#clojure – 2014-07-30

2014-07-30

00:00	Balveda	no clojurebot?
00:01	Balveda	this if tells me that it has too few arguments
00:01	jeremyheiler	need a comman before the sexp
00:01	jeremyheiler	,(inc 1)
00:01	clojurebot	2
00:01	jeremyheiler	Balveda: count your parens ;-)
00:02	jeremyheiler	Balveda: balance them correctly, rather.
00:03	Balveda	thanks
00:04	jeremyheiler	np
00:05	Balveda	oh
00:05	Balveda	nm
00:06	Balveda	'cause the main issue is I was having some issues with this func
00:06	Balveda	(if (< 3 (count data)) ((doall (take 3 data)) ((doall (drop 3 data))))) where data is a vector of vectors
00:07	Balveda	i'm putting the doalls (not to well admittedly) because it's giving me that a LazySeq can't be casted to IFn
00:07	amalloy	that is way too many parens, Balveda
00:07	jeremyheiler	yeah
00:08	Balveda	agreed
00:08	gws	it's not like other languages where they're just used for grouping - you're actually changing the semantics of your program by adding parens
00:10	Balveda	How should I deal with the LazySeq issue?
00:10	jeremyheiler	Balveda: the problem is different than you thinking
00:11	jeremyheiler	LazySeq cannot be cast to an IFn because you're trying to invoke the lazy sequenece as a function because you have too many parens.
00:11	Balveda	Mind you that these doalls are me trying to fix thsi
00:11	Balveda	this*
00:12	jeremyheiler	why did you do that?
00:12	Balveda	As I understand it, the doall evaluates the stuff I'd be passing, so it's not a LazySeq
00:13	jeremyheiler	so, what is it after you doall it?
00:13	jeremyheiler	,(type (doall (map inc [1 2 3])))
00:13	clojurebot	clojure.lang.LazySeq
00:14	Balveda	hm.
00:14	Balveda	What would be a good way to deal with this issue?
00:14	jeremyheiler	lets back track and look at the original error
00:15	TEttinger	,(def data [1 2 3 4 5])
00:15	clojurebot	#'sandbox/data
00:15	TEttinger	,(if (< 3 (count data)) (doall (take 3 data)) (doall (drop 3 data)))
00:15	clojurebot	(1 2 3)
00:15	TEttinger	,(if (> 3 (count data)) (doall (take 3 data)) (doall (drop 3 data)))
00:15	jeremyheiler	when you see "LazySeq can't be casted to IFn", it's saying that whatever you're trying to use (LazySeq in this case) is being used as a function.
00:15	clojurebot	(4 5)
00:16	jeremyheiler	and... that it's wrong to do so
00:16	TEttinger	I suspect you may have meant > instead of <
00:16	Balveda	True
00:16	jeremyheiler	why do you think it's trying to call a lazyseq as a function?
00:16	Balveda	,(doc take)
00:16	clojurebot	"([n coll]); Returns a lazy sequence of the first n items in coll, or all items if there are fewer than n."
00:16	Balveda	,(doc drop)
00:16	clojurebot	"([n coll]); Returns a lazy sequence of all but the first n items in coll."
00:17	TEttinger	no, that's not it at all
00:17	jeremyheiler	yes, those fns returns lazy seqs
00:17	jeremyheiler	look at your code. why is it trying to invoke those lazy seqs as functions?
00:17	TEttinger	,(+ 1 2 3)
00:17	clojurebot	6
00:17	TEttinger	,((+ 1 2 3))
00:17	clojurebot	#<ClassCastException java.lang.ClassCastException: java.lang.Long cannot be cast to clojure.lang.IFn>
00:18	Balveda	I'm not too sure.. if you're saying it's a parens issue I imagine I mucked it up somewhere but I can't quite see it
00:18	jeremyheiler	look at TEttinger's example carefully.
00:18	jeremyheiler	what's different between teh two expressions?
00:19	Balveda	ah
00:20	Balveda	god it
00:20	Balveda	got it*
00:20	Balveda	thanks!
00:20	jeremyheiler	:-)
00:20	Balveda	,(inc jeremyheiler)
00:20	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: jeremyheiler in this context, compiling:(NO_SOURCE_PATH:0:0)>
00:20	Balveda	heh
00:20	jeremyheiler	i don't actually exist
00:20	Balveda	Spooky
00:21	jeremyheiler	(inc Balveda)
00:21	lazybot	⇒ 1
00:21	TEttinger	(inv jeremyheiler)
00:21	TEttinger	(inc jeremyheiler)
00:21	lazybot	⇒ 1
00:22	TEttinger	the (inc) thing isn't actually evaluating code, I think it actually allows spaces in names
00:22	TEttinger	(inc #())
00:22	lazybot	⇒ 1
00:22	Balveda	oh, the , is for evaluatinc code?
00:22	TEttinger	(dec #())
00:22	lazybot	⇒ 0
00:22	Balveda	jeez i'm typing bad
00:22	TEttinger	yep
00:22	jeremyheiler	the inc thing is just a custom lazybot plugin
00:22	Balveda	I just hooked up this usb keyboard to my laptop and I'm readapting
00:22	Balveda	(inc jeremyheiler)
00:22	lazybot	⇒ 2
00:23	Balveda	all these parens and brackets were killing my right hand, heh
00:23	TEttinger	and #() is not a valid clojure identifier, which is why I picked it to check if the thing was doing any code stuff
00:23	kelseygi	CookedGryphon i'm totally creeping on the logs right now but did you ever get your emojis encoded right?
00:23	TEttinger	are the logs up again?
00:23	Balveda	Is the take/drop method proper, though?
00:23	TEttinger	what are you trying to do though?
00:23	Balveda	It seems a little brute force imo knowing that things like reduce exist, but I'm not too acquainted yet
00:24	kelseygi	i was looking at this http://clojure-log.n01se.net/date/2014-03-07.html
00:24	TEttinger	woah, march
00:24	Balveda	As a whole, I'm taking an excel and making pdf tables so as to make cards, right? I have a 3x3 grid per page, so I need to split the data I'm recieving
00:24	kelseygi	yeah i'm tryna do exactly what they were doing there
00:24	TEttinger	(doc partition-all)
00:24	clojurebot	"([n coll] [n step coll]); Returns a lazy sequence of lists like partition, but may include partitions with fewer than n items at the end."
00:25	TEttinger	(doc partition)
00:25	clojurebot	"([n coll] [n step coll] [n step pad coll]); Returns a lazy sequence of lists of n items each, at offsets step apart. If step is not supplied, defaults to n, i.e. the partitions do not overlap. If a pad collection is supplied, use its elements as necessary to complete last partition upto n items. In case there are not enough padding elements, return a partition with less than n items."
00:25	TEttinger	##(partition-all (range 50)) ; the ## is to eval with lazybot, which prints full sequences
00:25	lazybot	clojure.lang.ArityException: Wrong number of args (1) passed to: core$partition-all
00:25	TEttinger	##(partition-all 3 (range 50)) ; the ## is to eval with lazybot, which prints full sequences
00:25	lazybot	⇒ ((0 1 2) (3 4 5) (6 7 8) (9 10 11) (12 13 14) (15 16 17) (18 19 20) (21 22 23) (24 25 26) (27 28 29) (30 31 32) (33 34 35) (36 37 38) (39 40 41) (42 43 44) (45 46 47) (48 49))
00:26	TEttinger	Balveda, does that look like the kind of output that you want?
00:26	Balveda	Yes, it's perfect
00:27	Balveda	Or, hang on;
00:27	TEttinger	ha, I'm notostraca, kelseygi
00:27	Balveda	I suppose I could have the whole set of data partitioned into threes and iterate over it with the row generator
00:27	TEttinger	indeed
00:28	Balveda	Rather than feed a list into the function and have it exhaust the initial coll
00:28	Balveda	List, vector, same thing
00:28	kelseygi	heh TEttinger, i was secretly hoping someone had been inspired to create an emoji-handling library in the time since that convo happened
00:28	TEttinger	partition guarantees items of 3, partition-all will give items of 1, 2 or 3 as the last thing just to use up the whole collection
00:29	Balveda	Partition would return nils though for the empty items in the list?
00:29	TEttinger	##(print (partition-all 3 (range 8))" " (partition-all 3 (range 8)))
00:29	lazybot	⇒ ((0 1 2) (3 4 5) (6 7)) ((0 1 2) (3 4 5) (6 7))nil
00:29	TEttinger	err
00:29	TEttinger	##(print (partition3 (range 8))" " (partition-all 3 (range 8)))
00:29	lazybot	java.lang.RuntimeException: Unable to resolve symbol: partition3 in this context
00:29	TEttinger	##(print (partition 3 (range 8))" " (partition-all 3 (range 8)))
00:29	lazybot	⇒ ((0 1 2) (3 4 5)) ((0 1 2) (3 4 5) (6 7))nil
00:29	TEttinger	I'm not typing well either
00:29	Balveda	What's your timezone?
00:30	TEttinger	pacific standard, GMT - 8 or so
00:30	Balveda	Ah, alright; I'm GMT -3, so it's getting late
00:30	Balveda	not like I have a schedule anyways
00:30	TEttinger	partition will just not use elements if they wouldn't fill up 3, partition-all will use them up and make less-than-3-many colls
00:31	TEttinger	which you want depends a lot on whether you're expecting exactly 3 (or any number) or 3 or less
00:31	Balveda	It's the same thing either way, I'd still be making empty cards
00:31	Balveda	It's not a guarantee I'll have a multiple of 3
00:31	TEttinger	partition-all then, I'd say
00:32	TEttinger	you can trim unneeded ones later
00:32	Balveda	Yeah, either way it's if count is < 3 or if item is nil, make an empty card
00:32	TEttinger	cool
00:34	Balveda	thanks for all the help!
00:34	Balveda	I'll be sure to pester you soon enough
00:34	Balveda	I like how clojure has these functions for handling things rather than just having to iterate al ot
00:35	TEttinger	oh yeah, you should see the stuff I can pull off in one IRC message using "clever" partition and other seq stuf
00:36	Balveda	coming from enterprise java it's a breath of fresh air
00:36	Balveda	the less I can touch that stuff the better, the things i've seen.. ugh
00:36	TEttinger	##(clojure.string/join " "(repeatedly 2000(fn [](apply str(concat[(rand-nth ["rh""s""z""t""k""ch""n""th""m""p""b""l""g""phth"])](take(+ 2(* 2(rand-int 2)))(interleave(repeatedly #(rand-nth ["a""a""o""e""i""o""au""oi""ou""eo"]))(repeatedly #(rand-nth ["s""p""t""ch""n""m""b""g""st""rst""rt""sp""rk""f""x""sh""ng"]))))[(rand-nth ["os""is""us""um""eum""ium""iam""us""um""es""anes""eros""or""ophon""on""otron"])])))))
00:36	lazybot	⇒ "zeoxitum thaurkum gausteochus phthaustachium kourkofon neopaunor moganium thomon geortoutus koneosis paspanes zaumus zemouxophon rhastenis chagaufophon chourtor rhoges gamiam phthesaspus leochortus zoupeobor somanes somiam rhangiam tachertos poirkasophon soibum logi... https://www.refheap.com/88708
00:36	TEttinger	forgive the ugliness, but lazybot only takes single lines
00:37	Balveda	nice
00:37	TEttinger	that's within about 30 chars of the limit for Freenode messages IIRC
00:38	Balveda	I'm getting on a project where I can choose what I want to dev in and despite me not having much experience with Clojure I'd rather use it than anything else
00:39	aeikenberry	can i ask a few more?
00:39	aeikenberry	i've changed my macro to accept any number of argument
00:39	aeikenberry	and i want to do something like this:
00:40	aeikenberry	(list 'cond test (list 'map 'do (:then chunkmap))
00:40	TEttinger	Balveda, it's an awesome language
00:40	aeikenberry	but i guess you can no map do
00:40	aeikenberry	do i need to do a for?
00:40	TEttinger	that looks like a loop/recur or reduce problem; how did you solve the first, aeikenberry?
