqr	all I can think of is how notepad fails badly when it comes to LF line endings
15:32	justin_smith	I actually logged into and tried the vms system hosted by sdf, and used TECO just to see what it would be like
15:32	emaphis	wow, $171, it's a thing.
15:32	justin_smith	I actually liked the vms versioning system (but hated the shell)
15:33	rhg135	i hope it's a joke
15:33	noprompt	bbloom, dnolen_: i'm trying to write a test for the patch we talked about earlier but i run in to an error: IllegalStateException Can't change/establish root binding of: cljs-ns with set clojure.lang.Var.set (Var.java:221)
15:33	bbloom	,(def x 1)
15:33	clojurebot	#'sandbox/x
15:33	bbloom	,(set! x 2)
15:33	clojurebot	#<IllegalStateException java.lang.IllegalStateException: Can't change/establish root binding of: x with set>
15:34	bbloom	,(alter-var-root #'x inc)
15:34	clojurebot	2
15:34	bbloom	,x
15:34	clojurebot	2
15:34	bbloom	noprompt: ^^
15:34	noprompt	bbloom: so should i patch the analyzer to use alter-var-root instead of set!
15:34	bbloom	noprompt: i have no idea. depends on what it's using set! for
15:34	noprompt	cljs-ns is a dynvar so it seems like alter-var-root is a bad idea.
15:35	dnolen_	noprompt: no
15:35	dnolen_	noprompt: we use set! just fine it's dynvar
15:36	dnolen_	noprompt: do you have the patch somewhere? or a gist?
15:36	TimMc	By the way, this is the editor I used to transition to writing Clojure: http://www.prodevtips.com/2010/05/03/my-clojure-editor/
15:36	noprompt	dnolen_: right. from the doc on set! that makes sense. i guess what i'm asking is is there anything else i need to do when trying to analyze the ns form in the repl.
15:36	hiredman	likely you are doing some lazily and it is escaping the scope of the binding
15:36	danielszmulewicz	noprompt: Currently, it is an error to attempt to set the root binding of a var using set!, i.e. var assignments are thread-local.
15:36	noprompt	dnolen_: cljs.analyzer-tests> (a/analyze ns-env '(ns foo {:baz "woz"}))
15:36	TimMc	It's fragile and low on features, but it is sexp-aware and that's just what I needed to make the jump.
15:36	dnolen_	noprompt: so you need to establish a binding around that first
15:37	noprompt	dnolen_: doh.
15:38	bridgethillyer	hlprmnky: I had that exact problem this morning
15:39	arrdem	https://github.com/clojure/clojure/pull/17#issuecomment-53628365 hooboy
15:40	justin_smith	arrdem: razum2um1 is the same person who considers the lack of ubiquitous usage of ansi colors in clojure output to be a usability bug
15:40	razum2um1	yep
15:40	joegallo	'oh man, what a beautiful hornet's nest this is that i have come across. i shall kick it heartily!'
15:41	catern	git send-email/git format-patch should be mandatory in all git tutorials
15:41	bridgethillyer	hlprmnky: Or, rather, no, misreading your question
15:41	catern	alternatively, github shouldn't be so lame and use its own proprietary method for pull requests
15:41	catern	should stop being so lame and should stop using its own*
15:41	bridgethillyer	hlprmnky: There is a merge, also, in Clojure core
15:42	danielszmulewicz	TimMc: If you ever want to make the jump: https://github.com/pierre-lecocq/emacs4developers
15:42	campeterson	Hey all, got a question about fn args.
15:42	joegallo	you may ask it
15:42	campeterson	:)
15:43	campeterson	The function takes a keyword as the first arg, and n vectors
15:43	campeterson	I want to programatically build the "arg" and send it in
15:43	justin_smith	campeterson: sounds like you may want apply
15:44	joegallo	(apply your-fn the-keyword seq-of-the-vectors)
15:44	rhg135	,(apply + 1 [1 2 3])
15:44	clojurebot	7
15:44	rhg135	thank you mr. bot
15:44	campeterson	Ah, I think that might do it
15:45	danielszmulewicz	Please someone help me put an end to my misery: I'm trying to retract a fact in datomic that involve a ref, and I can't wrap my head around it. In my schema, I have :video/user, it's a ref with cardinality many. I want to remove the link of that particular video with a particular user. I've tried this: (d/transact conn [[:db/retract 17592186045835 :video/user 17592186045424]]) where the first id is the id of the video, and the second
15:45	danielszmulewicz	the id of the user. This is wrong, apparently.
15:45	schmee	am I correct in thinking that `apply` is basically argument unpacking ala Python?
15:45	joegallo	i don't know python, but, uhhhh... probably.
15:45	danielszmulewicz	schmee: same here. but yes.
15:46	joegallo	yeah, it looks sortof similar
15:46	rhg135	yes it is schmee
15:46	schmee	nice, that'll be handy!
15:46	rhg135	except it's only on the last arg to apply
15:47	lemonodor	foo(x, y, *z) in python is basically (apply foo x y z) in clojure.
