#clojure – 2014-06-15

2014-06-15

00:01	catern	umpa: er, which values in particular are you trying to get?
00:02	catern	the first and last of the first in the list, and the first and last of the last in the list?
00:02	umpa	catern: yes
00:02	catern	no offense, but it sounds like you're doing something horribly wrong...
00:02	catern	why do you need those values in particular?
00:03	umpa	I use them as arguments
00:03	catern	for what?
00:04	umpa	for an intersection with a dpendency
00:06	catern	if you say so. you could write a function: (defun bounds [seq] (list (first seq) (last seq))) and then do (map bounds (bounds x))
00:09	TEttinger	why do you need the last of the first list?
00:09	TEttinger	don't you just need ffirst and (last (last ...)) ?
00:10	johnwalker	you really like partition
00:11	umpa	catern: nice
00:11	umpa	TEttinger: its a grid with corners
00:11	TEttinger	oh I get it
00:12	catern	ah, that's not that bad then
00:19	deathknight	been at this for hours...would probably take one of you guys about 5-10 minutes to help me do this
00:19	deathknight	trying to auth with google with this tutorial: http://blog.knoldus.com/2014/04/05/google-sign-in-using-clojure/
00:19	deathknight	When I add that to a new compojure project, I get an http 400 error
00:24	benkay	is there a dumb-dumb test-profiler kicking around?
00:26	johnwalker	by google sign-in, do you mean openid?
00:27	johnwalker	deathknight: or are those separate things entirely?
00:27	deathknight	Sorry for not being clear
00:27	deathknight	Google uses OAuth 2
00:28	johnwalker	ahh
00:29	deathknight	As of 1 minute ago I just found an example on github that has a Google Oauth2 login handler that uses the friend lib!!
00:30	deathknight	so pumped
00:30	johnwalker	oh nice
00:31	deathknight	whats nice is the 3 energy drinks
00:31	johnwalker	caffeine is bad for you
00:32	johnwalker	,(def caffeine-overflow "8mg")
00:32	clojurebot	#'sandbox/caffeine-overflow
00:33	deathknight	keep going
00:33	benkay	,(println caffeine-overflow)
00:33	clojurebot	8mg\n
00:33	deathknight	mmmm
00:34	benkay	hm.
00:34	noonian	,caffeine-overflow
00:34	clojurebot	"8mg"
00:34	benkay	a yeah.
00:35	benkay	thanks, noonian
00:38	noonian	hmm, 8mg is not much caffeine at all
00:39	johnwalker	lol
00:41	benkay	noonian: http://goo.gl/oWcMzB
00:42	noonian	haha, that looks dangerous
00:42	johnwalker	related: https://twitter.com/dozba/status/477917824870912000
00:43	benkay	tchyeah
00:43	systemfault	The LD50 of caffeine in humans is dependent on individual sensitivity, but is estimated to be about 150 to 200 milligrams per kilogram of body mass or roughly 80 to 100 cups of coffee for an average adult
00:44	benkay	when i was working out in the burbs and the coffee was miserable and likely moldy i was in the habit of dropping 200 mikes into 2 liters of water, and drinking one of those concoctions in the morning and after lunch
00:44	benkay	'twas GREAT
00:44	TEttinger	mikes?
00:44	TEttinger	http://www.amazon.com/ProLab-Caffeine-Maximum-Potency-100-Count/dp/B0011865IQ/ref=pd_sbs_hpc_17?ie=UTF8&refRID=0AA4EDD7RYQK16J99SJF these are not that different from 2 cups coffee
00:45	benkay	TEttinger: mg
00:45	TEttinger	(taking one now)
00:45	benkay	that's a 200 mike pill as well :)
00:46	TEttinger	the 2 liters water is odd though
00:46	TEttinger	it seems like a ton
00:46	deathknight	has anyone used friend-oauth2 to communicate with google?
00:48	luxbock	I'm having some trouble trying to merge two project files generated from templates to get the best out of both
00:48	TEttinger	luxbock, got them in a gist maybe?
00:48	benkay	TEttinger: 4L/day is a superb hydration level.
00:48	luxbock	https://gist.github.com/anonymous/70c4e04b84c1197e756e
00:49	luxbock	this is what I have
00:49	TEttinger	benkay, no doubt, I drink water constantly
00:49	TEttinger	but finishing a 1.5 liter bottle in a sitting would be a challenge
00:49	benkay	TEttinger: i rarely made it through the whole second bottle.
00:49	luxbock	but when I start try to use cider-repl-jack-in it complains about the austin class file missing
00:50	benkay	ah well it was over the course of the pre-lunch drag and the post-lunch drag
00:50	luxbock	so basically it's not finding the stuff I have defined in the dev-profile
00:50	benkay	that job was a real drag
00:50	benkay	i've graduated to a .75L bottle always at hand. people think i'm a wino when they first see this in practice!
00:50	luxbock	reading the Leiningen profile docs it says :dev profile is automatically merged into the default profile, so I'm confused why it's not finding the :dev :plugins/dependenceis
00:51	TEttinger	WARNING Austin is not compatible with the (currently) latest ClojureScript release. (context) Use ClojureScript 0.0-2156 until the a new ClojureScript release is cut that includes the resolution of the underlying issue.
00:51	luxbock	ahhh
00:51	luxbock	yeah I just did `lein ancient update` before
00:51	luxbock	alright makes sense, thanks
00:52	TEttinger	so you would change the dep on line 8 to 0.0-2156 , luxbock
00:53	TEttinger	no prob, I actually only found it because I didn't know what austin was
00:54	luxbock	hmm, I still have the same problem
00:54	luxbock	I think these issues might be unrelated
00:54	TEttinger	gah
00:55	TEttinger	maybe https://gist.github.com/anonymous/70c4e04b84c1197e756e#file-project-clj-L40 doesn't have it loaded by the highlighted point?
00:56	luxbock	the two templates that I'm trying to merge are https://github.com/magomimmo/om-start-template and https://github.com/bhauman/devcards-template
00:56	umpa	,(defn coffee [sip] (coffee (inc sip)))
00:56	clojurebot	#'sandbox/coffee
00:56	luxbock	in om-start-template the :dev profile is defined in a profiles.clj file
00:56	luxbock	which I moved into the project.clj file for my project, thinking it would not make a difference
00:57	luxbock	but maybe that's the issue
00:57	umpa	,(coffee 0)
00:57	clojurebot	#<StackOverflowError java.lang.StackOverflowError>
00:58	TEttinger	I dunno, I've barely touched profiles
00:59	luxbock	yeah works now, though I'm not exactly sure what's actually different from leins point of view
01:02	deathknight	friend-oauth2-example's google example doesn't work -- it's not up to date with friend-oauth2's new :credential-fn. Is anyone familiar with how I can adapt this (https://github.com/ddellacosta/friend-oauth2-examples/blob/master/src/friend_oauth2_examples/google_handler.clj) to work correctly with this (https://github.com/ddellacosta/friend-oauth2/blob/master/test/friend_oauth2/test_helpers.clj#L65-L67)?
01:11	johnwalker	deathknight: if you can figure it out, it's good blog post material
01:12	deathknight	:D
01:12	johnwalker	i would retweet it to my fifteen followers who favorite but never retweet
01:12	deathknight	Haha!
01:12	deathknight	It's super over my head at the moment in the sense that I'm not sure where to start looking for the answer. Is there a direction your finger can point in so I can understand the problem?
01:14	johnwalker	does ddellacosta's link work?
01:14	johnwalker	er, the first one?
01:15	johnwalker	oh
01:15	deathknight	his google example does not work
01:15	deathknight	he even addressed it in an issue on the corresponding github page
01:16	johnwalker	but the oauth2 does?
01:16	deathknight	so it successfully takes me to google's oauth2 page where I give consent to the app
01:16	deathknight	and then upon redirect, there is a failure of permission authentication
01:16	deathknight	(even though I gave permission)
01:17	johnwalker	does this pr help?
01:17	johnwalker	https://github.com/ddellacosta/friend-oauth2/pull/18
01:17	deathknight	not sure -- looking in to it
01:18	deathknight	also, relevant to the issue: https://github.com/ddellacosta/friend-oauth2-examples/issues/4
01:18	johnwalker	oh, thats good stuff
01:19	johnwalker	i don't know but i hope you figure it out
01:19	deathknight	thank you John!
01:20	johnwalker	you're welcome, sorry i can't help
01:20	johnwalker	friend confuses me in general
02:07	benkay	I don't use Friend, tbh.
02:08	benkay	it's a little more work, but the flexibility and power of writing my own auth handlers is pretty great.
02:08	benkay	definitely not what people coming from the webdev community at large are expecting though vis-a-vis authentication libraries.
02:38	ambrosebs	a seq can only be infinite if (every-pred #(instance? clojure.lang.IPending %) (complement realized?))
02:38	ambrosebs	right?
02:39	ambrosebs	I don't think that make sense.
02:39	ambrosebs	,((every-pred #(instance? clojure.lang.IPending %) (complement realized?)) (seq (range)))
02:39	clojurebot	false
02:40	ambrosebs	so I guess all bets are off if you're holding an arbitrary seq.
02:44	dbasch	ambrosebs: it’s the halting problem essentially
02:45	amalloy	ambrosebs: IPending is very much a "well, maybe, whatever" - it doesn't really mean anything reliable
02:47	amalloy	and the ISeq type is open, so i can easily create an instance of it which is infinite but doesn't implement IPending, with something like (reify ISeq (first [this] 1) (more [this] this))
02:57	amalloy	&(take 10 (reify clojure.lang.ISeq (first [this] 1) (more [this] this) clojure.lang.Seqable (seq [this] this) clojure.lang.Sequential)) ;; to be specific, this is apparently the minimum list of interfaces you need to implement to be printable
02:57	lazybot	⇒ (1 1 1 1 1 1 1 1 1 1)
03:10	ambrosebs	amalloy: interesting.
03:12	ambrosebs	assuming seqs are infinite is probably a cleaner solution for me anyway.
03:51	sveri	Hi, I am struggling to check if a function call generates an assertion with test.check. In a normal test I would call (is (thrown? Assert (function... But how do I do that with test.check?
03:57	visof	Hello
03:58	visof	,(clojure.string/replace (clojure.string/replace "\"Hello\nWorld\"" #"\"" "") "\n" " ")
03:58	clojurebot	"Hello World"
03:58	visof	is there a better way to do this?
04:04	TEttinger	it's two separate replacements?
04:13	danielcompton	visof: so the original string is "\"Hello\nWorld\"
04:13	danielcompton	,(println "\"Hello\nWorld\")
04:13	clojurebot	#<RuntimeException java.lang.RuntimeException: EOF while reading string>
04:13	danielcompton	,(println "\"Hello\nWorld\"")
04:13	clojurebot	"Hello\nWorld"\n
04:24	sveri	reiddraper: pint
04:24	sveri	*ping
04:59	ptcek	Datomic users here? How to upsert entity when having :db/id stored in var? I am getting symbol cannot be cast to number error... https://www.refheap.com/86602
06:08	toxmeister	with fressian, how important is it to use write-footer? can't seem to find any docs and it seems to work without… do I only need to use it when writing more than a single object?
07:45	kzar	$findfn {:a {:b 2} :c 3} :a :b {:c 3}
07:45	lazybot	[clojure.core/dissoc]
07:45	kzar	$findfn {:a {:b 2} :c 3} :a :b {:a {} :c 3}
07:46	lazybot	[]
07:46	kzar	$findfn {:a {:b 2 :g 4} :c 3} [:a :b] {:a {:g 4} :c 3}
07:46	lazybot	[]
07:47	kzar	$findfn {:a {:b 2 :g 4} :c 3} :a :b {:a {:g 4} :c 3}
07:47	lazybot	[]
07:49	kzar	,(update-in {:a {:b 2 :z 4} :c 3} [:a] dissoc :b)
07:50	clojurebot	{:c 3, :a {:z 4}}
07:52	kzar	,(defn dissoc-in [m keys]
07:52	kzar	(update-in m (butlast keys) dissoc (last keys)))
07:52	clojurebot	#<RuntimeException java.lang.RuntimeException: EOF while reading>
07:52	kzar	,(defn dissoc-in [m keys] (update-in m (butlast keys) dissoc (last keys)))
07:52	clojurebot	#'sandbox/dissoc-in
07:52	kzar	,(dissoc-in {:a {:b 2 :z 4} :c 3} [:a :b])
07:52	clojurebot	{:c 3, :a {:z 4}}
07:53	lvh	Is it okay for the result-expr of a case to have side effects?
07:53	lvh	are they only eval'd for the particular case?
08:42	Jarda	hm
08:58	visof	is there a fast and good way to make values of hash as a keys? as example {1 2 3 4} tobe {2 1 4 3} ?