00:41	jeremyheiler	aeikenberry: mapping do doesn't make much sense
00:41	aeikenberry	yeah, i'm not making a new coll
00:41	aeikenberry	i just want to iterate over and do that stuff
00:42	jeremyheiler	aeikenberry: remember, macros jobs are to spit out code.
00:42	jeremyheiler	do you really need to "iterate" over anything?
00:44	jeremyheiler	(cons 'do (:then chunkmap))
00:45	jeremyheiler	if the value of :then in the map is a list of expresions, you can just cons do to the front
00:45	jeremyheiler	(you could also dive into using syntax-quote and unquote-splicing to create a more literal macro)
00:45	aeikenberry	https://dpaste.de/cKOb
00:46	jeremyheiler	so,
00:46	jeremyheiler	,(list 'do '(1 2 3))
00:46	clojurebot	(do (1 2 3))
00:46	jeremyheiler	vs
00:46	jeremyheiler	,(cons 'do '(1 2 3))
00:46	clojurebot	(do 1 2 3)
00:47	aeikenberry	ah
00:47	aeikenberry	thanks jeremyheiler
00:47	jeremyheiler	np!
00:49	aeikenberry	if you are stuck in repl
00:50	TEttinger	ctrl-d to exit
00:50	aeikenberry	thanks
00:50	TEttinger	might be z or c
00:50	jeremyheiler	burn your computer and buy a new one
00:51	aeikenberry	my exp isn't terminating
00:51	aeikenberry	no number of ))'s or "'s
00:51	aeikenberry	will let me finish
00:51	aeikenberry	how do i scrap that but not leave repl?
00:52	jeremyheiler	ctrl+c if you're just in a lein repl
00:53	aeikenberry	doesn't seem to work for me
00:56	aeikenberry	thanks for all the help
00:58	aeikenberry	https://dpaste.de/srFZ
00:58	aeikenberry	last one
01:00	aeikenberry	,(do (quote ok) (quote cool))
01:00	clojurebot	cool
01:02	aeikenberry	,(do (println 'ok) (println 'cool))
01:02	clojurebot	ok\ncool\n
01:02	aeikenberry	what's going on there?
01:02	TEttinger	aeikenberry, there's no way to do an early terminate and stay in repl
01:03	aeikenberry	TEttinger, oh ok. thanks
01:03	TEttinger	like if you try to assign a variable to the evaluation of an infinite sequence, terminating and staying in repl would be bad
01:03	TEttinger	(unless you have infinite ram)
01:03	aeikenberry	that makes sense
01:03	jeremyheiler	,(println "hi")
01:03	clojurebot	hi\n
01:03	jeremyheiler	aeikenberry: so, clojurebot is hiding it, but println returns nil
01:03	TEttinger	aeikenberry, also, do returns the last element
01:04	TEttinger	but it evals the others, it doesn't print them unless you tell it to print
01:04	aeikenberry	ah
01:08	aeikenberry	off to bed. you guys rock. great first day of clojure
01:08	aeikenberry	thanks again
01:08	TEttinger	no prob, glad to help
01:08	jeremyheiler	good night!
02:48	swgillespie	hi all, is there a way to read a single character from stdin without calling into java?
02:48	TEttinger	well seeing as... #(class in)
02:48	TEttinger	well seeing as... ##(class in)
02:48	lazybot	⇒ clojure.lang.LineNumberingPushbackReader
02:49	TEttinger	interesting
02:49	TEttinger	it's different here than at a command line repl
02:49	TEttinger	,(class in)
02:49	clojurebot	clojure.lang.LineNumberingPushbackReader
02:49	TEttinger	maybe not
02:50	TEttinger	swgillespie, is this at the repl or in a running application like a jar?
02:50	swgillespie	Shouldn't it be the same in both scenarios?
02:50	swgillespie	I'm writing a brainfuck interpreter as an intro to clojure
02:50	TEttinger	repl I think uses a different class for in, likely has the same API
02:52	TEttinger	,(.read in)
02:53	TEttinger	n
02:53	swgillespie	TEttinger: (read in) reads until a newline for me
02:53	clojurebot	eval service is offline
02:53	TEttinger	oh whaaaaat
02:53	TEttinger	no, .read
02:53	TEttinger	read is ##(doc read)
02:53	lazybot	⇒ "([] [stream] [stream eof-error? eof-value] [stream eof-error? eof-value recursive?]); Reads the next object from stream, which must be an instance of java.io.PushbackReader or some derivee. stream defaults to the current value of in ."
02:53	TEttinger	an object
02:53	swgillespie	oh jeez
02:53	TEttinger	but .read is a method on Reader, which is what in implements
02:54	swgillespie	gotcha
02:54	TEttinger	it is java, but the only clue is that there's a .
02:54	swgillespie	I see, I didn't catch the dot
02:58	swgillespie	oh, I see - even if you put in multiple characters to the terminal, (.read in) pops them off one at a time
02:58	swgillespie	thanks TEttinger
03:01	TEttinger	no prob, swgillespie
03:28	blunte	Q: is it possible to do something like this? (defrecord RecBar [a b]) (defn foo [rectype a b] (->rectype a b)) ?
03:29	blunte	it's not able to compile the function (because it has no idea what a "rectype" is)
03:32	blunte	I'm trying to make a function that can accept any type of defined record, read a file, and populate records from the file. But I can't figure out how to define that function (how to tell the compiler that it should expect some generic "class")
03:32	pyrtsa	blunte: If it helps, you could construct a RecBar instance with assoc, like this: (assoc (map->RecBar {}) :a 1 :b 2), or just use map->RecBar.
03:33	pyrtsa	Does the returned type of record depend on the contents of the file somehow? Or only the contents?
03:33	blunte	yes, I'm making a file reader that should populate and return a set of the appropriate type of record
03:33	blunte	so I want to call load-file with the RecBar or RecBoo or whatever the appropriate record type is for the associated file
03:34	pyrtsa	So the caller tells that "read this file, it should return a RecBar", right?
03:34	blunte	I need some way of saying (defn foo [this-is-a-class-that-has-arrow-operator filename] ...)
03:34	blunte	yes
03:35	pyrtsa	Then assoc could just work.
03:35	blunte	I'll read about assoc. I'm quite new, so it doesn't mean much to me yet. Thanks!
04:38	engblom	What free and open-source data storage library would you use if you do not want a separate database server (like mysql, postgresql)? I would even prefer something not being SQL at all as that requires a lot of mapping between clojure data-structures and SQL.
04:39	ssideris	engblom: I happily use honeysql, but that might be too close to SQL for your liking
04:40	ssideris	but I don't end up doing many conversions really
04:40	hyPiRion	engblom: So what's the use case? That's seems pretty important when deciding on DB.
04:40	hyPiRion	I mean, do you have structural data? Or is it not that important with schemas etcs?
04:40	Glenjamin	what sort of load, do you need to access from multiple processes
04:40	ssideris	engblom: sorry, I mis-read your question, ignore me
04:44	TEttinger	engblom, I have seen an article that mentioned how easy it is to use clojure's built-in transactional memory to make in-memory DBs
04:45	TEttinger	http://blog.malcolmsparks.com/?p=67
04:47	engblom	The typical projects would be in size of a personal project manager keeping track of how many hours I have been working on different projects and a short description.
04:47	TEttinger	oh wow, hat one isn't in-memory
04:48	TEttinger	heh, engblom: 30 lines enough? I'm curious if anyone has improved on that example
04:48	engblom	In memory DB would be great if it had some durability by writing in the background to disk.
04:49	engblom	TEttinger: That one I found earlier today when I googled for this question
04:50	TEttinger	yeah, it's number one on my search, but it is good. I will check github for improvements other people made
04:50	Glenjamin	spit on change / slurp on boot might be enough for that case
04:52	Glenjamin	oh right, thats what this does
04:52	hyPiRion	engblom: You could probably use Monger, which has this automatic conversion to and from Clojure data structures if I read it right
04:53	TEttinger	there's also https://github.com/ibdknox/simpledb/blob/master/src/simpledb/core.clj
04:53	hyPiRion	I've never used MongoDB though, so take that with a grain of salt.
04:53	TEttinger	I have, it does need a server
04:54	hyPiRion	Oh, so we're talking bundling everything up in the same application?
04:54	Glenjamin	there are few, if any, good reasons to use mongo in my experience
04:55	Glenjamin	either your app is too small to get any benefits over a normal DB, or its too big and scaling is a pain
04:55	TEttinger	Glenjamin, yeah I'm curious what the better alternatives are?
04:55	Glenjamin	me too
04:55	Glenjamin	we use it quite a bit here, and we all hate it now
04:56	engblom	I wonder if MongoDB is a server-client database or just a library storing to a local file?
04:56	hyPiRion	It is a "proper" database.
04:57	engblom	OK, thanks!
04:57	engblom	I might still look into it as monger seem to do what I wish to do, from the short description.
05:00	TEttinger	but yeah mongodb does have a server (daemon) running
05:00	engblom	The reason I wish it would not be a server-client database is because I would want to be able to upload whatever I make to any java-server and know it will run regardless of if they provide mysql-, postgresql-, whatever-server
05:00	TEttinger	mongod
05:01	TEttinger	it might be the most flexible to use something like what ibdknox provides in that sample
05:01	TEttinger	slurp and spit can serialize non-binary data
05:01	TEttinger	or rather with the right toString things for the types you use
05:11	engblom	What do you think about this one: https://github.com/gcv/cupboard ?
05:11	TEttinger	that looks excellent
05:12	TEttinger	but I know nothing about Berkeley DB
05:12	engblom	Sadly it seem to not be actively developed anymore
05:13	TEttinger	ah, AGPL
05:24	teemu_f	ClojureScript q: I'd like to filter data based on search term (basic search box). How do I use js/RegExp properly? Basic (filter #(term (:name %)) data-chunk) works but accepts only direct matches.
05:34	ebaxt	I'm looking for a light-weight solution for scheduling jobs across multiple clojure apps. I need to make sure only one application runs the job for each trigger. I've looked at Quartzite, but it only supports this via JDBC-JobStore, and were not using any RDMS. Does anyone here use any other solution for this type of scenario?
05:47	gws	maybe it wouldnt' be too hard to plug yours in, or maybe a library already exists
05:51	ebaxt	gws: I noticed that as well. Currently the mongodb adapter doesn't support clustering (https://github.com/michaelklishin/quartz-mongodb/issues/66), so I'm trying to figure out if there are other options that already support this before adding the feature myself
05:59	gws	ah, i see - yeah given your requirements either etcd/zookeeper for coordination or adding that feature to the existing mongodb adapter seem like the paths of least resistance
06:00	gws	there are some lighter weight schedulers but none attempt to run "singleton" jobs across a cluster, at least not that i'm aware of
06:04	ebaxt	gws: ok, I'm currently using chime, but this is the first time I need synchronization across multiple servers. We also use redis so I guess I could use a global lock as well. Anyways, I was hoping there was a small idiomatic clojure lib that I didn't know about that solved it for me :)
06:06	schmir	ebaxt: RabbitMQ may be an option. but that's not really light-weight
06:11	ebaxt	schmir: Thanks, but I would prefer not to add another pice of infrastructure :)
06:17	sveri	Hi, I am using http-kit and lein cljsbuild on windows 7. Now, when I run the cljsbuild in advanced\|whitespaced mode and start up my server the rebuild of the cljs file fails because the server or leiningen or whatever still has a handle on it, causing that the cljsbuild cannot write the js file. Did someone experience the same? Any Ideas how to solve it (switching to linux is not rly a valid proposal in my case ;-))
06:30	engblom	In *nix you have dot-folders in $HOME for storing data. In Windows you have AppData under your profile. Other OS might have other places... How do you best store data belonging to the user? Do I need manually to detect OS and take different actions depending on OS?
06:31	engblom	Is there a ready library for this?
06:32	imperman	engblom: OS checking is the safest way, I think. Check this also - https://github.com/derekchiang/Clojure-Watch
06:34	katratxo	sveri: i guess that is a common windows issue and not a lein cljsbuild specific problem, right? https://en.wikipedia.org/wiki/File_locking#In_Microsoft_Windows
06:35	imperman	engblom: here is also some platform-neutral examples - http://stackoverflow.com/questions/21198361/clojure-file-system-portability
06:38	sveri	katratxo: Yea, that's my guess too, I was just wondering if anyone knows a workarund, besides that, It works if I compile the cljsbuild without any optimization
06:49	teemu_f	Okay, having lunch helped with my regex problem..