15:48	schmee	I was converting some old Python code I wrote to clojure yesterday, and by god Python looks horrible when you've gotten used to clojure
15:48	schmee	everything is so... inconsistent
15:48	campeterson	joegallo: so, say I don't need to apply a function, but I have my keyword and my seq-of-the-vectors
15:48	rhg135	yup
15:49	joegallo	okay, you've got your keyword and the seq-of-the-vectors... what of them?
15:50	campeterson	In need to build "args" and pass it into the function
15:50	joegallo	the way to do that is with apply
15:51	SagiCZ1	hey how do i watch this video? cant find any play button http://www.infoq.com/presentations/Value-Identity-State-Rich-Hickey
15:51	joegallo	there is no (my-fn n-arguments-that-have-been-prepacked). there's only (apply mf-fn n-arguments-that-have-been-prepacked)
15:51	razum2um1	justin_smith: btw, it's not only me, see there is already some progress :) https://github.com/AvisoNovate/pretty but I didn't mentioned it has to be in the core, no. but there should be some way to get newbie-freindly REPL. imagine you're life without readline library inside bash/other shell-or-repl?
15:52	SagiCZ1	schmee: python does look horrible and i have never thought different.. even java looks better
15:52	danielszmulewicz	Is there any blog post/ tutorial about retracting facts in datomic?
15:52	joegallo	or maybe you haven't fully conveyed the situation to me, campeterson
15:52	nullptr	SagiCZ1: you might need to install flash
15:53	campeterson	joegallo: yeah, sorry
15:53	TimMc	danielszmulewicz: Oh, I already made the jump; I'm using a big-boy text editor now.
15:53	danielszmulewicz	TimMc: Oh, I must have misread. All Good.
15:53	SagiCZ1	nullptr: yeah it was flash.. works in IE
15:54	campeterson	joegallo: the light just came on
15:55	joegallo	excellent! http://xkcd.com/224/
15:55	TimMc	However, this might be good to skim through.
15:56	campeterson	joegallo: yeah, stupid rookie mistake. but I see it now. Apply take the function you want the args to go to, then the args
15:57	joegallo	exactly
15:57	kqr	woo! i finally have reason to write my first macro
15:57	campeterson	joegallo: thanks again
15:57	danielszmulewicz	You know, people, for a dummy like me, Clojure has been a boon. Suddenly I could write web applications with minimum effort, and I went on to master every piece in the so-called "immutable stack". I'm writing 100% Clojure for a year now. There's only one piece of the stack that eludes me, and that is datomic. I get the concept of facts accretion, but it has never played out well for me. I think I'm about to give up on datomic. Thoughts?
15:58	joegallo	campeterson: you're welcome!
15:58	rhg135	ask yourself 'do i need a db'
15:59	danielszmulewicz	I need a db I can understand.
15:59	rhg135	ah
15:59	rhg135	never give up
16:00	technomancy	a db you can understand is easier when the license doesn't prevent you from examining the implementation
16:00	danielszmulewicz	rhg135: but is it worth it?
16:00	danielszmulewicz	technomancy: Everything in the datomic experience is a turn off.
16:01	danielszmulewicz	Docs, installation, default memory settings.
16:01	technomancy	for many use cases, it's not that hard to treat postgres as an immutable event log
16:01	tuft	that's what i've been thinking lately
16:01	danielszmulewicz	technomancy: exactly.
16:01	tuft	also postgres has real serializable transaction isolation, which datomic only has half of
16:02	tuft	the rest you have to do yourself with functions that run in the transactor
16:02	tuft	just need a way to get the transactions and queries as data part -- still like that =)
16:03	rhg135	danidk if it is i've never used it
16:03	rhg135	danielszmulewicz, ***
16:03	rhg135	srsly i need to type out nicks
16:05	TimMc	Get a better chat client!
16:05	rhg135	i'm using hexchat
16:05	akurilin	quick question: how do you guys measure the load on your web applications ? I have a ring app that's getting a lot of traffic, but I'm not sure how much is too much. I have munin set up to monitor/track things like IO, cpu and mem use. I could also analyze the logs and see how long requests take on average
16:05	rhg135	too many ppl in this chan
16:06	tuft	rhg135: no tab completion in that client?
16:07	schmee	tuft: there sure is, works for me at least
16:10	loliveira	when should I use datomic over postgres?
16:11	danielszmulewicz	tuft: to be honest, even mongodb would fit my use case. Heresy maybe, but good enough for me. And monger just redesigned their api to be more "functional".
16:11	loliveira	did this question make any sense?
16:11	technomancy	loliveira: when does vendor-lock-in not matter?
16:12	loliveira	technomancy: Do you know how does the vendor-lock-in is implmented in datomic?
16:12	technomancy	when you work for Cognitect I guess? =)
16:12	kqr	if I have a list of statements in a macro, how do I expand them one by one?
16:12	dnolen_	danielszmulewicz: you know there's #datomic and an active mailing list for datomic questions :)
16:13	danielszmulewicz	dnolen_: Oh, thanks. I forgot about that.
16:13	kqr	ah I found it, ~@
16:14	danielszmulewicz	So Clojure was a gift to the world, but the plan to monetize it was datomic. Result: the world adopts Clojure and rejects datomic? How accurate a description is this?