09:06	sandbags	,(into {} (map (fn [[k v]] [v k]) {:a 1 :b 2 :c 3})
09:06	clojurebot	#<RuntimeException java.lang.RuntimeException: EOF while reading>
09:06	sandbags	,(into {} (map (fn [[k v]] [v k]) {:a 1 :b 2 :c 3})))
09:06	clojurebot	{3 :c, 2 :b, 1 :a}
09:06	sandbags	visof: ^
09:06	visof	sandbags: well done
09:07	sandbags	i suspect there's an even easier way, there usually is :)
09:48	gfredericks	with prismatic/plumbing it'd be (for-map [[k v] m] v k)
09:51	mi6x3m	anyway of determining if a text is a valid html?
09:53	sveri	mi6x3m: there are some java libraries for this which I would use
09:55	mi6x3m	sveri: yeah, I know, was wondering if there is something more clojure-ish
09:57	sveri	mi6x3m: not that I know of, but that doesnt mean anything. Maybe enlive provides something?
09:57	mi6x3m	checking it out just now :)
10:01	deathknight	what is the name of -> so that I may look up what it does?
10:05	ambrosebs	deathknight: http://blog.fogus.me/2010/09/28/thrush-in-clojure-redux/
10:06	ambrosebs	or thread-first
10:06	deathknight	you rock
10:06	deathknight	thats the author o JOC, right?
10:06	ambrosebs	yes
10:06	deathknight	:D
10:07	deathknight	Today is the day I replace who sits on the throne - err, communicate with google calendar's API
10:16	umpa	(require 'clojure.set)
10:16	umpa	,(require 'clojure.set)
10:16	clojurebot	nil
10:16	umpa	,(set/difference #{1 4} #{1 4 7})
10:16	clojurebot	#<CompilerException java.lang.RuntimeException: No such namespace: set, compiling:(NO_SOURCE_PATH:0:0)>
10:16	umpa	,(clojure.set/difference #{1 4} #{1 4 7})
10:16	clojurebot	#{}
10:17	umpa	,(clojure.set/difference #{1 4 7} #{1 4})
10:17	clojurebot	#{7}
10:17	michaniskin	,(clojure.set/difference #{1 4 7} #{1} #{4})
10:17	clojurebot	#{7}
10:18	umpa	How do I get the difference no matter the order ?
10:19	engblom	Is there a ready function in core that would be similar to map, but only changing the first element matching a condition. For example if you want to decrese the first non-zero element in a collection.
10:20	ambrosebs	,(require '[clojure.set :as set])
10:20	clojurebot	nil
10:21	ambrosebs	,(let [ss [#{1 4 7} #{1} #{4}]] (set/difference (apply set/union ss) (apply set/intersection ss)))
10:21	clojurebot	#{7 1 4}
10:21	ambrosebs	some combination of set operations..
10:21	michaniskin	heh
10:22	michaniskin	perhaps you want the complement of the intersection?
10:22	michaniskin	i guess thats what ambrosebs just did
10:23	ambrosebs	probably the union of all the combinations of calling difference.
10:24	ambrosebs	,(require 'clojure.math.combinatorics)
10:24	clojurebot	#<FileNotFoundException java.io.FileNotFoundException: Could not locate clojure/math/combinatorics__init.class or clojure/math/combinatorics.clj on classpath: >
10:24	michaniskin	i mean order is kind of important for difference, like minus: (- 1 3) is not the same as (- 3 1)
10:26	engblom	,(Math/abs (- 3 1))
10:26	clojurebot	2
10:26	engblom	,(Math/abs (- 1 3))
10:26	clojurebot	2
10:26	engblom	michaniskin: ^
10:27	ambrosebs	(apply union (map (fn [[s1 s2]] (union (difference s1 s2) (difference s2 s1)) (comb/combinations ss 2))
10:27	michaniskin	engblom: i don't understand?
10:27	ambrosebs	something like that?
10:27	engblom	<michaniskin> i mean order is kind of important for difference, like minus: (-
10:27	engblom	1 3) is not the same as (- 3 1)
10:29	michaniskin	engblom: sure, or (Math/min -10000 (- 3 1))
10:29	ambrosebs	funny enough, I just needed this kind of set operation. Useful if you already know the differences in advance.
10:29	michaniskin	lots of things you can do to wrap minus in something lol
10:31	john2x	does `& rest` arguments work with multi-arity functions?
10:31	umpa	ambrosebs: the first one works
10:31	ambrosebs	umpa: oh. Yes you're right :P
10:32	michaniskin	`(defn ff ([x] :one-arg) ([x & more] :more-args))
10:32	michaniskin	,(defn ff ([x] :one-arg) ([x & more] :more-args))
10:32	clojurebot	#'sandbox/ff
10:32	michaniskin	,(ff :x)
10:32	clojurebot	:one-arg
10:32	michaniskin	,(ff :x :y)
10:32	clojurebot	:more-args
10:33	michaniskin	john2x: is that what you mean?
10:33	john2x	michaniskin: oh, I mean both definitions having `& more`..
10:35	michaniskin	john2x: that would be kind of crazy, no?
10:35	michaniskin	also
10:35	john2x	yeah..
10:36	michaniskin	,(defn ff ([x] :one-arg) ([[x & rest] & {:keys [opt] :as options}] :bad))
10:36	clojurebot	#'sandbox/ff
10:36	michaniskin	,(ff [:a])
10:36	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: ff in this context, compiling:(NO_SOURCE_PATH:0:0)>
10:38	engblom	Assuming I have [0 0 3 2 2] and I want to decrease the first non-zero element to get [0 0 2 2 2], is there any ready function I can pass #(not (zero? %)) and dec and the vector to and it will take care of it?
10:49	gfredericks	engblom: no, that sounds pretty specific to me
10:50	gfredericks	,(defn decfirst ([v] (decfirst v 0)) ([v i] (if (pos? (v i)) (assoc v i (dec (v i))) (recur v (inc i)))))
10:50	clojurebot	#'sandbox/decfirst
10:50	gfredericks	,(decfirst [0 0 3 2 2])
10:50	clojurebot	[0 0 2 2 2]
10:57	sandbags	maybe using partition-by and zero?
10:58	michaniskin	,(defn replacefirst [pred f coll] (let [head (take-while (complement pred) coll), [item & tail] (drop-while pred coll)] (into (empty coll) (concat head (cons (f item) tail)))))
10:58	clojurebot	#'sandbox/replacefirst
10:58	michaniskin	,(replacefirst (complement zero?) dec [0 0 3 2 2])
10:58	clojurebot	[0 0 -1 0 3 ...]
10:58	michaniskin	whoa nevermind
10:58	michaniskin	lol
11:00	michaniskin	,(replacefirst zero? dec [0 0 3 2 2])
11:00	clojurebot	[2 2 2]
11:00	michaniskin	,(replacefirst pos? dec [0 0 3 2 2])
11:00	clojurebot	[0 0 -1 0 3 ...]
11:00	johnwalker	whats the idea? update the first element in a collection that matches pred?
11:01	michaniskin	indeed
11:02	michaniskin	,(defn replacefirst [pred f coll] (let [head (take-while (complement pred) coll), [item & tail] (drop-while (complement pred) coll)] (into (empty coll) (concat head (cons (f item) tail)))))
11:02	clojurebot	#'sandbox/replacefirst
11:03	michaniskin	,(replacefirst (complement zero?) dec [0 0 3 2 2])
11:03	clojurebot	[0 0 2 2 2]
11:03	michaniskin	whew
11:05	michaniskin	,(replacefirst #(pos? (val %)) dec (sorted-map :a 0 :b 0 :c 3 :d 3 :e 2))
11:05	clojurebot	#<ClassCastException java.lang.ClassCastException: clojure.lang.PersistentTreeMap$BlackVal cannot be cast to java.lang.Number>
11:06	michaniskin	,(replacefirst #(pos? (val %)) (juxt identity dec) (sorted-map :a 0 :b 0 :c 3 :d 3 :e 2))
11:06	clojurebot	#<ClassCastException java.lang.ClassCastException: clojure.lang.PersistentTreeMap$BlackVal cannot be cast to java.lang.Number>
11:12	johnwalker	(defn update-first [pred f coll] (let [[x [t & r]] (split-with (complement pred) coll)] (concat x [(f t)] r)))
11:12	mi6x3m	hey, can anyone tell if this looks like something a sane clojure programmer would produce? http://pastebin.com/KdqeTb7p
11:12	mi6x3m	Trying to get a feeling for these things :)
11:13	johnwalker	yeah looks good
11:14	mi6x3m	johnwalker: thanks
11:14	michaniskin	ah split-with, that's right
11:15	michaniskin	it would be useful to have the FL "construct" function in clojure: (construct [inc dec] [10 10]) => [11 9]
11:15	johnwalker	whats FL?
11:16	johnwalker	i see what construct does, but haven't heard of FL
11:17	michaniskin	johnwalker: http://www.thocp.net/biographies/papers/backus_turingaward_lecture.pdf
11:17	michaniskin	FL is pretty mindblowing
11:18	michaniskin	basically in FL the expression [inc dec] is like in clojure (partial construct [inc dec])
11:18	michaniskin	square brackets indicate construction
11:19	ohpauleez	michaniskin: Have you played with J?
11:19	michaniskin	ohpauleez: hey dude! no, is it good?
11:19	ohpauleez	Dude, you need to
11:19	ohpauleez	basically all the good ideas from FP and FL
11:20	ohpauleez	plus some extra solid stuff
11:20	ohpauleez	Hmm, a quick Google search shows me this: http://www.infoq.com/presentations/j-language
11:21	ohpauleez	And, like other programming languages, good history on wikipedia: http://en.wikipedia.org/wiki/J_%28programming_language%29
11:21	michaniskin	i was really into FL at one point, and i was a little obsessed with forth, but i find that applications need to be expressed as trees
11:22	michaniskin	it's not really easy to do that in FL or stack based languages
11:22	johnwalker	(map #(% %2) [inc dec] [10 11])
11:22	johnwalker	,(map #(% %2) [inc dec] [10 11])
11:22	clojurebot	(11 10)
11:22	kzar	Is there a way to set jruby as a dependency in my project.clj? (Not sure how to add non-clojars dependencies.)
11:22	hyPiRion	,(map deliver [inc dec] [10 11])
11:22	clojurebot	(11 10)
11:23	johnwalker	ahh, nice
11:23	johnwalker	,(doc deliver)
11:23	clojurebot	"([promise val]); Delivers the supplied value to the promise, releasing any pending derefs. A subsequent call to deliver on a promise will have no effect."
11:23	hyPiRion	except never use deliver like that
11:23	michaniskin	whoa!
11:23	michaniskin	(inc hyPiRion)
11:23	lazybot	⇒ 37
11:23	hyPiRion	&(source deliver)
11:23	lazybot	java.lang.RuntimeException: Unable to resolve symbol: source in this context
11:23	hyPiRion	,(source deliver)
11:23	clojurebot	Source not found\n
11:23	hyPiRion	$source deliver
11:23	lazybot	deliver is http://is.gd/U3dg84
11:24	michaniskin	ohpauleez: do you program in J?
11:24	catern	michaniskin: (def construct (partial map apply)) ?
11:24	ohpauleez	kzar: You can add any clojars or maven central dep. You can also specify other maven repositories with :repositories
11:25	ohpauleez	michaniskin: Nope, just briefly studied some time ago (I was digging through FL and FP, and stumbled upon it when I asked, "Has anyone taken these in a new direction?"
11:25	ohpauleez	hyPiRion: How the hell did you discover that abuse?
11:26	kzar	ohpauleez: Oh cool [org.jruby/jruby "1.7.12"] did it, thanks
11:26	ohpauleez	kzar: No problem, happy to help!
11:26	hyPiRion	ohpauleez: you don't want to know
11:26	michaniskin	,(def construct (partial map deliver))
11:26	clojurebot	#'sandbox/construct
11:26	michaniskin	,(construct [inc dec] [10 10])
11:26	clojurebot	(11 9)
11:27	michaniskin	,(replacefirst #(pos? (val %)) (construct identity dec) (sorted-map :a 0 :b 0 :c 3 :d 3 :e 2))
11:27	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: replacefirst in this context, compiling:(NO_SOURCE_PATH:0:0)>
11:28	michaniskin	,(update-first #(pos? (val %)) (construct identity dec) (sorted-map :a 0 :b 0 :c 3 :d 3 :e 2))
11:28	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: update-first in this context, compiling:(NO_SOURCE_PATH:0:0)>
11:32	umpa	(update-first #(pos? (val %)) (construct identity dec) (sorted-map :a 0 :b 0 :c 3 :d 3 :e 2))
11:32	umpa	(update-first #(pos? (val %)) (construct identity dec) (sorted-map :a 0 :b 0 :c 3 :d 3 :e 2))
11:33	umpa	sorry about that
11:33	michaniskin	it won't work because construct should really return a vector, not a seq
11:33	michaniskin	,(def construct (comp ver (partial map deliver)))
11:33	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: ver in this context, compiling:(NO_SOURCE_PATH:0:0)>
11:33	michaniskin	,(def construct (comp vec (partial map deliver)))
11:33	clojurebot	#'sandbox/construct
11:34	umpa	,(update-first #(pos? (val %)) (construct identity dec) (sorted-map :a 0 :b 0 :c 3 :d 3 :e 2))
11:34	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: update-first in this context, compiling:(NO_SOURCE_PATH:0:0)>
11:35	umpa	$source identity
11:35	lazybot	identity is http://is.gd/iexysV
11:38	johnwalker	\join #org-mode
11:39	johnwalker	woops
12:16	johnny_	hello!