06:49	teemu_f	(filter #(not (nil? (re-matches (re-pattern search-term) (:name %)))) (:persons app))
06:52	nobodyzzz	(not (nil? ...)) is not nesseccary i suppose
06:53	teemu_f	yup, I just removed them
06:54	teemu_f	first you make it work, then clean up afterwards :)
06:54	ro_st	you can just use re-find, btw
06:55	ro_st	(->> app :persons (filter (comp (partial re-find (re-pattern search-term) :name))))
06:56	ro_st	erk
06:56	ro_st	(->> app :persons (filter (comp (partial re-find (re-pattern search-term)) :name)))
06:57	blunte	Is it possible to pass in a class (such as a defined record) to a function, and then within the function use operations from that class?
06:58	blunte	(defrecord R [a b]) (defn foo [a-rec-class val1 val2] (->a-rec-class val1 val2))
07:01	teemu_f	ro_st: thanks for alternative solution. I'm not that familiar with comp function yet, have to try it
07:04	scape_	why do I get a "no matching field for class" error? https://www.refheap.com/88710
07:05	ro_st	init is a static method
07:05	ro_st	Bullet/init
07:05	scape_	public static
07:05	scape_	yes
07:05	scape_	oh ok
07:07	scape_	thanks
07:14	clgv	blunt: just pass in the appropriate function
07:14	blunte	clgv: I'm not sure I understand?
07:15	clgv	blunte: you can pass any function as parameter to any other function in clojure
07:16	blunte	clgv: so I pass ->R instead of R ?
07:16	clgv	blunte: yes
07:17	clgv	,(do (defrecord R [a b]) (defn foo [f a b] (f a b)) (f 1 2))
07:17	clojurebot	#<CompilerException java.lang.RuntimeException: Unable to resolve symbol: f in this context, compiling:(NO_SOURCE_PATH:0:0)>
07:18	clgv	,(do (defrecord R [a b]) (defn foo [f a b] (f a b)) (foo ->R 1 2))
07:18	clojurebot	#sandbox.R{:a 1, :b 2}
07:18	blunte	clgv: that's great, and simple
07:18	clgv	blunte: but that minimal function is actually pretty useless^^
07:19	blunte	clgv: yes, there's actually more going on in my foo. data will be read from a file, and records will be generated. the details come down to the filename (what's in the file) and the type of record I'm stuffing it into
07:19	blunte	clgv: much thanks :)
07:19	clgv	np#
07:59	razum2um	how to load a jar with -javaagent options with lein ?
08:01	clgv	razum2um: https://github.com/technomancy/leiningen/blob/stable/sample.project.clj#L252
08:01	razum2um	clgv: thanks
08:21	mbac	http://dev.clojure.org/display/community/Clojure+Success+Stories
08:21	mpenet	how do you even put stuff in a c.core.async/sliding-buffer?
08:21	mpenet	>!! >! put! seems all to fail
08:22	mbac	that success stories link is a bit hand wavey. e.g. i believe citigroup usees clojure somewhere but it doesn't say what for
08:22	ro_st	mpenet: don't you pass the sliding buffer as an arg to chan, and write to the chan?
08:23	mpenet	ah, might be!
08:23	mpenet	right, chan!=buffer
08:26	AeroNotix	gfredericks: can I depend on this code for production uses? https://github.com/gfredericks/vcr-clj
08:26	AeroNotix	looks like it would be useful for some stuff we're doing here
08:27	teemu_f	Hooray, there's a new euroclojure 2014 video on vimeo! Hope the rest will be there soon..
08:28	ro_st	link teemu_f?
08:28	teemu_f	http://vimeo.com/euroclojure
08:29	mbac	is there a macro like defn, but where if you call the function with too few arguments it returns a function that takes the remaining arguments?
08:30	mbac	that is, it auto partially applies the function?
08:30	ro_st	no. not even slightly.
08:30	mbac	it seems like it should be doable
08:30	ro_st	oh, wait. i read that as 'is macro like defn, but'
08:30	ro_st	isn't that like fnil?
08:31	ro_st	http://grimoire.arrdem.com/1.6.0/clojure.core/fnil
08:31	ro_st	it's kinda similar
08:32	clgv	mbac: https://github.com/gfredericks/currj ?
08:32	ohpauleez	mbac: You want to curry
08:33	mbac	right
08:33	ohpauleez	there are a few curry macros out there, including one in definition of reducers (called defcurry)
08:33	pyrtsa	I think #'clojure.core.reducers/defcurried does it but it's unfortunately declared private.
08:34	ohpauleez	It's also fun to write that macro, so as a thought exercise, you can take a stab at writing it before looking at others ( mbac )
08:34	mbac	basically i have a form like this:
08:34	mbac	(defn date-to-string [date] (time-format/unparse (time-format/formatters :year-month-day) date))
08:34	mbac	and i want to be able to factor out the date
08:34	mbac	(def date-to-string (time-format/unparse (time-format/formatters) :year-month-day))
08:35	mbac	and i can, by using partial, but that's almost as long as the original
08:36	gfredericks	AeroNotix: by "production" do you mean "testing code used in production"?
08:37	AeroNotix	gfredericks: yeah
08:37	AeroNotix	same diff :)
08:37	gfredericks	clgv: mbac: currj is like three things at once, and currying is one of them; it's not complete though
08:38	gfredericks	AeroNotix: I've been using it for that for over a year now
08:38	AeroNotix	gfredericks: ok, I'll take 'er for a spin
08:38	AeroNotix	thanks :)
08:38	gfredericks	np
08:41	gfredericks	AeroNotix: I'm reminded now of the "library extensibly serializing" problem that I didn't know how to solve; now I'm wondering if transit could do it, but of course we're not supposed to persist transit...
08:42	gfredericks	AeroNotix: be aware of issue #2; there's a separate release described there with a haxy fix
08:43	gfredericks	might have to write my own damn edn printer to fix it properly
08:45	mbac	http://stackoverflow.com/questions/5329079/higher-order-functions-in-clojure
08:45	mbac	the answer below the accepted answer, defn-ho, seems like the business
08:49	mbac	oh currj. hmm.
08:49	H4ns	so i have this clojure application packed in an uberjar and now i wonder where i would put a suitable log4j.properties file so that it is seen when i run the application. can anyone help?
08:50	gfredericks	,(defn curry [f arg-num] (if (pos? arg-num) (fn [& args] (curry (apply partial f args) (- arg-num (count args)))) f))
08:50	clojurebot	#'sandbox/curry
08:50	gfredericks	,(defn plus (curry + 3))
08:50	clojurebot	#<IllegalArgumentException java.lang.IllegalArgumentException: Parameter declaration curry should be a vector>
08:50	gfredericks	,(def plus (curry + 3))
08:50	clojurebot	#'sandbox/plus
08:50	gfredericks	,(plus 8)
08:50	clojurebot	#<sandbox$curry$fn__25 sandbox$curry$fn__25@c5a44e>
08:50	gfredericks	,(plus 8 9 10)
08:50	clojurebot	#<core$partial$fn__4232 clojure.core$partial$fn__4232@1b66659>
08:51	gfredericks	,(plus 8 9 10 11)
08:51	clojurebot	#<core$partial$fn__4234 clojure.core$partial$fn__4234@1ccad9c>
08:51	gfredericks	oops
08:51	ohpauleez	close
08:52	gfredericks	I guess if you call curry with a non positive number it should just call your function :D
08:52	gfredericks	,(defn curry [f arg-num] (if (pos? arg-num) (fn [& args] (curry (apply partial f args) (- arg-num (count args)))) (f)))
08:52	clojurebot	#'sandbox/curry
08:52	gfredericks	,(def plus (curry + 3))
08:52	clojurebot	#'sandbox/plus
08:52	gfredericks	,(curry 8 9 10)
08:52	clojurebot	#<ArityException clojure.lang.ArityException: Wrong number of args (3) passed to: sandbox/curry>
08:52	gfredericks	,(plus 8 9 10)
08:52	clojurebot	27
08:52	gfredericks	,(plus 8 9)
08:52	clojurebot	#<sandbox$curry$fn__166 sandbox$curry$fn__166@1d57c7>
08:52	gfredericks	,((plus 8 9) 10)
08:52	clojurebot	27
08:52	gfredericks	,(((plus 8) 9) 10)
08:52	clojurebot	27
08:52	gfredericks	,((plus 8) 9 10)
08:52	clojurebot	27
08:52	gfredericks	,(curry + 0)
08:52	clojurebot	0
08:53	gfredericks	could also get it to support extra args getting called on the return value
08:53	drbobbeaty	H4ns: I put mine in the resources/ directory at the top-level of the leiningen project, and it is (by definition) in the classpath, and will get picked up.
08:53	H4ns	drbobbeaty: ah, thanks, i'll try that!
08:54	drbobbeaty	H4ns: let me know if it fails for you - but it shouldn't.
08:57	H4ns	drbobbeaty: it appears not to be picked up. i probably make a very basic mistake. maybe i'm not even using log4j, but java.utils.logger. i'll have a look.
09:00	Mandar	hello
09:00	Mandar	quick leiningen question
09:00	Mandar	i've just started a new project with 'lein new sg
09:00	drbobbeaty	H4ns: when I `lein uberjar` it places my log4j.properties file from my resources/ directory at the top-level of the generated jar in the target/ directory - both the stand-alone jar, and the non-standalone version.
09:01	Mandar	added [org.clojure/math.combinatorics "0.0.8"] to my project.clj file
09:01	Mandar	and ran lein deps
09:01	Mandar	but looks like nothing got downloaded
09:01	H4ns	drbobbeaty: yeah, i think what i was missing was log4j itself. trying that right now.
09:01	H4ns	tada!
09:02	H4ns	drbobbeaty: thanks! i've added it to my project.clj and the props from the resources directory are picked up.
09:02	Mandar	and when adding (:require [clojure.math.combinatorics :as combo]) to my core.clj file, I only get an exception (java.io.FileNotFoundException)
09:02	hyPiRion	Mandar: does project.clj look like this? https://www.refheap.com/88713
09:03	drbobbeaty	H4ns: Great! :)
09:03	Mandar	hyPiRion, yes it does
09:03	hyPiRion	Mandar: well, that's strange, because it works fine over here.
09:04	Mandar	hmm
09:04	Glenjamin	anyone know where i can find a decent levenstein distance function?
09:05	Mandar	hyPiRion, lein deps :tree shows the dependency as well
09:05	hyPiRion	Glenjamin: there's one in Lein main. It's levensthein-damerau though, and not particularly performance tuned
09:06	hyPiRion	Mandar: ohh
09:06	hyPiRion	Mandar: wait no, I thought I understood why, but I don't.
09:07	hyPiRion	Mandar: this happens when you do `lein repl`, right?
09:08	Mandar	it happens when I eval the require line in Lighttable
09:08	hyPiRion	Hm. I'm not sure how Lighttable and Lein works together, so I'm not sure how I could help you there
09:09	hyPiRion	Other than just restart Lighttable and/or ensure that the :require call is within a file in the project.
09:09	Mandar	I'll try with the repl first, you're right
09:10	hyPiRion	At least we can trace it down to some specific software
09:11	Mandar	heh, restarting LightTable worked
09:11	Mandar	thank you hyPiRion
09:12	hyPiRion	You're welcome. I guess LT has to reload the dependencies somehow – I wouldn't be surprised if there is some feature which makes you able to do so without restarting it
09:12	hyPiRion	But I'm no LT person, so that's just a guess
09:23	teslanick	If you disconnect from the connected client and then reconnect, it will re-retrieve dependencies.
