#clojure – 2012-07-14

2012-07-14

00:04	xeqi	how you, uh... how you coming on that clojure -> elisp you working on?
00:08	duck1123	If I'm getting a "Don't know how to create ISeq from: clojure.lang.Symbol" in cljs, how can I find where it's happening? I don't se line numbers for any of my code from cljsbuild
00:09	technomancy	xeqi: compiling to elisp?
00:09	technomancy	xeqi: I don't think it can happen; at least until elisp gets native immutability for at least strings; hopefully for lists too.
00:10	technomancy	more ranting on the topic here: http://technomancy.us/159
00:11	duck1123	I would love to see a good editor taking all the best parts from emacs but built with clojure
00:12	duck1123	clooj didn't cut it
00:12	technomancy	the best parts from emacs are the decades worth of elisp that's accumulated around it
00:13	technomancy	you could make something that works really well for editing clojure, but you'd have to give up on magit, tramp, emacsclient, etc
00:13	duck1123	true, but the best of those modules could be converted
00:13	technomancy	sure; how much time you got? =)
00:14	akhudek	probably not worth reinventing what already works
00:15	akhudek	what would emacs in clojure ultimately gain?
00:15	akhudek	I guess being able to write plugins with clojure...
00:15	duck1123	a more powerful lisp system at the heart of the editor
00:16	technomancy	it's much more likely that emacs would start running on guile and you could compile to scheme instead
00:16	technomancy	by which I mean not very likely, but at least possible
00:16	sjl	s/with clojure/with all of the JVM ecosystem/
00:18	eggsby	hey guys, in paredit mode how do I add a surrounding paren if I forgot one? every time I try to enter one it just puts it outside the form :/
00:18	ibdknox	Light Table = Clojure!
00:19	ibdknox	but she is teh young
00:19	sjl	ibdknox: call me when it can run through tmux ;)
00:19	clojurebot	excusez-moi
00:19	eggsby	I wonder if light table works in lynx...
00:19	ibdknox	lol
00:20	sjl	that's the main problem really -- there are lots of new generation editors popping around but none run in a command line
00:20	ibdknox	why would they?
00:20	sjl	so I can't use any of the pretty things like vico, st2, etc
00:20	eggsby	yes but luckily you have vi for that sjl
00:20	sjl	eggsby: heh, yeah
00:20	duck1123	So will Light Table ever get an application window, or will it always be strictly browser based?
00:21	technomancy	ibdknox: you can't run a distributed team with anything but SSH
00:21	ibdknox	technomancy: huh?
00:21	sjl	ibdknox: editing a file in a command line is not going away any time soon (when ssh'ed into a server, pairing through tmux, or just for those that live in tmux locally)
00:21	ibdknox	duck1123: yes, for mac there already is one
00:21	ibdknox	sjl: none of those things are strictly reliant on a command line
00:22	sjl	ibdknox: well yeah, once someone implements all their replacements we can ditch it, heh
00:22	andres-v	eggsby: M-(
00:22	technomancy	even inclusion-transformation-based GUI collaborative systems (which are really hard to get right) so far only work on a document-by-document basis
00:22	technomancy	which makes them useless for pairing
00:23	sjl	I gave money to light table because I want to see where it goes and hopefully it'll kick the other editing communities a bit and prod them into making awesome stuff too
00:23	andres-v	eggsby: or mark a region and then: (
00:23	ibdknox	technomancy: a document-by-document basis?
00:23	duck1123	I was too broke to consider donating, but I wish you well, and that hast to count for something
00:24	ibdknox	technomancy: as in the entire environment isn't shared?
00:24	technomancy	ibdknox: with things like gobby you have a set of docs you're subscribed to, but it doesn't coordinate focus or repl/shell buffers
00:24	eggsby	ah, thanks andres-v
00:24	ibdknox	technomancy: sure, that's a limitation in the way people have though about the problem, not an inherent problem with a GUI solution
00:25	ibdknox	which is not to say I'll be working on that any time soon, but I don't think the argument should be "x can't be run on the command line" - that's not a real solution
00:26	ibdknox	it's a workaround
00:27	akhudek	anyone remember the days of X11 application forwarding?
00:27	technomancy	it's possible; it's just a lot of work and nobody's gotten it right yet
00:27	technomancy	a lot of work for something you can already get for free
00:27	ibdknox	technomancy: no argument there :)
00:27	ibdknox	eh
00:28	ibdknox	in a limited fashion yes
00:28	duck1123	That's how I develop. My server is in the house and I'm forwarding my emacs to my mac out in the garage
00:28	ibdknox	doesn't mean we should just say fuck it, let's not innovate anymore
00:28	akhudek	ibdknox: completely agree, light table looks interesting
00:29	ibdknox	light table is infantile :)
00:30	ibdknox	It'll be very interesting to see where it ends up
00:30	ibdknox	tons of learning to do
00:31	akhudek	speaking of learning, I found the recent hacker news discussion on MVC versus finite state machines pretty interesting
00:32	sjl	so I'm guessing I just reimplemented something built-in, and probably did it poorly
00:33	sjl	https://github.com/sjl/caves/commit/cf1c6b442a70e7b207d2fef8d35a41da112a4c5f
00:33	sjl	Basically I wanted a way to give default implementations when defining a protocol
00:33	sjl	but not just define the protocol for Object, because I only want the types that specifically implement the protocol to (maybe) get the default methods
00:34	sjl	so, for example, in my game: lots of things are going to implement Mobile, and need a can-move? function. Most of those all have the same movement rules, but occasionally some will want to override
00:36	Nican	I knew I read that somewhere else, is this reference to this? http://stevelosh.com/blog/2012/07/caves-of-clojure-04/#movement
00:36	sjl	But I'm not entirely sure whether this pair of macros is going to blow up in my face later... so hopefully someone can confirm whether they're sane or not
00:37	sjl	Nican: yeah. For that entry I just defined the functions when adding the Mobile aspect to the Player
00:37	sjl	since at that point I only had a player and nothing else
00:37	sjl	but if I wanted to add another creature that moved, I'd need to copy and paste the code
00:37	sjl	or add a helper function somewhere and manually type it out each time
00:39	sjl	that (do (defprotocol Foo ...) (def Foo ...)) feels particularly filthy.
00:39	akhudek	sjl: so the only problem with using Object is that it won't throw an error if you pass it something that doesn't implement the protocol?
00:40	sjl	akhudek: well, no, implementing it on Object won't let me override just one of the methods -- either it gets all the defaults or none of them
00:40	Nican	http://clojure.org/protocols "No implementations are provided"
00:40	sjl	Nican: yeah that's what I'm trying to hack around, heh
00:41	sjl	akhudek: because if the type you give it implements Foospect, it uses the methods defined for the thing you're giving it. If the type doesn't implement Foospect at all, then it falls back to Object's implementations
00:41	sjl	I think.
00:43	akhudek	sjl: I bet you could get the behaviour you want with a multimethod
00:45	Nican	Only interesthing things I found: http://fulldisclojure.blogspot.com/2010/08/thoughts-on-protocols.html http://stackoverflow.com/questions/10732355/clojure-defprotocol-default-implementaion-not-inlined/10732895#10732895
00:45	sjl	akhudek: here's the example showing the problem with Object https://www.refheap.com/paste/3606
00:47	akhudek	I see. If hello and world were both multimethods with default implementations, then it would be easy to override just one.
00:47	sjl	Nican: yeah, that first link goes about halfway to where my code goes (my code uses the merging maps into extend trick, but defines the defaults as metadata on the protocol itself)
00:48	sjl	akhudek: yeah, but then they're not really grouped together in any special way
00:49	sjl	akhudek: plus, it doesn't solve the case of me wanting to distinguish "something that is Mobile, regardless of whether it uses the default impls" and "something that is not Mobile at all", right?
00:49	sjl	akhudek: it's too permissive, i.e. those default implementations catch too much
00:50	akhudek	sjl: that's true, but you could consider writing functions like "is-mobile?" that has no default implementation
00:50	sjl	sure... that seems a bit ugly too though
00:50	akhudek	I mean, it sounds like you are partly using protocols as a type checking system
00:51	sjl	effectively, yes
00:51	akhudek	which as I understand it, wasn't really the intention of protocols
00:51	akhudek	would be nice to have an optional type system for clojure, that's for sure
00:51	sjl	https://github.com/frenchy64/typed-clojure
00:51	akhudek	yeah, I know of that, haven't checked on the progress though
00:52	sjl	seems to be active development, which is a good sign
00:52	akhudek	yeah, very!
00:52	sjl	I agree that optional static typing would be awesome
00:52	sjl	Haskell's typechecker would catch so much annoying crap
00:54	sjl	okay time to get to sleep, later
00:54	akhudek	night sjl
00:54	Nican	later
02:19	mindbender	I'm getting this error when trying to run `foreman start` from my project: java.lang.IllegalArgumentException: No implementation of method: :as-file of protocol: #'clojure.java.io/Coercions found for class: java.lang.Character
02:20	mindbender	I read something similar @ https://groups.google.com/forum/?fromgroups#!topic/leiningen/59655dfCNJg. I'm using lein 2
02:21	mindbender	technomancy: any suggestions with ^-^
02:23	tomoj	mindbender: it would help if you pasted (not in here) the full stack trace
02:23	tomoj	..probably
02:30	mindbender	tomoj: just did. Here https://gist.github.com/55acd9099967140b6605.
02:33	tomoj	try :source-paths ["src/app/clj"]
02:37	metajack	Are PersistentTreeMaps not comparable? If not, why?
02:40	metajack	,(compare (sorted-map) (sorted-map))
02:40	clojurebot	0
02:41	metajack	,(compare (sorted-map-by compare) (sorted-map-by compare))
02:41	clojurebot	#<ClassCastException java.lang.ClassCastException: clojure.lang.PersistentTreeMap cannot be cast to java.lang.Comparable>
02:41	tomoj	,(identical? (sorted-map) (sorted-map))
02:41	clojurebot	true
02:42	metajack	so sorted-map would fail if they weren't identical?
02:42	tomoj	,(let [x (sorted-map-by compare)] (compare x x))
02:43	tomoj	yeah
02:43	tomoj	how would you compare sorted maps?
02:44	tomoj	assume both the keys and values are comparable?
02:44	mindbender	tomoj: worked. Thanks
02:44	metajack	keys are vectors of length 2 of longs, and values are keywords
02:44	metajack	so i believe so
02:45	tomoj	but if the values were not comparable, should PersistentTreeMap.compareTo throw an exception?
02:45	tomoj	I don't know actually the answer to your question ("why?")
04:40	AtKaaZ	hi, is there a program for 3D representation of clojure programs?
04:41	AtKaaZ	rather than you know textual
04:47	AtKaaZ	some kind of 3D graph visualization of the clojure code ? possibly with editing it too
05:02	TEttinger	AtKaaZ, I think there was something done like this
05:02	AtKaaZ	I would've hoped so, can you remember more precisely ?
05:03	AtKaaZ	a name of a program
05:05	AtKaaZ	I would think that this could be applied right after the Reader part of clojure, also I've no experience with clojure just starting to read about it
05:10	AtKaaZ	TEttinger did you mean clojureatlas ?
05:10	TEttinger	sorry AtKaaZ, was helping in another channel
05:10	AtKaaZ	it's quite ok
05:10	TEttinger	I haven't tried clojureatlas, i think you need money for it
05:10	AtKaaZ	yeah looks like it :/
05:10	TEttinger	https://github.com/quil/quil maybe?
05:13	AtKaaZ	is that like clojureatlas ? it seems like it's only a 3D library for clojure?
05:13	AtKaaZ	or something to help you draw
05:15	TEttinger	http://keminglabs.com/c2/ maybe could be handled with some macros
05:16	TEttinger	lynaghk: ping
05:17	AtKaaZ	do you mean that by using macros I can make c2 draw graphs of the clojure code ? sort of like using the clojure source code as data for c2 to draw it?
05:17	AtKaaZ	(which is what I want btw :D)
05:17	TEttinger	yep, code is data
05:18	TEttinger	but you obviously need some way for it to know how to display different kinds of code
05:18	AtKaaZ	yeah for sure
05:18	AtKaaZ	can clojure create code on the fly at runtime ?