16:15	danielszmulewicz	Is Cognitect making money with Datomic?
16:15	technomancy	http://p.hagelb.org/mystery.gif
16:16	TimMc	I hope so.
16:17	danielszmulewicz	technomancy: :-)
16:18	justin_smith	kqr: that splices them - everything returned by your macro gets expanded to create the form that gets run
16:19	dnolen_	danielszmulewicz: seems pretty inaccurate. also there's a bunch of people in this room with Clojure products making money.
16:19	kqr	justin_smith, i'm not sure I follow
16:20	TimMc	That doesn't monetize it for the Core folks, though.
16:20	TimMc	I mean, except via consulting, I guess.
16:21	danielszmulewicz	dnolen_: I also make money with a Clojure product. I'm asking about Cognitect's product, Datomic.
16:21	kqr	justin_smith, if xs is (a b), will not `(x ~@xs x) be (x a b x)?
16:21	justin_smith	kqr: that is the splicing I was talking about
16:21	justin_smith	but whatever you return is expanded, and then evaluated
16:21	kqr	ah yeah
16:22	Jumblemuddle	Let's say that I wanted to create a route that allows for forward slashes. (e.g. site.tld/a/b/c would return a (list?) of the arguments) I want it to allow for an 'infinite' amount of 'sub-folders' and return them as a list. Any way to do this with compojure?
16:23	amalloy_	Jumblemuddle: it's certainly possible, if a little bit weird. you'd probably want to define a recursive route
16:24	Jumblemuddle	Interesting, I hadn't even thought about recursion. I was thinking I'd just have to do a regex that allowed for '/' then parse it myself.
16:28	TimMc	Jumblemuddle: I think that makes far more sense.
16:30	Jumblemuddle	I wasn't quite sure if it'd work, because '/' is a 'special' character, so I figured I'd ask here in case there was some 'pretty' way to do it.
16:40	noprompt	dnolen_: could you take a look at a gist real quick and give me some feedback about this patch? https://gist.github.com/noprompt/130197e3c1ede9567bc3
16:40	noprompt	bbloom: you're welcome to it as well.
16:42	noprompt	bbloom, dnolen_: i'm unsure what makes sense to test. for my use case i'm positive i want to look at cljs-ns when the macro is being compiled but as far as clojurescript goes i'm not sure what's relevant for the test.
16:42	hiredman	a2enmod proxy
16:42	hiredman	whoops pardon
16:43	Jaood	now everyone knows you use apache instead of nginx
16:43	dnolen_	noprompt: is that how ns metadata works, a map after the symbol is supported?
16:43	noprompt	dnolen_: according to what the current code says.
16:44	noprompt	dnolen_: https://github.com/clojure/clojurescript/blob/245de2e752f638f5d06527a1cafed26b0d8d0da8/src/clj/cljs/analyzer.clj#L1132-L1134
16:44	timothyw	hey all. when trying to do a `lein release` the input screen doesn’t respond to my inputs (http://bit.ly/1q5Ucmp)
16:44	timothyw	I think input is "gpg-agent > pinentry-curses"
16:44	amalloy_	does anyone have a recommendation for a thrift library to use from clojure? i'll be working with some java code so just a clojure-looking layer on top of the generated java classes would be nice
16:44	timothyw	any ideas? (probably missing something simple)
16:44	noprompt	dnolen_: and from the look of it in the repl, yes.
16:44	noprompt	it's news to me. :P
16:45	dnolen_	noprompt: huh, ok. looks good to me, ticket + patch in JIRA then.
16:45	noprompt	dnolen_: what about the tests? should i leave all of that in there?
16:46	dnolen_	noprompt: tests are fine
16:46	dnolen_	noprompt: all that matters is a squashed patch with an appropriate commit message
16:46	dnolen_	noprompt: https://github.com/clojure/clojurescript/wiki/Patches
16:46	noprompt	dnolen_: thanks! i'll get that started.
16:59	kqr	is there any sort of style guide for clojure (or some project which I can imitate?)
16:59	kqr	wait
17:00	kqr	I just googled "clojure style"
17:00	kqr	first hit seems promising
17:00	justin_smith	kqr: there are even automated linters - like kibit and eastwood
17:00	Jumblemuddle	Alright, I'm probably doing this wrong. I'm trying to create a compojure route that will allow for '/'s in the route '(GET ["/:route", :route #"[A-z,a-z,/]"] [] (default))
17:00	timothyw	kqr: I’ve seen this one reference a few times: https://github.com/bbatsov/clojure-style-guide
17:00	justin_smith	kqr: also, I like this one, though it is not wholly uncontroversial https://github.com/bbatsov/clojure-style-guide
17:01	amalloy_	that regex is bananas, Jumblemuddle. you probably want [A-Za-z/]+
17:01	Jumblemuddle	I knew I'd end up doing something stupid... thanks
17:03	Jumblemuddle	It seems to work fine, until I try doing a route with a '/' in it. Is it even possible to make compojure treat routes with them as a single route?
17:04	Jumblemuddle	I'm an idiot... It works fine, I was trying localhost:3000/test/test3 (I don't have numbers in the regex) :/ Thanks, guys!