09:23	teslanick	(usually)
09:24	teslanick	(with relevant caveats about disconnecting from running state possibly screwing up your development process)
09:31	Mandar	teslanick, thanks
09:31	razum2um	I wonder why do all lein plugins exist if putting a lib into ~/.lein/profiles.clj under :dependencies is way better than that. why should i start another jvm and force my clojure to be scripty-like when i can require lib (not plugin!) in my repl an feed it with any clojure structure as argument
09:35	TimMc	razum2um: Well, for one thing, that would require having a buildable project.
09:36	TimMc	For another, some tasks require varying the dependencies, so you wouldn't want the plugin in that same dependency space.
09:36	razum2um	TimMc: plugins like expectations or midje (for continous testing) don't do any good if project isn't buildable anyway
09:37	TimMc	I mean, I agree that if your plugin could just be a library, maybe it should be, but that's just not always the case.
09:37	clgv	razum2um: if you want to customize build related stuff, then you'll probably need a leiningen plugin
10:58	tvanhens	x
11:11	technomancy	hm; he left. I was going to tell him that he's right, and most plugins being created nowadays should just be libs.
11:34	raj91	Is there any rationale for why `for` doesn't wrap body-expr in an implicit do?
11:34	jcromartie	raj91: because `for` is a sequence comprehension and not for side effects
11:35	jcromartie	also, it basically does
11:35	jcromartie	it uses let
11:35	jcromartie	and do
11:35	jcromartie	so
11:36	jcromartie	,(first (for [i (range 100000)] (println "i =" i) i))
11:36	clojurebot	#<CompilerException clojure.lang.ArityException: Wrong number of args (3) passed to: core/for, compiling:(NO_SOURCE_PATH:0:0)>
11:37	jcromartie	,(first (for [i (range 100000)] i))
11:37	clojurebot	0
11:37	clgv	the ultimative laziness check is with (range) ;)
11:39	AWizzArd	clgv: as soon you want to produce the full range the jvm can be everything, but lazy ;)
11:41	raj91	jcromartie: thanks your explanation makes sense -- but as you showed, only one body-expr is allowed
11:42	clgv	AWizzArd: I always do that to heat my office ;)
11:42	jcromartie	clgv: oh neat, I didn't know there was a 0-arity version
11:58	elben	when running (clojure.core.typed/check-ns), I get the message below, even though (meta ns) is nil:
11:58	elben	Finished collecting myapp.core
11:58	elben	Collected 1 namespaces in 1021.73551 msecs
11:58	elben	Not checking myapp.core (tagged :collect-only in ns metadata)
11:58	elben	sorry for the multi-line. any ideas why it would say “not checking myapp.core”?
12:04	ambrosebs	elben: can you share the ns form?
12:04	elben	ambrosebs: (ns myapp.core (require [clojure.core.typed :refer [check-ns ann] :as t]))
12:05	ambrosebs	elben: should be :require instead of require
12:05	ambrosebs	elben: core.typed uses tools.namespace, which is (correctly) rather picky about the ns form
12:06	ambrosebs	elben: what happened here is core.typed doesn't pick up the ns dependncy of clojure.core.typed, which means the namespace isn't type checked.
12:06	elben	interesting—so some other module is the one that is “non-picky” and allows ‘require?
12:06	stuartsierra	The `ns` macro is notoriously lax about what it accepts.
12:12	edw	In CIDER, is there a(n easy) way to get C-u C-x C-e to automagically put some text (e.g. ";; => ") in between point and the inserted result?
12:19	TimMc	,(ns sandbox (:+ 1 2))
12:20	clojurebot	nil
12:20	TimMc	,(ns sandbox (:println hello))
12:20	clojurebot	hello\n
12:21	technomancy	~egalitarian-ns
12:21	clojurebot	Gabh mo leithscéal?
12:21	technomancy	dang it clojurebot
12:21	technomancy	http://p.hagelb.org/egalitarian-ns.html
12:22	Raynes	technomancy: We should storm into Rich Hickey's home playing Nirvana with that patch in hand.
12:22	technomancy	clojurebot: egalitarian-ns is seizing power from those who have held it exclusively for too long: http://p.hagelb.org/egalitarian-ns.html rise up, workers
12:22	clojurebot	Ok.
12:23	Raynes	Phil and I have grand plans for if that damned thing is ever accepted.
12:23	Raynes	Grand plans, I say.
12:23	technomancy	well it would have to be submitted before it could be accepted
12:23	Raynes	Conch, for starters, would become orders of magnitudes more often.
12:23	stuartsierra	Making tools.namespace even harder to maintain? :P
12:23	Raynes	technomancy: It says something that neither of us is brave enough to do it :(
12:24	Raynes	stuartsierra: I also maintain a similar library. I'm cool with it :p
12:24	Raynes	s/often/awesome/
12:24	Raynes	not enough coffee, can't brain
12:24	stuartsierra	I guess it wouldn't be too bad, but why do you want arbitrary code in the ns declaration?
12:24	technomancy	Raynes: I read that as "conch would come into being orders of magnitude more often" and started wondering if you have some paranormal ability to cross into multiple time streams
12:25	Raynes	Well, yeah, but not relevant.
12:25	technomancy	stuartsierra: because clojure.core/require is not the be all and end all of code loading?
12:25	technomancy	https://code.google.com/p/clj-nstools/ for one
12:25	technomancy	the :like clause is really nice
12:25	stuartsierra	technomancy: I don't claim it is, but why does the alternative have to be inside the ns form?
12:25	technomancy	plus it's a requirement for experimenting with nix-like namespaces
12:26	technomancy	stuartsierra: it just looks gross otherwise
12:26	Raynes	stuartsierra: Yes.
12:26	Raynes	:p
12:26	technomancy	nstools has existed for years, and no one uses it because it's so hideous to hack around not having it in proper ns
12:26	Raynes	I think conch is a pretty great example of why this would be awesome. It'd make it darn near as seamless as amoffat's Python lib.
12:27	Raynes	Without silly metahax
12:27	gfredericks	Raynes: referring shellout commands?
12:27	Raynes	gfredericks: ?
12:27	gfredericks	(:me.raynes.conch/require-commands [ls])
12:27	Raynes	Sure
12:28	gfredericks	wasn't sure if that's what you meant by "it"
12:28	stuartsierra	What's wrong with just putting stuff at the top-level after `ns`? It's all side-effects anyway, even `ns`.
12:28	Raynes	What's wrong with just having this three line change?
12:28	Raynes	:p
12:29	Raynes	I mean, we could do without (ns ..) entirely.
12:29	technomancy	stuartsierra: it's ugly
12:29	technomancy	there's value in maintaining the illusion of declarativity
12:29	Raynes	If this wasn't so trivial, I'd be more opposed.
12:30	technomancy	literally all of functional programming is just making changes to bits in memory somewhere
12:30	technomancy	but we like to be able to pretend it's clean and pure and purty
12:31	tbaldridge	My favorite comeback to those who claim functional purity "yeah, but you malloc call has side effects...."
12:31	technomancy	anyway I've come to grips that it's never going to happen, but it's an interesting thought experiment
12:32	stuartsierra	I'll be the first to line up and say I don't like the syntax of `ns`, but I'm not convinced that adding arbitrary code execution would make it better. :)
12:32	Raynes	stuartsierra: I'll fight you.
12:32	technomancy	allowing experimentation in userspace is always preferable to requiring patches to get applied to clojure itself
12:32	gfredericks	it already has mostly arbitrary code execution
12:32	gfredericks	it's just ugly
12:33	technomancy	gfredericks: (ns my.thingy (:eval (my.other.ns with-args)))
12:33	technomancy	wooo
12:33	gfredericks	right
12:33	stuartsierra	heh
12:34	Raynes	I need a new editor color scheme
12:35	technomancy	actually kinda wish I had somewhere I could sneak this :eval ns thing in
12:35	technomancy	somewhere that isn't leiningen I mean
12:38	gfredericks	technomancy: huh?
12:39	gfredericks	technomancy: oh like you just want to use :eval so that people can see you using it?
12:40	technomancy	"We've secretly replaced these clojure programmers' reasonable ns forms with something containing :eval. Let's see if they notice."
13:06	mirf	hi jimmybot
13:07	mirf	I'm pleased to meet you
13:28	daGrevis	yelp! https://gist.github.com/876ad28c7d70d70bbefa
13:30	stuartsierra	daGrevis: something called `concat` in a loop.
13:30	daGrevis	hmm
13:31	daGrevis	https://gist.github.com/10dc0db756867c4662bf
13:31	daGrevis	only place I'm using concat is build-chain and it always should get the same input
13:31	daGrevis	so why is it failing randomly
13:32	jcromartie	daGrevis: inside build-words you have your loop binding backwards
13:32	jcromartie	I think
13:32	daGrevis	what does it mean?
13:33	jcromartie	sorry, I misinterpreted the loop binding
13:34	stuartsierra	It's `merge-with concat` that's the problem. `concat` is lazy, so each time you call it you build a deeper "stack" of lazy sequences.
13:35	daGrevis	<daGrevis> only place I'm using concat is build-chain and it always should get the same input
13:35	daGrevis	<daGrevis> so why is it failing randomly
13:35	stuartsierra	When one of those lazy sequence nestings is deeper than the max Java stack size, you get a stack overflow.
13:36	daGrevis	ok I guess I understand
13:36	daGrevis	how could I fix it?
13:36	stuartsierra	Replace the sequences with vectors, and use `into` instead of `concat`.
13:37	daGrevis	great! you're awesome :)
13:39	stuartsierra	daGrevis: Thanks! Good luck.
13:59	seangrove	dnolen_: What do you think about a cljs patch that lets users supply a munging function? Then when exporting a js-oriented library, you could export updateIn and swapDanger instead of update_in and swap_BANG_?
14:01	dnolen_	seangrove: I don't see how this isn't already solved by existing export support in Closure and thus ClojureScript
14:01	gfredericks	clojurebot: swapDanger \|would be\| a decent name for a character in an adventure story
14:01	clojurebot	c'est bon!
14:01	dnolen_	this types of things inline well
14:02	seangrove	dnolen_: Thinking specifically about Mori and the complaints about the naming
14:03	seangrove	dnolen_: What's the existing option that closure supports?
14:03	dnolen_	^:export works
14:03	seangrove	dnolen_: Ah, you mean one-by-one, rather than a systemic approach
14:04	dnolen_	for methods you can use externs.
14:04	dnolen_	seangrove: something like that is super low priority
14:04	Fare	hi
14:04	seangrove	Yeah, I'm not saying it's on the critical path for cljs support
14:05	dnolen_	seangrove: feel free to make a ticket and low priority ticket + patch - I'll take a look if you do
14:05	dnolen_	er, you know what I mean
14:05	seangrove	Heh, sure
14:05	Fare	is it possible to have resources that are only available during testing?
14:05	mdrogalis	dnolen_: Saw the Facebook immutable JS library you tweeted about. Pretty neat that it came out of there.
14:05	Fare	i.e. test files for my compiler to execute?
14:05	dnolen_	mdrogalis: very neat, React may very well bring immutability to masses
14:05	mdrogalis	Fare: Check out Leiningen profiles. You can change the resource path.
14:06	mdrogalis	dnolen_: For sure. :)
14:06	ro_st	dnolen_: you must be stoked :-)
14:07	dnolen_	ro_st: exciting times
14:07	Fare	(inc mdrogalis)
14:07	lazybot	⇒ 5
14:08	Fare	ls test_resources a good name for such a resources directory?
14:08	lazybot	home mnt proc sbin srv usr var
14:09	Fare	is, not ls
14:09	arrdem	as good a name as any..
14:09	arrdem	/resources/{test,production} could work too..
14:09	ro_st	o/ arrdem. round two begins
14:10	arrdem	ro_st: o/
14:11	arrdem	ro_st: I'm gonna grab some food and get started on Oxcart for the day, hit me up if I can help
14:13	ro_st	for sure
14:49	ben_vulpes	i'd like to do (clojure.java.shell/sh "cp" "place/dir/*"), but the wildcard isn't hitting bash correctly. how should i use a wildcard with c.j.s/sh?
14:49	amalloy	ben_vulpes: sh does not go through your shell at all
14:49	amalloy	if you want to call bash instead of calling cp, then shell out to bash instead of to cp
14:50	ben_vulpes	gotcha.