05:18	TEttinger	I think yes
05:19	AtKaaZ	is there a simple example of that maybe? that I could peek
05:19	TEttinger	not clojurescript, but clojure can
05:19	AtKaaZ	yeah the real clojure
05:19	TEttinger	well macros are "compile-time", but it compiles continuously IIRC
05:19	AtKaaZ	c2 seems definitely something I want to look into
05:19	TEttinger	you need to talk to someone who knows how to use macros, there are more qualified people here
05:20	AtKaaZ	alright thanks so far, I'll alt-tab and try read more
05:20	TEttinger	lynaghk wrote C2, might be good to stay and wait for a response
05:20	AtKaaZ	yeah I'll definitely hang in here
05:20	AtKaaZ	btw, what ide/editor do you use?
05:20	AtKaaZ	something with colors :D
05:21	TEttinger	yeah, I am having a mess of a time in #coffeescript right now explaining what I hope is the right thing to someone with very poor English-language skills
05:21	TEttinger	oh I use Eclipse with rainbow parens via the main plugin, CCW
05:22	TEttinger	but I mostly use eclipse for editing and stick to the lein2 repl
05:22	TEttinger	I mean Leiningen v2
05:22	TEttinger	the repl is better in that version, it auto-completes and other amazing stuff
05:22	AtKaaZ	i have eclipse and some clojure plugin cross something...
05:22	TEttinger	yep
05:22	TEttinger	counterclockwise
05:23	AtKaaZ	yeah that one
05:23	AtKaaZ	what else would I need ? rainbow parens?
05:23	TEttinger	it has it
05:23	TEttinger	or at least it should
05:23	AtKaaZ	ok
05:23	TEttinger	outer level is red, second nest is green, etc. (colors may have changed with newer versions)
05:23	AtKaaZ	that Liningen v2 is something outside eclipse?
05:24	TEttinger	oh yes
05:24	TEttinger	leiningen is the best way for managing clojure projects across OS's and IDEs
05:25	TEttinger	it is also one of two de facto dependency managers for clojure
05:26	TEttinger	https://github.com/technomancy/leiningen/ is the main tool, http://timpner.com/computing/clojure-importing-a-leiningen-project-into-eclipse/ is how to make your leiningen project work in eclipse (not the other way around though)
05:26	AtKaaZ	oh i see, btw, did I notice right that eclipse doesn't auto build my clojure project (or recompile it) when I modify the source, so I cannot know right away if I made mistakes (Project->Build auto... is on)
05:26	TEttinger	yeah, I just go into the command line i keep open at all times and just lein run
05:26	AtKaaZ	oh I see I need lein-eclipse
05:26	TEttinger	it recompiles if needed and runs
05:27	TEttinger	yeah, in your leiningen project file
05:27	TEttinger	it will automatically download it if it is a dependency or dev-dependency in your project file and you then run "lein deps"
05:28	TEttinger	no need to handle standalone downloads of most libs
05:29	AtKaaZ	alright so if a project doesn't have lein-eclipse as dep in the project.clj then to make it work in eclipse correctly i need to add it as dep ?
05:30	TEttinger	yup
05:30	TEttinger	then run lein deps and it will grab it
05:30	AtKaaZ	can I have two lines with ":dev-dependencies" or do I have to append them in one line ?
05:31	TEttinger	:dev-dependencies [[lein-eclipse "1.0.0"]]
05:31	TEttinger	note the nested vectors
05:32	TEttinger	:dev-dependencies [[lein-eclipse "1.0.0"] [some-other-thing "1.2"]]
05:32	AtKaaZ	i understand
05:32	TEttinger	lines don't matter
05:32	AtKaaZ	however I'm curious what would happen if :dev-dependencies appears twice in project.clj
05:32	AtKaaZ	error?
05:33	TEttinger	probably an error when it tries to assign different values to one map entry
05:33	AtKaaZ	it acts as a clojure keyword? the ":" ones
05:33	TEttinger	like {:a "alpha" :b "beta" :a "alpha again"}
05:33	TEttinger	yes
05:33	AtKaaZ	makes sense, I like it
05:33	TEttinger	leiningen projects are clojure code
05:34	AtKaaZ	so that would err then.
05:34	TEttinger	and once you can visualize clojure code you can visualize lein files :D
05:34	AtKaaZ	sounds good to me
05:35	AtKaaZ	how do I get that prompt to run lein deps
05:38	msappler	hi. It seem to be bad for transactions if I have refs which have other refs in them as pointers?
05:38	AtKaaZ	oh i see I shouldn't be in eclipse while running lein deps
05:41	msappler	I have a grid data structure (for example width and height 100) so a grid of vector 100 of vector 100 cells
05:42	msappler	and each cell is a ref of a map which also holds :adjacent-cells the other cells'refs
05:47	y3di	does clojure not have a distinct else-if form?
05:47	TEttinger	y3di: maybe you want one of the other conditional macros?
05:48	TEttinger	http://clojuredocs.org/quickref/Clojure%20Core#OperationsFlow+Control
05:48	AtKaaZ	i managed to run lein deps and lein eclipse :)
05:48	TEttinger	woohoo!
05:48	AtKaaZ	lein.bat was in my case on windows
05:48	TEttinger	yep
05:49	TEttinger	it should create the needed .project and .classpath files
05:49	AtKaaZ	there are only 3 colors in clojure code, is that right ?
05:49	TEttinger	hm?
05:49	AtKaaZ	maybe 4 with the parenthesis
05:49	AtKaaZ	yeah it probably did, I had eclipse closed just for that
05:51	y3di	TEttinger, yea i'm not seeing any elseif forms. I guess I should just use another if statement as the last argument of the first if?
05:51	TEttinger	or clojuredocs.org/clojure_core/clojure.core/cond
05:58	AtKaaZ	how do you make it show errors in eclipse ?
05:59	AtKaaZ	like there's an error on line x and marked as red
06:01	AtKaaZ	switching to edit mode restricted does help some
06:03	TEttinger	AtKaaZ: you can change colors in Eclipse with Window->Preferences->Clojure
06:03	TEttinger	I don't know if the clojure plugin does that
06:03	TEttinger	the error markers
06:04	y3di	paredit is sposed to handle closing parens for me right?
06:04	TEttinger	y3di, I don't use it, but I think so yes
06:04	AtKaaZ	y3di it does for me if I switch from unrestricted edit
06:06	AtKaaZ	TEttinger thanks, I notice the rainbow parens are also there
06:06	TEttinger	yep
06:06	TEttinger	they won't show in the preview thoughh
06:06	y3di	hm, i don;t think i've installed it right then
06:06	TEttinger	you probably also want to make strings and symbols not the same color
06:07	y3di	tho i swear i did
06:07	y3di	is there a way to test whether i have a certain package installed in paredit
06:17	msappler	how can i see number of retries of my transactions
06:21	y3di	does every clojure file need a namespce
06:21	AtKaaZ	i think it's the package declaration in java
06:21	AtKaaZ	like*
06:25	y3di	nvm it was an issue with swankclojure
06:25	y3di	i have a notherquestion though
06:25	y3di	currently im printing a really larger number
06:25	y3di	and its displayed 297392^M or something like that
06:25	y3di	how can I get it to print the actual number
06:26	AtKaaZ	^M makes me think of \n\r displayed in vim
06:26	AtKaaZ	or was it \r\n :D
06:28	AtKaaZ	if Chas Emerick here?
06:28	AtKaaZ	is*
06:30	y3di	AtKaaZ, do you know how to print it normally?
06:31	AtKaaZ	I don't sorry, I'm too new to clojure to even know how to code hello world:)
06:32	AtKaaZ	but I take it the number has more digits than 297392 ?
06:32	y3di	yea
06:32	AtKaaZ	no ideas then
06:40	y3di	this seems like it should be an easy problem to solve, but ive been googling and i cant find an answer
06:41	AtKaaZ	I don't suppose you could pastebin the program code?
06:41	AtKaaZ	or a simplified version of it:)
06:45	y3di	https://gist.github.com/3110511
06:46	AtKaaZ	I get the same numbers without the ^M
06:47	y3di	oh
06:47	y3di	maybe i was just being a dumbass this entire time
06:47	AtKaaZ	lol
06:47	y3di	i tried entering that number as the solution, and it didn't work
06:48	AtKaaZ	well maybe indeed those numbers have no more digits
06:48	AtKaaZ	but the ^M seems to be the new line
06:49	AtKaaZ	if I do (println (reduce sum-if-divisible (range 1 10000))) I get 23331669
06:50	AtKaaZ	or (println (reduce + (range 1 100000))) to get 4999950000
06:50	AtKaaZ	(granted I've no idea what the program does)
06:52	y3di	nvm i figure itd out
06:53	AtKaaZ	I like knowing what the fix is, should you wish to share it:)
06:54	y3di	i did (range 1 1000) instead of (range 0 1000)
06:54	y3di	the problem was
06:54	y3di	summing up the integers between 1 and 1000 that are divisible by 3 or 5
06:55	AtKaaZ	I eventually figured from the code, tho I am very new to clojure
06:55	AtKaaZ	thanks
07:26	evildaemon	Okay, sorry but I HAD to show this to somebody.
07:26	evildaemon	http://forums.thedailywtf.com/forums/p/26379/297191.aspx#297191
07:59	zoldar	Hello. Why do I get integer overflow when running that taking larger sequences from that collection? https://www.refheap.com/paste/3607 (that's problem #147 from 4clojure). Shouldn't clojure automatically promote to long when needed? shouldn't it be long in the first place?
08:09	zoldar	ok, found it, had to use math operator with apostrophe: http://dev.clojure.org/display/doc/Documentation+for+1.3+Numerics
08:10	zoldar	seems it was going even beyond long range
08:21	daniel___	is there a function which will give me this? "123456789" -> ("147" "258" "369") ?
08:21	daniel___	like partition with interleave
08:24	eagleflo	daniel___: can't you compose exactly those two functions, partition and interleave, to achieve that?
08:26	daniel___	,(map #(apply str %) (partition 3 (apply str (apply interleave (partition 3 "123456789"))))
08:27	daniel___	that should achieve it i think
08:27	daniel___	&(map #(apply str %) (partition 3 (apply str (apply interleave (partition 3 "123456789"))))
08:27	lazybot	java.lang.RuntimeException: EOF while reading, starting at line 1
08:27	daniel___	&(map #(apply str %) (partition 3 (apply str (apply interleave (partition 3 "123456789")))))
08:27	lazybot	⇒ ("147" "258" "369")
08:38	wingy	what does Attempting to call unbound fn: #'user/zipmap2 mean?
08:42	wingy	oh nvm
08:50	wingy	i wonder a thing .. looking at "and" i see its calling itself
08:50	wingy	,(source and)
08:50	wingy	&(source and)
08:50	lazybot	java.lang.RuntimeException: Unable to resolve symbol: source in this context
08:51	wingy	here the source: https://gist.github.com/3111123
08:51	wingy	could/should the self call be replaced by recur to introduce TCO?
08:52	wingy	if one calls and with a large nr of arguments that would eat up stack space with the current implementation?
08:55	wingy	yepp .. i got a stackoverflow error when running it with a long arg list :/
09:05	wingy	isn't it possible to use recur instead in and macro?
09:07	rfgpfeiffer	has anybody got libgdx to play nicely with leiningen
09:10	yonatane	in (defmacro with-open... why does it call with-open recursively to resolve bindings? Couldn't it just call let with the entire bindings vector at once?
09:10	yonatane	https://github.com/clojure/clojure/blob/d0c380d9809fd242bec688c7134e900f0bbedcac/src/clj/clojure/core.clj#L3376
09:23	yonatane	wingy, since i'm new as well i can only guess, so i guess there's no recur point (e.g loop or function definition) in the and macro. Maybe having it there would cause other problems, and isn't worth the trouble since having so many arguments is bad style.
09:27	wingy	yonatane: a fn or loop could always be used as recur point
09:27	wingy	so if you call recur in a fn where there is no loop the fn is the recur point
09:28	wingy	i tried to use recur but an error said the argument list didnt match
09:28	yonatane	but and is a macro. i'm not sure it's the same
09:28	wingy	it expected 0 args .. i guess the first argument list (it had multiple ones) was used and that was for no args []
09:31	yonatane	well, i'm in the process of reading the clojure source. I'll tell you next year when i'm done ;)
09:31	wingy	perhaps you are right
09:32	wingy	https://gist.github.com/3111353
09:32	wingy	the fn worked but not the macro
09:33	wingy	but the error message is strange: java.lang.IllegalArgumentException: Mismatched argument count to recur, expected: 0 args, got: 1, compiling:(NO_SOURCE_PATH:2:30)\|\|
09:33	wingy	anyway :)
09:35	yonatane	read this: https://github.com/clojure/clojure/blob/d0c380d9809fd242bec688c7134e900f0bbedcac/src/clj/clojure/core.clj#L406
09:36	yonatane	i can't understand it yet
09:37	yonatane	maybe macroexpand it to get a hint
09:47	yonatane	wingy: (macroexpand '(recurme false))
09:47	yonatane	;=> (if false (recur false) "done")
09:54	yonatane	ok i just understood why with-open is recursive - to match the .close to the resource and close only opened resources.