17:05	justin_smith	Jumblemuddle: it seems like for a route like that it would be easier to make your own middleware layer, that detects the regex for the deep path you want, and calls your handler, and only invokes the real compojure router if there is no match
17:05	justin_smith	oh, so it worked, never mind then
17:06	Jumblemuddle	Ya, I was just being an idiot. Thanks, anyways!
17:08	PigDude	if a Java method takes a Map, can you provide a Clojure map to it?
17:08	PigDude	walking my way from IPersistentMap on up, I don't think so.
17:08	PigDude	is that right?
17:09	technomancy	,(instance? java.util.Map {})
17:09	clojurebot	true
17:09	justin_smith	,(isa? (class {}) java.util.Map)
17:09	clojurebot	true
17:09	gtrak	technically that's an array map :-)
17:09	technomancy	psh; isa?
17:09	PigDude	hm, what did i miss?
17:09	amalloy_	,(isa? clojure.lang.IPersistentMap java.util.map)
17:09	clojurebot	#<CompilerException java.lang.ClassNotFoundException: java.util.map, compiling:(NO_SOURCE_PATH:0:0)>
17:09	amalloy_	,(isa? clojure.lang.IPersistentMap java.util.Map)
17:09	clojurebot	false
17:10	amalloy_	PigDude: you didn't miss anything - the interfaces are separate, it's just that every implementation of c.l.map stuff also implements j.u.map
17:10	technomancy	(isa? clojure.lang.TerriblyNamedFunction isa?)
17:10	PigDude	amalloy_: that's what i missed :) i'm not java guy, forgot to look for the implementation
17:10	PigDude	amalloy_: i see APersistentMap extends Map
17:10	amalloy_	it really should be a subinterface, because stuff in clojure.core breaks if you create classes that implement clojure stuff but not java stuff
17:11	PigDude	i agree
17:12	tuft	danielszmulewicz: what would mongo offer you over postgres? i'm not so familiar with it, just curious
17:12	justin_smith	tuft: unreliability, and speed
17:13	justin_smith	~mongodb
17:13	clojurebot	mongodb is an ORM for mmap
17:14	TimMc	~mmap
17:14	clojurebot	Huh?
17:14	TimMc	~ORM
17:14	clojurebot	Excuse me?
17:17	tuft	haha
18:00	myazaki	I have a question about java interop, I have a java library which returns classes of primitive types (int.class, float.class) and I want to use that as a dispatch value for a multi method
18:00	noprompt	dnolen_: i opened the issue (#846) and attached the patch. thanks for your help on it.
18:01	myazaki	so for instance I got a float.class, but I can't seem to make it work
18:01	amalloy	myazaki: Float/TYPE
18:02	amalloy	(is one place where the class for the float primitive can be found)
18:02	myazaki	I tried that...
18:03	myazaki	I probably must be doing something wrong :)
18:03	dnolen_	noprompt: thanks will take a look later
18:04	justin_smith	,(class (float 1.0)) ; a more hacky way to get it
18:04	clojurebot	java.lang.Float
18:04	noprompt	dnolen_: awesome. really appreciate it.
18:04	lemonodor	i’m getting “java.lang.IllegalAccessError: in-seconds does not exist, compiling:(cookies.clj:1:1)” with ring 1.3.1, which supposedly has to do with pulling in an old clj-time, but i’ve put [clj-time “0.8.0”] in my project.clj and still get the error. lein deps :tree shows only that clj-time.
18:04	justin_smith	oh, never mind, for a primitive
18:04	lemonodor	anyone else run into this with ring 1.3.1 (released this week, i think)?
18:06	amalloy	justin_smith: that doesn't work at all?
18:06	justin_smith	amalloy: I misunderstood what was wanted, thus my "never mind" followup
18:06	amalloy	okay you said that already. but also: if the goal were to get at the class for Float, it's just...Float
18:07	amalloy	(class (float 1.0)) is silly
18:07	justin_smith	yes, it is
18:07	myazaki	I'm dispatch on Float/TYPE, for instance like this (defmulti blah (fn [arg1 dt] dt)) and (defmethod blah Float/Type [arg1 dt] ...)
18:08	myazaki	when I use blah with the float.class returned from the java library it says that it can't find a method for dispatch value: float
18:15	hiredman	myazaki: does (= Float/TYPE dt) return true?
18:16	myazaki	yes!
18:16	myazaki	I'll try isa?
18:17	myazaki	because I think multimethods use that
18:17	hiredman	did you load the code that defines the method for Float/TYPE ?
18:18	hiredman	(require or use of the namespace, direct call to load-file or load, etc)
18:18	danielszmulewicz	tuft: Oh, what mongodb gives me over postgres? The luxury to be lazy. No schemas in mongodb. I can shoot myself in the foot as much as I want. It's a document store, so at development time I can experiment with how I want my data to be structured without formalizing. Maybe that's why so many people hate it. It seems that people who understand databases recommend against it. Since I don't understand them, I use it happily.
18:19	myazaki	most definitely, hiredman
18:19	myazaki	damn it, such a strange error
18:19	hiredman	myazaki: what happens when you call the mutlimethod at the repl with (blah :foo Float/TYPE) ?