14:50	amalloy	(eg, with bash -c)
14:50	ben_vulpes	(sh "bash" "-c" "cp")
14:50	amalloy	yep
14:50	ben_vulpes	thanks, amalloy !
14:50	ben_vulpes	derp dorp o'clock
14:51	amalloy	guys you know what would be amazing? a bash history file that's context-sensitive, so i can see what i did last time i was in this stupid directory a month ago
14:51	technomancy	oh man
14:51	technomancy	I bet that would be pretty easy to write in eshell
14:52	technomancy	as horrible as hash tables are in elisp, it'd be a lot nicer than bash
14:53	hiredman	https://github.com/robbyrussell/oh-my-zsh/tree/master/plugins/per-directory-history
14:53	hiredman	links to some bash versions I guess
14:58	justin_smith	you'd think someone would have made a shell with a decent simple language underlying it (something like sh with extensions using scheme or lua) ... then again I bet it exists and nobody uses it. Hell, it's probably been done a few hundred times by now.
15:02	technomancy	justin_smith: that's kind of exactly what eshell is?
15:02	technomancy	also scsh, but that seems to have stalled out
15:03	xeqi	sounds like a repl
15:03	justin_smith	did scsh ever have proper sh syntax reading though?
15:03	tuft	amalloy: maybe this? http://www.compbiome.com/2010/07/bash-per-directory-bash-history.html
15:03	justin_smith	xeqi: repl with convenient process invocation / chaining, yeah
15:06	arrdem	justin_smith: seems like you just need a faster starting interpreter with a \| operator defined...
15:06	llasram	There's some interesting ideas in the Python `sh` library: http://amoffat.github.io/sh/
15:07	llasram	Aside from the fact that Python's syntax makes REPL usage kind of awkward
15:08	justin_smith	llasram: yeah, there are a bunch of languages that can almost do it, but they make normal shell usage awkward / don't fully support \| > >> & etc. functionality
15:09	justin_smith	even eshell doesn't have all the redirects
15:10	llasram	justin_smith: One of the interesting thing about the Python `sh` library is that it instead has pretty Pythonic equivalents
15:11	llasram	e.g. instead of having a literal "\|" pipe syntax, the return values of subprocess functions are objects which cause the `sh` library to setup process pipelines when passed as input to other subprocess functions
15:13	jinks_	I guess what people are really looking for is not "a shell implemented in $language and leveraging those features" but more of a "bash linked against $language" so they can carry 100% of their bash-isms across
15:13	llasram	Yeah, maybe so
15:14	justin_smith	jinks_: yeah, maybe an escape syntax to embed the extension language, and then define extensions as standard compatible bash functions
15:14	jinks_	shells like bash and zsh have decades of accumulated features and use-cases, so every "new" player on the field has to deal with a torrent of "that's nice, but can it do X like in bash"
15:15	justin_smith	because of course eshell can take input from a file or put output in a file in many elaborate ways (elisp is an editor extension language after all), it just isn't ready for some of my first attempt command line building syntaxes
15:25	amalloy	gfredericks: i need your damn system-exit library, so i can hook it to fire whoever is calling System/exit without even logging a warning saying where or why
15:25	gfredericks	system-slash-exti
15:25	gfredericks	s/exti/exit/
15:26	Raynes	amalloy: grep . -R 'System/exit'
15:26	Raynes	stfu gtfo
15:26	amalloy	Raynes: sure, bro. let me know when you've got that searching all the libraries in my project
15:26	justin_smith	Raynes: I bet there is some git command that will search commits for a string, and report the usernames of people who have checked in said commits
15:27	arrdem	Bronsa: how did you get GitHub to email you diffs on commits?
15:27	arrdem	Bronsa: I can't find that option
15:29	alandipert	amalloy, lein cp \| tr : '\n' \| grep 'jar$' \| xargs grep -R 'System/exit'
15:30	arrdem	Ah. email service. got it.
15:31	amalloy	justin_smith: git log -SSystem/exit --pretty=%an
15:35	amalloy	git log -p -SSystem/exit '--pretty=format:%h %an' is a little better so you can tell why they touched System/exit
15:37	justin_smith	cool, I figured it was possible
15:37	amalloy	today i learned about -S
15:42	Bronsa	arrdem: AFAIK github doesn't offer a way to email all the commits, I just use the dashboard rss feed + rss2email
15:42	arrdem	Bronsa: that works. there's also an email service you can enable per-repo that sends an email with the message and link on commits pushed
15:42	Bronsa	rly
15:52	ro_st	my kingdom for a way to add deps to a lein project without restarting my jvm
15:52	ro_st	ok, maybe not my kingdom. this cookie, perhaps?
15:52	arrdem	have karma need libs
15:54	llasram	ro_st: alembic? https://github.com/pallet/alembic
15:55	llasram	ro_st: And... you mentioned something about a cookie?
15:55	ro_st	:cookie:
15:55	arrdem	(inc llasram)
15:55	lazybot	⇒ 30
15:55	ro_st	if your client supports emoji, eat up :-)
15:55	ro_st	thanks llasram i'll try that!
15:55	arrdem	I need to post my erc-emoji mode..
15:55	ro_st	-adds yak to queue-
15:56	justin_smith	🍪
16:04	mi6x3m	evening, can someone tell me of a project using expectations?
16:06	schmee	mi6x3m: using what?
16:06	mi6x3m	schmee: the library expectactions...
16:06	schmee	ahh, didn
16:06	schmee	't know about that, looks cool
16:07	mi6x3m	schmee: popular and nice testing library :)
16:07	mi6x3m	but cant figure out the naming scheme for test files
16:09	justin_smith	mi6x3m: what kind of content?
16:11	mi6x3m	justin_smith: well give a namespace x.y.z how would you name the test file?
16:11	mi6x3m	x.y.z-expectations?
16:13	justin_smith	x.y.test.z
16:13	justin_smith	under a parallel source tre under test/ rather than src/
16:13	justin_smith	*tree
16:14	mi6x3m	justin_smith: yeah that one's clear :)
16:15	mi6x3m	justin_smith: I would rather use the same tree just append "-test" to every leaf
16:16	justin_smith	sure, that's another option
16:25	mi6x3m	justin_smith: what is the difference between use :only and require :refer ?
16:27	justin_smith	mi6x3m: off the top of my head, I forget, I do know that require :refer is preferred though - it may have something to do with the namespace mappings that are generated
16:28	mi6x3m	justin_smith: seems to be equivalent http://stackoverflow.com/questions/10358149/in-clojure-1-4-what-is-the-use-of-refer-within-require
16:28	arrdem	gfredericks: how is lein lein possibly useful
16:28	justin_smith	mi6x3m: yeah, I just experimented, and they led to identical ns mappings it seems
16:28	technomancy	arrdem: 無
16:28	gfredericks	arrdem: so say you needed to run a leiningen task
16:29	gfredericks	but for some reason you were using leiningen
16:29	gfredericks	lein-lein lets you tie those two together
16:29	gfredericks	we all love using leiningen, but sometimes you just need to run a leiningen task
16:29	technomancy	"It's synergy"
16:29	gfredericks	clojurebot: lein-lein is https://github.com/gfredericks/lein-lein
16:29	clojurebot	Ok.
16:29	arrdem	http://www.arrdem.com/i/upset.gif
16:31	technomancy	gfredericks: what the heck did you do to your readme
16:31	justin_smith	technomancy: it bothers me that emacs tells me that is "CJK IDEOGRAPH-7121" rather than anything mentioning "mu"
16:31	gfredericks	technomancy: huh?
16:32	ro_st	technomancy: where can i find intel on getting a list of all the jars that will load in a lein project?
16:32	technomancy	gfredericks: the copyright sign
16:32	Raynes	what
16:32	technomancy	justin_smith: oh, that is regrettable
16:32	Raynes	I stop in here for 5 minutes
16:32	Raynes	And I get lein-lein
16:32	technomancy	gfredericks: it copies as :copyright: instead of ©
16:32	gfredericks	technomancy: the plugin profile does that
16:33	justin_smith	RaynesRaynes: that's what you get for joining #clojureclojure
16:33	technomancy	huh... maybe github screwed up their rendering
16:33	technomancy	the raw version is fine
16:33	mikerod	gfredericks: is lein-lein a serious thing?
16:33	technomancy	they always seemed a bit loopy for emoji
16:33	mikerod	I'm debating if it is a joke
16:33	technomancy	ro_st: I'm not sure what you mean
16:34	gfredericks	mikerod: I'm interested to see how people use it.
16:34	justin_smith	ro_st: perhaps you want "lein deps :tree"
16:34	amalloy	gfredericks: yes lein \| xargs lein
16:35	mikerod	hmmm
16:35	mdrogalis	gfredericks: Gonna fork it to lein-lein-lein. An improvement. One more, you see.
16:35	mikerod	`lein lein lein lein lein lein lein with-profile -user repl` hmmm
16:35	mikerod	I can see that being useful!
16:35	technomancy	mdrogalis: careful... it's a slippery slope
16:35	mikerod	:P
16:35	arrdem	just in case you type lein a bunch of times thoughtfully before you select a command..
16:36	mdrogalis	technomancy: I don't have a problem! Just one more!!
16:36	amalloy	i could stand to have git-git
16:36	mikerod	arrdem: yes I was thinking if you accidentally forget how many times you've typed lein
16:36	mdrogalis	For hesitation?
16:36	mdrogalis	`git`... "Ah, what did I want to type?.." `-git`....
16:36	arrdem	amalloy: if you build a git-git lemme know..
16:36	arrdem	or I may later today
16:36	amalloy	arrdem: it's easy to build; the trouble is i don't know how to make the tab-completion work for it
16:36	ro_st	technomancy: so i have a project.clj in a lein project. i want to load that project with lein.core, and generate a list of the jars on disk that would result from the dependencies in that project
16:37	technomancy	arrdem: https://twitter.com/technomancy/status/327234371813781505
16:37	amalloy	without which it's pretty pointless
16:37	ro_st	so that i can walk through them doing… things
16:37	arrdem	technomancy: my god
16:37	mdrogalis	Soo good
16:38	justin_smith	(inc technomancy)
16:38	lazybot	⇒ 127
16:38	justin_smith	that command line was fucking amazing
16:39	justin_smith	and congrats, you are almost too big for a byte now
16:40	aeikenberry	hello again!
16:40	aeikenberry	I have my macro working for the most part
16:40	aeikenberry	https://dpaste.de/Ou0T
16:40	aeikenberry	but getting an exception afterwards that I can't seem to shake
16:40	aeikenberry	any ideas?
16:41	aeikenberry	(noob)
16:42	technomancy	ro_st: sounds like `lein classpath` with the directory entries filtered out
16:42	TimMc	justin_smith: Nah man, he's unsigned.
16:42	technomancy	not on the JVM!
16:43	TimMc	:-)
16:43	TimMc	I wish Java allowed you to use char as an unsigned int or whatever.
16:44	justin_smith	aeikenberry: have you tried macroexpanding your call?
16:44	aeikenberry	yeah
16:45	aeikenberry	(if (= 1 8) ((println (quote ok)) (println (quote cool))) ((println (quote nooooo)) (println (quote noooo)) (println (println (quote wellok)))))
16:45	justin_smith	,((println "OK")) aeikenberry:
16:45	clojurebot	OK\n#<NullPointerException java.lang.NullPointerException>
16:45	justin_smith	it prints, that returns nil, then it tries to call nil as a function
16:45	justin_smith	you want do
16:46	justin_smith	(cons 'do (for ...))