09:56	wingy	yonatane: are you new to clj as well?
09:56	wingy	just trying to get it?
10:03	yonatane	i'm not totally new. there are entire parts that i don't know at all, like macros which i just now read enough to answer you
10:06	yonatane	but you can do a lot without knowing much. I have a pet facebook app using lein, noir, jdbc, sandbar, ring etc.
10:13	wingy	yeah .. but i love to get more about programming so you can create powerful, less buggy systems than your competitors
10:13	wingy	features added each week others could just dream of
10:15	wingy	this was a really good video http://blip.tv/clojure/clojure-concurrency-819147
10:15	wingy	all rich's videos are of high quality
10:18	yonatane	have you watched the reducers video?
10:19	yonatane	http://vimeo.com/45561411
10:34	wingy	dont think so
10:34	wingy	ill watch it later
10:36	wingy	persistent data structures remind me about how git is working
10:42	cshell	I have a namespace (:require [google.identity :as id]) - what's the correct way to reference keywords in that namespace? :id/foo doesn't work - but :google.identity/foo does
10:50	andres-v	cshell: are you writing :id/foo or id/foo? (it shhould work with id/foo)
10:51	cshell	if i do id/foo I get a no such var
10:51	cshell	exceptiong
10:53	andres-v	cshell: I don't know that library, but what about: (require '(google [identity :as id])) and then use it like (id/foo ...)
10:53	cshell	it's just a custome on in there
10:53	cshell	that I created
10:54	cshell	it's not bound to a var so I don't think i can access it without a :
11:00	andres-v	and what is the name of the namespace, (ns google.identity) ?
11:01	inkybutton	hi everyone, i'm using js->clj on a json feed in clojurescript, but the result is a vector containing single characters of the feed string - a bug?
11:03	cshell	andres-v: yes, google.identity is the namespace
11:23	andres-v	cshell: :( I am testing a couple of simple namespaces and referencing a def from each other with require from the repl and it works but I don't know where is the difference with your code
12:04	samaaron	is it possible to catch reader exceptions when evaling arbitrary strings?
12:06	wingy	why do we buy computers with multiple cores if softwares/games dont use them
12:07	samaaron	wingy: so we can play multiple games at the same time?
12:07	wingy	:)
12:10	antoineB	hello, i don't manage to use a simple defmulti example http://pastebin.com/cVMD9rt3
12:15	duck1123	No one would write multi-core software if there were no multi-core computers
12:18	jkkramer	antoineB: that should work. if you're developing interactively and changed the definition of defmulti, you may be getting bit by its defonce semantics -- i.e., re-evaling from emacs won't change the definition
12:19	wingy	duck1123: smart :)
12:19	jkkramer	antoineB: you could try (ns-unmap ns 'emit) from within example.a namespace, and re-eval. or just restart repl/swank
12:21	wingy	i dont get this one: (defn map2 [f coll] (cons (f (first coll)) (map f (rest coll)))) isn't the recursion infinite since there is no base case to get out of it?
12:23	antoineB	jkkramer: i use C-c C-k i see the result expected, but when i type (emit "def") in the repl i get NullPOinterException
12:24	duck1123	wingy: there was no lazy-seq around that ?
12:24	antoineB	if i eval the last expression of the file it works (see the last expression in the paste)
12:25	wingy	duck1123: its just an example code how map works .. nothing tied to clj
12:25	jkkramer	antoineB: is your repl in the right namespace and/or requiring the right stuff?
12:25	antoineB	no
12:25	wingy	but it works .. i just dont get how it gets out of the recursion
12:26	jkkramer	wingy: the defn you pasted works, because it's not recursive: map2 called regular map
12:26	yonatane	wingy: i don't see a recursion
12:26	wingy	oh
12:26	wingy	yeah typo
12:26	zdennis	I've got two functions in my point namespace, but I always receive the warnings that I am overriding clojure-core classes, but when I require them I am excluding them. could any one help me understand what I'm doing wrong? https://gist.github.com/89d7095f77f1b7d96801
12:27	zdennis	*err. not core clases, core functions
12:27	wingy	ok it didnt work
12:27	wingy	(defn map2 [f coll] (cons (f (first coll)) (map2 f (rest coll)))) fails
12:27	wingy	nvm
12:27	wingy	an example from http://vimeo.com/45561411
12:30	wingy	wii .. starting to get recursion .. you have to have a base case, this works: (defn map2 [f coll] (if (empty? coll) '() (cons (f (first coll)) (map2 f (rest coll)))))
12:30	zdennis	the core functions I am overriding are max/min
12:32	zdennis	ah… I think I needed (:refer-clojure :exclude [max min]) in my point.clj ns
12:32	zdennis	is that bad form to define functions that share names and then to exclude core functions?
12:32	gfredericks	zdennis: yessir
12:32	gfredericks	zdennis: nossir
12:33	zdennis	gratzi @gfredericks
12:35	duck1123	Does anyone know about waltz? I'm trying to find out if it's possible to define the machine with an initial state, or if I have to just set it afterwards
13:13	cshell	andres
13:13	antoineB	(case (class "aa") clojure.lang.PersistentList "p" java.lang.String "s" "A")
13:13	antoineB	what is wrong with my switch?
13:16	AWizzArd	antoineB: what result do you expect?
13:17	antoineB	"s"
13:18	antoineB	java.lang.String isn't a symbol
13:18	AWizzArd	case is a macro
13:19	antoineB	AWizzArd: see the doc for availble type you can check on
13:19	antoineB	they talk about "(Clojure) composites thereof" i don't know what it is
13:21	antoineB	(case (str (class "aa")) "class clojure.lang.PersistentList" "p" "class java.lang.String" "s" "A")
13:21	antoineB	this work
13:21	antoineB	notice the "class ..."
13:22	gfredericks	,(doc case)
13:22	gfredericks	&(doc case)
13:22	lazybot	⇒ "Macro ([e & clauses]); Takes an expression, and a set of clauses. Each clause can take the form of either: test-constant result-expr (test-constant1 ... test-constantN) result-expr The test-constants are not evaluated. They must be compile-time literals, and ne... https://www.refheap.com/paste/3608
13:22	AWizzArd	It must be constants.
13:23	gfredericks	antoineB: so apparently classes aren't in the set of constants you can use
13:23	AWizzArd	But fortunately there are mechanisms to do what you want.
13:23	antoineB	i hope
13:23	gfredericks	like multimethods
13:23	antoineB	my solution is very elegant
13:24	gfredericks	or cond
13:24	AWizzArd	You can try condp for example
13:24	antoineB	in fact i use this "case" in a multi method context
13:25	wingy	does anyone know why this doesnt return 6
13:25	wingy	,(-> 1 inc (partial * 3))
13:25	AWizzArd	,(condp instance? "aa" clojure.lang.PersistentList "p" java.lang.String "s")
13:25	cshell	it returns a function
13:26	AWizzArd	wingy: (partial (inc 1) * 3) is what you got
13:26	gfredericks	wingy: ##(-> 1 inc ((partial * 3)))
13:26	lazybot	⇒ 6
13:26	gfredericks	or ##(-> 1 inc (* 3))
13:26	lazybot	⇒ 6
13:27	wingy	i thought that the args should return a fn
13:27	gfredericks	$google fogus thrush
13:27	lazybot	[fogus: Thrush in Clojure – Redux] http://blog.fogus.me/2010/09/28/thrush-in-clojure-redux/
13:27	wingy	inc returns a fn
13:27	wingy	why shouldn't partial?
13:27	gfredericks	wingy: ^ that's a good read on what's going on
13:27	AWizzArd	inc should return a number
13:27	cshell	inc returns a function?
13:27	cshell	,(inc 1)
13:27	wingy	i mean inc alone
13:27	AWizzArd	Not clojure.core.inc
13:27	cshell	inc is a function
13:27	wingy	so i guess the last one should run the function?
13:28	AWizzArd	(-> 1 inc) ==> (inc 1) ==> 2
13:28	gfredericks	wingy: -> is a macro that works syntactically; it doesn't know if something is a function or not
13:29	AWizzArd	-> takes the first argument and puts it on position two in the following list
13:29	AWizzArd	(-> 1) take the one and return it
13:29	AWizzArd	(-> 1 inc) take the one and “put it into inc”
13:30	AWizzArd	(-> 1 inc dec) take the one and “put it into inc”. Take that result again and put it into dec.
13:30	AWizzArd	(-> 1 inc dec (* 10) ) take the one and “put it into inc”. Take that result again and put it into dec. The result from that put as first arg into (* 10), making it (* 1 10)
13:31	gfredericks	,(macroexpand '(-> 1 inc dec (* 10)))
13:31	gfredericks	&(macroexpand '(-> 1 inc dec (* 10)))
13:31	lazybot	⇒ (* (clojure.core/-> (clojure.core/-> 1 inc) dec) 10)
13:31	gfredericks	&(macroexpand '(-> 1 inc dec (partial * 10)))
13:31	lazybot	⇒ (partial (clojure.core/-> (clojure.core/-> 1 inc) dec) * 10)
13:32	gfredericks	oh those are not too helpful
13:32	gfredericks	&(clojure.walk/macroexpand-all '(-> 1 inc dec (* 10)))
13:32	lazybot	⇒ (* (dec (inc 1)) 10)
13:32	gfredericks	&(clojure.walk/macroexpand-all '(-> 1 inc dec (partial * 10)))
13:32	lazybot	⇒ (partial (dec (inc 1)) * 10)
13:32	wingy	i see
13:35	wingy	makes more sense now that i know -> is a macro
13:36	wingy	it gives a total new syntax
13:36	wingy	so if the 2nd form is a function it just wraps that one around () so it will be called with the value
13:36	wingy	(fn1 val)
13:36	gfredericks	$google github swiss-arrows
13:36	lazybot	[rplevy/swiss-arrows · GitHub] https://github.com/rplevy/swiss-arrows
13:37	AWizzArd	wingy: you can think of it as „bash pipe”
13:37	wingy	but if its already in a run form (fn1 arg2) it will put the val into it as the 1st arg (fn1 val arg2)
13:37	wingy	i thought that every fn should return itself but that wasnt the case
13:37	AWizzArd	take a thing t1 and pass it into the next function to get a thing t2. Take t2 and pass it into the next function to get thing t3. Take t3 and pass it into the next function, etc…
13:38	AWizzArd	A fn does typically not return itself.
13:38	wingy	but it could in this case with inc
13:38	AWizzArd	The symbol clojure.core.inc when evaled returns a function.
13:38	AWizzArd	inc is a symbol
13:38	wingy	i know
13:38	AWizzArd	no a function
13:39	AWizzArd	not
13:39	wingy	this was the confusion: (-> 1 inc (inc) ((partial * 3)))
13:39	AWizzArd	yes
13:39	wingy	diff between inc and (inc) .. they both work
13:39	bpr	does anyone know how feature-ful nrepl/emacs is compared to slime/swank?
13:39	gfredericks	that's because -> supports both
13:39	wingy	but if you have args like in partial you have to wrap it around ()
13:39	gfredericks	it converts inc into (inc) before doing anything else
13:39	wingy	so it will know its one fn with args
13:39	wingy	yea
13:39	AWizzArd	Yes, the -> macro is nice and allows you to omit the parens, if the fn takes only one arg.
13:40	wingy	confusion gone :)
13:41	gfredericks	&((apply comp (repeat 10000 inc)) 1)
13:41	lazybot	⇒ 10001
13:41	gfredericks	lazybot: move this pile of bricks to the other side of the yard
13:41	gnarmis	maybe it needs sudo in there?
13:42	gfredericks	&(format "%s is not in the sudoers file. This incident will be reported." "gfredericks")
13:42	lazybot	⇒ "gfredericks is not in the sudoers file. This incident will be reported."
13:42	gnarmis	hahaha
13:49	wingy	reducers are really hard to get
13:50	aphyr	Resurrecting a very old question: does anyone know why clojure.lang.Var is not backed by an InheritableThreadLocal?
13:51	aphyr	Seems like in almost every case where using dynamic bindings, I really want bound-fn/future's ability to carry the bindings onto threads I might spawn.
13:55	clojure-newcomer	hey guys, its a bit quiet over in the #datomic channel, can anyone help with an ec2 DDB query ?