12:16	johnny_	anyone around?
12:17	engblom	johnwalker: In your solution for update-first, how is that let working as t and r are not yet set at that moment? (defn update-first [pred f coll] (let [[x [t & r]] (split-with (complement pred) coll)] (concat x [(f t)] r)))
12:17	engblom	johnny_: Yep
12:18	johnny_	yay! I'm a clojure newbie; I've done a fair bit if LISPing before... is this the place for me?
12:18	engblom	johnny_: Yes, this is the place
12:18	johnwalker	engblom: it's wacky destructuring
12:18	johnny_	engblom: grand!
12:19	johnwalker	engblom: x t and r are the left values for split-with
12:20	engblom	johnwalker: So even the binding in the let is not happening until you have some values?
12:22	johnwalker	i'm not sure if i understand the question, but this should help:
12:22	johnwalker	http://blog.jayfields.com/2010/07/clojure-destructuring.html
12:24	engblom	johnwalker: This is the first destructuring I simply have really difficulties to grasp. I will look at that page to see if I get the point.
12:25	johnwalker	the idea is that x gets the first list
12:25	johnwalker	and then [t & r] get the second
12:26	johnwalker	t is the first element of the second list
12:26	johnwalker	and r is the rest of the second list
12:26	engblom	I always assumed the [] in let was evaluated before the body of let. Not having f & t when let begins causes my trouble to understand
12:26	johnwalker	split-with isn't being evaluated in the body
12:27	johnwalker	it's the right value for [x [t & r]]
12:27	engblom	Ah! Now I see
12:27	engblom	Thanks, that was what I failed to notice
12:27	johnwalker	cool. yeah, it's hard to parse on one line
12:28	visof	Hi
12:31	visof	if i can a string like "Hello_$002C_$0021", and i want to replace all unicode chars like $002C which is "," and $0021 "!" that's easy to do with 2 replace statements one inside another, but how can i do this for undetermined n of unicodes ?
12:32	visof	,(clojure.string/replace #"$0021" (clojure.string/replace #"002C" "Hello_$002C_$0021" ",") "!")
12:32	clojurebot	"$0021"
12:32	visof	,(clojure.string/replace #"$0021" (clojure.string/replace #"$002C" "Hello_$002C_$0021" ",") "!")
12:32	clojurebot	"$0021"
12:32	visof	,(clojure.string/replace #"$002C" "Hello_$002C_$0021" ",")
12:32	clojurebot	"$002C"
12:33	visof	,(clojure.string/replace (clojure.string/replace "Hello_$002C_$0021" #"$002C" ",") #"$0021" "!")
12:33	clojurebot	"Hello_$002C_$0021"
12:33	visof	,(clojure.string/replace (clojure.string/replace "Hello_$002C_$0021" #"\$002C" ",") #"\$0021" "!")
12:33	clojurebot	"Hello_,_!"
12:35	johnwalker	you can do it with reduce, but it won't be very efficient
12:37	visof	johnwalker: https://www.refheap.com/86611 to replace all codes by symbols, i created two lists as you can see
12:38	visof	johnwalker: maybe define empty atom and change it when replace string
12:38	johnwalker	no thats not a good idea
12:38	visof	johnwalker: why?
12:39	johnwalker	how is using an atom in this manner different from regular mutation?
12:40	johnwalker	or a global variable in general?
12:40	visof	so you think reduce is better?
12:40	johnwalker	a lot better
12:40	johnwalker	but still not ideal
12:41	johnwalker	one thing you might try is
12:41	johnwalker	actually nevermind, that's not a good solution either
12:42	johnwalker	you can do this and it's basically ok for small strings
12:42	johnwalker	http://stackoverflow.com/questions/9568050/in-clojure-how-to-write-a-function-that-applies-several-string-replacements
12:46	johnwalker	a better way would be to compile an fsm to handle multiple replacements
12:47	johnwalker	java probably has a standard solution, but i'd also be interested in seeing a clojure implementation
12:51	umpa	How do i group list elements ((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)) given intersect dependency (0 2 6 8) to arrive at ( 0 ((0 1 2) (0 3 6) (0 4 8)) 2 ((0 1 2)(2 5 8) etc..))
12:51	justin_smith	(apply str (replace {"$002C", \,, "$0021", \!} "Hello_$002C_$0021")) ;; visof: depending on how many replacements, this may make sense
12:52	justin_smith	,(apply str (replace {"$002C", \,, "$0021", \!} "Hello_$002C_$0021")) ;; visof
12:52	clojurebot	"Hello_$002C_$0021"
12:52	justin_smith	err...
12:52	justin_smith	oh, if the things to replace were chars and not strs it would work
12:53	justin_smith	,(apply str (replace {\☃, \,, \✈, \!} "Hello_☃_✈"))
12:53	clojurebot	"Hello_,_!"
12:55	hyPiRion	you guys
12:55	hyPiRion	,(clojure.string/replace "Hello_$002C_$0021" #"\$(\p{XDigit}{4})" (fn [b] (-> b second (Integer/parseInt 16) char str)))
12:55	clojurebot	"Hello_,_!"
12:56	hyPiRion	replace can take a function as last argument
12:57	justin_smith	hyPiRion: cool, I think I misunderstood the task here - and I think you can eliminate the str from that last function
12:57	justin_smith	nope, str is needed
12:59	johnwalker	thats badass
12:59	hyPiRion	visof: ^ does that solve your problem?
13:00	bbloom	(doc clojure.string/replace)
13:00	clojurebot	"([s match replacement]); Replaces all instance of match with replacement in s. match/replacement can be: string / string char / char pattern / (string or function of match). See also replace-first. The replacement is literal (i.e. none of its characters are treated specially) for all cases above except pattern / string. For pattern / string, $1, $2, etc. in the replacement string are substituted with the string that
13:02	alandipert	hyPiRion, mindblown.gif
13:02	bbloom	,(clojure.string/replace "Hello_X_X" "X" (fn [_] "Y"))
13:02	clojurebot	#<ClassCastException java.lang.ClassCastException: sandbox$eval25$fn__26 cannot be cast to java.lang.CharSequence>
13:02	bbloom	hyPiRion: seems like an implementation detail ^^
13:03	bbloom	,(-> #'clojure.string/replace-by meta :private)
13:03	clojurebot	true
13:03	bbloom	i wouldn't rely on that...
13:05	hyPiRion	bbloom: oh what. I honestly thought it was documented that you can pass functions as replacement values for regexes.
13:06	johnwalker	hold on now
13:07	johnwalker	,(clojure.string/replace "Hello_X_X" #"X" (fn [_] "Y"))
13:07	clojurebot	"Hello_Y_Y"
13:07	bbloom	johnwalker: yeah, it only works if the match argument is a regex
13:07	hyPiRion	johnwalker: It doesn't make sense for a string anyway. You will always get the string as argument.
13:08	hyPiRion	Well, unless you use mutation under covers.
13:08	johnwalker	oh, you're saying that this works too
13:08	bbloom	hyPiRion: if you want to rely on that, i'd submit a patch with a new doc string ;-)
13:08	johnwalker	,(clojure.string/replace "Hello_X_X" "X" "Y")
13:08	clojurebot	"Hello_Y_Y"
13:08	bbloom	johnwalker: right
13:09	johnwalker	still, that's inconsistent
13:09	hyPiRion	bbloom: Well, I rely on that at least (for regexes). I was just surprised that it wasn't documented.
13:09	hyPiRion	I have used clojure.string/replace wrong all this time.
13:09	hyPiRion	"wrong"
13:11	bbloom	hyPiRion: like i said, a doc string change would probably fix that right up :-)
13:11	johnwalker	shouldn't it also be extended to accept functions as the second argument for strings?
13:12	johnwalker	it might not be useful but it's certainly more regular
13:13	hyPiRion	bbloom: Sure, I'd be happy if someone were to file a patch for that. Seems super useful to let other people know about it
13:21	umpa	How do I group-by (0 7) coll ((0 2)(1 3)(0 5)(1 7)) ?
13:23	justin_smith	umpa: what does (group-by (0 7) mean? group by the presence of 0 vs. 7?
13:24	umpa	justin_smith: in a map form maybe {0 ((0 2)(05)) 7 ((1 7))}
13:25	justin_smith	,(group-by #(#{0 7} (into #{} %)) '((0 2)(1 3)(0 5)(1 7)))
13:25	clojurebot	{nil [(0 2) (1 3) (0 5) (1 7)]}
13:25	justin_smith	err
13:26	justin_smith	,(group-by #(clojure.set/intersection #{0 7} (into #{} %)) '((0 2)(1 3)(0 5)(1 7))) closer
13:26	clojurebot	#<CompilerException java.lang.ClassNotFoundException: clojure.set, compiling:(NO_SOURCE_PATH:0:0)>
13:26	justin_smith	,(require 'clojure.set)
13:26	clojurebot	nil
13:26	justin_smith	,(group-by #(clojure.set/intersection #{0 7} (into #{} %)) '((0 2)(1 3)(0 5)(1 7))) closer
13:26	clojurebot	{#{0} [(0 2) (0 5)], #{} [(1 3)], #{7} [(1 7)]}
13:27	umpa	justin_smith: neat
13:29	umpa	,(group-by #(clojure.set/intersection #{0 2 6 8} (into #{} %)) (((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6))))
13:29	clojurebot	#<ClassCastException java.lang.ClassCastException: java.lang.Long cannot be cast to clojure.lang.IFn>
13:30	umpa	,(group-by #(clojure.set/intersection #{0 2 6 8} (into #{} %)) '(((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6))))
13:30	clojurebot	{#{} [((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) ...)]}
13:31	justin_smith	umpa: that happened because you changed the list of lists to a list of list of lists
13:31	umpa	,(group-by #(clojure.set/intersection #{0 2 6 8} (into #{} %)) '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)))
13:31	clojurebot	{#{0 2} [(0 1 2)], #{} [(3 4 5) (1 4 7)], #{6 8} [(6 7 8)], #{0 6} [(0 3 6)], #{2 8} [(2 5 8)], ...}
13:32	justin_smith	maybe you want something that takes the intersection, and decides what the "most relevant" key out of that is, and returns that alone
13:33	justin_smith	and you would wrap the intersection call in that extractor / orderer
13:34	justin_smith	something like #(or (% 6) (% 8) (% 2) (% 0))
13:35	justin_smith	,(group-by #( (fn [s] (or (s 6) (s 8) (s 2) (s 0))) (clojure.set/intersection #{0 2 6 8} (into #{} %))) '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)))
13:35	clojurebot	#<CompilerException java.lang.ClassNotFoundException: clojure.set, compiling:(NO_SOURCE_PATH:0:0)>
13:36	justin_smith	,(require 'clojure.set)
13:36	clojurebot	nil
13:36	justin_smith	,(group-by #( (fn [s] (or (s 6) (s 8) (s 2) (s 0))) (clojure.set/intersection #{0 2 6 8} (into #{} %))) '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)))
13:36	clojurebot	{2 [(0 1 2)], nil [(3 4 5) (1 4 7)], 6 [(6 7 8) (0 3 6) (2 4 6)], 8 [(2 5 8) (0 4 8)]}
13:38	umpa	justin_smith: nice
13:40	justin_smith	that fn could be (fn [s] (first (keep s [6 8 2 0]))) actually
13:44	umpa	,(group-by #((fn [s] (first (keep s [6 8 2 0]))) (clojure.set/intersection '(0 2 6 8) (into #{} %))) '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6))) like that ?
13:44	clojurebot	#<IllegalArgumentException java.lang.IllegalArgumentException: contains? not supported on type: clojure.lang.PersistentList>
13:45	umpa	(doc keep)
13:45	clojurebot	"([f coll]); Returns a lazy sequence of the non-nil results of (f item). Note, this means false return values will be included. f must be free of side-effects."
13:46	justin_smith	umpa: the reason I was using sets was to be able to do the intersection / contains? stuff, which does not work on seqs
13:46	justin_smith	well, get rather than contains? but you know
13:47	justin_smith	you replaced a # with a ' to get a seq rather than a set, which made that code fail
13:49	umpa	justin_smith: you are abs right
13:50	umpa	getting tangled
13:50	umpa	(umpa untangle)
13:51	umpa	,(umpa untangle)
13:51	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: umpa in this context, compiling:(NO_SOURCE_PATH:0:0)>
13:51	justin_smith	right, it's a one liner for clojure bot demo sake, but better to actually make it readable
13:53	umpa	justin_smith: whats (first (keep s []
13:53	justin_smith	,(keep #{0 1 2} (range 10))
13:53	clojurebot	(0 1 2)
13:53	justin_smith	bad example
13:53	justin_smith	,(keep #{10 1 42} (range 100))
13:53	clojurebot	(1 10 42)
13:54	justin_smith	,(first (keep #{10 1 42} (range 100)))
13:54	clojurebot	1
13:54	justin_smith	just gives you the first member of the set, found in the collection
13:54	justin_smith	,(first (keep #{10 1 42} (reverse (range 100))))
13:54	clojurebot	42
13:56	justin_smith	so the ordering of the collection, decides which of its members will be returned if the set contains multiple of its members
13:56	justin_smith	(misplaced ,)
14:01	mi6x3m	hey clojure, is it idiomatic to use clojure.repl/source in other scenarios than the REPL
14:01	justin_smith	what's the context?