18:20	hiredman	myazaki: why are you so sure you have the defmethod loaded?
18:20	myazaki	man I think I found the error
18:21	myazaki	I'm an idiot
18:21	myazaki	lol
18:24	myazaki	thanks hiredman
18:25	myazaki	Actually, the error was inside a let statement of a method that called one of the multimethod
18:25	myazaki	I ultimately passed a wrong argument
18:26	myazaki	wow, it works now
18:27	myazaki	Lesson #928: Be careful with refactoring
18:40	TimMc	&(sort ["foo" :foo])
18:40	lazybot	java.lang.ClassCastException: java.lang.String cannot be cast to clojure.lang.Keyword
18:40	TimMc	:-(
18:41	TimMc	For a moment I was living in a utopia where Clojure implemented universal sorting.
18:41	TimMc	like in Python
18:42	Scorchin	Evening, something that might be relevant to folks in this channel is a "SICP Distilled" course that's being run in Clojure: https://www.kickstarter.com/projects/1751759988/sicp-distilled
18:42	mdeboard	I thought that strict ordering was like a requirement for type systems or something
18:42	mdeboard	TimMc:
19:39	danielszmulewicz	I'm offering a macro in a library for convenience. What is more idiomatic? (defystem name [args]) or (defsystem name {args}). A vector or a map?
20:00	bogdan	can somebody tell me if cider in any of version have something like clj browsing feature like speedbar or i only limited to jump to source?
20:00	celwell	How can I 'unwrap' a collection and just have each of its elements in place of it in the function? (I assume this needs some sort of macro) (I don't think I can use apply because of the details of the situation)
20:04	verma	celwell, apply?
20:04	verma	,(apply + [1 2 3 4])
20:04	clojurebot	10
20:04	aperiodic	celwell: why don't you think you can use apply?
20:05	verma	celwell, oh didn't see you comment for "apply"
20:11	danielszmulewicz	bogdan: M-. ?
20:50	lpvb	right now I have a function that synchronously loads sound data from an audio stream freezing the UI
20:50	lpvb	what's the right clojure concurrency way to do this?
20:50	lpvb	future? java threads? agents?
20:51	TEttinger	lpvb, so it's synchronous now but you want it to be async?
20:51	TEttinger	sounds like future would be easiest
20:52	TEttinger	I've used it before for very similar tasks, just not from disk
20:53	lpvb	TEttinger: I read that it will block when I deref it
20:53	lpvb	but for this case doing IO I don't need a value back so I don't ahve to worry about it right?
20:54	lpvb	will the future thread be terminated once it's done doing its work?
20:54	bogdan	ty daniel
20:55	alandipert	lpvb: another thing to look into would be agents + send-off
21:00	lpvb	alandipert: what would I put in the agent?
21:01	alandipert	i usually put something in there to store any information i want about how the processing went
21:01	lpvb	I'm confused
21:01	lpvb	how does that help me load audio data into a buffer asynchronously
21:04	alandipert	it’s just another way to make a thread
21:08	celwell	Hi, How can I unwrap a coll and have the elements in its place? (to anyone who recognizes this question: sorry, got disconnected from irc)
21:09	akhudek	yay, immutant 2 finally hit alpha!
21:09	lpvb	seems like future spawns an agent send off pool thread anyway
21:09	lpvb	celwell: what do you mean elements in its place?
21:09	celwell	I don't know how I would be able to use apply because it is part of the with-query-results macro.
21:09	celwell	lpvb: (1 2 3 4) -> 1 2 3 4
21:11	celwell	It's going into this macro in place of 'params' for [stmt & params] https://github.com/clojure/java.jdbc/blob/97be163884d402c4506e2397a528b15a23867a8f/src/main/clojure/clojure/java/jdbc/deprecated.clj#L680
21:11	celwell	But I have params as a vector and it has to 'unwrapped' I assume.
21:12	akhudek	celwell: you could do it with another macro I guess. Try searching for “using apply with a macro”
21:12	akhudek	celwell: It should cover the same topic
21:12	turbofail	[stmt [a b c d & more]]
21:12	amalloy	celwell: sql-params just wants a vector anyway
21:12	amalloy	turbofail: well, it's actually [stmt a b c d & more]
21:12	verma	so, (put! c val close!) is a common thing to do for one shot async calls?
21:14	celwell	amalloy: sorry, are you saying it's fine to just give it a vector? im confused
21:15	akhudek	celwell: if you are trying to give sql-params, yes
21:15	amalloy	celwell: all it wants is a vector to begin with
21:17	celwell	amalloy: but it wants this vecotr [stmt & params], but I'm giving it this vector [stmt [blah blah blah]]. I need to figure how to give it [stmt blah blah blah]
21:17	amalloy	(apply vector stmt [blah blah blah]) or (into [stmt] [blah blah blah])
21:19	amalloy	if the entire vector is this opaque thing you can't control at all, you can do something like (into [(first v)] (rest v))
21:19	celwell	amalloy: ok, i see what you mean. ok I'll work on this, thanks
21:19	celwell	amalloy: I was getting tunnel-visioned into a certain vector
21:20	lpvb	I used a future to execute loading the sound buffer, but if the user clicks another sound while the original future is still running it will continue to play the previous future
21:22	TEttinger	lpvb, then agents may be better, I think you can cancel both, but I'm more certain that you can with agents
21:22	TEttinger	most of my clojure does not use concurrency :\|
21:37	verma	how do I put a value into an async channel and close it rightaway, I guess I cal do (do (put! c v) (close! c)) .. but was wonder if there's a thing for this
21:40	lpvb	Maybe to solve my problems I need to send my buffer loading function a continuation so it can halt itself when it notices the user has changed selection?