14:04	wingy	data structures imply the collections and not the literals/symbols?
14:05	cshell	wingy: what do you mean?
14:07	wingy	when someone says data structure he means a collection (either a list, vector, map or set) and not a literal (string, boolean, number, keyword and character) right?
14:12	uvtc	wingy: "data structure" would also mean collection of collection ...
14:12	wingy	uvtc: right but never literals right?
14:13	uvtc	wingy: I think of strings, booleans, numbers, keywords, etc. as "scalars".
14:13	wingy	literls is another word for scalars?
14:14	uvtc	wingy: note that `{:a 1 :b 2}` is a map literal.
14:14	wingy	i see
14:14	wingy	then the word scalar is better to use i think
14:14	wingy	scalar vs data structures
14:14	uvtc	wingy: As opposed to (hash-map :a 1 :b 2), which is not a map literal.
14:16	daniel___	how can i calculate a sequence 0000 0001 0010 0100 1000 0011 0101... all possibilities
14:17	uvtc	wingy: Heh. It's difficult for me to put into words the definition of "literal". It means, "You can tell what this is by looking at it --- it literally is what I just typed" :)
14:17	daniel___	i suppose its binary increments, which i can just convert to a string
14:17	gfredericks	daniel___: ##(let [bs [0 1]] (for [a bs b bs c bs d bs] (str a b c d))) could work
14:17	lazybot	⇒ ("0000" "0001" "0010" "0011" "0100" "0101" "0110" "0111" "1000" "1001" "1010" "1011" "1100" "1101" "1110" "1111")
14:18	wingy	uvtc: yeah thats a good perspective
14:18	uvtc	,(for [i (range 2) j (range 2) k (range 2) m (range 2)] (str i j k m))
14:19	uvtc	,(println (for [i (range 2) j (range 2) k (range 2) m (range 2)] (str i j k m)))
14:19	uvtc	&(for [i (range 2) j (range 2) k (range 2) m (range 2)] (str i j k m))
14:19	lazybot	⇒ ("0000" "0001" "0010" "0011" "0100" "0101" "0110" "0111" "1000" "1001" "1010" "1011" "1100" "1101" "1110" "1111")
14:19	daniel___	oh lovely
14:20	uvtc	Whoops. Mine was the same as gfredericks , only I kept repeating myself. :)
14:20	uvtc	clojurebot: are you there?
14:20	uvtc	lazybot: what happened to clojurebot?
14:20	lazybot	It's AWWWW RIGHT!
14:22	gfredericks	&(for [i (range 16)] (subs 1 (Long/toString (+ 16 i) 2)))
14:22	lazybot	java.lang.ClassCastException: java.lang.Long cannot be cast to java.lang.String
14:22	gfredericks	&(for [i (range 16)] (subs (Long/toString (+ 16 i) 2) 1))
14:22	daniel___	java.lang.Long/toBinaryString?
14:22	lazybot	⇒ ("0000" "0001" "0010" "0011" "0100" "0101" "0110" "0111" "1000" "1001" "1010" "1011" "1100" "1101" "1110" "1111")
14:22	gfredericks	I win
14:22	uvtc	kahnnnnnn!
14:23	hyPiRion	##(for [i (range 16)] (clojure.pprint/cl-format nil "~4,'0B" i))
14:23	lazybot	⇒ ("0000" "0001" "0010" "0011" "0100" "0101" "0110" "0111" "1000" "1001" "1010" "1011" "1100" "1101" "1110" "1111")
14:24	lazybot	⇒ this syntax nil
14:24	gfredericks	is it worth going to the conj if I'm also going to strange loop?
14:26	daniel___	the number of permutations when everything is distinct is factorial
14:26	daniel___	in this case, what is it?
14:27	hyPiRion	exponential
14:27	gfredericks	2^n
14:27	gfredericks	or m^n
14:27	uvtc	&(for [j (range 16)] (apply str (for [i (range 4)] (rand-int 2))))
14:27	lazybot	Execution Timed Out!
14:28	uvtc	Was going to joke, "close enough". Odd that it timed out though...
14:29	hyPiRion	uvtc: Clojail is weird like that.
14:29	uvtc	&(for [j (range 16)] (apply str (for [i (range 4)] (rand-int 2))))
14:29	lazybot	Execution Timed Out!
14:30	uvtc	hyPiRion: Ok. :)
14:31	hyPiRion	&(for [i (range 16)] (apply str (repeatedly 4 #(rand-int 2))))
14:31	lazybot	⇒ ("0000" "1000" "0101" "0101" "1100" "1011" "0101" "1100" "1111" "0101" "1110" "1000" "0011" "0010" "0111" "0011")
14:31	hyPiRion	It's something about the for comprehension I believe.
14:33	gfredericks	&(frequencies (repeatedly 16 (rand-int 16)))
14:33	uvtc	Ah. `repeatedly`. Thanks, hyPiRion.
14:33	lazybot	java.lang.ClassCastException: java.lang.Integer cannot be cast to clojure.lang.IFn
14:33	gfredericks	&(frequencies (repeatedly 16 #(rand-int 16)))
14:33	lazybot	⇒ {2 1, 3 2, 4 3, 7 1, 8 3, 11 2, 12 1, 13 1, 15 2}
14:50	daniel___	what are the alternatives to associative arrays? i suppose nested vectors is one way, what about if you want a non integer key?
14:50	daniel___	a map of lists?
14:50	daniel___	a map of vectors*?
14:52	gfredericks	vectors are somewhat like maps with keys of (range n); so in that sense maps will always be more flexible. I don't quite understand the details of your question.
14:52	gfredericks	I must not know what an associative array is.
14:53	cshell	I'm confused too
14:56	yonatane	Is there a use-case for using protocols when java interop is not needed?
14:57	gfredericks	yes
14:58	gfredericks	actually I don't think protocols have very much at all to do with interop
14:58	gfredericks	only that you can extend a protocol to a java type
15:02	tmundal	(+ 1 2)
15:02	gfredericks	3
15:02	tmundal	:D
15:03	gfredericks	:D
15:05	yonatane	gredericks: do you have an example?
15:06	gfredericks	sorta
15:06	yonatane	Can I define a protocol with common function names like doseq, and implement it in some random crazy way, and so existing code would behave crazily when passed my implementation?
15:07	tmundal	(loop [] (drink-more-beer) (recur))
15:07	gfredericks	the function names that belong to your protocol are vars in the namespace you define them in. So (ns foo.bar) (defprotocol FooBar (doseq [x y])) will just create foo.bar/doseq, which is not the same as clojure.core/doseq
15:08	gfredericks	so...no
15:08	yonatane	so an unqualified doseq outside my namespace will behave normally
15:08	gfredericks	yep
15:08	yonatane	that's no fun
15:09	gfredericks	you can redefine doseq directly if you just really want clojure to crash
15:09	aphyr	daniel___: Why can't you use a map?
15:09	gfredericks	but since it's a macro you'll only effect code that's loaded after you redefine it
15:09	yonatane	oh, doesn't matter. any other function then.
15:10	bpr	hey, i'm using lein2 and i want to do lein2 new foo and have the files/directories dumped into my current dir. I used to be able to do lein new foo . with v1.7.0, is there a way to do this in lein2?
15:10	gfredericks	yonatane: with-redefs will let you redefine stuff temporarily
15:10	gfredericks	yonatane: the only protocol example I care to bother dredging up is https://github.com/fredericksgary/ranguages/blob/master/src/ranguages/core.clj#L10
15:10	bpr	i clearly could just use a tmp dir and copy everything over, but why take the extra step if i don't have to :)
15:12	daniel___	sorry, something came up
15:12	daniel___	aphyr: i can use a map :)
15:13	uvtc	bpr: I think `lein new foo` has always created the "foo" directory.
15:13	uvtc	bpr: At least, I've never seen it do anything different from that.
15:14	yonatane	gfredericks: if you've done it without a protocol, how would it be worse?
15:14	daniel___	im used to thinking of everything in loops and arrays
15:15	gfredericks	yonatane: well you can't really use types/records without protocols; so I would've been using raw maps, and my single-location implementation would have a giant cond checking the types of the maps
15:15	gfredericks	it gives you the interface/implementation separation that a lot of people think is clean but I'm not smart enough to argue convincingly for
15:17	yonatane	i'm coming for OO, and had the impression that a functional language does not need OO features
15:17	yonatane	*coming from
15:18	gfredericks	some people find some of the features useful. I don't think it's expected that you use protocols/records/types for most application-level stuff
15:18	gfredericks	they were first introduced to make clojure-in-clojure feasible
15:18	gfredericks	certainly clojurescript makes heavy use of them in its foundational stuff
15:18	gfredericks	as does core.logic
15:19	bpr	uvtc: "lein new foo ." used to put the new files in the current dir
15:19	gfredericks	I think for regular business data maps are still recommended
15:19	bpr	uvtc: this is because in v1.7.0 the third parameter was an optional one used to specify the directory to put the new project
15:19	uvtc	bpr: Oh. Didn't know that. Thanks.
15:20	bpr	yup
15:20	cshell	has anyone ever seen the IllegalArgumentException: Can't define method not in interfaces. when used in protocols?
15:22	yonatane	gfredericks: how about when you want a common persistency protocol, where you don't know if it will go to mysql or to text files, would you use the protocol mechanism?
15:22	gfredericks	cshell: you're just defining a protocol and don't have any types/records yet?
15:23	gfredericks	yonatane: that sounds reasonable
15:23	gfredericks	yonatane: I think I've done that before
15:23	cshell	I have a defrecord for the protocol
15:23	cshell	implementation
15:23	yonatane	gfredericks: any other idiomatic way to do it?
15:23	cshell	I've done a lein clean
15:23	gfredericks	yonatane: not really
15:23	cshell	i found something that said to disable don't-recompile-if-same-source but I don't know where that is
15:24	gfredericks	yonatane: protocols are also great when you want something to be extensible
15:24	daniel___	ach, earlier i meant multidimensional arrays, not associative arrays
15:24	gfredericks	daniel___: I think I've used flat maps and nested vectors in different contexts depending on exactly what sort of interface I wanted
15:24	cshell	gfredericks: it's the same problem as listed here: http://bit.ly/MpC3hL
15:25	gfredericks	ah ha
15:25	gfredericks	but this is presumably fixed and you're still having the problem?
15:25	cshell	yep
15:26	cshell	although a clean and repl restart isn't working
15:26	cshell	so I can't get past it
15:26	cshell	where else would lein put the compiled protocol def?
15:26	gfredericks	cshell: what's your protocol def look like?
15:26	daniel___	gfredericks: ok
15:26	yonatane	gfredericks: i recently asked how to implement an api that requires a client token throughout, and i didn't want all the functions to have the token as a parameter.
15:27	cshell	gfredericks: https://www.refheap.com/paste/3609
15:27	daniel___	im trying to write a chess engine, and generate a table of file attacks: http://chessprogramming.wikispaces.com/Rotated+Bitboards (file_attacks[R][10011011])
15:27	daniel___	in this case, i suppose i want a vector of maps
15:28	cshell	it's complaingin about the update! method, that's what I just added today
15:28	gfredericks	yonatane: ...so you're asking the question again?
15:28	daniel___	[{:10010011 <attacks>}{:10100000 <attacks>}..]
15:28	cshell	yonatane: you can create a closure with the token bound
15:28	daniel___	[{:10010011 <attacks> :10100000 <attacks>}{}{}..] in fact
15:28	yonatane	gfredericks: Yeah, in light of records :). maybe a record with the token as ctor argument
15:29	gfredericks	daniel___: what does each map signify?
15:29	daniel___	the ocupancy on a given file or rank
15:29	gfredericks	yonatane: I think that kind of thing is normally handled with dynamic vars
15:29	daniel___	the key signifies that, and it maps to a bitboard with the possible attack squares
15:29	yonatane	cshell: yeah, but if i have several functions, do i create each one, or a map of functions, or some other thing?
15:29	gfredericks	daniel___: so why do you have multiple maps instead of one?
15:30	daniel___	(file_attacks[R][10011011]) the first dimension is the square that a rook is on for example, and inside that element i have a map of all the occupancies
15:30	gfredericks	cshell: looks like the sig for update! doesn't match
15:31	cshell	gfredericks: omg, that's it
15:31	cshell	thanks g
15:31	cshell	slaps forehead
15:31	cshell	yonatane: what I do is make the protocol with the methods and parameters (but not the token)
15:31	cshell	then I create a deftype which has a constructor that has the token
15:32	daniel___	so it will be 8 maps in a vector i suppose [{}{}{}{}{}{}{}{}]
15:32	daniel___	the index signifying the square that the attacking piece is on
15:32	daniel___	and then inside each map, i have an occupancy mapped to an attack board
15:33	gfredericks	cshell: doesn't that mean you have to pass around a type instead of a token?