16:46	justin_smith	on each branch
16:46	justin_smith	and after that, if you use macros again, learn ` / ~ / ~@
16:46	aeikenberry	that's what i had before, but it would only print the last quote
16:46	aeikenberry	i need a way to ensure they all get printed
16:47	justin_smith	do will not prevent that
16:47	aeikenberry	i still plan on using the those symbols once it's 100%
16:48	aeikenberry	need to learn more about them though
16:48	aeikenberry	like the difference between ~ and ~@
16:48	justin_smith	,(do (defn first-if-key [coll] (if (or (= (first coll) :then) (= (first coll) :else)) (first coll) coll)) (defmacro multi-if [test & args] (let [chunks (partition-by #{:then :else} args)] (let [chunkmap (apply hash-map (map first-if-key chunks))] (list 'if test (cons 'do (for [x (:then chunkmap)] (list 'println x))) (cons 'do (for [x (:else chunkmap)] (list 'println x))))))) (multi-if (= 1 8) :then 'ok 'cool :else 'nooooo 'noooo
16:48	justin_smith	(println 'wellok)))
16:48	clojurebot	#<RuntimeException java.lang.RuntimeException: EOF while reading>
16:48	justin_smith	too big :(
16:49	justin_smith	anyway, that version (which should be reformatted of course) does work
16:49	aeikenberry	ah! i see what you mean now
16:52	aeikenberry	thanks man
16:52	justin_smith	also, you could simplify first-if-key: (if (or (= coll [:then]) (= coll [:else])) ...)
16:52	justin_smith	np
16:52	justin_smith	at least I find the above simpler, that's a style thing I guess
16:57	dagda1	anybody here familiar with the devcards source?
17:01	justin_smith	bhauman is sometimes on this channel (isn't here now it seems)
17:01	justin_smith	$seen bhauman
17:01	lazybot	bhauman was last seen quitting 3 weeks and 4 days ago.
17:10	gfredericks	$ lein do lein do lein do lein do lein do version
17:10	gfredericks	Leiningen 2.4.2 on Java 1.7.0-release OpenJDK Client VM
17:10	gfredericks	^ protip
17:10	aperiodic	gfredericks: time that
17:12	gfredericks	okay
17:12	gfredericks	sys 0m0.190s
17:12	gfredericks	real 0m1.729s
17:13	gfredericks	is there a lein-time
17:13	aperiodic	that's not too bad!
17:14	mdrogalis	Are we still going on about lein-*? D:
17:14	mdrogalis	The revenge of lein-lein-lein.
17:14	gfredericks	aperiodic: it's all just one jvm
17:16	aperiodic	huh, is `do` implemented in the shell wrapper?
17:17	gfredericks	no...that's why it's one jvm
17:19	aperiodic	well, if it were in the shell bit, then wouldn't the only jvm it would need to launch be the one needed for the version task?
17:20	gfredericks	leiningen tasks can run other leiningen tasks without needing another jvm
17:20	gfredericks	like so: https://github.com/technomancy/leiningen/blob/master/src/leiningen/do.clj#L39-40
17:21	aperiodic	and so lein-lein treats the `lein` task as a no-op since there's already a lein running?
17:23	gfredericks	well it's not a noop
17:23	gfredericks	it's the exact same thing that happens when you use `lein do` without any commas
17:25	aperiodic	ah, that makes sense
17:27	hyPiRion	but guess what happens when you do `lein do repl, trampoline repl, repl, trampoline repl`
17:27	gfredericks	O_O
17:28	technomancy	it tears a hole in the fabric of spacetime
17:29	amalloy	i bet lein doesn't like to have some tasks trampolined and others not
17:30	adamt	Hi. Does anybody know when clojars.org will be back up again?
17:30	hyPiRion	amalloy: that's the "tears a hole in the fabric of spacetime" part, along with some repl bugs during shutdown
17:30	technomancy	adamt: it's just the splash page that's down
17:30	aperiodic	heh, `lein trampoline trampoline repl` works
17:30	technomancy	https://clojars.org/cheshire
17:31	amalloy	do you hear that, adamt? that's the sound of hundreds of programmers frantically checking that their builds still work
17:31	aperiodic	for certain definitions of "works"
17:31	adamt	amalloy: most of all i just hear the noise of the fan, trying to keep this room calm..
17:32	adamt	technomancy: thanks :-)
17:32	technomancy	https://github.com/ato/clojars-web/issues/235
17:32	adamt	technomancy: and thanks for all the work you've put into various projects i depend on.
17:33	technomancy	wish I had the time to actually fix it
17:34	gfredericks	(inc technomancy)
17:34	lazybot	⇒ 128
17:36	technomancy	sweet; I made it.
17:36	technomancy	well, it's been fun. catch you cats and dogs later.
17:37	gfredericks	wait wait but what about unicode
17:37	amalloy	technomancy has ascended to join me in #two-byte-clojure
17:38	justin_smith	amalloy: I prefer to call it #short-clojure
17:38	amalloy	justin_smith: i like that better too, but i decided it wasn't quite as obvious, and misses the nice pun on two-bit
17:38	justin_smith	oh, I missed the two-bit pun
17:38	technomancy	also I'm tall
17:39	amalloy	hah
17:39	technomancy	then again, tall is what they call the smallest drinks at starbucks
17:39	amalloy	justin_smith: insensitive clod
17:42	ro_st	Raynes: does refheap expect pygmentize to be executable at the project root?
17:42	ro_st	i'm grabbing your highlight code for grimoire
17:43	Raynes	https://github.com/Raynes/refheap#usage
17:43	ro_st	oh, nm. i see :dir says where to do things
17:43	Raynes	bootstrap snatches pygments
17:43	Raynes	Puts it in the correct place
17:43	ro_st	cool, tx
17:44	Raynes	ro_st: All you really need is pygmentize though
17:44	gfredericks	technomancy: they have short drinks too it's just a secret
17:44	Raynes	ro_st: So if you can just pip install a recent version of pygments, you should be fine.
17:44	ro_st	ok, cool, ta
17:44	technomancy	gfredericks: inconceivable.
17:44	Raynes	ro_st: The reason I'm fetching from bitbucket is because I needed to make patches frequently at first.
17:45	technomancy	gfredericks: some places will even make a cup of french press for you if you ask nicely
17:45	technomancy	what's weird is going to a place where you ask that and they're like "what's a french press?"
17:45	justin_smith	technomancy: I wonder if they will do pour-overs
17:45	technomancy	justin_smith: a bridge too far
17:46	technomancy	french presses are pretty hard to screw up, even if you're not trained. but it's too easy to make a bad pour over.
17:47	technomancy	I had an indie place near where I live stop doing pour overs since there wasn't demand for it... I have to do puppy-dog eyes to get it now, and even then it only sometimes works.
17:47	gfredericks	they've been doing pour-over every time I ask for a decaf for years
17:48	technomancy	no kidding?
17:48	technomancy	like on one of those ceramic cone things?
17:49	hyPiRion	technomancy: what? I think French presses are really easy to work with
17:49	gfredericks	technomancy: eeh, it mighta been plastic
17:49	gfredericks	no guarantees
17:49	amalloy	hyPiRion: that's what he said
17:49	technomancy	hyPiRion: yeah, hard to screw up -> easy to work iwth
17:50	technomancy	gfredericks: still... hum. I will check next time I am forced to fall back on starbucks
17:50	ro_st	thanks Raynes i'm sorted
17:50	amalloy	i figured every building in seattle was a starbucks
17:50	technomancy	amalloy: that's what makes avoiding them such a challenge!
17:50	hyPiRion	oh. I have no idea why I thought you meant they were hard to work with. hrm
17:51	gfredericks	Starbucks: Fall back on us.
17:51	technomancy	gfredericks: it's an acceptable plan B
17:51	Chousuke	I've only drunk coffee at starbucks once. In Japan :P
17:52	technomancy	actually they've started rolling out the clover brewer now supposedly, which might make things more interesting
17:52	Chousuke	they pretty much don't exist in Finland
17:52	technomancy	I guess it all depends on whether they're featuring any blends that aren't dark-roasted to hell on that particular week
17:52	hyPiRion	Chousuke: It's the same over all of Fennoscandia
17:52	gfredericks	technomancy: I use starbucks when I get out of the city
17:52	hyPiRion	No Starbucks here either
17:53	gfredericks	Starbucks: Let us be your lower bound
17:53	technomancy	Chousuke: are there strong existing chains that have filled the gap?
17:53	gfredericks	maybe they don't have the gap
17:54	Chousuke	Also the best coffee ever I had in Japan too. some random old (like, 70-90 years old) guy kept a cafe on the Hachijoujima island. best coffee ever.
17:54	technomancy	Chousuke: from what I can tell the pour-over renaissance in the US is largely fueled by imports from japan
17:54	Chousuke	technomancy: I don't think we have a single chain so much as random coffee shops everywhere.
17:54	hyPiRion	There are incredibly many small ones over here, but no big one.
17:55	technomancy	kind of like how the west lost access to aristotle and only rediscovered his works through arabic translations
17:55	technomancy	huh
17:55	technomancy	but yeah, all the pour over gear is Hario this, Yama that.
17:56	gfredericks	chemex
17:56	technomancy	hyPiRion: is there a general preference for supporting independent shops vs megacorps?
17:56	Chousuke	I make my own coffee with an aeropress mostly because it's decent and very easy.
17:56	Chousuke	it's also trivial to clean, which is really nice.
17:57	technomancy	yeah, taste:effort ratio of aeropress can't be beat =)
17:57	hyPiRion	technomancy: No, I don't think it is intentional.
17:57	technomancy	curious
17:58	Chousuke	there might not be enough people to support massive chain franchises
17:58	Chousuke	every place needs a coffee shop in finland anyway
17:59	hyPiRion	technomancy: I think it's just that Norwegians drink insanely much coffee, and that small companies can actually live fine after bootstrapping
17:59	hyPiRion	I think every person over 15 year old drink 4 cups a day on average.
18:00	technomancy	yeah my grandmother is swedish and she guzzles it like water
18:00	Chousuke	Finns held the world record at some point. don't know if we still do
18:01	amalloy	http://en.wikipedia.org/wiki/List_of_countries_by_coffee_consumption_per_capita
18:01	nathan7	so have y'all built a Clojure-powered coffee maker yet
18:01	hyPiRion	Chousuke: you still do by a large margin
18:02	Chousuke	seems so
18:02	technomancy	sweden's gotta step up their game
18:02	amalloy	nathan7: we can make a mean cup of java
18:02	technomancy	make for an all-scandanavian top-10
18:02	nathan7	amalloy: <3
18:02	Chousuke	I wonder why coffee is so popular in the nordic countries
18:03	technomancy	I'd be interested in seeing it by percentage of adult population that drinks coffee
18:03	technomancy	rather than by kg
18:03	hyPiRion	Chousuke: yeah, it's strange that it's such a dramatic difference
18:04	hoverbear	Hey folks, I'm getting a cljs error (I think?) https://gist.github.com/Hoverbear/3bd455c73786c7ffb70f#file-gistfile1-txt-L66 any suggestions on how to dig into this? My project.clj is at the bottom in comment
18:06	Guest58805	Hi, I have a question about def: https://gist.github.com/kurtenbachkyle/866136fba285385badc9
18:06	Guest58805	Am I missing something about how def works?
18:07	gfredericks	nope
18:07	gfredericks	you're missing something about how quote works
18:07	gfredericks	,(def foo-1 {:a 1})
18:07	clojurebot	#'sandbox/foo-1
18:07	gfredericks	,(def foo-2 {:a 2})
18:07	clojurebot	#'sandbox/foo-2
18:07	gfredericks	,(def foos '(foo-1 foo-2))
18:07	clojurebot	#'sandbox/foos
18:07	gfredericks	,foos
18:07	clojurebot	(foo-1 foo-2)
18:07	catern	foos is a list of the symbol "foo-1" and the symbol "foo-2"
18:08	gfredericks	,(str foo-1)
18:08	clojurebot	"{:a 1}"
18:08	gfredericks	,(str (first foos))
18:08	clojurebot	"foo-1"
18:09	catern	if you want instead to make foos be a list of the values of those variables, use ` and ~
18:10	catern	,`(~foo-1 ~foo-2)
18:10	clojurebot	({:a 1} {:a 2})
18:10	justin_smith	or you could call list
18:10	justin_smith	or use a vector literal
18:10	catern	but preferably call list, yeah
18:10	Guest58805	ok. That makes sense.