21:57	whatisthewhat	My HDD crashed earlier. After recovering, I can't do lein repl anymore. Instead, I get the this Java error http://pastebin.com/2M7buaBJ. Tried reinstalling without success. Any thoughts what's causing this?
22:01	verma	which "transformer" is being referred to here: https://clojure.github.io/core.async/#clojure.core.async/filter<?
22:01	verma	a transducer?
22:02	beamso	whatisthewhat: https://github.com/technomancy/leiningen/issues/1321 ?
22:02	verma	oh
22:02	verma	https://github.com/clojure/core.async/commit/81fc946a4d80dd57f641282831e72ae0d829f1b0
22:03	verma	so I can use transducers with 1.6.0?
22:05	whatisthewhat	beamso: Not clear but I don't think that's my issue. The error code on that page shows the nREPL server starting. My error occurs before the server starts.
22:06	TimMc	technomancy: I'm thinking of buying a Maltron ergonomic keyboard. Ever used one of those?
22:07	whatisthewhat	This also seemed relevant, but my /etc/hosts file looks fine: http://stackoverflow.com/questions/22114716/lein-repl-errorconnection-refused?rq=1
22:09	beamso	i checked my hosts file on two machines. both had a '127.0.0.1 localhost' entry.
22:10	beamso	i also note from your stackoverflow post that 2.4.2 isn't the latest one... but there is a annoying bug in 2.4.3 anyway.
22:10	whatisthewhat	yeah i intentionally was running 2.4.2 because of that bug
22:23	TimMc	whatisthewhat: What does your hosts file look like?
22:24	TimMc	The selected answer on SO hinges on the hostname not being in /etc/hosts.
22:28	gastove	Buh: I have a project that declares dev dependencies in its project.clj. On one of my computers, my lein repl loads with dev dependencies on the classpath; on my other computer, dev dependencies are omitted from the classpath. Does anybody... have any suggestions where to start looking to fix this?
22:29	xeqi	gastove: lein versions?
22:29	gastove	Even more bizarre to me: lein with-profile dev repl still doesn't start a repl with dev dependencies loaded. So something... is wacky. _But what?_
22:29	TimMc	gastove: Do you have a ~/.lein/profiles.clj on either machine?
22:30	gastove	I have a profiles.clj on both machines; they're identical. And: I updated lein on both of 'em this morning.
22:30	gastove	So both are on 2.4.3
22:32	hiredman	gastove: I would check lein classpath
22:32	whatisthewhat	TimMc: my hosts file is normal, and has a 127.0.0.1 localhost entry
22:33	xeqi	gastove: add lein-pprint and see what it tells you about the :dev profiles
22:33	TimMc	whatisthewhat: Right, but does it have an entry for your hostname?
22:34	TimMc	It shouldn't matter, but it apparently did for the SO person.
22:34	whatisthewhat	when I type hostname into terminal it says it's simply localhost
22:35	TimMc	huh
22:36	TimMc	At this point I would try changing/setting the hostname and setting it in /etc/hosts under 127.0.1.1, but that's because I'm a little fuzzy on this stuff.
22:37	gastove	xeqi: I'm not perfectly sure what I'm looking for. (I'm on the broken computer, and don't have access to the working computer at the moment, so I can't compare and contrast.)
22:38	gastove	xeqi: but it does show the dev dep in the dev profile: `:dependencies ([midje/midje "1.6.3"])`
22:39	hiredman	:(
22:39	hiredman	I blame midje what the problem is
22:39	gastove	heh
22:39	hiredman	it is terrible
22:40	xeqi	gastove: and `lein deps :tree` and `lein with-profile +dev deps :tree` don't show midje?
22:41	gastove	xeqi: nope, neither of them.
22:44	gastove	hiredman: (or anybody else for that matter) is "which testing framework is good" going to be the kind of question I regret asking, or is there one-or-another that's emerging as Generally Pretty Solid? (I'm not super attached to midje, just like it better than clojure.test.)
22:45	TimMc	Midje is batteries-included, but the error messages can be pretty difficult to parse.
22:46	gastove	TimMc: good to know, actually -- that's rather a pet peeve of mine.
22:46	hiredman	midje is also written by people who have demonstrated time and time again that they don't know clojure
22:46	gastove	heh.
22:46	gastove	Any love for Speclj? Was also looking at it.
22:46	hiredman	clojure.test
22:47	xeqi	gastove: running out of ideas debugging without pastes
22:47	xeqi	test.check and clojure.test have been sufficent for my purposes
22:48	gastove	xeqi: I appreciate the help. I'm going to put the working computer and the broken computer in the same place tomorrow, see if one is config'd differently than the other.