14:01	justin_smith	if you are recompiling something from a new ns based on the output of clojure.repl/source, that is just wrong
14:03	mi6x3m	justin_smith: I am translating a program showing java examples to a library
14:03	mi6x3m	and I shows controls + a "source" tabg
14:03	mi6x3m	clicking on the source tab fetches the .java file's contents
14:03	justin_smith	ahh, that should be OK :)
14:08	umpa	,(group-by #((fn [s] (first (keep s [6 8 2 0]))) (clojure.set/intersection #{0 2 6 8} (into #{} %))) '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6))) some matches are missing
14:08	clojurebot	#<CompilerException java.lang.ClassNotFoundException: clojure.set, compiling:(NO_SOURCE_PATH:0:0)>
14:09	umpa	,(group-by #((fn [s] (first (keep s [6 8 2 0]))) (clojure.set/intersection #{0 2 6 8} (into #{} %))) '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)))
14:09	clojurebot	#<CompilerException java.lang.ClassNotFoundException: clojure.set, compiling:(NO_SOURCE_PATH:0:0)>
14:09	justin_smith	you need to require clojure.set
14:09	justin_smith	,(require 'clojure.set)
14:09	clojurebot	nil
14:09	umpa	,(group-by #((fn [s] (first (keep s [6 8 2 0]))) (clojure.set/intersection #{0 2 6 8} (into #{} %))) '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)))
14:09	clojurebot	{2 [(0 1 2)], nil [(3 4 5) (1 4 7)], 6 [(6 7 8) (0 3 6) (2 4 6)], 8 [(2 5 8) (0 4 8)]}
14:09	mi6x3m	justin_smith: thanks but when is the source available, to begin with?
14:10	mi6x3m	is it always included with the build artifacts?
14:10	justin_smith	the nil key shows you which items did not contain any of the keys you were looking for
14:10	justin_smith	mi6x3m: most clojure code is distributed as source, or at least includes source alongside the compiled classes
14:11	umpa	the 2 should be {2 [(0 1 2)(2 5 8)(2 4 6] ...
14:11	justin_smith	umpa: than change [6 8 2 0] to [2 6 8 0]
14:11	justin_smith	or do you want every permutation present?
14:12	umpa	^
14:17	justin_smith	,((fn [ks ls] (let [lsets (map set ls)] (reduce (fn [m v] (assoc m v (filter #(contains? (set %) v) ls))) {} ks))) [0 2 6 8] '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)))
14:17	clojurebot	{8 ((6 7 8) (2 5 8) (0 4 8)), 6 ((6 7 8) (0 3 6) (2 4 6)), 2 ((0 1 2) (2 5 8) (2 4 6)), 0 ((0 1 2) (0 3 6) (0 4 8))}
14:18	justin_smith	hmm, but no more nil key this way
14:18	justin_smith	oh and I bound lsets and never used it anyway :)
14:22	umpa	(,)(,)
14:46	benkay	does anyone know how to add metadata to clojure.test tests in such a way that midje will pick up on that metadata? i've tried wrapping the whole deftest in a (with-meta... :kw), wrapping just the (is...), wrapping the (test...), and i'm fully stumped.
14:49	gfredericks	benkay: you've tried (deftest ^{:kw ...} test-name ...)?
14:49	justin_smith	(deftest foo ...) (def (new meta) foo foo) maybe? deftest creates a var, that is effectively a defn with args of []
14:49	gfredericks	,(require '[clojure.test :as t])
14:49	clojurebot	nil
14:50	gfredericks	,(deftest ^:foo bar (is true))
14:50	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: deftest in this context, compiling:(NO_SOURCE_PATH:0:0)>
14:50	gfredericks	,(t/deftest ^:foo bar (t/is true))
14:50	clojurebot	#'sandbox/bar
14:50	gfredericks	,(meta #'bar)
14:50	clojurebot	{:ns #<Namespace sandbox>, :name bar, :file "NO_SOURCE_PATH", :column 0, :line 0, ...}
14:50	gfredericks	,(:foo (meta #'bar))
14:50	clojurebot	true
14:51	justin_smith	oh, well that's much easier isn't it
14:51	justin_smith	d'oh, I even do that already for test selectors
14:51	benkay	<gfredericks> benkay: you've tried (deftest ^{:kw ...} test-name ...)? // yup, i have a brick for a brain.
14:51	gfredericks	any well-behaved macro that def things will let you put metadata on the symbol
14:54	johnwalker	what are you using metadata here for?
14:55	benkay	johnwalker: epic refactor in proecss. using midje to only run one test at a time so i can focus on one problem at a time.
14:55	benkay	now, if only i knew enough to get clojure-test-mode to only evaluate the test with metadata...
14:55	benkay	test(s)
14:57	gfredericks	dangit people stop putting arglists on the same line as the function name
14:57	technomancy	gfredericks: no.
14:57	gfredericks	technomancy: it leads to accidental docstring misplacement
14:57	gfredericks	and it's not compatible with docstrings at all
14:57	johnwalker	it looks nice
14:57	technomancy	gfredericks: sure; I don't do that on nontrivial functions
14:58	gfredericks	technomancy: I'll allow it only if the whole function is one line
14:58	technomancy	that does bug me though... you have to choose between having documentation and looking tidy.
14:58	gfredericks	nothing untidy about args having their own line
14:58	gfredericks	my opinions are more valid than yours
14:58	technomancy	it counts against your LOC
14:58	gfredericks	why on earth would you LOC
14:59	johnwalker	why isn't it defn foo [] "docstring" like in elisp?
14:59	technomancy	gfredericks: for Internet Points
14:59	gfredericks	technomancy: we should count tokens
14:59	dbasch	johnwalker: I don’t know, but having defn foo “several lines of docstring
14:59	dbasch	“ and then [args] can be hard to read
14:59	technomancy	gfredericks: I need to keep Leiningen under an order of magnitude less than sbt
14:59	technomancy	it's important!
14:59	gfredericks	I guess this is more of a linter issue eh?
14:59	johnwalker	dbasch: agreed
15:00	technomancy	gfredericks: clojure-mode will highlight docstrings same as comments
15:00	technomancy	so if you see a string that highlights as a string, it's not a docstring
15:00	gfredericks	technomancy: that does not happen for me
15:00	dbasch	for that matter, (> x y) is always harder for me to read that (x > y)
15:01	gfredericks	do I need to go force-delete my clojure-mode from my ~/.emacs/elpa and cross my fingers and hope I won't get a Bad Gateway from the emacs repo again?
15:01	technomancy	gfredericks: huh... it used to
15:01	dbasch	*than
15:01	technomancy	dbasch: > is pronounced "descending"
15:01	gfredericks	and con> is of course "condescending"
15:01	technomancy	dbasch: the problem isn't prefix notation; the problem is calling it "greater than"
15:02	johnwalker	(inc gfredericks)
15:02	lazybot	⇒ 68
15:02	gfredericks	I bet 75% of my karma comes from jokes
15:02	technomancy	clojurebot: 正名?
15:02	clojurebot	If language is not correct, then what is said is not what is meant; if what is said is not what is meant, then what must be done remains undone; if this remains undone, morals and art will deteriorate; if justice goes astray, the people will stand about in helpless confusion. Hence there must be no arbitrariness in what is said. This matters above everything.
15:02	technomancy	dbasch: ^
15:02	gfredericks	technomancy: that first character means Internet Explorer I posit
15:03	Frozenlock	johnwalker: I think it's because you can have functions with variable number of arguments
15:03	gfredericks	Frozenlock: good point
15:03	dbasch	so, is clojure correct or will we end up standing about in helpless confusion?
15:04	technomancy	,(doc >)
15:04	clojurebot	"([x] [x y] [x y & more]); Returns non-nil if nums are in monotonically decreasing order, otherwise false."
15:04	johnwalker	ehh
15:04	technomancy	unfortunately clojure doesn't ship with an IPA pronunciation guide =)
15:04	johnwalker	what do variable number args have to do with it?
15:04	Frozenlock	johnwalker: https://www.refheap.com/86637 Where would you put the docsctring otherwise?
15:04	johnwalker	oh, i see what you mean now
15:04	johnwalker	yeah thats a really good point
15:04	benkay	gfredericks: it appears that midje-mode wants nothing to do with selecting only clojure.test tests.
15:05	technomancy	,(alter-var-root #'> assoc :pronunciation "dɪˈsɛndiŋɡ")
15:05	clojurebot	#<ClassCastException java.lang.ClassCastException: clojure.core$_GT_ cannot be cast to clojure.lang.Associative>
15:06	technomancy	,(alter-meta #'> assoc :pronunciation "dɪˈsɛndiŋɡ")
15:06	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: alter-meta in this context, compiling:(NO_SOURCE_PATH:0:0)>
15:06	technomancy	,(alter-meta! #'> assoc :pronunciation "dɪˈsɛndiŋɡ")
15:06	clojurebot	{:added "1.0", :ns #<Namespace clojure.core>, :name >, :file "clojure/core.clj", :inline-arities #{2}, ...}
15:06	technomancy	third time's the charm
15:06	gfredericks	clojurebot: the third time?
15:07	clojurebot	Pardon?
15:07	gfredericks	clojurebot: the third time is the charm
15:07	clojurebot	Ok.
15:07	gfredericks	clojurebot: the charm?
15:07	clojurebot	No entiendo
15:08	justin_smith	clojurebot: the third time?
15:08	clojurebot	the third time is the charm
15:09	umpa	justin_smith: why does this not work ? (reduce (assoc {} ks (filter (contains? (set ls) ks) ls)) {} ks)
15:10	justin_smith	reduce takes a function as the first argument
15:10	benkay	"tests?Note: Fixtures and test-ns-hook are mutually incompatible. If you
15:10	benkay	are using test-ns-hook, fixture functions will never be run."
15:10	benkay	so...
15:11	benkay	now i guess i feel like an idiot for using clojure.test.
15:14	benkay	well fml.
15:14	benkay	i guess i need to refactor my test suite to use a different test runner to refactor my app.
15:15	Frozenlock	I've only ever used clojure.test. Is it worth taking the time to look at other tools?
15:16	benkay	well i can't specify a single test to run so i guess the answer is a resounding "yes"
15:16	gfredericks	what's up with test-ns-hook? midje uses it?
15:16	benkay	gotta wonder what's more insane: refactoring everything to use midje or writing some macros to get the fixtures that i already have to run around the tests.
15:16	benkay	test-ns-hook is a hook for specifying which tests to run. it ignores fixtures.
15:17	gfredericks	yes but why are you using it?
15:17	justin_smith	I think the problem is trying to use clojure.test and midge at the same time - they are different and incompatible test runners
15:17	benkay	so that i can focus on one failing test at a time, gfredericks.
15:17	Raynes	benkay: lein test :only your.test.ns/your-test
15:17	benkay	justin_smith: midje is capable of running clojure tests.
15:18	justin_smith	benkay: but as we've learned, the test runners are incompatible
15:18	gfredericks	benkay: does the tactic Raynes just mentioned not work?
15:19	Raynes	I sure hope it does.
15:19	Raynes	I wrote it specifically to do this exact thing. :(
15:19	Raynes	(run individual tests with fixtures)
15:19	justin_smith	benkay: also, as I mentioned above, a deftest form creates a var that is effectively a function of 0 arguments, you could simply write a function that calls that wrapped in your fixtures
15:20	benkay	gfredericks: i'm working on it. now i have to endure the jvm startup for each run of...one test.
15:20	gfredericks	doesn't clojure 1.6 have a function for this too?
15:20	gfredericks	Raynes and I tag-teamed on it
15:20	Raynes	gfredericks: Yes.
15:20	Raynes	We wrote test-vars.
15:20	Raynes	Also to do this exact thing.
15:21	Raynes	We've got so many things to do this thing, man.
15:21	Raynes	:P
15:21	gfredericks	hopefully clojure-test-mode does that, but who knows
15:22	benkay	justin_smith: that's probably what i'll end up doing.
15:23	benkay	although quite frankly i'm now contemplating refactoring everything into midje facts.
15:25	Frozenlock	Please tell me someone made a function to english-capitalize-title strings...
15:25	gfredericks	to do what?
15:26	Frozenlock	Capitalize words in a title. You know, only 'some' words.
15:26	gfredericks	like ignoring minor words?
15:27	Frozenlock	Yes.