15:33	daniel___	dont have a clue how im going to generate such a structure
15:33	gfredericks	daniel___: why 8 maps instead of 64 then?
15:34	gfredericks	since there are 64 squares
15:34	daniel___	because im just concentrating on a rank or file
15:34	cshell	I just call a factory method when I need the functionality and then pass that in as the this parameter on the protocols
15:34	gfredericks	cshell: right. So you have to pass in something, whether the type or the token; I can't see why that's any easier
15:34	daniel___	the attacks are going to be the same for each rank or file, given the same occupancy
15:36	yonatane	gfredericks: I should read about bindings more. I asked before about a case where i have multiple tokens to handle. Maybe I can change the binding of the token var for each one. I wonder how it handles concurrency.
15:36	cshell	yonatane: dynamic vars can get thread local bindings - so that helps with concurrency
15:36	gfredericks	yonatane: dynamic vars have thread-local values
15:36	yonatane	that's great
15:37	cshell	gfredericks: Hmm, that's a good poing gfredicks - I always did it because it took that off of the interface - the this has to be there anyway
15:37	cshell	gfredericks: and maybe I have a stubbed implementation or something which doesn't need the token to return values
15:37	yonatane	so I can call a bunch of functions with a specific value of the var.
15:37	yonatane	setting the var only once
15:37	gfredericks	yonatane: ofter the dynamic var is accompanied by a with-foo macro that sets it for you
15:38	gfredericks	s/ofter/often/
15:38	gfredericks	yonatane: e.g., clojure.whatsit.jdbc's with-connection
15:38	yonatane	gfredericks: exactly!
15:39	gfredericks	it's a mildly controversial tactic since it kills function purity. but you can only put up with so much ceremony in the name of purity
15:39	gfredericks	also if you use a monad people stop being your friend
15:39	cshell	haha
15:40	gfredericks	on the other hand, some times you can do similar things quite succinctly using ->
15:42	uvtc	Regarding the url-shortener example in the "Clojure Programming" book (Oreilly), (page 537),
15:42	uvtc	~CPOR
15:42	uvtc	it uses a Ref to refer to its `mappings` data structure (mappings of string id's to long urls).
15:42	uvtc	Why does it use a Ref? My understanding is that refs are for when you want to coordinate changes to multiple refs that are connected or interdependent in some way. But here we've only got one ref; the mapping. It doesn't need to coordinate with any other data structures.
15:42	uvtc	Why not instead use an atom here?
15:44	gfredericks	I THINK that there is a use case for a single ref which is
15:44	gfredericks	you want to read it, update it based on what you read, and also do something else with the read value
15:44	gfredericks	where on earth did I see this
15:44	gfredericks	it was in one of hiredman's codez
15:45	gfredericks	uvtc: imagine you wanted to use an atom to implement a queue
15:45	gfredericks	and have worker threads pulling stuff off to work on
15:45	gfredericks	think about how (defn pull-thing-off-queue [] ...) would be defined
15:46	uvtc	Ah. That's what's going on in the example I noted. You're checking the contents of `mappings`, and if a given value isn't yet in it, you assoc it.
15:46	gfredericks	well that sounds doable actually
15:46	gfredericks	with an atom
15:46	gfredericks	unless you're trying to return info on whether it was there or not
15:47	uvtc	The check to see what the mapping contains, and also the alteration, are both inside a dosync.
15:47	gfredericks	(swap! a #(if (% k) % (assoc % k v)))
15:47	gfredericks	I'm not sure why ^ that wouldn't work
15:48	uvtc	gfredericks: Will think on that. Thanks.
15:48	gfredericks	next time cemerick is around ask him what on earth
15:49	uvtc	Will do.
15:50	wingy	what is the :test meta doing in the meta data of a function?
15:50	gfredericks	pointing to a test?
15:51	wingy	whe is it executed?
15:51	wingy	when
15:51	wingy	manually by user?
15:52	gfredericks	or by any framework that looks for that sort of thing
15:52	gfredericks	I dunno which ones do
15:52	wingy	ok
15:56	dgrnbrg	is it possible to have a type hint for an array of IFns?
15:57	dgrnbrg	I tried ^"[Lclojure.lang.IFn", but it said classnotfound
15:57	gfredericks	I don't remember the syntax but I'm sure it is
15:57	dgrnbrg	i can't find it anywhere on google
15:57	dgrnbrg	i found the syntax for primitive arrays
15:58	gfredericks	$google clojure array type-hints
15:58	lazybot	[Array Type Hints in Clojure] http://asymmetrical-view.com/2009/07/02/clojure-primitive-arrays.html
15:58	dgrnbrg	that post only talk about primitive arrays
15:58	gfredericks	it says primitive but it looks like he has an example with String
15:59	dnolen	^"[Lclojure.lang.IFn;"
15:59	daniel___	with a vector like [\X \X \O \X \O \O \X \X \X], how would i iterate over it and return two sets X #{0 1 3 6 7 8} & O #{2 4 5} with the indexes?
15:59	dnolen	,(type (make-array clojure.lang.IFn 2))
16:00	dgrnbrg	oh, semicolon
16:00	dgrnbrg	i missed that
16:00	gfredericks	&(doc map-indexed)
16:00	lazybot	⇒ "([f coll]); Returns a lazy sequence consisting of the result of applying f to 0 and the first item of coll, followed by applying f to 1 and the second item in coll, etc, until coll is exhausted. Thus function f should accept 2 arguments, index and item."
16:00	dgrnbrg	hooray!
16:00	daniel___	map-indexed, thats the one
16:00	dgrnbrg	it was the semicolon
16:03	dgrnbrg	i just got a job writing clojure all day
16:03	dgrnbrg	it's pretty awesome
16:03	gfredericks	that sounds pretty awesome
16:05	dgrnbrg	i get to integrate things like storm, cascalog, leiningen, and incanter with our java code base
16:07	eggsby	sounds like a lot of work
16:07	eggsby	but congrats :)
16:10	uvtc	dgrnbrg: nice. New job, or current one switching you to working on Clojure stuff?
16:10	dgrnbrg	it's actually the same job
16:10	dgrnbrg	i was previously working on our in-house distributed analysis framework
16:10	dgrnbrg	i made the case that clojure brings a lot to the table
16:11	dgrnbrg	so we're building an experimental system in clojure to iterate faster and explore the possibilities
16:11	uvtc	Hooo doggie. :)
16:11	dgrnbrg	that's built on our analysis framework
16:11	dgrnbrg	it's pretty awesome
16:11	dgrnbrg	we write emacs plugins, compilers, and think about new modes of interaction
16:11	dgrnbrg	like light table, but in our domain
16:13	eggsby	nice
16:13	eggsby	I'm working on some analytics/continuous delivery stuff in clj for my work, a lot more fun than maintaining old groovy and spring mvc apps...
16:14	antoineB	is there a function to get the number of a list ? example: (get 3 '(a b c d)) => c
16:15	uvtc	antihero: `nth`
16:15	dnolen	antoineB: nth
16:15	dgrnbrg	,(nth 2 '(a b c d))
16:15	uvtc	Whoops. Sorry antihero --- premature autocomplete. :)
16:15	dgrnbrg	&1
16:15	lazybot	⇒ 1
16:15	antoineB	thanks
16:15	dgrnbrg	&(nth 2 '(a b c d))
16:15	lazybot	java.lang.ClassCastException: clojure.lang.PersistentList cannot be cast to java.lang.Number
16:15	dgrnbrg	&(nth '(a b c d) 2)
16:15	lazybot	⇒ c
16:20	antoineB	a function for the size or lenght of a list?
16:21	eggsby	count
16:21	piotr	antoineB: you should have a look at http://clojure.org/cheatsheet
16:21	eggsby	you don't use get and count as often as you'd think antoineB
16:22	gfredericks	$findfn [1 2 3 40] 4
16:22	lazybot	[clojure.core/count]
16:22	eggsby	they aren't used as often as .length and someArray[nth]
16:22	gfredericks	,(.length '(a b c d))
16:22	gfredericks	&(.length '(a b c d))
16:22	lazybot	java.lang.IllegalArgumentException: No matching field found: length for class clojure.lang.PersistentList
16:23	antoineB	what is the difference with .length ?
16:23	eggsby	rather, as often as .length and array accessors in java/imperitive langs :p
16:23	eggsby	imperative even
16:23	antoineB	ok
16:23	wingy	when creating a var with def .. is that var called a variable?
16:23	wingy	or just a var
16:24	gfredericks	&(.length '[a b c d])
16:24	lazybot	⇒ 4
16:24	gfredericks	wingy: a var
16:24	wingy	there are no variables in clj?
16:24	gfredericks	not local variables
16:24	wingy	what is a global variable?
16:25	wingy	atom, ref, var and agent?
16:25	gfredericks	well the changey things that exist are
16:25	aphyr	gfredericks: what about let frames?
16:25	uvtc	gfredericks: it seems to me that there are variables in Clojure, except that they're not called "variables" per se, and you don't vary them in-place, but rather modify them to refer to different values when you need to change them.
16:25	eggsby	ya, I like calling them values instead of variables
16:25	gfredericks	uvtc: how do you do int x = 12; x = 13;?
16:26	antoineB	(def x 12)
16:26	uvtc	gfredericks: I suppose create an atom with the value 12, then later swap in 13 for it.
16:26	eggsby	lol, (let [x 12] (let x [13] ;; do stuff))
16:26	eggsby	it's beautiful
16:26	gfredericks	eggsby: those are two different x's
16:26	gfredericks	the first one doesn't change
16:26	eggsby	right
16:26	gfredericks	so clojure has explicit reference types
16:27	gfredericks	and the java types can have mutable fields
16:27	wingy	(i think i shall just forget the word variable)
16:27	gfredericks	but you cannot create a locally-scoped name that points to one thing and then change it to point to another thing
16:27	wingy	references are bound to values if you wanna simulate changing states/values
16:28	eggsby	ya
16:28	eggsby	well you were right stuff stuff like 'atom' is closer to a var wingy
16:28	daniel___	clojure makes my brain hurt
16:28	gfredericks	ruby makes my fingers hurt
16:29	wingy	another question regarding a comment i read in the reference regaring descructuring "There are similar :strs and :syms directives for matching string and symbol keys."
16:29	eggsby	ya
16:29	uvtc	daniel___: feel the burn. Only a few ... more ... reps.
16:29	eggsby	so a hashmap can have anything as its key wingy
16:29	eggsby	like strings or symbols
16:29	eggsby	{"myString" 1, 'mySym 2}
16:29	antoineB	perl makes me confuse
16:29	eggsby	you could get out strings with :strs or symbs, similar to how you would get :keywords out with :keys
16:30	wingy	eggsby: ah .. i think i know what the problem was with :syms .. i didnt quote
16:30	wingy	yepp that was the problem
16:30	wingy	eggsby: thx
16:30	uvtc	gfredericks: But do you see what I mean? I think it's confusing to say that Clojure doesn't have variables. Seems more accurate to say that it has no local variables, but it's got "global variables" --- with the caveat that you can't modify them in-place; you instead change which value they refer to.
16:31	dgrnbrg	does anyone know how to get lein to stop saying "Could not find artifact org.clojure:clojure:pom:1.+ in central (http://repo1.maven.org/maven2)" ...
16:31	wingy	clojure makes my eyes, heart and brain happy
16:31	antoineB	clojure have name wich refer to an object
16:31	uvtc	antoineB: funny thing about Perl though, it seems to have at least one thing in common with Clojure: practicality.
16:31	wingy	i have become a happier person overall since 2 weeks ago
16:32	wingy	thats weird a lang can have that effect on you
16:32	gfredericks	uvtc: the question doesn't mean too much until you rigorously define variable. I think the concept I was going for is what most people mean when asking such a question.
16:32	akhudek	uvtc: you wouldn't say that if you've ever tried to build a relatively large code base in perl that remotely needs nice data structures
16:33	uvtc	akhudek: don't worry --- I stop the comparison at that. :)
16:33	eggsby	uvtc: it's fine so long as you acknowledge that global mutable state isn't bad so long as it isn't concurrently muteable, which clj enforces
16:34	eggsby	s/mutea/muta
16:34	wingy	off topic question: is there anyone here from Stockholm?