22:48	tbaldrid_	gastove: let's just say that everytime I've run into a mature midje codebase, the authors have either said "we should have written that in clojure.test" or "we're rewriting that in clojure.test"
22:48	gastove	nods I'm interested in test.check; seems solid.
22:48	gastove	tbaldrid_: ha! That's also good to know.
22:49	tbaldrid_	clojure.test has it's warts, but it's so simple it's fairly easy to understand.
22:49	gastove	I just wish there were a better way to do setup/teardown/test ordering than test-ns-hook.
22:49	gastove	(Maybe there is and I just haven't come across it?)
22:49	xeqi	gastove: I was considering multiple profiles in the ~/.lein directory merging weird, but I'd have expected lein-pprint to get the same merger
22:50	hiredman	what is test-ns-hook?
22:50	xeqi	something beyond `use-fixture`?
22:50	hiredman	^-
22:51	gastove	If I've got this right: test-ns-hook lets you specify the order in which the functions in a test ns get called. It's incompatible with fixtures, though, which is a shame.
22:51	hiredman	never heard of it, so you don't need it
22:52	xeqi	hmm, I always considered the randomness in the ordering a positive
22:52	xeqi	test independent/idempotency and what now
22:52	hiredman	xeqi: it isn't random enough to be a positive
22:52	hiredman	it will run in the same order for ever run
22:52	hiredman	every
22:53	hiredman	if it was really random that would be great
22:53	hiredman	our big test suite at work ends up having problems with state leaking between tests in a namespace, and even across namespaces
22:54	gastove	Eh, I'm doing integration testing. Idempotence/commutativity is lovely, but I also need to do things like "definitely add 5 things to a DB, make sure code FNs remove them correctly." You need at least a little ordering for that.
22:54	hiredman	gastove: that sounds like one deftest
22:55	hiredman	and the body of deftest is just code that runs in order as written
22:55	technomancy	TimMc: Maltrons seem pretty hardcore. a lot like the kinesis advantage, but harder to find.
22:55	TimMc	You can build anything you want out of clojure.test, but the question is do you want to.
22:56	hiredman	TimMc: there are no alternatives that are good enough to trade off the simplicity
22:57	gastove	Well. Thanks for the time, all.
22:57	technomancy	TimMc: IMO the main flaw of the kinesis advantage (beyond size) is it's only available with quiet or linear switches. I don't know if the Maltron has the same problem or not off the top of my head.
22:58	TimMc	hiredman: You always build your own testing utils on top of clojure.test with each new project?
22:58	TimMc	technomancy: What key switch do you prefer?
22:58	hiredman	TimMc: not really, I just wrote code that does things and asserts things
22:59	technomancy	TimMc: the cherry blue switches are the best IMO, unless you are in an office where you can't make a lot of noise.
22:59	hiredman	with-redefs here and there
23:00	technomancy	TimMc: hm; one source says that maltron only has blacks, which are mechanical, but not tactile. (mostly used in gaming keyboards)
23:00	technomancy	seems like an odd choice
23:00	TimMc	technomancy: I bet Maltron would use different key switches for little to no extra cost.
23:00	TimMc	They hand-assemble them anyhow...
23:00	technomancy	TimMc: considering they're all hand-wired anyway, it's definitely a possibility
23:01	technomancy	at that cost level there should be room for customization =)
23:01	technomancy	TimMc: curious if you've seen this? http://atreus.technomancy.us
23:02	technomancy	can't remember if I showed you that or not
23:03	xeqi	technomancy: does it come in bluetooth?
23:03	TimMc	Oh, I vaguely recall this...
23:04	technomancy	xeqi: the base design doesn't. it might be possible to hack it in, but the case clearance is pretty tight.
23:04	technomancy	maybe if you increased the width of the spacer
23:04	technomancy	it's an open source design, so ... anything is possible =)
23:04	technomancy	personally I have enough batteries to keep charged every day as it is
23:05	TimMc	I think I'd want something with more keys.
23:05	technomancy	it definitely takes a lot of getting used to
23:07	TimMc	Not having arrow keys at the base layer is a no-go for sure.
23:07	TimMc	If I spent all my time in Emacs, maybe not...
23:07	technomancy	it's not for everyone; just thought I'd mention it
23:08	TimMc	nod
23:08	technomancy	the maltron looks neat if you don't need to take it with you
23:08	TimMc	heh
23:08	TimMc	What's wrong with Cherry MX Brown?
23:08	technomancy	I'd just watch out for the switches. maybe you'd like the ones they use by default, but I'd see if you can try it out in person before you pull the trigger.
23:09	TimMc	You can "rent" the Maltrons for a month if you live in the UK.
23:09	technomancy	huh, that's pretty cool
23:09	TimMc	I'd want to do some remapping to make Ctrl, Alt, and ( really easy to hit.
23:09	technomancy	makes a lot of sense; people would be reluctant otherwise. and you need about a month to get used to a layout like that.