15:27	gfredericks	I cannot tell you whether someone made such a function
15:27	justin_smith	,(clojure.string/join " " (map clojure.string/capitalize (clojure.string/split "this is the title of my hit movie" #" ")))
15:27	clojurebot	"This Is The Title Of My Hit Movie"
15:27	gfredericks	,(def title "this is the title of my hit movie")
15:27	justin_smith	plus a blacklist of articles and such I guess
15:27	clojurebot	#'sandbox/title
15:28	gfredericks	,(def minor-words #{"is" "the" "of" "my"})
15:28	clojurebot	#'sandbox/minor-words
15:28	Frozenlock	http://english.stackexchange.com/a/34 :-(
15:28	Frozenlock	I should just write everything in uppercase.
15:28	Frozenlock	THIS WAY I CAN'T MAKE A MISTAKE.
15:29	kelseygi	it’d be a dang bazooka to hit a fly, but you could use https://github.com/eandrejko/clj-tokenizer
15:29	gfredericks	,(defn titlecap [s] (->> (clojure.string/split s #" ") (map #(cond-> % (not (minor-words %)) (clojure.string/capitalize))) (clojure.string/join " ")))
15:29	clojurebot	#'sandbox/titlecap
15:29	gfredericks	(titlecap title)
15:29	gfredericks	,(titlecap title)
15:29	clojurebot	"This is the Title of my Hit Movie"
15:29	justin_smith	IN A WORLD WHERE THE CLOJURE LIBRARIES LACK NATURAL LANGUAGE PROCESSING SUPPORT, FROZENLOCK MUST RESORT TO EXTREME MEASURES... TO WRITE HIS CODE
15:30	kelseygi	there’s also a section on stop words in this nlp-focused clojure tutorial http://writingcoding.blogspot.com/2008/06/clojure-series-table-of-contents.html
15:30	justin_smith	though you could likely fake it with gfredericks' approach (good luck detecting infinitives to get the "to"
15:30	justin_smith	rule right)
15:31	gfredericks	clojurebot: you \|could likely fake it\| with gfredericks' approach
15:31	clojurebot	Roger.
16:00	amalloy	gfredericks: is "my" a minor word for title casing? i would have guessed not
16:01	gfredericks	amalloy: I haven't been speaking english as long as you have
16:03	amalloy	really?
16:03	gfredericks	I'm the youngest person in this channel excepting Raynes
16:03	Raynes	Are you certain?
16:03	Raynes	arrdem is pretty young.
16:03	gfredericks	I'm young so of course I'm certain
16:03	Bronsa	me too
16:03	Bronsa	I'm between arrdem and Raynes
16:03	amalloy	how do you even know how long i've been speaking english though? i could be a well-spoken russian import
16:04	arrdem	Raynes: older than you, whippersnapper
16:04	amalloy	Bronsa: we don't need to hear about your fantasies
16:04	gfredericks	I'm young so I make all sorts of assumptions about other people
16:04	arrdem	(inc amalloy)
16:04	lazybot	⇒ 129
16:04	technomancy	OVARFLOW
16:04	Jarda	,amalloy
16:04	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: amalloy in this context, compiling:(NO_SOURCE_PATH:0:0)>
16:04	Jarda	:(
16:05	Raynes	It's not really executing that code :P
16:05	gfredericks	in complete overhonest frankness I am yes probably rather older than all of the college kids hanging around
16:05	Raynes	It's a well crafted illusion.
16:05	technomancy	I'll grant well-crafted when you can juxt it =)
16:05	justin_smith	$karma Jarda
16:05	lazybot	Jarda has karma 0.
16:06	justin_smith	((juxt inc dec) technomancy) ; standard response to a laughably bad joke
16:06	Bronsa	amalloy: I don't even know how to reply to that
16:06	arrdem	(let [arrdem 21.3 Raynes 20 Bronsa (fresh)] (run (x) (< Bronsa arrdem) (> Bronsa Raynes)))
16:07	arrdem	,(/ 1.0 365)
16:07	clojurebot	eval service is offline
16:07	arrdem	clojurebot is a steaming pile of fail
16:08	Bronsa	the last thing it evaluates before breaking is "amalloy". sure this proves he's all evil.
16:08	arrdem	Bronsa: WE DO NOT SPEAK THE NAME
16:09	Raynes	On June 15, 2014, Alan Malloy made Bronsa speechless.
16:09	justin_smith	arrdem: cool trick is (/ 365.0) does the same thing
16:09	arrdem	justin_smith: yeah, I have mixed feelings about the implicit 1
16:09	justin_smith	multiple arity is so fun
16:10	justin_smith	arrdem: I mean what else would you want implicitly if / had one arg? taking an inverse is pretty basic
16:11	arrdem	justin_smith: I'd accept a type or arity error
16:11	arrdem	,(doc /)
16:11	clojurebot	eval service is offline
16:11	arrdem	##(doc /)
16:11	lazybot	⇒ "([x] [x y] [x y & more]); If no denominators are supplied, returns 1/numerator, else returns numerator divided by all of the denominators."
16:11	arrdem	at least it _is_ documented clearly.
16:12	justin_smith	arrdem: it's an old behavior, common among other lisps
16:12	arrdem	justin_smith: yay tradition
16:13	dbasch	amalloy: pronouns are typically capitalized in titles
16:15	justin_smith	hlship is not much older than me, but the oldest clojurian whose age I know (we actually met in person, or I wouldn't even know)
16:18	dbasch	justin_smith: I’m guessing Rich Hickey must be around 50
16:19	justin_smith	dbasch: well, I was talking about on this channel
16:21	dbasch	(> 30 (apply max (filter #(knows-irl dbasch %) clojure-irc)))
16:39	technomancy	anyone have last-minute issues to raise before lein 2.4.1 is released?
16:42	gtrak	technomancy: Tom Cruise was in an action movie at 51 years of age.
16:43	technomancy	is this a bug report or a feature request
16:44	Raynes	It's a glitch in the Matrix is what it is.
16:46	arrdem	gtrak: and Arnold is still doing them. next!
16:46	gtrak	we should probably clear some caches somewhere to prevent further damage
16:47	gtrak	arrdem: he was _literally_ jumping around in robot suits killing things
16:47	arrdem	gtrak: that's Tony Stark's job
16:47	dbasch	technomancy: ship it!
16:48	technomancy	dbasch: hokay
16:48	gtrak	Robert Downey Jr is only 49
16:48	arrdem	technomancy: send it
16:49	arrdem	clojure core's whitespace is a work of modern art...
16:52	technomancy	arrdem: http://nonadventures.com/2013/07/20/a-whiter-shade-of-fail/
16:52	technomancy	arrdem: aka M-x whitespace-mode
16:52	arrdem	technomancy: hehe
16:54	arrdem	technomancy: http://nonadventures.com/2013/08/17/wooly-bullied/ this is an awesome webcomic. thank you.
16:54	technomancy	arrdem: it is a gift
16:55	justin_smith	(inc palindromidary)
16:55	lazybot	⇒ 1
16:55	justin_smith	(inc palindromedary)
16:55	lazybot	⇒ 1
16:55	justin_smith	(it was spelled wrong the first time)
16:56	metellus	(dec palindromidary)
16:56	lazybot	⇒ 0
16:58	technomancy	http://nonadventures.com/2014/05/17/gravys-anatomy/
16:59	umpa	,(defn nice [m v] (assoc m v (filter #(contains? (set %) v) '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)))))
16:59	clojurebot	#'sandbox/nice
16:59	umpa	,(reduce (nice {} [8 2 6 0]) [8 2 6 0])
16:59	clojurebot	0
16:59	umpa	justin_smith: why does this happen ? Its something with the map
17:00	justin_smith	does (nice ...) return a function?
17:00	justin_smith	you are still not using reduce properly
17:01	umpa	(nice {} [8 2 6 0])
17:01	umpa	,(nice {} [8 2 6 0])
17:01	clojurebot	{[8 2 6 0] ()}
17:01	justin_smith	so you don't get an error, because {... ...} is callable as a function
17:02	justin_smith	it just probably won't return anything you find interesting
17:03	umpa	what kind of function does this return then ? (fn [m v] (assoc m v (filter #(contains? (set %) v) ls)))
17:03	justin_smith	that is a function
17:03	justin_smith	and it returns an updated map
17:04	justin_smith	that's why it is a valid first arg to reduce
17:05	justin_smith	,(reduce (fn [m [k v]] (assoc m k v)) {} [[:a 0] [:b 1]]) ; not the best way to do this task, but it should help clarify what the first arg to reduce does
17:05	clojurebot	{:b 1, :a 0}
17:06	dbasch	umpa: may I ask what you’re trying to do?
17:07	justin_smith	he wants a map from elements to collections that contain those elements
17:08	umpa	^
17:09	justin_smith	umpa: you took two inputs to reduce, and threw them into one defn body for some reason
17:10	justin_smith	if you took that literal list off the end of "nice", nice could be the first arg to reduce
17:10	justin_smith	and then the list would be the last arg, the one you are reducing over
17:11	umpa	dbasch: coll: ((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)) elements: [8 2 6 0] output form {0 ((0 1 2) (0 3 6) (0 4 8)), 2 ( ...
17:11	arrdem	so I'm looking at this: https://github.com/clojure/clojure/blob/master/src/clj/clojure/core.clj#L5351
17:12	arrdem	this being with-loading-context: https://github.com/clojure/clojure/blob/master/src/clj/clojure/core.clj#L5295
17:12	arrdem	is there a reason for the anonymous function call? seems to me like it's just there for no reason when a do would suffice.
17:15	justin_smith	,(do (def coll '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6))) (defn nice [m v] (assoc m v (filter #(contains? (set %) v) coll))) (reduce nice {} [8 2 6 0])) umpa
17:15	clojurebot	{0 ((0 1 2) (0 3 6) (0 4 8)), 6 ((6 7 8) (0 3 6) (2 4 6)), 2 ((0 1 2) (2 5 8) (2 4 6)), 8 ((6 7 8) (2 5 8) (0 4 8))}
17:15	justin_smith	but really those things should all be in one function, and bound in a let
17:21	umpa	,(def coll '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)))
17:21	clojurebot	#'sandbox/coll
17:22	umpa	(contains? (set coll) 0)
17:22	umpa	,(contains? (set coll) 0)
17:22	clojurebot	false
17:22	justin_smith	,(contains? (set coll) '(0 1 2))
17:22	clojurebot	true
17:23	umpa	,(contains? (set coll) '(0))
17:23	clojurebot	false
17:23	dbasch	umpa: (set coll) is a set of triples, it’s not going to contain 0
17:23	dbasch	or not even triples, a set of sequences
17:24	dbasch	you have nested sequences in there, are you trying to see if 0 is in any of your nested sequences?
17:25	technomancy	TimMc: hey
17:25	technomancy	TimMc: I think we have to revert the pom.properties change; it adds 300ms to lein startup time
17:26	umpa	dbasch: ,(filter #(contains? (set %) [0]) coll)
17:26	umpa	,(filter #(contains? (set %) [0]) coll)
17:26	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: coll in this context, compiling:(NO_SOURCE_PATH:0:0)>
17:26	justin_smith	umpa: filter takes each element of the coll one by one
17:27	justin_smith	umpa: that's why that code works
17:27	dbasch	umpa: where are you trying to find 0? what would be your desired output?
17:27	justin_smith	except you want 0, not [0]
17:27	justin_smith	dbasch: see my paste above (do (def coll ...))
17:27	justin_smith	that does what he wants, he is just figuring out how it works
17:29	umpa	^ debugging justin_smith's code
17:29	justin_smith	debugging? it has no bugs!
17:30	justin_smith	though it could be omptimized better
17:31	umpa	justin_smith: debugging understanding
17:32	umpa	dbasch: ((0 1 2)(0 4 7) ... etc
17:32	justin_smith	(def coll '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)))
17:32	justin_smith	,(def coll '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)))
17:32	clojurebot	#'sandbox/coll
17:32	justin_smith	,(reduce (fn [m el] (reduce (fn [m e] (update-in m [e] (fnil conj []) el)) m el)) {} coll)
17:32	clojurebot	{8 [(6 7 8) (2 5 8) (0 4 8)], 7 [(6 7 8) (1 4 7)], 6 [(6 7 8) (0 3 6) (2 4 6)], 5 [(3 4 5) (2 5 8)], 4 [(3 4 5) (1 4 7) (0 4 8) (2 4 6)], ...}
17:32	justin_smith	better version
17:33	justin_smith	though maybe not easier to read
17:33	justin_smith	oh wait, that gets all keys, not just the ones specified
17:35	justin_smith	,(def wanted #{0 2 4 6 8})
17:35	clojurebot	#'sandbox/wanted
17:35	justin_smith	,(reduce (fn [m el] (reduce (fn [m e] (if (contains? wanted e) (update-in m [e] (fnil conj []) el) m)) m el)) {} coll)
17:35	clojurebot	{8 [(6 7 8) (2 5 8) (0 4 8)], 6 [(6 7 8) (0 3 6) (2 4 6)], 4 [(3 4 5) (1 4 7) (0 4 8) (2 4 6)], 2 [(0 1 2) (2 5 8) (2 4 6)], 0 [(0 1 2) (0 3 6) (0 4 8)]}
17:38	umpa	justin_smith: nice
17:40	catern	so many numbers aieee
17:57	blaenk	I'm reading about clojure's STM and I'm kind of confused with the concept of deref prompting a transaction retry and the notion of a history of values somehow mitigating that
17:58	blaenk	at first my understanding was that at the beginning of the transaction, a snapshot of all (?) refs' states was taken, the so called in-transaction values
17:58	blaenk	but now the book I'm reading is saying that deref can cause a retry, because "if a new value is committed by another transaction since the beginning of the current transaction, the value of the ref as of the start of the transaction cannot be provided"
17:59	blaenk	I'm not sure I understand why it's saying that that value cannot be provided, if my understanding was that the value of that ref was 'snapshotted' at the beginning of the transaction
18:00	blaenk	what I'm confused about is mainly why it's saying that the value of the ref as of the start of the transaction cannot be provided
18:01	llasram	blaenk: If I read the implementation correctly, a ref will maintain a history of values, but not necessarily a complete history
18:01	justin_smith	blaenk: cannot be provided while also being correct
18:01	llasram	blaenk: So one condition which can cause a restart is a sufficiently-old ref value not being available
18:01	justin_smith	blaenk: that's how the whole thing can be non-locking, by retrying if something else finalizes in the meantime
18:02	justin_smith	llasram: sufficiently-old ref value?