16:35	gfredericks	eggsby: it can still be bad to the extent that it keeps you from using mostly pure functions
16:35	antoineB	uvtc: in perl you loose so many time to learn the many syntaxicaly ways of doing something instead of learning the different semantics ways
16:35	eggsby	gfredericks: I haven't heard a compelling argument about why pure is better than impure, is there anything I should read? :p
16:36	Squee-D	Side effects are the hardest things to debug?
16:36	gfredericks	bam
16:36	wingy	eggsby: easier to test since all the thing a pure fn need is passed as args
16:36	gfredericks	yeah, for debugging and testing, pure functions are huge
16:36	Squee-D	and im new to lisp, but thats what 'compels' me to learn it.
16:38	wingy	eggsby: i think it's also better for concurrency since now you can run that fn not caring about place and time it is run
16:38	eggsby	ya but i'm not going to start using the state monad in exchange for an atom
16:38	gfredericks	so there's this piece of code I wrote that's in production, but I'm not actively working on it. The ops guy says it crashes, and is crashing in a part of the code that is pure functions. So rather than doing all sorts of poking around on production, I just have him send me a log of the input to the pure part of the code, and I have a fully reproducible failing test in 2 seconds.
16:38	Squee-D	Otherwise you may as well be using scala :D
16:38	eggsby	I don't see practical benefit, your transformation functions can still be pure functions...
16:38	wingy	fns dependent on global states is spaghetti
16:38	wingy	bad abstractions
16:38	gfredericks	eggsby: I didn't say all pure, I said mostly pure
16:39	Squee-D	you cant do all pure
16:39	daniel___	im hating this function but im struggling to come up with something better: https://gist.github.com/3113254
16:39	Squee-D	lol
16:39	daniel___	can anyone think of a clever way of detecting a winner (and even better, detecting the player \X or \O) in a game of tictactoe
16:40	daniel___	the board is a vector of length 9: [\space \X \space \O ... ]
16:40	antoineB	i never understand the monad stuff
16:41	antoineB	is it a way to force the programmer to test all the possble case?
16:41	antoineB	like Option[] in scala?
16:41	eggsby	it's just a way to capture the entire notion of a computation antoineB
16:41	gfredericks	daniel___: (def in-a-rows (let [idx (partition 3 (range 9))] idx (map list idx) [[0 4 8] [2 4 6]]))
16:41	gfredericks	^ then use that
16:42	gfredericks	antoineB: just figuring out what the possible cases are is rather difficult in a dynamic language
16:42	antoineB	what is "the entire notion of a computation"?
16:42	gfredericks	qubits!
16:43	eggsby	antoineB: http://www.scribd.com/doc/72134705/Comprehending-Monads can explain monads far better than I ever could
16:44	daniel___	gfredericks: not sure what that does exactly
16:44	daniel___	,(def in-a-rows (let [idx (partition 3 (range 9))] idx (map list idx) [[0 4 8] [2 4 6]]))
16:44	daniel___	,in-a-rows
16:45	gfredericks	&(let [idx (partition 3 (range 9))] idx (map list
16:45	gfredericks	idx) [[0 4 8] [2 4 6]])
16:45	lazybot	java.lang.RuntimeException: EOF while reading, starting at line 1
16:45	gfredericks	&(let [idx (partition 3 (range 9))] idx (map list idx) [[0 4 8] [2 4 6]])
16:45	lazybot	⇒ [[0 4 8] [2 4 6]]
16:45	gfredericks	mrrrrmrmmm
16:45	gfredericks	ah I forgot the concat
16:45	gfredericks	&(let [idx (partition 3 (range 9))] (concat idx (map list idx) [[0 4 8] [2 4 6]]))
16:45	lazybot	⇒ ((0 1 2) (3 4 5) (6 7 8) ((0 1 2)) ((3 4 5)) ((6 7 8)) [0 4 8] [2 4 6])
16:45	antoineB	eggsby: espected on irc the font isn't so uggly :)
16:45	gfredericks	man I really screwed that up
16:45	daniel___	:/
16:46	gfredericks	&(let [idx (partition 3 (range 9))] (concat idx (apply map list idx) [[0 4 8] [2 4 6]]))
16:46	lazybot	⇒ ((0 1 2) (3 4 5) (6 7 8) (0 3 6) (1 4 7) (2 5 8) [0 4 8] [2 4 6])
16:46	gfredericks	okay that's what I meant
16:46	gfredericks	then you can write functions that iterate over all 3-in-a-rows without thinking about what that structure actually is
16:48	daniel___	why are the last two vectors
16:48	daniel___	the diagonals
16:48	gfredericks	cuz I gave them explicitly; isn't supposed to make a diff
16:49	daniel___	i cant think hwo to iterate over them, without using a load of if/or/and statements like before
16:50	gfredericks	(some (fn [[a b c]] (= (board a) (board b) (board c))) three-in-rows)?
16:51	daniel___	what about detecting whether it's a \X or a \O
16:52	gfredericks	use one of the indices in the result of that expression?
16:52	gfredericks	depends on what kind of output you're looking for exactly
16:53	daniel___	i guess either \O \X or nil
16:53	daniel___	then i can test for not nil
16:53	daniel___	and whatever is returned is the winner
16:53	daniel___	i lack a serious amount of creativity or something
16:54	daniel___	how do you arrive at the point where you can see a good solution to every problem? just doing loads of 4clojure type challenges
16:54	daniel___	?
16:54	gfredericks	(some (fn [[a b c]] (if (= (board a) (board b) (board c)) (board a))) three-in-rows)
16:54	emezeske	daniel___: Nobody can see a good solution to every problem ^_^
16:55	gfredericks	daniel___: I can see a good solution to these kind of silly problems because I've done them over and over.
16:55	wingy	i wanna get to that point i can read code very fast
16:55	wingy	to be thinking in code
16:56	emezeske	wingy: There's only one way to do that, AFAIK: read more code. :)
16:56	wingy	yeah
16:56	gfredericks	daniel___: where by "silly" I mean "not directly practical"
16:56	akhudek	wingy: think in datastructures, not code
16:56	wingy	akhudek: right .. everything is data
16:56	wingy	that's what you wanna say?
16:56	akhudek	pretty much
16:57	akhudek	I always think in trees, lists, vectors, etc
16:57	akhudek	translates to many languages easily
16:57	gfredericks	daniel___: https://github.com/fredericksgary/tictactoe/blob/master/src/ttt/core.clj#L69
16:57	akhudek	functional languages just make it easier to directly implement this kind of thinking
16:57	akhudek	e.g. "transform each element in the list with function X"
16:57	wingy	akhudek: one of the fundamentals of FP is to manipulate these data structures right
16:57	wingy	i read somewhere that algorithms are overrated
16:58	akhudek	well, algorithms are not overrated
16:58	wingy	most of the things you do is to manipulate data
16:58	akhudek	but to create good algorithms you need to think about data structures
16:59	emezeske	wingy: I think the "algorithms are overrated" thing means something to the effect of "once you have the right datastructure, the right algorithm is much more obvious"
16:59	antoineB	akhudek: algorithm go with datastructure
16:59	akhudek	that's exactly right
17:00	antoineB	there is no datastructure on a side and algorithm on another side
17:00	emezeske	wingy: But the reality is that datastructures and algorithms are really just two aspects of the same thing
17:00	gfredericks	datastrorithms
17:00	wingy	hmm interesting
17:00	wingy	they go hand in hand
17:00	emezeske	gfredericks: haha
17:00	antoineB	algotucture
17:01	akhudek	and they are language independent too
17:01	akhudek	for the most part
17:01	wingy	emezeske: what do you mean two aspects of the same thing
17:01	hyPiRion	emezeske: I'd say algorithms are "operations" on datastructures.
17:01	akhudek	obviously language can make particular algorithms and data structures easier or harder to write (or sometimes impossible)
17:01	gfredericks	never impossible!
17:01	wingy	hyPiRion: that seems fair
17:02	wingy	yeah .. its not a lot of fun to do FP in javascript
17:02	daniel___	my version so far: https://github.com/danielstockton/naughts-and-crosses/blob/master/src/naughts_and_crosses/core.clj
17:02	emezeske	wingy: Well, I guess I just mean that some algorithms can't be described without describing their underlying datastructure
17:02	wingy	; function {} nightmare
17:02	wingy	i realized clj is so concise so it really fits very well as a cli lang
17:02	antoineB	hyPiRion: flat data itself a datastructure
17:03	antoineB	*is itself
17:03	gfredericks	wingy: the underscore lib helps slightly
17:03	wingy	gfredericks: yeah .. but i ended up switching to livescript
17:03	cheater__	but is the data structure data?
17:03	arohner	is there a built-in fn for casting a byte to an unsigned byte? i.e. (unsigned-byte -106) => 153
17:03	antoineB	its not a lot of fun to do javascript
17:03	wingy	a better coffeescript with FP features
17:03	gfredericks	$findfn -106 153
17:03	lazybot	[]
17:04	gfredericks	wingy: but now you're using cljs?
17:04	cheater__	$findfn 153 -106
17:04	lazybot	[]
17:04	cheater__	interesting.
17:04	antoineB	cheater__: it's meta data or not :)
17:04	akhudek	arohner: java doesn't have unsigned types :-(
17:04	antoineB	or meta meta data
17:04	wingy	gfredericks: once i mastered server side with clj/datomic i plan to use cljs for frontend
17:04	akhudek	I think there are some convenience functions to simulate them anyways though
17:05	arohner	akhudek: I know, but I'd still like to treat it as unsigned, turning into a short or int if necessary
17:05	akhudek	arohner: I'm trying to remember where I read this, maybe it is Java7 that introduced something to handle this
17:06	ivan	Java 8 has some real unsigned ints
17:08	akhudek	ack, sorry, I can't seem to find it, but I remember that it was specific either to 1.7 or 1.8
17:08	daniel___	arohner there is unchecked-byte
17:08	daniel___	oh, hes gone
17:09	hyPiRion	daniel___: that wouldn't convert it to an unsigned byte, though.
17:09	hyPiRion	,(unchecked-byte -106)
17:10	gfredericks	lazybot should watch for , messages and respond if clojurebot doesn't.
17:10	daniel___	hyPiRion: nope
17:10	gfredericks	maybe they should just keep tabs on eachother in the background
17:12	wingy	so once you have mastered all basic stuff in clj and cljs .. how do you ensure that you develop your skills to read and write idiomatic code. should one just pick a random lib in github and read the source code? that sounds boring
17:13	wingy	doing tasks like 4clojure all day?
17:13	gfredericks	use it for stuff and read/participate-in discussions
17:13	antoineB	try to write your own dsl in clojure?
17:13	cheater__	what is cljs?
17:14	wingy	clojurescript
17:14	cheater__	what's that do?
17:14	antoineB	it compile to javascript
17:14	cheater__	neato
17:14	emezeske	wingy: You could always contribute to an open source project, which would force you to read source code but might not be boring due to having a purpose
17:14	cheater__	does it work well in browsers?
17:14	cheater__	or is it mostly for v8 & co?
17:15	wingy	antoineB: DSLs using macro seem fun .. perhaps i could write a DSL for my business domain
17:15	gfredericks	cheater__: it works well in browsers
17:15	emezeske	wingy: Plus if you contribute something useful, you might get a free code review :)
17:15	cheater__	gfredericks: even in ie7?
17:15	wingy	emezeske: yeah :)
17:15	gfredericks	cheater__: I'd be surprised if it didn't
17:16	emezeske	cheater__: CLJS uses the google closure compiler and library, which I'm pretty sure supports really old browsers
17:16	wingy	hope more books will be available .. i like reading books
17:17	wingy	would be cool if they had lynda videos for clojure .. i searched but they arent any
17:17	cheater__	lynda? hahah.
17:17	wingy	cheater_: thats what i used to learn programming back in the days :)
17:18	cheater__	at least it's not w3schools.
17:18	wingy	i just think fun stuff makes the entry easier
17:18	wingy	like head first series
17:18	wingy	so great
17:19	antoineB	i fail the first problem of 4clojure :/
17:19	wingy	like when we have light table ready to rock .. i think it will convert a lot of people to clj
17:19	wingy	the tooling is really important for a lang's survival
17:20	antoineB	anyone who wants do an interactive test in swank? (you change the code, it change the result of your tests)
17:21	cheater__	huh?
17:21	cheater__	antoineB: what problem did you fail at?