23:10	TimMc	and maybe set up sticky keys or something
23:11	technomancy	http://deskthority.net/workshop-f7/anybody-ever-change-the-switches-on-their-maltron-t4675.html
23:12	technomancy	TimMc: the brown switches are supposed to be tactile, but they just feel weak to me; like a token effort at tactility.
23:12	technomancy	try it for yourself I guess, but IMO nothing is as nice as the clicky blues
23:13	TimMc	How loud are blues?
23:13	technomancy	kinda like clicking a pen
23:13	TimMc	Will my coworkers kill me?
23:14	technomancy	now the wisdom of try-before-you-buy becomes more apparent =)
23:14	TimMc	I'm going to see if any of my coworkers have one.
23:14	mdeboard	Does flipping the brown switch activate the brown note
23:15	technomancy	I'm working on a prototype that uses Matias quiet switches, which are dampened but still have the good tactile feedback.
23:15	technomancy	unfortunately they take a different style of keycap that is very difficult to source
23:16	technomancy	I had to steal caps from a thrift store mid-90s Apple keyboard for the prototype
23:21	TimMc	I should also get ahold of a set of Cherry MX switches to poke at.
23:22	technomancy	if you do need something quiet, the MX clear are better than brown IMO
23:25	TimMc	Maybe I can play with one of these: http://www.cmstore-usa.com/mechanical-key-switch-demo-board/
23:25	TimMc	(What I should really do is go to bed.)
23:26	amalloy	nonsense, TimMc. it's only like 5:30, in hawaii
23:38	ifiht	Hey I have a crazy question, any takers?
23:40	catern	don't ask to ask, just ask
23:40	catern	clojurebot? do you have that memorized?
23:40	catern	~ask
23:40	clojurebot	The Ask To Ask protocol wastes more bandwidth than any version of the Ask protocol, so just ask your question.
23:40	ifiht	kk
23:40	ifiht	I want to redifine the result of an equation
23:40	ifiht	*define
23:40	ifiht	i.e. 1+2=7
23:41	ifiht	HOW?
23:41	amalloy	you want to make three equal to seven? this question is too vague to answer
23:42	ifiht	I want to specify a specific instance of incongruity in the math library
23:42	ifiht	not 3=7
23:42	ifiht	but 8-3 = 7
23:43	ifiht	or the like
23:43	tbaldridge	you want to break math?
23:43	ifiht	yes, but only s little
23:43	ifiht	*a
23:43	tbaldridge	I may regret this next question....but why?
23:43	catern	(defn + [& _] 7)
23:43	catern	there you go
23:43	ifiht	it's important for working with lots of zeros
23:44	catern	,(defn + [& _] 7)
23:44	clojurebot	#<CompilerException java.lang.SecurityException: denied, compiling:(NO_SOURCE_PATH:0:0)>
23:44	ifiht	hold on, repl tiem
23:44	catern	:)
23:44	ifiht	NICE!
23:44	tbaldridge	ifiht: what does "working with lots of zeroes" mean?
23:44	catern	lol
23:44	catern	i may have just done something horrible
23:44	catern	~xy
23:44	clojurebot	xy is http://mywiki.wooledge.org/XyProblem
23:44	ifiht	8/0 comes up more than once
23:44	ifiht	rather n/0
23:45	ifiht	I have to define them
23:45	catern	ifiht: ~xy
23:45	catern	clojurebot dang it
23:45	ifiht	clojurebot?
23:45	amalloy	use a function different from clojure's built-in /. attempting to redefine how / works for all other code that exists is just doomed
23:45	catern	anyway, I feel a strong xy vibe
23:45	clojurebot	clojurebot is not a benchmarking platform
23:45	tbaldridge	I think there's got to be a much better way to avoid division by zero errors instead of redefining math
23:46	ifiht	I'm actually actively using n/0 to make it work.
23:46	amalloy	maybe if you're w WIMP, tbaldridge. i refuse to be bound by the chains of mathematics
23:46	ifiht	But the separate operator is a good idea
23:46	ifiht	THanks
23:55	technomancy	what just happened
23:55	ifiht	@catern Hey about the xy thing, the main instance that led to this is 0/0
23:56	ifiht	I need it to return a value, and I can do it in c, but Java stops it from happening in clojure
23:56	ifiht	THAT's x
23:56	amalloy	,(/ 0 0.0)
23:56	clojurebot	NaN
23:57	ifiht	yep
23:57	amalloy	that's a value
23:57	ifiht	,(/ 0 0)
23:57	clojurebot	#<ArithmeticException java.lang.ArithmeticException: Divide by zero>
23:57	ifiht	Oh, i see
23:57	ifiht	*numerical value
23:57	amalloy	if you divide an integer 0 by an integer 0 in C, i don't think anything good happens to you after that
23:58	amalloy	in either language, you can do it with ieee floats, and you get back NaN
23:58	ifiht	it can, but it's pretty processor dependant
23:58	ifiht	nul = (0 / 0); if (c == (1 \|\| nul)) <- like that
23:59	amalloy	then it's not "in C", it's "with my particular C compiler on this one computer", because your program isn't standards-compliant
23:59	ifiht	I'm dividing by 0, and you bring up standards compliance?