18:02	blaenk	so for example, with alter, if the value it altered isn't the same at the beginning of the transaction as it is now at point of committing, then it retries, I understand that part, because then the work would have been done on an outdated value
18:03	blaenk	so this is basically saying that the deref is another such check for the transaction being correct?
18:03	llasram	blaenk, justin_smith: https://github.com/clojure/clojure/blob/master/src/jvm/clojure/lang/LockingTransaction.java#L397-L422
18:03	llasram	blaenk: Note that if you're just learning Clojure, you can probably ignore refs
18:03	llasram	The STM was a big exciting feature when Clojure was getting going, but I've never seen production code legitimately use refs
18:03	blaenk	thanks, but I'd prefer to learn them. I'm not too new to stm, I've done some of it in haskell
18:03	blaenk	ah okay
18:04	blaenk	it's mainly that I was following along up to this point
18:04	llasram	(Not that they don't work as advertised -- just I haven't seen any code which actually needed them)
18:04	justin_smith	llasram: oh, shows how little I know about STM, thanks
18:05	blaenk	my main confusion is the following: it makes it sound like, if the value being derefernced has changed mid-transaction, then clearly that won't be correct, because it'll be computing off of an outdated value, no? but then, how would maintaining a history of this value mitigate this? if, again, it would be being based off of an outdated value?
18:05	blaenk	*changed mid-transaction by another transaction, then clearly that won't be correct, hence the retry
18:05	justin_smith	blaenk: I wonder if this may be related to functions like commute
18:05	llasram	Exactly
18:06	justin_smith	with commute, even if your order is off, your calculation can be correct, since the f is commutative
18:06	blaenk	right
18:06	llasram	Function which don't update or only commute don't need the most recent value, just consistent values from all refs read
18:07	blaenk	and that's the case where maintaining a history would help that situation
18:07	llasram	Exactly (is my understanding)
18:07	blaenk	that's what I was thinking, but why would a history need to be maintained, and not just the snapshot of the ref as it was at the beginning of the transaction, like it does with other things?
18:08	blaenk	I mean, why isn't a history of size 1 enough
18:08	blaenk	because it's talking about tuning the history size etc
18:08	llasram	Oh, because ref access is dynamic. When building the code for a transaction the compiler doesn't know ahead of time which refs are accessed
18:09	blaenk	ohhh it doesn't, ok
18:09	blaenk	somehow I got the impression that it somehow knew which ones were accessed, and took a snapshot of them at the beginning of the transaction
18:09	blaenk	"All reads of Refs will see a consistent snapshot of the 'Ref world' as of the starting point of the transaction (its 'read point'). The transaction will see any changes it has made. This is called the in-transaction-value."
18:09	llasram	The Clojure compiler is about as dumb as you can get, but intentionally so as a design commitment to radical simplicity
18:09	blaenk	http://clojure.org/refs
18:10	blaenk	I understand
18:10	blaenk	I'm just not sure what that quote means then
18:10	llasram	Yep. It's an illllusion. It uses tricks like maintaining history and restarting transactions to allow produce the effect described in that quote
18:10	blaenk	ohhh okay
18:12	blaenk	since it just now introduced the concept of history of values, with deref in a transaction, I take it that modification operations aren't affected by this somehow?
18:12	blaenk	in some other way then, they just know that the ref's state has changed, and retry the transaction?
18:12	blaenk	or do they also run into the problem of not having enough history
18:13	llasram	blaenk: That I don't know off the top of my head. If you're really interested, I'd suggest digging into the implementation
18:13	blaenk	because the way I had previously (mistakenly) understood it was that they would take a snapshot of the state at the start of the transaction, and if that differed from its state at the point of committing, then it would retry
18:13	blaenk	alright, thanks
18:13	llasram	The Compiler classes are unnecessarily hairy, but most of the core stuff like the STM is pretty approachable Java
18:13	llasram	One day we will be self-hosting...
18:13	llasram	(inc Bronsa)
18:13	lazybot	⇒ 23
18:14	blaenk	that'd be cool
18:14	llasram	Yah. And Bronsa is working hard on the Clojure-in-Clojure implementation which will lead us to the promised land
18:14	llasram	No pressure
18:15	blaenk	haha
18:16	_JokerDoom	(inc Bronsa) -> 23?
18:17	llasram	karma mechanism
18:17	llasram	$karma lazybot
18:17	lazybot	lazybot has karma 26.
18:17	llasram	(inc lazybot)
18:17	lazybot	⇒ 27
18:17	llasram	$karma lazybot
18:17	lazybot	lazybot has karma 27.
18:17	_JokerDoom	fun
18:17	_JokerDoom	$karma _JokerDoom
18:17	lazybot	_JokerDoom has karma 0.
18:17	_JokerDoom	lol
18:18	justin_smith	as is apropriate, the "inc" for karma has nothing to do with clojure's inc
18:18	justin_smith	(well, maybe half a thing to do with it)
18:18	llasram	I think of it as part of a macro-based DSL which just happens to use s-expressions
18:18	justin_smith	right
18:20	blaenk	llasram: would you happen to know which part of the code you linked has to do with alter?
18:21	llasram	blaenk: I'm afraid I do not off the top of my head. You can probably find it about as quickly as I can recall it :-)
18:21	blaenk	no worries, thanks
18:21	blaenk	looks like it delegates to doSet
18:28	martinklepsch	I want to "merge" two maps like so: {:a 1 :b 2} {:a 15 :b 16} ;; => {:a {:file 1 :database 15} :b {:file 2 :database 16}}
18:31	Raynes	&(merge-with #(zipmap [:file :database] %&) {:a 1 :b 2} {:a 15 :b 16})
18:31	lazybot	⇒ {:a {:database 15, :file 1}, :b {:database 16, :file 2}}
18:32	Raynes	martinklepsch: ^
18:33	blaenk	Raynes: what's that &() syntax?
18:33	blaenk	is it specific to lazybot?
18:33	Raynes	That's not syntax.
18:33	blaenk	alright, what's that not-syntax
18:33	Raynes	The & is the evaluation prefix for lazybot
18:33	blaenk	ah thanks, figured
18:33	Raynes	It wouldn't know what to evaluate otherwise.
18:33	blaenk	got it
18:33	catern	why use lazybot over clojurebot?
18:33	Raynes	Because I wrote one of them and not the other.
18:34	catern	good reason
18:34	Raynes	Either one of them generally works fine.
18:34	blaenk	he wrote clojurebot, is ashamed
18:34	catern	haha
18:34	Raynes	lol
18:34	Raynes	lazybot can ##(println "also evaluate stuff in the middle of messages, which clojurebot cannot do")
18:34	lazybot	⇒ also evaluate stuff in the middle of messages, which clojurebot cannot do nil
18:35	Raynes	So if you want to do that, that's an excellent reason to use lazybot.
18:36	blaenk	nice
18:37	noonian	testing #10 lazybot
18:37	noonian	testing ##10 lazybot
18:37	Raynes	noonian: Only matches ()
18:37	Raynes	You'd be incredibly likely to set it off accidentally otherwise.
18:37	Raynes	Because ##channels are often prefixed with two hashes to indicate that the person who created them don't actually own the name of the channel.
18:38	noonian	hmm, i had no idea
18:38	justin_smith	also, there's ##:meta-hashtags
18:38	justin_smith	which are hashtags about hashtags
18:39	Raynes	##hashtag
18:41	noonian	#poundsign
18:42	justin_smith	I like ##:octothorpe
18:43	mdeboard	This is kind of a terrible question, but I'd like to know if there's a more concise way of expressing this logic here https://github.com/mattdeboard/ticket-to-ride/blob/master/src/ttr/board.clj#L329 `@cards` derefs into a vector of keywords like `[:red :red :blue :yellow :white :blue]`
18:44	mdeboard	(I'm asking about line 329-355)
18:44	Raynes	https://github.com/mattdeboard/ticket-to-ride/blob/master/src/ttr/board.clj#L329-L355
18:45	mdeboard	thanks
18:45	Raynes	mdeboard: shift click a second line number to highlight a section
18:45	mdeboard	gotcha, thanks
18:46	mdeboard	I really don't like the `(apply count (rest %))` bit
18:46	mdeboard	,(apply count (rest [:red [:red :red :red]]))
18:46	clojurebot	3
18:46	mdeboard	idk if that's idiomatic-enough
18:47	justin_smith	(filterv (fn [[k v]] (and (>= (count v) cost) (#{(name k) :gray} color))) (group-by identity @cards))
18:47	justin_smith	maybe soemthing like that?
18:47	mdeboard	I find it p hard to remember what is happening there, it's pretty dense syntax. "Get the last element of this vector, then unpack that and count the elements in it"
18:47	justin_smith	it's more concise, and unless I made a thinko has the same semantics
18:47	mdeboard	justin_smith: well fortunately I have tests for this function, so I'll give it a whirl
18:47	Raynes	&(count (last [:red [:red :red :red]]))
18:47	lazybot	⇒ 3
18:48	mdeboard	,(rest [:a [:a :a :a]])
18:48	clojurebot	([:a :a :a])
18:48	mdeboard	aha
18:48	mdeboard	Raynes: nice, that's good too
18:48	justin_smith	oh, not fn [k v], fn [k & v]
18:48	justin_smith	oh, never mind, k v was right
18:48	justin_smith	bleh
18:49	martinklepsch	Raynes, thanks that's working great and is nice and simple :)
18:49	justin_smith	I think bindings like k (for key) and v (for value) are more clear than first rest when mapping across a map
18:49	mdeboard	Raynes: thanks, that's working great and is nice and simple :(
18:49	mdeboard	:)*
18:49	Raynes	lol
18:50	mdeboard	justin_smith: Oh right, I didn't notice that destructuring in yours
18:50	mdeboard	I like that muhc better too
18:50	martinklepsch	Raynes, what does the %& do?
18:51	justin_smith	martinklepsch: it is like [arg & rest] in a normal fn
18:51	Raynes	martinklepsch: (fn [& args] (zipmap [:file :database] args))
18:51	martinklepsch	justin_smith, ah ok. thought so thanks!
18:51	martinklepsch	have a good day/night! cheers!
18:52	mdeboard	fedora*
18:52	Raynes	He only likes me for my merge-withs
18:53	mdeboard	btw if anyone here is going to Austin Clojure meetup tomorrow, I'm gonna head up there
18:53	mdeboard	(I'm on a trip here atm)
18:55	mdeboard	justin_smith: To use `fn [k v]` I'd have to call `apply` first
18:55	mdeboard	oh maybe it's `fn [[k v]]` ?
18:55	Raynes	mdeboard: I totally would have done (map (partial hash-map :name) ["Atlanta" "Boston" ...])
18:55	justin_smith	mdeboard: why apply? I did have it as fn [[k v]] right?
18:55	Raynes	Because it would have made me feel lazier.
18:55	mdeboard	justin_smith: Oh, if you did I misread, sorry
18:56	xeqi	devn: losingkeys = bendyworks person?
18:56	Raynes	Also holy God did you associate a ref with every one of these
18:56	Raynes	mdeboard: plz no
18:56	mdeboard	every one of what
18:56	Raynes	edges
18:56	Raynes	Every edge has a ref
18:56	mdeboard	ya
18:56	mdeboard	don't hurt my feelings
18:58	Raynes	mdeboard: In these kinds of situations, usually you want one ref/atom holding all state rather than one for every single piece of state.
18:58	Raynes	It's generally better to have as little state as possible.
18:58	Raynes	Also, you don't seem to be using the features of refs that warrant their usage.
18:58	Raynes	atoms would suffice here as far as I can tell
18:59	mdeboard	Yeah, probably, not updating several in a single transaction
18:59	mdeboard	not now, anyway
18:59	Raynes	Unless you need to do coordinated updates of more than one ref inside of transactions, not very useful.