17:21	antoineB	write your tests, write your function and have directly the result of your tests
17:22	antoineB	change your function, modify the result and so on
17:22	cheater__	i don't understand :)
17:23	antoineB	when you write, your function you see the tests are executed (using the current function)
17:24	antoineB	and you see the result
17:24	antoineB	you don't need compile/eval + run the test
17:25	cheater__	i'm sorry i see words but i don't know what they refer to
17:25	wingy	im so glad i moved on from php when i read this http://www.slideshare.net/fabpot/design-patternrevisitedphp53
17:25	cheater__	wingy: was going to click, but noticed "php" in url
17:26	gnarmis	anyone else doing a deep dive into emacs? just setup swank and all…things seem to work!
17:26	gnarmis	coming from vim
17:27	gfredericks	I did that 4 months ago maybe
17:27	antoineB	i just do clojure-jack-in
17:27	antoineB	that all
17:31	wingy	wow, never knew about condition-map in fns .. are there good use cases for those?
17:32	wingy	when would you wanna have a certain check in a condition-map instead of putting in the function itself
17:32	gfredericks	when you want the program to crash due to programmer error
17:32	gfredericks	rather than some other sort of exceptional case
17:33	gfredericks	they're the kind of thing you should be comfortable disabling in production
17:34	cheater__	hey guys can someone explain this to me?#(+ % 5)
17:34	cheater__	is % the "unshift"?
17:34	gfredericks	it's the arg
17:34	Raynes	(fn [x] (+ x 5))
17:34	wingy	cheater__: the first arg
17:34	cheater__	what if i have multiple args?
17:34	gfredericks	%1 %2
17:34	wingy	the same as %1
17:34	Raynes	#(+ % %2)
17:34	cheater__	ah k
17:35	cheater__	is #() the same as function definition?
17:35	gfredericks	man we can't resist easy questions apparently
17:35	Raynes	I feel insulted when people don't answer in code.
17:35	gfredericks	cheater__: it's a reader macro; you can use read-string to see what it expands to
17:35	gfredericks	&(read-string "#(+ %1 %2)")
17:35	lazybot	⇒ (fn* [p1__69903# p2__69904#] (+ p1__69903# p2__69904#))
17:35	gfredericks	Raynes: ^ there you go :P
17:36	cheater__	ok so fn defines an anonymous function?
17:36	gfredericks	yeah
17:36	cheater__	what if i do (fn add-five [x] (+ x 5)) ?
17:36	cheater__	is it still anonymous?
17:36	wingy	cheater__: you should pick up the book "Programming Clojure" .. it will walk you through all this stuff
17:36	Raynes	Yes.
17:37	Raynes	You can't access add-five outside the scope of add-five.
17:37	gfredericks	Raynes: say that in code
17:37	cheater__	i can't wait till a book arrives from amazon
17:37	cheater__	i need to know clojure now !
17:37	cheater__	:)
17:37	cheater__	Raynes: thanks
17:37	Raynes	&(do (fn add-five [x] (+ x 5)) (add-five 5))
17:37	lazybot	java.lang.RuntimeException: Unable to resolve symbol: add-five in this context
17:37	Raynes	gfredericks: ^
17:37	cheater__	Raynes: i was wondering why they ever give names to anonymous functions, i guess it's only for recursion?
17:37	gfredericks	Raynes: phew
17:38	wingy	cheater__: there is an ebook i think
17:38	gfredericks	cheater__: alse I think it will leak into the compiled class name, which can be nice
17:38	cheater__	wingy: i'm just going through 4clojure since you guys linked it.
17:38	wingy	cheater__: i think basic knowledge before doing 4clojure is better
17:38	Raynes	&((fn to-five [x] (if (< 5 x) (to-five (inc x)) x)) 1)
17:38	lazybot	⇒ 1
17:38	cheater__	i thought 4clojure was for learning basic knowledge
17:38	Raynes	I lose.
17:39	Raynes	&((fn to-five [x] (if (< x 5) (to-five (inc x)) x)) 1)
17:39	lazybot	⇒ 5
17:39	Raynes	&((fn [x] (if (< x 5) (recur (inc x)) x)) 1)
17:39	lazybot	⇒ 5
17:39	cheater__	is "partial" the same as currying?
17:39	Raynes	No.
17:39	cheater__	what's the diff?
17:40	Raynes	One is called partial application and the other is called currying.
17:40	wingy	gfredericks: cant still create a use case in my head regarding pre and post condition map .. would you mind give me a concrete example?
17:40	cheater__	is there an actual difference?
17:40	wingy	eg. if they don't pass a wrong data type?
17:40	wingy	i mean if they pass a wrong data type?
17:40	wingy	but that would be like type checking
17:41	gfredericks	you can check arbitrary things
17:41	gfredericks	so it's kind of stronger than type checking
17:41	wingy	like no trailing whitespace in strings?
17:41	gfredericks	sure
17:41	emezeske	cheater__: There is a difference. Currying would mean that ((fn [a b] (+ a b) 1) would return a function like (fn [b] (+ 1 b))
17:41	Raynes	cheater__: http://stackoverflow.com/questions/218025/what-is-the-difference-between-currying-and-partial-application
17:42	emezeske	cheater__: Clojure does not support that, though, it would complain about wrong arity to the first fn
17:42	gfredericks	emezeske: http://github.com/fredericksgary/currj
17:43	wingy	gfredericks: I thought that pre condition is used for fn parameter list while post condition is used for return value .. but after reading this line "pre-expr and post-expr are boolean expressions that may refer to the parameters of the function." post condition seems to be used for parameter list as well
17:43	emezeske	gfredericks: that is super cool.
17:43	ldh	i'm struggling to configure jetty from lein-ring. I've got:
17:43	ldh	:ring {:handler my-app.core/app
17:43	ldh	:adapter { :configurator (fn [& more] (println "Configuring"))}}
17:43	ldh	which gives me : java.lang.ClassCastException: clojure.lang.PersistentList cannot be cast to clojure.lang.IFn
17:43	gfredericks	wingy: I imagine they're available in the post conditions, but probably most postconditions just refer to %, the return value
17:45	ldh	if i unquote the configurator function (as suggested here:http://stackoverflow.com/questions/10289617/clojure-ring-jetty-i-am-using-lein-ring-server-how-do-i-configure-the-jetty) i get ReaderException: java.lang.Exception: Unreadable form
17:45	wingy	gfredericks: i created an example: (defn constrained-sqr [x] {:pre [(pos? x)] :post [(= x 1)]} (* x x))
17:46	weavejester	ldh: Embedding functions in the project.clj file is not really what it's designed for...
17:46	wingy	what does the word :pre and :post refer to?
17:46	wingy	they both check the parameter list
17:46	weavejester	ldh: What do you need to configure?
17:46	gfredericks	wingy: :pre gets run before, :post gets run after
17:47	ldh	weavejester: i'm playing with configuring log handlers. i'd prefer not embedding it in project.clj actually, but had no luck trying to refer to a function elsewhere either
17:47	wingy	whats the exact difference? and before/after what btw? the fn body execution?
17:47	gfredericks	yes
17:47	gfredericks	that is the exact difference
17:47	gfredericks	and of course :pre can't access the return value
17:48	wingy	i see .. putting % in pre generates error
17:48	gfredericks	quite
17:48	weavejester	ldh: Hm, you might need to use the adapter directly. If you need to do a lot of work with a specific adapter, then lein-ring is probably not the right tool to use.
17:49	wingy	but is there a reason for wanting to assert the argument list again in a post conditin?
17:49	wingy	condition
17:49	wingy	seems like you are really interested in the return value
17:49	gfredericks	wingy: right; probably the only reason you can use the args at all is if you want to say something about the args and the return value together
17:50	wingy	yeah
17:50	wingy	this was a quite hidden feature .. haven't encountered it yet in 2 books
17:51	oskarth	:q
17:51	oskarth	wrong winodw
17:51	ldh	weavejester: ok, thanks. i can just use 'lein run' and call run-jetty from within my main function and configure it that way, but then i lose the nice code reloading
17:51	weavejester	ldh: You can apply the "wrap-reload" middleware directly.
17:52	weavejester	ldh: It's in ring.middleware.reload in the ring/ring-devel package
17:52	ldh	weavejester: i thought as much. fantastic, thanks.
17:52	Squee-D	oskarth oh i though you were licking your nose
17:52	Squee-D	:P
17:52	gfredericks	lol
17:53	Squee-D	thanks, I'll be here all week. With very little useful to say :D
17:53	gfredericks	,:q
17:53	gfredericks	&:q
17:53	lazybot	⇒ :q
17:54	Squee-D	im really enjoying Land Of Lisp. If anyone asks for a noob primer, this thing is great
17:55	gfredericks	apparently cljs does not have pre/post conditions
17:55	wingy	oh
17:56	wingy	i hope cljs could be as near clj as possible though
17:56	wingy	so there is no surprise
17:56	wingy	pre/post seems possible to do at least
18:07	arohner	$findfn 154 -106
18:07	lazybot	[]
18:22	bpr	is there a more succinct way to do: https://www.refheap.com/paste/3610
18:25	gfredericks	bpr: that looks like a reduce
18:25	bpr	ah blah
18:25	bpr	yes
18:45	oskarth	Squee-D: lol :p
18:45	oskarth	using vim with clojure is kind of like that
18:52	Squee-D	i believe it :D
18:55	wingy	seems that data structures refer to not only collections but to scalars as well: http://clojure.org/data_structures
19:47	wingy	oh my bad i think
19:47	wingy	data structures don't refer to scalars .. weird that they are mentioned in the same section though
19:49	duck1123	well, do you consider 7 to be a data structure?
19:50	wingy	watching a video where he explains data structures
19:50	gfredericks	I'm using 7 as a data structure in my cljs program :)
19:50	wingy	it seems that scalars are data structures as well
19:50	duck1123	I don't see the point of 7
19:50	duck1123	and they eat all the cake
19:50	gfredericks	(extend-type js/Number IBasisVector ...)
19:51	wingy	i just wanna know if the term "data structure" refer to scalar values as well
19:52	wingy	could someone confirm that
19:52	gfredericks	I would think not
19:52	duck1123	what does it matter
19:52	duck1123	clojure doesn't really concern itself with "scalar"
19:53	duck1123	I mean there's things that return sequences and what not, but the concept of scalar isn't really special
19:54	duck1123	If the function doesn't work with non-collections or non-seqable things, it'll say as such
20:04	wingy	but watching this video rich says that the reader reads in data structures
20:04	wingy	when you type 1 in the repl you get returned the data structure of it
20:04	wingy	which is just 1
20:05	wingy	so it seems that all numbers, strings, booleans, keywords etc are called data structures as well
20:06	duck1123	sure
20:09	gfredericks	I think he was probably being loose with his language
20:10	gfredericks	precisely defining data structures was not his intention at that moment
20:10	wingy	so you are saying that by data structures he mean collections
20:10	wingy	list, vector, map and set only
20:11	gfredericks	yeah
20:12	gfredericks	he was probably describing homoiconicity
20:12	wingy	what do we call the other ones?
20:12	gfredericks	scalars
20:12	wingy	and both scalars and data structures are forms right
20:12	gfredericks	yeah; a form is usually something evalable; or at least code-like
20:13	gfredericks	rather than data-like
20:13	gfredericks	it's a distinction of purpose
20:13	wingy	'(1 2 3) is a form?
20:13	gfredericks	probably not; depends on the purpose
20:13	wingy	but then you can probably say it evals to itself
20:13	wingy	just like :keyword is
20:13	gfredericks	no it doesn't
20:14	wingy	oh
20:14	wingy	but :keyword is?
20:14	gfredericks	:keyword does eval to itself
20:14	wingy	ok .. thought that what you saw returned in the repl was the evaluation
20:15	wingy	so its confusing the scalar values are listed in the Data Structures section
20:15	gfredericks	if '(1 2 3) evaled to itself then you could run (eval *1) all day and get the same thing
20:15	wingy	incidental confusion!
20:15	gfredericks	wingy: it's not a word meant to be stressed over
20:16	LesZedCB	why is it not a form? is it not equivalent to (quote (1 2 3))?
20:16	wingy	but when the same term is mentiond all over in books and in blogs etc you kinda wanna know what they exactly mean :)
20:16	gfredericks	it is
20:17	gfredericks	LesZedCB: I was using a somewhat constrained definition of form; if you meant "anything output by the reader" then yes it is
20:17	LesZedCB	well, isn't a form technically just any valid clojure data structure?
20:17	gfredericks	wingy: if you want something generally accepted look at the wikipedia article
20:17	xumingmingv	I'm trying to build this lein plugin: https://github.com/febeling/lein-junit
20:17	gfredericks	LesZedCB: it doesn't have a techincal definition that I'm aware of
20:18	xumingmingv	but seems leiningen:leiningen:jar:1.7.0 can not be found in clojars.org
20:18	xumingmingv	anybody know why?