18:10	mdrogalis	technomancy: Your cousin, I presume? https://github.com/technoweenie
18:10	gfredericks	,(def other-foos [foo-1 foo-2])
18:10	clojurebot	#'sandbox/other-foos
18:10	gfredericks	,other-foos
18:10	clojurebot	[{:a 1} {:a 2}]
18:10	technomancy	mdrogalis: yeah, we used to get confused for each other all the time
18:10	technomancy	back when I did ruby all the time
18:11	mdrogalis	technomancy: I can't tell if that's serious or not.
18:11	technomancy	no, for reals
18:11	amalloy	technomancy: "get confused for each other" sounds vaguely flirtatious
18:11	mdrogalis	Oh man, haha.
18:11	gfredericks	(inc amalloy)
18:11	lazybot	⇒ 152
18:11	Guest58805	thanks for the help.
18:11	mdrogalis	Pretty good way to hint at someone that you've had an affair.
18:12	mdrogalis	"We were uh.. Confused for each other. Or something."
18:12	amalloy	like, not that people confuse you with him, but that you two run into each other and are like...whoa. hey. sorry if i'm uh. not very articulate. you just confuse me
18:12	technomancy	I blame cycling tab completion
18:12	technomancy	haha
18:12	amalloy	technomancy: i wish my bike had tab completion
18:12	mdrogalis	Heh
18:13	technomancy	for changing gears?
18:13	amalloy	just because technology is fun
18:13	amalloy	i'd be riding along, hit tab, and be content to know that my bike is thinking of a file, or a user in #clojure or something, even if i don't know who
18:13	gfredericks	yeah like when I get on the lake front trail I would hit tab and I'd be at the end
18:14	hyPiRion	nothing like good old M-x bicycle-mode
18:14	amalloy	hyPiRion: i don't think it's possible to resist seeing if that exists
18:14	justin_smith	I tried it, but got pushed off the road by eclipse
18:14	amalloy	but even M-x bic<tab> doesn't work. you'd think there'd be a bic-lighter-mode at least
18:15	technomancy	if there's not a guy out there with a raspberry pi duct-taped to his water bottle holder then I don't even know what the point is
18:15	amalloy	there are probably enough such people to form a small city
18:17	justin_smith	https://sites.google.com/site/unipiper/ speaking of cities with weirdoes and foot powered wheeled vehicles
18:18	justin_smith	https://sites.google.com/site/unipiper/_/rsrc/1402983008858/home/gandalf_banner.jpg
18:23	technomancy	classic
18:33	arohner	tab completion is magical though. I started getting into the habit of hitting tab for things that couldn't possibly be completed, like this text box
18:33	arohner	"tab completion is mag<tab>"
18:34	technomancy	hehe
18:34	justin_smith	arohner: the "it's on the tip of my tongue" button
18:35	hiredman	I had an irssi plugin that would tab complete based on the system dictionary
18:35	catern	tab completion for the english language. driven by a markov chain generator that analyzes all your emails
18:35	catern	good thought
18:41	arrdem	every android keyboard ever tho
18:43	catern	yes
18:45	TimMc	arohner: That's completeable.
18:45	danielcompton	arrdem: and iOS8 keyboards soon
18:45	arohner	TimMc: that specific example was
18:45	arohner	but what about "TimcMc: <tab>"?
18:45	TimMc	wasn't there an XKCD about this...
18:45	arohner	probably
18:45	amalloy	TimMc: tab completion is magenta
18:45	arohner	<tab>
18:47	justin_smith	,(partial \tab)
18:47	clojurebot	\tab
18:47	justin_smith	lol
18:47	TimMc	arohner: http://xkcd.com/1068/
18:48	hoverbear	Are NREPLs TCP or UDP based?
18:49	hiredman	I believe the bencoded nrepl transport is tcp based
18:50	hiredman	but nrepl sits on top of a transport, so there are http transports, message queue transports, etc
18:50	hiredman	the bencoded nrepl transport is what lein uses I think
18:50	technomancy	UDP nrepl transport sounds super sketch
18:50	hoverbear	Yeahh....
18:51	justin_smith	well, localhost only would be fine
18:51	hoverbear	Having a hell of a time connecting to an nrepl in a docker container :S
18:51	justin_smith	hoverbear: by default it is localhost only
18:51	justin_smith	it probably thinks your host container is not "local"
18:51	hoverbear	justin_smith: That'd be a problem.
18:51	justin_smith	hoverbear: you could use ssh -X to make a forwarded port
18:52	justin_smith	or maybe there is a config hiding somewhere to open it to outside hosts (with the scary implications that come with that...)
18:52	hoverbear	justin_smith: I can probably bind it to 0.0.0.0
18:52	rlb	does anyone know to what extent the EPL and LGPL are (in)compatible? i.e. could I use EPLed code to create a guile module, for example?
18:52	hiredman	hoverbear: how are you starting the nrepl server?
18:53	justin_smith	hoverbear: since the idea of opening nrepl to other boxes seems terrifying to me, I always just make an ssh tunnel, but when connecting to a vm, you may want to do things differently, dunno
18:53	hoverbear	hiredman: Via immutant
18:53	hoverbear	https://gist.github.com/tobias/24770e9514cf60752f7c
18:53	hoverbear	justin_smith: For sure, thankfully it's a private link to my host behind a firewall
18:54	amalloy	justin_smith: -X? why would you want to forward the X server?
18:54	amalloy	i think you mean -L or -R, but if i'm wrong i'd be curious to hear why
18:54	hoverbear	justin_smith: You were right, +1!
18:54	sritchie	does anyone know how to do a nested validation with validateur?
18:55	justin_smith	amalloy: yeah, sorry, brain-fart, I meant -L
18:58	danielcompton	justin_smith: https://github.com/mtnygard/ssh-repl
18:58	justin_smith	danielcompton: nice!
18:59	danielcompton	hoverbear: you could use https://github.com/mtnygard/ssh-repl but don't deploy to prod
18:59	rksm	hoverbear: with vagrant I usually have vbox setup a private network and start repl on 0.0.0.0 (config.vm.network :private_network, ip: "192.168.50.4")
18:59	danielcompton	hoverbear: just thinking about it makes me cringe a little
19:03	rlb	After some advice from technomancy I started working on an authenticated repl (shared random token for now), but only got to the server-side before being side-tracked. With any luck I'll get back to it soon.
19:03	rlb	But you'd still need an ssh tunnel (or similar) for remote use.
19:03	hoverbear	danielcompton: Hey, I understand. It's behind a firewall though and the docker link is 127.0.0.1 so it can't be touched outside.
19:04	rlb	This would mostly be to keep from opening up your account to everyone else on the machine when you run the repl.
19:06	amalloy	i look forward to seeing 'privilege escalation via nrepl' hit the front page of hacker news. that's when we'll know we've made it big
19:07	arrdem	lol
19:08	danielcompton	amalloy: "How to take control of any Clojure programmer's computer"
19:09	amalloy	danielcompton: tell him about this new shell that's purely functional. he downloads and runs it. remote code execution
19:09	danielcompton	hahaha
19:10	mthvedt	purely functional shell?
19:10	mthvedt	don’t give haskell programmers ideas
19:10	hoverbear	amalloy: It's not purely functional then
19:11	hoverbear	Pure programs are useless, theres no side effects, like output.
19:11	hoverbear	(Snicker)
19:11	amalloy	mthvedt: to install it, run `echo 'rm -rf --no-preserve-root /' > funshell`; then `sudo funshell` to start it up and you'll be a cool kid in no time!
19:12	mthvedt	amalloy: can’t infer instance of monad
19:12	amalloy	(for the safety of those following at home by just typing in random crap they see, this command will not actually work)
19:12	danielcompton	amalloy: that looks like it would work?
19:13	amalloy	danielcompton: there's no chmod +x
19:13	amalloy	plus . shouldn't be in your PATH for sudo to find
19:13	danielcompton	amalloy: don't worry, I chmod -r +x / in the background on every shell exeution
19:14	mthvedt	amalloy: i automatically set everything to +x for convenience
19:14	amalloy	danielcompton: -R; -r is different
19:14	mthvedt	that way i never have annoying chmod messages
19:14	danielcompton	mthvedt: I know right?
19:16	danielcompton	you know what they say, everything's a hoot when you're root
19:27	DomKM	Design/performance question: If you have a function that returns an object that implements a protocol using the function's arguments, would you use reify within that function or create a type with deftype that implements the protocol? I realize that you can do either, and I'd personally prefer reify if the type isn't used otherwise, but I don't fully
19:27	DomKM	understand the performance implications of reify. Here's an example: https://www.refheap.com/88733
19:28	amalloy	DomKM: reify
19:28	amalloy	the performance difference is basically zero: reify is much like an anonymous deftype
19:29	DomKM	Great, reify it is. Thanks amalloy
19:29	mlb-	How might I find the directory my uberjar is running from?
19:29	technomancy	mlb-: (System/getProperty "user.dir") iirc
19:30	mlb-	err, I mean to say, the directory the uberjar is in, not the working directory where the java -jar incantation takes place
19:30	justin_smith	,(.getCanonicalPath (java.io.File. ".")) here's a fun one
19:30	clojurebot	#<SecurityException java.lang.SecurityException: denied>
19:30	justin_smith	clojurebot: you're no fun
19:30	clojurebot	Gabh mo leithscéal?
19:32	mlb-	technomancy: but thanks, I might be able to make some headway with that =]
19:33	justin_smith	mlb-: well, that gets neither, it gets the directory the user who ran the jar is currently in (the working directory)
19:34	justin_smith	(as does my example)
19:36	mlb-	I'm aware. At $DAYJOB, I'm shipping Clojure code to Windows machines, but the "java -jar" command will be issued by a batch file. Provided the batch file sets the working directory to where the uberjar lives, it should provide the path I'm looking for.
19:40	justin_smith	as for the location of the jar itself... (-> #'myns.core/init meta :file java.io.File. .getCanonicalPath) is one hack
19:40	justin_smith	or wait, that won't get the path to a jar containing myns.core...
19:41	justin_smith	mlb-: bigger picture, an invocation of java could have multiple jars, in multiple directories, all in classpath
19:41	amalloy	i'm not really sure you're entitled to know what jar you're running from
19:41	amalloy	you might not be running from a jar at all
19:43	mlb-	Sorry to be asking an XY Problem. What I've got is my uberjar and a separate jar containing resources I want to access. When deployed upon the end user's machine, they'll reside in the same directory.
19:44	hiredman	is there a reason not to just put the second jar on the classpath?
19:44	amalloy	or indeed include it in the uberjar
19:46	mlb-	hiredman: unless I'm reading wrong, the -jar option clears both the -cp and -classpath options, and the CLASSPATH env var.
19:46	hiredman	mlb-: sure, just don't use the -jar option
19:46	hiredman	java -cp uberjar.jar:other.jar mainclass
19:46	mlb-	amalloy: It's intentionally decided to be kept out of the uberjar to allow a 3rd party service to replace the secondary jar
19:47	mlb-	hiredman: oh? I'll have to try that!
19:47	hiredman	'java -cp uberjar.jar:other.jar clojure.main -m main.namespace' is good too
19:52	mlb-	hiredman: okay, so I certainly see the other jar in the classpath, but how do I access the resources in it? I've tried (clojure.java.io/resource "other.jar"), but no luck :(
19:53	amalloy	(resource "whatever.txt"). you don't need to know what jar it's in
19:53	justin_smith	mlb-: if other.jar is in the classpath, you can access its contents relative to its top level
19:55	mlb-	that's useful!
19:56	mlb-	wait, what if there are resources by the same name in different jars in the classpath?
19:57	hiredman	resources will return a list
19:57	technomancy	huh, I didn't know that was built-in
19:57	amalloy	i didn't either
19:58	hiredman	tada
19:59	amalloy	hiredman: where is resources?
19:59	technomancy	um ... yeah
19:59	technomancy	about that
20:00	hiredman	damn
20:00	hiredman	really?
20:00	hiredman	sorry, I could have sworn that existed
20:00	technomancy	it should exist
20:01	hiredman	https://gist.github.com/hiredman/aeb55c7b54c4a6b48ca1
20:01	hiredman	err
20:01	justin_smith	hiredman: man, I was gonna say, mind=blown because I assumed first resource matching a path hid all else...