18:59	Raynes	And if you only have one atom instead of a billion refs, it'll never be necessary! :p
19:00	mdeboard	but that's a little legacy of first iteration, I had state being managed all over the place. Once I corralled all the state manipulation into one place I found I didn't need refs after all... but i've got other stuff that I need to work on before I optimize
19:01	mdeboard	e.g. https://github.com/mattdeboard/ticket-to-ride/blob/master/src/ttr/cards.clj#L94-L125
19:02	mdeboard	so once I do make time for replacing with atoms, it'll be really easy since all the state manipulation is handled in very few places
19:03	Raynes	How about not using atoms or refs! :D
19:03	Raynes	Just pass state from function to function.
19:03	Raynes	Purely functional ticket to ride
19:03	Raynes	You can write your own ticket, mdeboard.
19:04	mdeboard	because I'm just trying to get a full-featured simulation of the game running before I start making optimizations, otherwise I'll just tinker forever
19:04	mdeboard	and the thing that actually interests me about this is using neo4j, but I need a working game before I can start doing meaningful stuff there
19:05	mdeboard	na'mean
19:06	Raynes	I refuse to accept your excuses.
19:06	Raynes	off*
19:06	noonian	+1 million on the purely functional implementation
19:07	kelseygi	irc commands: not mutate-able!
19:13	mdeboard	Raynes: pull requests gladly accepted btw
19:13	Raynes	I've never played ticket to ride nor do I have the foggiest clue how it is played.
19:14	mdeboard	Oh, well, you're missing out.
19:14	Raynes	I'm certain
19:18	justin_smith	mdeboard: what Raynes describes with passing state is a code stability / design optimization
19:18	justin_smith	in other words, its something that pays off most if you do it from the beginning
19:18	justin_smith	and the longer you wait to switch to the technique, the more you pay for it
19:19	justin_smith	(I have switched a mature codebase to arg passing - I was glad I did it, but it would have been much easier if I had done things that way in the first place)
19:19	mdeboard	I se
19:19	mdeboard	e
19:20	mdeboard	I've never actually done anything with games or anything, I just have always understood them to require a lot of state
19:20	mdeboard	I'm definitely forcing some OOP concepts into this design
19:20	mdeboard	but trying to keep it in check wihtout sacrificing gtd
19:20	justin_smith	mdeboard: passing an arg can provably do anything state can (plus things state can't do easily)
19:21	mdeboard	Can you give an example? I have trouble modeling this in my head
19:21	mdeboard	especially in a "loop" like a game
19:21	justin_smith	well, loop creates a binding, and recur passes a binding
19:22	justin_smith	so when you recur, you pass in the updated "state of the world", which is actually an immutible object
19:22	mdeboard	orite
19:22	mdeboard	I see, I see
19:22	justin_smith	all "modifications of state", are new versions of the state of the world (usually efficient, because we are using persistent data structures)
19:22	noonian	right, instead of modifying the contents of the deck you could write a function that takes a deck and returns a different deck
19:22	mdeboard	Yeah that was enough for a light bulb to come on
19:23	justin_smith	right
19:23	mdeboard	right
19:23	mdeboard	right
19:23	mdeboard	Anyway, I understand. I may revisit that once I figure out the thing I'm trying to solve now
19:23	justin_smith	advantages including the fact that you can save and restore and replay games trivially
19:24	justin_smith	since it is right there in one immutible thing already, why not fork it if you want to play things out multiple directions
19:24	justin_smith	playing the game in reverse, etc. etc.
19:25	justin_smith	things that are very hard in the mutable version become trivial in this one
19:26	mdeboard	justin_smith: Interesting
19:26	mdeboard	very interesting
19:26	mdeboard	That gets at an annoying issue I have had to deal with in tests :P
19:27	mdeboard	justin_smith: line 16, 28, 40
19:27	mdeboard	resetting state
19:27	mdeboard	It's one of those deals wher eyou know it's dumb
19:27	mdeboard	like as you're typing it
19:30	mdeboard	I think I've lessened the impact on refactoring to something less stateful by making sure state's not being mutated all over the place
19:30	justin_smith	also, add a constrained tree search on moves with simulated opponent moves, and you get an ai for playing the game
19:30	mdeboard	Yeah AI is the second thing I'm working on
19:31	justin_smith	yeah, another bonus of the "state" just being a function argument is that suddenly testing is much easier
19:31	justin_smith	right, and for the ai to branch, you either have to play crazy games with state (very brittle), or just have a proper state argument with no global binding
19:43	umpa	this filter (filter #(contains? (set %) [0 1) ((1 2 3)(4 5 6))) why you no work ?
19:44	llasram	umpa: Well, syntax errors at least
19:44	Raynes	Well, that's not valid Clojure for starters.
19:45	technomancy	(1 2 3) is not a function you can call
19:45	umpa	,(filter #(contains? (set %) [0 1]) ((1 2 3)(4 5 6)))
19:45	clojurebot	#<ClassCastException java.lang.ClassCastException: java.lang.Long cannot be cast to clojure.lang.IFn>
19:45	justin_smith	also the nesting is off
19:46	umpa	,(filter #(contains? (set %) [0 1]) '((1 2 3)(4 5 6)))
19:46	clojurebot	()
19:46	justin_smith	and you probably want to test intersection rather than membership
19:46	llasram	umpa: What do you expect to happen here?
19:46	justin_smith	I think he wants to filter the lists that contain a 0 or 1
19:46	llasram	Ok...
19:47	umpa	^
19:47	llasram	umpa: `contains?`'s second argument is a single value
19:47	llasram	So ##(contains? #{[0 1]} [0 1])
19:47	lazybot	⇒ true
19:47	llasram	But not ##(contains? #{0 1} [0 1])
19:47	lazybot	⇒ false
19:49	justin_smith	,(do (require '[clojure.set :as s]) (filter #(s/intersection #{0 1} (set %)) '((1 2 3) (4 5 6) (0 1 2))))
19:49	clojurebot	((1 2 3) (4 5 6) (0 1 2))
19:49	llasram	,(filter (partial some #{0 1}) '((1 2 3) (4 5 6)))
19:49	justin_smith	err
19:49	clojurebot	((1 2 3))
19:50	justin_smith	that's a good way to do it
19:51	umpa	justin_smith: why did the v in filter work ##((fn [ks ls](reduce (fn [m v] (assoc m v (filter #(contains? (set %) v) ls) )) {} ks))[8 2 6 0] '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)))
19:51	lazybot	⇒ {0 ((0 1 2) (0 3 6) (0 4 8)), 6 ((6 7 8) (0 3 6) (2 4 6)), 2 ((0 1 2) (2 5 8) (2 4 6)), 8 ((6 7 8) (2 5 8) (0 4 8))}
19:51	justin_smith	oh yeah, empty sets are truthy
19:52	justin_smith	because that was reducing across individual elements
19:52	umpa	(doc truthy)
19:52	clojurebot	No entiendo
19:53	justin_smith	umpa: http://blog.jayfields.com/2011/02/clojure-truthy-and-falsey.html
19:54	mdeboard	ugh now all I see is how OO I made this
19:54	mdeboard	THANKS
19:54	mdeboard	I'm disgusted with myself
19:54	mdeboard	It's like the Garden of Eden
19:54	justin_smith	umpa: because of other lisps, I sometimes lapse and expect empty things to act as false. It's a bad habit
19:54	mdeboard	I realize I'm naked, and I'm ashamed
19:54	justin_smith	mdeboard: hah
19:57	benkay	does anyone else have problems with running 2 repls in 2 projects at the same time?
19:57	benkay	using cider
19:57	mdeboard	benkay: I don't do that, as a rule. I had trouble with nrepl back in the day
19:58	mdeboard	running 2 repls
19:58	mdeboard	i don't remember what it entailed
19:58	benkay	buuuh
19:58	justin_smith	benkay: in nrepl I start each repl in a terminal, and do C-u m-x nrepl, so that each one gets its own repl
19:58	justin_smith	this is one of those things where jacking in just makes it more complex
19:58	benkay	ah yes, i suppose it's worth mentioning i'm working from emacs.
19:58	benkay	sigh
19:59	justin_smith	benkay: this is from emacs, just using nrepl.el
19:59	benkay	so the answer to "can i use code navigation on two projects at once?" is "no"?
19:59	justin_smith	cider.el should be similar
19:59	benkay	justin_smith: neat. that's a protip for sure.
19:59	mdeboard	justin_smith: like I get the loop example, and "recur until some condition is met", but I guess I don't understand how minimal state can get. I can't envision how to do this without each player having some "bag of state"
20:00	benkay	does nrepl (and hopefully cider, but i know you don't know) then pin the relevant connection to the relevant projects?
20:00	justin_smith	also, if you haven't tried it, try "C-u M-x cider-jack-in" to start your second repl
20:00	mdeboard	Or C-c M-j
20:00	justin_smith	mdeboard: can't player specific state be a key under the state object?
20:00	mdeboard	(by default, anyway)
20:00	umpa	##(def x nil)
20:00	lazybot	java.lang.SecurityException: You tripped the alarm! def is bad!
20:01	umpa	,(def x nil)
20:01	clojurebot	#'sandbox/x
20:01	umpa	(false? x)
20:01	umpa	,(false? x)
20:01	clojurebot	false
20:01	mdeboard	justin_smith: It can, I just am having a tough time conceptually :P
20:01	llasram	umpa: only `false` is `false?`. `nil` is false-y
20:01	llasram	,(boolean nil)
20:01	clojurebot	false
20:01	benkay	justin_smith: yeah, cider gets confused when i jack in to two separate projects.
20:02	llasram	,(false? (boolean nil))
20:02	clojurebot	true
20:02	justin_smith	well that sucks! nrepl handles it OK if I start the projects via lein repl and connect to that, at least
20:02	llasram	benkay: cider 0.5.0 is fine with multiple projects
20:02	benkay	grrrrrrrrr clojure startup time has ruined my weekend
20:02	llasram	at least in my experience
20:03	llasram	benkay: how so?
20:03	umpa	,(def x false)
20:03	clojurebot	#'sandbox/x
20:03	benkay	hm llasram how do i check a package version in emacs? list packages?
20:03	llasram	benkay: ya
20:03	umpa	,(false? x)
20:03	clojurebot	true
20:04	benkay	llasram: i've been doing a bunch of refactoring and having to restart the jvm proc and clojure with it is murdering my productivity.
20:04	benkay	implementing a reloaded workflow on something that was never meant to have one...
20:04	umpa	justin_smith: how does truthy apply to the filter ?
20:05	mdeboard	I guess https://gist.github.com/mattdeboard/c8b163b88152f2d3fb1f ? idk captain
20:05	justin_smith	umpa: because I wrongly expected an empty intersection of sets to act as a false value
20:05	justin_smith	wrapping it in seq fixes it though (that generates nil for an empty set)
20:05	justin_smith	,(do (require '[clojure.set :as s]) (filter #(seq (s/intersection #{0 1} (set %))) '((1 2 3) (4 5 6) (0 1 2))))
20:05	clojurebot	((1 2 3) (0 1 2))
20:05	llasram	benkay: What are you refactoring that's requiring restarts?
20:06	benkay	silly little webapp.
20:06	justin_smith	he's moving things between clojure.test and midge, iirc
20:06	benkay	no, i straight up gave up on midje.
20:06	llasram	Hmm. Still shouldn't be necessary to restart
20:06	llasram	You can just kill namespaces and reload them
20:06	justin_smith	benkay: good plan, imho
20:06	llasram	agreed
20:07	mdeboard	why so
20:07	mdeboard	too fat?
20:07	benkay	llasram: definitely in the workflow, and yet i keep getting SLIMEd.
20:08	llasram	mdeboard: Not exactly... It's got some nice ideas, but (a) I've never found it to play entirely nicely with a "normal" Clojure workflow, and (b) it is essentially it's own non-Clojure mini-language implemented by interpreting the s-expressions its macros get passed
20:08	llasram	benkay: Not quite following
20:08	benkay	https://github.com/clojure-emacs/cider-nrepl // build: failing
20:08	benkay	yay tooling
20:09	justin_smith	mdeboard: whereas clojure.test is built on functions and defs and metadata in a very idiomatic way, which means it is easy to use standard clojure composition / abstraction techniques to make it do what you want
20:09	mdeboard	I see
20:10	benkay	by the way, i also completely failed at running individual tests with clojure.test
20:10	benkay	(run-test 'my-test-ns/test-i-want-run-dammit) just returns nil
20:10	justin_smith	does it print anything?
20:11	benkay	lol nil
20:11	Raynes	benkay: What version of clojure are you using?
20:12	benkay	Raynes: 1.5.1
20:12	Raynes	benkay: Any reason you can't use 1.6.0?
20:13	benkay	aside from being *-deep in a refactor that's consuming my life already?
20:13	Raynes	benkay: https://clojure.github.io/clojure/clojure.test-api.html#clojure.test/test-vars
20:13	Raynes	test-var existed in clojure 1.5 though, so if you really only need to run one specific test, it may work fine for you.
20:14	benkay	ah yes, i did actually mean test-var, apologies.