20:18	gfredericks	LesZedCB: I can imagine a majority of people believing what you said though, so I might just be ignorant
20:21	LesZedCB	so, i have a question about '&' used in variadic functions.
20:22	LesZedCB	suppose (fn useless [x & rest] (println rest))
20:23	dnolen	gfredericks: wingy: LesZedCB: I think a "form" would refer to anything that is readable - that is will be accepted by the reader.
20:23	LesZedCB	how does & fit in with the language? for example I can't evaluate (type &)
20:23	LesZedCB	dnolen: That makes sense, thanks!
20:25	wingy	i just use (read) to know what is a form
20:25	wingy	& is a form as well since it can be read
20:25	lazybot	java.lang.RuntimeException: Unable to resolve symbol: is in this context
20:26	LesZedCB	& is not a form by itself, and [x & rest] is not a form by itself
20:26	lazybot	java.lang.RuntimeException: Unable to resolve symbol: is in this context
20:26	wingy	why can i do
20:26	wingy	(read)
20:26	wingy	then type &
20:26	LesZedCB	(read &) you mean?
20:26	wingy	i returns & which means it has to be a form
20:26	wingy	no
20:26	wingy	(read) .. then & and hit enter
20:27	wingy	everything (read) can read must be a form i guess
20:28	LesZedCB	hmm. interestingly, in the lein REPL, "(read) &" does not evaluate, and in the clojure repl, it does
20:30	wingy	im in lein repl
20:30	wingy	it does read for me
20:30	wingy	lein 2
20:31	LesZedCB	oh, I see what you did
20:31	LesZedCB	try (read-string "&")
20:31	wingy	yeah
20:31	wingy	the same
20:31	wingy	but reading a string instead of a stream
20:31	wingy	so that means & is a form right :)
20:32	LesZedCB	sorry, right. but try just "&"
20:33	dnolen	LesZedCB: my earlier definition may be too loose, seems like in other Lisps - "form" means acceptable to the reader as well being meaningful when evaluted (in terms of the core language).
20:34	wingy	ok so what gfredericks said
20:34	wingy	code like thing
20:34	LesZedCB	dnolen: yes that makes sence, if it can pass read and eval.
20:34	musicalchair	a form has evaluation semantics
20:35	wingy	could one say that a form is an expression?
20:35	wingy	on the other hand an expression could be composed of multiple forms .. i think i get it
20:35	musicalchair	I would say all forms are expressions but not all expressions are forms
20:35	wingy	yeah
20:37	gfredericks	dnolen: that's funny I would have assumed it would refer to post-read if anything
20:37	gfredericks	as in (defmacro fooball [form] ...)
20:38	gfredericks	and "s-expression" might refer to the string version
20:39	gfredericks	wingy: this just confirms that it doesn't matter all that much :)
20:39	wingy	yeah haha
20:40	gfredericks	&(eval (eval :foo))
20:40	lazybot	java.lang.SecurityException: You tripped the alarm! eval is bad!
20:40	gfredericks	aw poopsmith
20:41	wingy	clj is too cool
20:41	wingy	ill teach my son clj that's for sure
20:41	gfredericks	good luck explaining what "the jvm" is
20:42	wingy	why would that be a hard thing
20:43	wingy	its a platform for runing clj
20:43	gfredericks	that's what computers are
20:43	gfredericks	or maybe operating systems
20:43	wingy	but then jvm is an abstraction on top of OSes
20:43	wingy	so we don't have to care about the OSes
20:43	wingy	i think he'll get the pic
20:44	gfredericks	but we do have to care about the jvms
20:44	wingy	perhaps he won't even bother with understanding OSes
20:44	musicalchair	not if you're a little kid learning clojure
20:44	wingy	thats because i dont get it
20:44	gfredericks	in the future people will not be familiar with the existence of "operating systems"
20:44	wingy	:D
20:45	gfredericks	&:D
20:45	lazybot	⇒ :D
20:45	wingy	lol
20:45	LesZedCB	lol
20:45	gfredericks	&'<o.o>
20:45	lazybot	⇒ <o.o>
20:45	duck1123	I gave my son a Ruby shell to play with. I'll let him get a bit older before I try lisp
20:45	gfredericks	look son. if you give it a funny face it will give you one back.
20:45	LesZedCB	thanks for the response. cheers!
20:46	wingy	duck1123: why not clj immediately?
20:46	gfredericks	he has to learn about assignment and mutable data structures before he can learn to avoid them
20:46	duck1123	he was 5
20:46	wingy	does he? can't we just start correct
20:47	wingy	and then show him the OOP where everything is a mess
20:47	gfredericks	no. start with GOTO and malloc.
20:47	duck1123	He's 7 now, he may be able to grok it.
20:47	wingy	duck1123: i think people can get things when they are 4-5 really quickly .. if they have a good teacher
20:47	duck1123	Kids these days don't appreciate not needing to number their lines by 10
20:47	wingy	a 10 years old can be way smarter than a 30 years old if he started early
20:48	wingy	i won't put my boy through hell like the one i had to go through
20:48	duck1123	I was thinking that if I was going to teach him Clojure, I would use light table for that instant response
20:49	wingy	i teach him lisp immediately .. syntax comes and goes .. logic stays forever!
20:49	wingy	i want him to tell his teacher that
20:50	mefisto`	I agree... I want to do the same thing with my daughter (due in 1 month!)
20:51	gfredericks	man I should get my son to type his first paren before he gets out his first word
20:51	wingy	i wont dare to teach my girl computer
20:51	wingy	she can't be as geeky as me
20:51	wingy	i'll teach her design!
20:52	duck1123	I want to give my daughter as much opportunity as possible to learn science and technology
20:52	gfredericks	duck1123: teach your son design!
20:53	wingy	yeah who knows .. maybe it will be more trendy for girls to know programming
20:53	gfredericks	your mother and I will love you no matter what you do, as long as it's trendy.
20:53	wingy	i teach my girl design .. my boy programming .. they both do an app together by 15 years old and bring me money
20:54	mattmoss	Awesome sauce, duck1123: "Kids these days don't appreciate not needing to number their lines by 10."
20:55	gfredericks	if I had designed the language they would have been numbered by 7
20:55	mattmoss	Line numbers written in base 7 as well? Or is that too easy?
20:55	gfredericks	no too easy
20:55	gfredericks	base 10
20:56	mattmoss	20 GOTO 10 becomes 26 GOTO 13
20:56	duck1123	I still have my old Atari XE. I would ride the bus to school while reading my BASIC manual then go home and try out new programs
20:59	mattmoss	POKE 1024,42
21:00	wingy	what happens when i have (def a 1) (def a 2) .. is it creating 2 separate vars or 1 var but the value is changed?
21:01	gfredericks	wingy: same var
21:01	hyPiRion	You could try it out.
21:01	wingy	hyPiRion: didn't know how to compare them to see if they were the same
21:01	wingy	(identical? (def a 1) (def a 2)) ;=> true
21:04	duck1123	wingy: you're creating the var a and altering its root binding
21:06	wingy	is that idiomatic way of changing a value of a var?
21:06	wingy	or should i use binding or atom?
21:07	wingy	or let
21:07	gfredericks	you don't normally want to change the root binding
21:07	gfredericks	what are you trying to do?
21:08	wingy	nothing .. watching videos are questions arise :)
21:08	wingy	nvm
21:08	wingy	ill get to it later more in depth
21:08	quizme	i have an array of 10 things wrapped in an atom. How can i access those 10 things from another clojure process ?
21:09	gfredericks	deref the atom?
21:09	gfredericks	and access the 10 things?
21:09	duck1123	quizme: do you have the atom stored in a var that both threads can read?
21:10	quizme	i want to access it from a different process
21:10	duck1123	avout?
21:10	quizme	i mean i want to fire up two different leiningen projects
21:10	duck1123	https://github.com/liebke/avout
21:10	quizme	and have them share the same atom
21:11	duck1123	you'll need to store it somewhere outside of the jvms memory or send it to the other process somehow
21:11	gfredericks	ah ha
21:12	quizme	how can i send it to the other process ?
21:12	quizme	i guess i have to create a socket
21:13	duck1123	open a socket between the two, store it in a queue, a myriad of ways
21:13	duck1123	check out that link i just sen
21:13	quizme	duck1123 ok thank you
21:16	quizme	duc1123 looks awesome.
21:18	duck1123	I've only played with it a little bit. I want to make use of it, but just haven't had the time
21:18	duck1123	also, slacker is kind of cool for inter-process communication
21:21	duck1123	I really wish there was a decent websocket client with the features of readline. There's a chrome one, but it doesn't support command history
21:35	emezeske	duck1123: Is a websocket client something that runs in the browser?
21:35	duck1123	well it can, but it could be any connection really
21:36	emezeske	duck1123: I was just going to suggest "rlwrap" to get readline features, but it only works for command-line tools
21:36	duck1123	I found a swing one, but it was no good and the node.js stuff I found was more of a library
21:37	duck1123	I've been building out a command interface to my server that works over the open websocket connection
21:37	emezeske	You might try staring it via rlwrap
21:37	emezeske	It tends to have pretty decent results for normal REPL type things
21:38	duck1123	I could probably take my cljs code and make a node.js command
21:38	duck1123	never done that before
22:43	sjl	is there something special I need to do to include images in my /resources/ directory when hosting a Noir site on Heroku?
22:43	sjl	The rest of the /resources/ work, and the image resources work locally...
22:44	technomancy	with compojure IIRC the "site" handler does that for you
22:44	technomancy	or wrap-resource with raw ring
22:44	technomancy	not sure how noir does it
22:44	sjl	nevermind, git is being an asshole
22:45	sjl	git commit --all-but-not-really-all-lol
22:45	sjl	sorry for the noise
22:46	musicalchair	could anyone explain under what circumstances I would use clojure.contrib.graph/stratification-list? Google hasn't been particularly illuminating for me
22:47	sjl	musicalchair: well clojure.contrib is abandoned, so effectively: "None"
22:47	duck1123	Slightly OT. Does anyone know why firebug complains about {"a": "b"} saying it has an invalid label. Should I be doing all my communication as clj forms?
22:48	musicalchair	sjl: I'm aware it's not officially supported, but the code still works and presumably the concept still lives
22:48	sjl	musicalchair: sure
22:49	emezeske	duck1123: I think more context is needed to explain firebug's behavior
22:49	emezeske	duck1123: Regarding communications, I think using pr-str/read-string for (de)serialization is pretty common
22:50	emezeske	duck1123: Although I know some people use JSON and other formats
22:50	duck1123	I'm just trying to eval that form in the js console. I'm trying to figure out why my json communication isn't working
22:51	duck1123	what is the safe way to eval the string with cljs, looks like the jquery one didn't get ported to jayq, but I'm sure I could use that
22:51	duck1123	I'm hesitant to use clj forms as I want the websocket connection usable from outside of the site
22:52	sjl	duck1123: heh, you're going to cry when I tell you the answer.
22:52	sjl	duck1123: try {a = "b"}
22:52	duck1123	see, my issue is that clojure.data.json gives me the form I said above
22:53	emezeske	duck1123: Or try x = {"a" : "b"}
22:53	emezeske	duck1123: I think the {string : string} format only works for assignment
22:53	duck1123	wow
22:54	sjl	try { 1 + 3; }
22:54	sjl	lol
22:54	sjl	Firebug is parsing the braces as a block, not the start of an object literal
23:00	duck1123	Ok, I screwed it up, but if you go to renfer.name and send in your js console: jiksnu.websocket.send("whoami") you can see what's going on
23:01	duck1123	It's because I was using eval. Need to find the fix for tha
23:03	duck1123	got it
23:42	ned	i'm doing problem 16 on 4clojure but am having a bit of a challenge with it anyone know why (fn [x] (concat "Hello, " x "!")) won't work ?
23:45	jkkramer	,((fn [x] (concat "Hello, " x "!")) "ned")
23:45	ned	,(+ 9 9)
23:45	jkkramer	&((fn [x] (concat "Hello, " x "!")) "ned")
23:45	lazybot	⇒ (\H \e \l \l \o \, \space \n \e \d \!)
23:46	ned	oh hm, what's the appropriate concat method then
23:46	duck1123	just use str (sorry for spoiler)
23:46	ned	duck1123: haha no need to apologize, i thought i was doing something dumb and i was!
23:48	duck1123	if you have a seq of charachters like that... (apply str coll)
23:48	michaelr`	good morning!