20:01	amalloy	justin_smith: i think it does, really
20:01	technomancy	justin_smith: turns out I implement resources twice in Leiningen
20:02	hiredman	yeah, just not in clojure.java.io
20:02	technomancy	justin_smith: you can "see through" the first thing if you learn the secret
20:02	hiredman	but you can call getResources on a classloader
20:02	hiredman	I could have sworn that existed
20:02	hiredman	heh
20:03	hiredman	it does, in the utils namespace at work
20:03	hiredman	and getResources returns an enumeration yetch
20:03	amalloy	hiredman: at least there's enumeration-seq
20:03	hiredman	yep
20:03	technomancy	avert your eyes because this is terrible, but https://github.com/technomancy/leiningen/blob/master/src/leiningen/release.clj#L106
20:04	hiredman	http://stackoverflow.com/questions/948194/difference-between-java-enumeration-and-iterator
20:04	hiredman	so it sounds like enumerations are 100% better than iterators
20:05	amalloy	hah
20:05	mlb-	so.. the actual answer is that clojure.java.io/resource will return the file belonging to the last .jar in the classpath?
20:07	hiredman	mlb-: it is slightly more complicated than that, because in reality resources are loaded from classloaders, the most basic of uses this thing called a classpath, but there are many others that don't
20:07	mlb-	ah, okay. Good to know
20:08	hiredman	I could point a a url classloader to a jar on a webserver somewhere, set that as the context classloader, then io/resource would load from the jar file on the webserver
20:10	gfredericks	hiredman: ha re SO
20:12	amalloy	gfredericks: 'ha' is not a musical note
20:12	mlb-	hiredman: so then... I should do my own conflict resolution bit?
20:13	amalloy	why is there a conflict? this resource is supposed to be in other.jar; are you expecting it to be somewhere else as well?
20:13	mlb-	amalloy: Yes. I make mistakes and I've got coworkers.
20:15	amalloy	but you think you can write code that will automatically and correctly handle that conflict? i think you're better off just saying "don't have two of these"
20:18	mlb-	amalloy: Were I not working on a team, such a simple soltuion would suffice.
20:19	mlb-	instead, I'll just write code to compare the name of the jar that contains a resource.
20:27	technomancy	who was asking about names yesterday? http://random-name-generator.info/
20:27	technomancy	I still prefer gfredericks' approach if you need something in a jiffy, but there you have it
20:28	technomancy	[dakota/lightfoot "0.1.0"] ; works for me
20:29	justin_smith	(mana/quarles "0.0.1-SNAPSHOT")
20:29	technomancy	needs an api and a lazybot plugin
20:39	mlb-	technomancy: Now I want to know, what's the gfredericks' approach?
20:40	technomancy	mlb-: hit wikipedia's "random article" till you find a person; use their name
20:40	technomancy	or maybe it's just till you find a proper noun?
20:40	gfredericks	personal name
20:40	technomancy	there we go
20:41	gfredericks	I am known primarily for my library naming strategies
20:41	TEttinger	##(clojure.string/join " "(repeatedly 998(fn[](apply str(concat(rand-nth[[(rand-nth["B""D""G""C""F""T""K"])(rand-nth["r"""""""""])][(rand-nth["S""L""Th""Ch""Sh""L""Y""W""Z""V""M""N""T"])]])(take(+ 2 (rand-int 3))(interleave(repeatedly #(rand-nth ["a""o""e""i""u""a""o""e""i""au""oi""ou""oo""ai""ee"]))(repeatedly #(rand-nth ["s""p""t""n""m""b""mb""rd""g""st""f""rt""sp""ch""rl""x""sh""y""ng"])))))))))
20:41	lazybot	⇒ "Kourt Wey Frespo Gosh Tog Baifaust Dorda Chombi Diy Nost Thetosh Yoonasp Toirdord Leng Yoyi Mang Lerla Tispemb Soorlerd Bairtoy Cuchoo Ceebo Vespa Zarl Noim Mooti Lap Tusoof Gigort Wogit Yaub Moospu Wapamb Taum Bartoost Meeng Faufe Daitu Fif Moonging Chaub Goox Vegu... https://www.refheap.com/88735
20:42	TEttinger	star wars names!
20:42	technomancy	that's eerily good for something that looks like a swearjure program gone wrong
20:42	TEttinger	haha it really hits the limits of an IRC message
20:44	gfredericks	what's a swearjure program gone right look like?
20:44	TEttinger	(imv gfredericks)
20:44	TEttinger	(inc gfredericks)
20:44	lazybot	⇒ 78
21:05	technomancy	gfredericks: well, it has fewer alphanumerics than the above =)
21:13	gfredericks	technomancy: alphanumerics in a swearjure project are technical debt
21:15	gfredericks	omg guys these toy train sets are so topological
21:16	gfredericks	I just accidentally discovered a sort of semi-orientable track layout
22:17	TimMc	gfredericks: C-loop?
22:20	gfredericks	eh?
22:20	TimMc	Maybe that's not the name for it.
22:20	gfredericks	I'm thinking a track is orientable if a train, by only moving forward, cannot flip itself around
22:20	TimMc	yeah
22:21	TimMc	A C-loop allows a train to turn around.
22:21	gfredericks	and this track allows the train to flip one way but not back
22:21	TimMc	Maybe it's just called a reverse loop.
22:21	TimMc	heh, nice
22:22	TimMc	gfredericks: I had a moment the other day while trying to maneuver a large desk up a flight of stairs where I was thinking "OK, we have to make sure to turn it the right number of times so that it doesn't end up inverted front to back" and then I remembered that my house is at least locally an orientable space.
22:23	gfredericks	ha
22:23	gfredericks	here's the simplest version of the semi-orientable track: http://upload.gfredericks.com/butt.png
22:23	gfredericks	(it kinda looks like a butt)
22:23	TimMc	A housemate pointed out that this would have been OK with them (just turn it around), but I think it would have made any future disassembly with a screwdriver rather surprising.
22:24	gfredericks	if you're heading north up the middle you can't ever switch to heading south down the middle
22:24	gfredericks	but the opposite is the opposite
22:25	TimMc	gfredericks: And you can only traverse the top bar once.
22:26	gfredericks	true
22:26	gfredericks	I realized you can formalize it as a directed graph, where the track pieces between the intersections correspond to two vertices
22:26	gfredericks	one for heading one way and another for heading the opposite way
22:26	gfredericks	and the intersection specifies what points to what
22:28	gfredericks	so in this case the graph would have two connected components, where one has two edges into the other, both of which are the top bar
22:30	TimMc	gfredericks: So what is the simplest track such that for all points there is a direction a train may be placed on that point such that it has the possibility to traverse the entire track and also may make a decision that will restrict it from ever reaching certain portions of the track but will still ahve the opportunity for indefinite forward motion.
22:31	TimMc	I believe it is this: https://i.imgur.com/9LaZFDg.png
22:32	TimMc	Hmm, no. That violates the "any point" rule.
22:32	TimMc	How about "there exists a point and direction".
22:33	gfredericks	we need a word for point + orientation
22:34	TimMc	Let us definte point to include orientation. :-P
22:35	TimMc	"directed point"
22:36	mthvedt	i studied the topology of paths over topological spaces in colleg
22:36	mthvedt	e
22:36	mthvedt	but i can’t help because i forgot all of it
22:37	Jabberz	so question on functions that call databases... would it be better to pass in the db-spec/connection pool as a parameter, or retrieve it within the function via a var or calling another function
22:38	Jabberz	ie should something like (defn lookup-user..) be more pure, by taking a db-spec as a param
22:39	sritchie	dnolen_: for a case where a parent component needs to distribute some async state to a bunch of subcomponents,
22:39	sritchie	dnolen_: (thinking data for autocomplete fields),
22:39	sritchie	dnolen_: would you recommend a core.async mult, or just sticking a future in init-state?
22:40	sritchie	oh, I guess there’s no future in cljs
22:40	TimMc	gfredericks: Here's the graph of the "g" track: https://i.imgur.com/YrzZzZw.png Three pairs of points for the three intersection-divided segments.
22:41	jeremyheiler	Jabberz: pass the db in.
22:41	sritchie	dnolen_: I can jam it into the global state so everyone can just share
22:42	Jabberz	jeremyheiler: that's what I was leaning to, eliminate those internal dependencies within the functions
22:43	jeremyheiler	yep. if you really need to, you could close over the db with an anonymous fn and pass that somewhere that doesn't have access to it.
22:43	jeremyheiler	Jabberz: ^
22:49	Jabberz	jeremyheiler: does this naturally lead you towards something like component or trapperkeeper? I feel like as soon as you start managing abstractions with closures or partials, might as well formalize the dependency relationships between, for lack of a better term, modules or groups of functions
22:55	TimMc	sritchie: You heard it here, folks, "there’s no future in cljs".
22:55	sritchie	:)
23:02	jeremyheiler	Jabberz: at a system level, yes. internally, not really.
23:22	justin_smith	TEttinger: silly 80 column restriction of your star wars name generator https://www.refheap.com/88737
23:22	TEttinger	ha, 16 lines in a single message
23:23	justin_smith	well, 16 is one way to break it up
23:23	justin_smith	did you make it as a one liner, or do you have a more normal breakup of the original?
23:25	TEttinger	I made it as a one liner submitted to lazybot
23:25	TEttinger	lazybot is not here?
23:26	timsg	Hey, has anyone here used Clojure-CLR, or, better, have a Clojure-CLR project set up that they can run a repl against? I’m mutating records with set! and need a sanity check, maybe I messed up the build or something
23:28	justin_smith	timsg: how do I run it?
23:28	justin_smith	(I installed it ages ago, and I forget)
23:28	justin_smith	never mind, I have it
23:29	justin_smith	timsg: got a repl log I can follow along?
23:29	timsg	justin_smith: great, I’ll send you a refheap thing in a minit
23:30	timsg	justin_smith: https://www.refheap.com/88738
23:30	timsg	justin_smith: oops, typed too fast, that won’t work
23:30	timsg	justin_smith: https://www.refheap.com/88739
23:32	danielcompton	In Cider, is there a way to set the point back onto the repl entry point? So if I've scrolled up three pages and want to get back to the 'REPL' input line, what would I press?
23:34	jeremyheiler	danielcompton: M-> wil jump to the end of the buffer
23:34	danielcompton	jeremyheiler: thanks! That's just what I needed
23:34	jeremyheiler	danielcompton: that's a general emacs trick. not cider specific
23:35	justin_smith	timsg: I got identical results
23:35	timsg	justin_smith: o god that’s terrible
23:35	danielcompton	jeremyheiler: yeah, I was trying it on other buffers and noticed that.
23:35	timsg	justin_smith: I think this means all user-defined types in Clojure-CLR are mutable
23:37	timsg	justin_smith: what version of Clojure-CLR are you using, and what are you running it against (Mono or something else?) I’m using 1.5.0 and running it against Mono on OSX. I hope this is some undiscovered bug endemic to that configuration
23:39	timsg	mono 3.4.0
23:40	justin_smith	I have an old one, 1.4.1
23:40	timsg	bbloom: if you’re there, might need some clr insight
23:43	swgillespie	i've got clr insight but not clojure insight :p
23:45	mindbender1	How can I resolve the following clojure compilation errors: https://gist.github.com/johnbendi/9b4f5f1f46c270f63b51#file-clojure-compilation-failure ?
23:46	sritchie	can anyone help me figure out how to call this .typeahead function via cljs? https://github.com/twitter/typeahead.js/blob/master/doc/jquery_typeahead.md
23:46	sritchie	it looks like the jquery plugin has added the function to any object
23:47	sritchie	BUT I can’t just call .typeahead, that seems to fail
23:51	amalloy	it doesn't put .typeahead on every object (it can't); it just puts it on all jquery elements
23:51	DomKM	sritchie: Can you call it on any object from plain js instead of cljs? The readme leads me to believe it's available on jquery objects, not any object.
23:51	trptcolin	gfredericks: good job on lein-lein. you’re truly doing the Lord’s work
23:51	sritchie	looks like I had to wrap it in $, yeah, my bad
23:51	sritchie	okay, getting closer, thanks