20:14	benkay	(test-var 'my-test-ns/test-...)
20:15	benkay	holy heck does test-var not use fixtures?
20:15	mdeboard	I reckon the simplest refactor here would be to take the state off the edges/routes and just associate that state with the player
20:16	llasram	benkay: Fixures only applying when running the tests for an entire namespace are standing issue with clojure.test
20:16	Raynes	benkay: I suppose it doesn't, which I imagine is in fact the reason I added test-vars.
20:16	llasram	I personally just avoid clojure.test fixtures for that reason -- it's just as easy to write your fixtures such that you explicitly apply their functions/macros in each test
20:16	Raynes	It isn't a standing issue because test-vars exists now.
20:16	llasram	ah
20:17	Raynes	But prior to 1.6.0, it was indeed an issue.
20:17	llasram	I missed that
20:17	benkay	Raynes: thou art a scholar
20:17	llasram	Obvs didn't read the changelog closely enough :-)
20:17	Raynes	https://github.com/clojure/clojure/blob/919a7100ddf327d73bc2d50d9ee1411d4a0e8921/src/clj/clojure/test.clj#L710
20:17	technomancy	Raynes actually succeeded in getting a commit applied to clojure.test
20:17	technomancy	clearly he is the chosen one of the prophecy
20:17	Raynes	I'm pretty sure it took over a year.
20:17	Raynes	lol
20:17	Raynes	A year + gfredericks.
20:18	llasram	I probably saw the entry, but my mind elided it because it was obviously impossible
20:18	technomancy	some interpretations of the prophecy mention the chosen two
20:19	Raynes	I don't think 1.6 had actual breaking changes over 1.5.1, did it?
20:19	Raynes	I've been knee deep in Python for like 5 months though, so I could be wrong.
20:20	talios	I don't believe it do.
20:20	llasram	The hash changes lead to requiring updates of some Clojure-aware collection libraries
20:20	llasram	But I think that's it
20:22	benkay	well thanks, all.
20:25	Raynes	benkay: Sorry I'm being so aggressive with the test-vars. It's approximately the only thing I've felt strongly enough about to go through the contribution process to add :P
20:25	benkay	i now have an entirely different set of errors from my tests to look over, so i guess that's progress...
20:25	benkay	not at all, Raynes! and as a matter of fact thanks for your work on that.
20:25	benkay	i feel like a complete and total moron for having seen test-var and not test-vars
20:27	mdeboard	lol
20:28	gfredericks	benkay: that's a mistake a regular person would make, not a complete and total moron
20:28	mdeboard	benkay: If it makes you feel any better, I just spent 3 minutes trying to figure out why (get "foo" {"foo" "bar"}) kept returning nil (backward params)
20:28	Raynes	Yeah, it's not like you're gfredericks or anything
20:28	clojurebot	It's greek to me.
20:28	Raynes	<3
20:28	Raynes	jk ily gf
20:28	gfredericks	lol
20:29	benkay	i am extremely tempted to throw everything out and upgrade to 1.6.0 just for test-vars, actually.
20:30	gfredericks	clojurebot: test-vars is the killer app of clojure 1.6.0
20:30	clojurebot	c'est bon!
20:30	benkay	Raynes: does test-vars work as a ns-hook?
20:30	Raynes	I've never really used much in the way of fancy testing hooksy thingies.
20:31	Raynes	But all test-vars does is look up metadata on namespaces to get fixtures.
20:31	Raynes	So as long as it can do that, it should work anywhere.
20:32	gfredericks	Dear Randall Munroe: what would happen if you actually crossed every crossable thing?
20:32	benkay	lol okay 1.6 great looking fine but where'd the output from clojure test mode go
20:32	benkay	waaah
20:34	gfredericks	what would it take to start using maven for emacs packages
20:34	mdeboard	...
20:34	johnwalker	don't
20:34	clojurebot	excusez-moi
20:34	mdeboard	what.
20:34	johnwalker	please don't
20:34	mdeboard	srs
20:35	johnwalker	for your sake
20:35	gfredericks	my sake is already ragged from being on the bleeding edge all the time
20:35	johnwalker	bleeding edge 6 years ago?
20:35	johnwalker	come on man, you gotta do grails for emacs packages
20:36	gfredericks	I mean the normal emacs package manager
20:36	gfredericks	which only gives you the latest of everything
20:36	johnwalker	oh
20:36	johnwalker	well there's marmalade stable
20:37	johnwalker	err, melpa stable
20:53	umpa	,coll
20:53	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: coll in this context, compiling:(NO_SOURCE_PATH:0:0)>
20:54	umpa	,col
20:54	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: col in this context, compiling:(NO_SOURCE_PATH:0:0)>
20:54	mdeboard	umpa: coll just means collection
20:55	umpa	,(def col '((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) (0 4 8) (2 4 6)))
20:55	clojurebot	#'sandbox/col
20:55	umpa	,(def ks [8 2 6 0])
20:55	clojurebot	#'sandbox/ks
20:56	umpa	,((fn [ks ls] (reduce (fn [m v] (assoc m v (filter #(set/intersection (set %) (set v)) ls))) {} ks)) ks col)
20:56	clojurebot	#<CompilerException java.lang.RuntimeException: No such namespace: set, compiling:(NO_SOURCE_PATH:0:0)>
20:56	umpa	this is not working
20:56	akhudek	clojure.set
20:57	umpa	,(require 'clojure.set :as set)
20:57	clojurebot	#<ClassCastException java.lang.ClassCastException: java.lang.Boolean cannot be cast to clojure.lang.Symbol>
20:57	akhudek	just use it directly
20:57	umpa	,(require 'clojure.set as set)
20:57	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: as in this context, compiling:(NO_SOURCE_PATH:0:0)>
20:57	akhudek	(clojure.set/intersection … )
20:57	umpa	,((fn [ks ls] (reduce (fn [m v] (assoc m v (filter #(clojure.set/intersection (set %) (set v)) ls))) {} ks)) ks col)
20:57	clojurebot	#<IllegalArgumentException java.lang.IllegalArgumentException: Don't know how to create ISeq from: java.lang.Long>
21:03	umpa	isn't reduce suppose to take values from coll and ks one by one and apply the filter
21:07	technomancy	gfredericks: nix for emacs packages makes like seventeen times more sense
21:07	technomancy	err--guix rather
21:08	gfredericks	technomancy: hokay hokay I'll go for that
21:08	gfredericks	I just don't know it as good
21:09	technomancy	it's worth getting to know
21:09	gfredericks	oh I use it for at least half of my junk
21:35	technomancy	guix or nix?
21:35	Raynes	Both
21:48	gfredericks	technomancy: nix
21:48	gfredericks	I have no idea what guix is
21:48	gfredericks	so I probably shouldn't have responded so confidently
21:51	justin_smith	gfredericks: wait, do you actually use nixos?
21:55	Raynes	Somebody has to
21:56	justin_smith	it looks interesting
21:56	Raynes	So does space travel, but I ain't looking to go to Mars any time soon.
21:59	justin_smith	funny, that does sound like what an experimental Linux distro would do
22:00	johnwalker	do you like nixos?
22:08	dpathakj	there’s nix, the package manager, and nixos, the linux distribution
22:08	dpathakj	seems like you could get away with one without the ohter
22:10	technomancy	yeah, nix works great on debian
22:11	technomancy	gfredericks: guix is nix without the weird one-off package language; it uses guile instead
22:11	technomancy	and it also actually has standards for licensing instead of just accepting any random shit
22:15	gfredericks	justin_smith: no just the package manager
22:15	gfredericks	technomancy: "is nix" meaning it's part of their ecosystem or it just instantiated the core concepts in some other way?
22:16	technomancy	gfredericks: it's compatible with the same package sources and shares a good deal of code
22:16	gfredericks	doesn't sound absolutely terrible so far
22:16	technomancy	afaict it's like a new language that compiles to the same runtime, like clojure
22:16	gfredericks	metaphors how do they work
22:16	technomancy	but nix is already pretty obscure, and the guix community is smaller still
22:17	technomancy	anyway, as soon as I find something that wheezy's versions are too old for imma try guix, but it hasn't actually happened yet
22:17	technomancy	(modulo lein/emacs obvs)
22:17	gfredericks	I should stick to popular software communities like clojure
22:17	technomancy	clojure is "mature"
22:19	dpathakj	technomancy: re: ‘standards for licensing’: do you know if guix is gpl-only, or are more permissive but still free licenses like MIT allowed by guix’s maintainers?
22:19	technomancy	dpathakj: it's libre licenses only
22:20	technomancy	dpathakj: it just bans crazy stuff like the adobe flash player, which I was dismayed to learn that nix ships, apparently illegally
22:20	technomancy	iirc it's basically limited to dfsg-free plus gfdl
22:24	dpathakj	technomancy: thanks. makes sense.
22:25	johnwalker	technomancy: do you use a gnu distribution now?
22:26	johnwalker	it sounds like you use debian
22:32	mdeboard	justin_smith, Raynes: Well I hope you're happy.
22:32	Raynes	In general, yes.
22:32	justin_smith	mdeboard: things are pretty peachy
22:33	mdeboard	I've been refactoring for "pure functional" blah blah blah
22:33	justin_smith	so you made the switch to an explicit game state object passed between functions?
22:33	mdeboard	Lucky for you I wrote things in such a manner that it makes such refactoring straightforward
22:33	justin_smith	lucky for me
22:33	mdeboard	LUCKILY
22:34	mdeboard	You've squeaked by, by the skin of your teeth.
22:34	talios	mdeboard - are you doing any form of clojure based monad foo there as part of being pure?
22:34	mdeboard	talios: All I know about monads is that they're like astronauts eating a burrito
22:34	gfredericks	~monad
22:34	clojurebot	monad is monads
22:34	gfredericks	~monads
22:34	clojurebot	monads are #=(str "super" "awesome")
22:35	gfredericks	Supera Wes o' Me
22:36	justin_smith	talios: that's funny, because the purity talk we had at my first highschool was "I'm worth waiting for", which I could interpret to be relevant to the usefulness of laziness in pure functional programming
22:36	mdeboard	(defer)
22:37	justin_smith	or maybe the usage of delays or something
22:37	justin_smith	yeah
22:37	mdeboard	(force)
22:37	mdeboard	:o
22:37	justin_smith	oh yuck
22:37	mdeboard	(defn chastity-belt (conj force defer))
22:37	mdeboard	throw a [] in there
22:37	talios	force sounds rapey
22:37	justin_smith	fits with the whole "promise keepers" thing I guess
22:37	mdeboard	hahaa
22:37	mdeboard	nice
22:37	mdeboard	@ promise keeper
22:38	technomancy	johnwalker: yeah, I'm on debian
22:39	gws	,(def m1 {:k "v"}) (defmacro mac [{v :k}] `(println ~v)) (mac m1)
22:39	clojurebot	#'sandbox/m1
22:40	gws	,(def m1 {:k "v"}) (defmacro mac [{v :k}] `(println ~v)) (mac {:k "v"})
22:40	clojurebot	#'sandbox/m1
22:40	gws	so that's not gonna work in IRC, but can anyone tell me why the first one prints nil and the second prints "v"?
22:41	justin_smith	(do ,(def m1 {:k "v"}) (defmacro mac [{v :k}] `(println ~v)) (mac {:k "v"}))
22:41	justin_smith	,(do (def m1 {:k "v"}) (defmacro mac [{v :k}] `(println ~v)) (mac {:k "v"}))
22:41	clojurebot	v\n
22:41	gws	ha, thanks
22:42	gws	,(do (def m1 {:k "v"}) (defmacro mac [{v :k}] `(println ~v)) (mac m1))
22:42	clojurebot	nil\n
22:42	gws	i'm trying to wrap my head around the difference there
22:43	gfredericks	,(let [{v :k} 'm1] ["My v is" v])
22:43	clojurebot	["My v is" nil]
22:43	dbasch	gws: macroexpand-1
22:43	johnwalker	technomancy: cool
22:45	amalloy	gws: remember that a macro receives unevaluated symbols as its arguments
22:45	amalloy	not the values to which those symbols may be bound in some particular evaluation context
22:47	gws	ah, alright. i see it now
22:49	gws	dbasch: i had that, but I was missing was the "why" - unevaluated symbols made it click
22:50	gws	amalloy: thanks, i appreciate it
23:58	kelseygi	hey i’ve got a string
23:58	kelseygi	& i need to split it on a separator
23:58	mdeboard	nice
23:58	lazybot	java.lang.RuntimeException: Unable to resolve symbol: i in this context
23:58	kelseygi	but i need to do something different with the last substring depending on whether it ended in the separator or not
23:58	kelseygi	everything i can think of is super imperative-y
23:59	mdeboard	,(clojure.string/split "a,b,c,d," #",")
23:59	clojurebot	["a" "b" "c" "d"]
23:59	mdeboard	?
23:59	kelseygi	,(clojure.string/split "a,b,c,d" #",")
23:59	clojurebot	["a" "b" "c" "d"]
23:59	kelseygi	i wanna be able to tell the diff between those two