#clojure – 2016-03-15

2016-03-15

00:08	mmercer	i was banned in #haskell but not in here. if that doesn't prove that clojure is superior I don't know what does
00:19	ben_vulpes	patience is next to godliness or what was it
00:23	tolstoy	Should it be possible to (using component) do something like (merge (system/map :some thing) {:other thing} {:still-other thing})?
00:26	tolstoy	Hm. Seems to return a SystemMap, so that's a yes, I guess.
01:19	pmbauer_	somebody please ban mmercer; those comments are inappropriate, sexist, unprofessional and do not represent the Clojure community
01:20	tolstoy	Doesn't seem to be an op.
01:20	rhg135	I'd hope not
01:21	rhg135	I think they just don't idle as op
01:23	tolstoy	Plus it's late. I wonder how they'd get notified?
01:24	rhg135	PushX?
01:25	tolstoy	I think puredanger is one. amolloy another?
01:25	tolstoy	Or, amalloy.
01:26	rhg135	Yeah
01:26	rhg135	I don't think amalloy sleeps anyway
01:44	whunix	Hi, what's a good tutorial for beginners?
01:44	whunix	I am going to read about Racket HTDP book and then move on to Clojure from that
01:44	whunix	what book would be appropriate to read after that?
01:45	whunix	in terms of Clojure
01:45	faxmodem	whunix: clojure from the ground up
01:45	faxmodem	its online from aphyr
01:45	tolstoy	Just listened to a podcast with the author of: http://www.braveclojure.com. He said he approached it as he learned it being a Ruby programmer.
01:45	faxmodem	then maybe clojure for the brave and true
01:46	whunix	faxmodem, where's the book? Can't find it :(
01:47	tolstoy	https://aphyr.com/posts/301-clojure-from-the-ground-up-welcome
01:48	whunix	tolstoy, is that all?
01:48	tolstoy	There's a link at the bottom to the next chapter.
01:49	tolstoy	Summary: https://aphyr.com/tags/Clojure-from-the-ground-up
01:51	whunix	isn't it better to just jump into Braveclojure?
01:52	tolstoy	Sure. I think it depends on your familiarity with functional programming, or Lisp, etc, etc, or your general programming experience. Might start with one of the books. Clojure Programming, or what have you.
02:49	irctc	In http-kit I want to defer the answer of http request, like return a future or something, what is the idiomatic way to do that?
03:04	Empperi	so you mean when using http-kit client and you make a request and want to get the response via defer?
03:04	Empperi	it's built in http-kit client http://www.http-kit.org/client.html#sync
03:05	Empperi	if you are talking about having http-kit as a server and calling it from browser with clojurescript then the short answer is "you don't"
03:05	Empperi	the long answer is "use core.async" although that too won't be synchronous
03:05	clojurebot	Ik begrijp
03:06	Empperi	but close at least
03:06	Empperi	and wtf clojurebot
03:06	Empperi	you messing with me? lol
03:24	amalloy	~the long answer
03:24	clojurebot	the long answer is "use core.async" although that too won't be synchronous
03:28	Empperi	lol
03:34	opqdonut	:D
05:15	CStorm	so anyone have a read/video for me who are looking to get insight in different database options?
05:15	CStorm	reading "mongodb sucks" is getting old, so trying to get a more insight in alternatives.
05:16	CStorm	or atleast to why mongodb sucks ;)
05:21	Empperi	mongodb doesn't have relations in any way and it doesn't have transactions and it doesn't have any support for schemas
05:21	Empperi	that enough for you why mongodb sucks? :)
05:21	prohobo	mongodb sucks
05:21	Empperi	but in truth mongodb doesn't suck
05:22	Empperi	but it sucks as a general purpose database
05:22	prohobo	its webscale
05:22	Empperi	for specific use case scenarios mongo is awesome
05:23	CStorm	what db should i read up on as a general purpose database?
05:23	CStorm	specific for clojure
05:23	prohobo	CStorm: postgres prolly
05:23	ridcully_	CStorm: postgres
05:23	CStorm	haha
05:23	CStorm	:D
05:23	prohobo	:\|
05:23	CStorm	goodie, thanks guys.
05:24	Empperi	yeah, postgres is a great choice for a lot of stuff
05:24	CStorm	being at ios and osx for 6-7 years and shifting to clojure has it ups and downs :)
05:24	CStorm	one of them being a mess to figure out the whole db mess.
05:25	Empperi	you've never used databases?
05:25	Empperi	you're up for a rough ride my friend :)
05:25	CStorm	used coredata for ios
05:25	CStorm	but else no
05:26	CStorm	might be rough but looking forward to it :)
05:26	CStorm	gotta start somewere right
05:26	prohobo	whats so rough about using dbs
05:27	Empperi	well for starters, you have learn a new language: SQL
05:27	Empperi	and to understand how relational databases work
05:28	Empperi	also there's transactions and transaction isolation levels
05:28	Empperi	there is a reason why there are people whose sole purpose is to do stuff with relational databases :)
05:28	Empperi	same thing goes for other types of databases
05:29	Empperi	not that it won't be a truly enlightening journey to learn all that stuff but it's going to be a rough ride and there's tons of new stuff to learn
05:30	CStorm	well what kinda programmer would i be if i just would say "ugh im already behind the wagon i dont wanna learn it"
05:30	CStorm	im looking forward to it, no matter how "rough" it will be :)
05:31	i-blis	CStorm: if you're looking for a good read on DBs, give "seven databases in seven weeks" a shot; it gives a nice overview on what you get with the different paradigms out there
05:31	CStorm	thank you i-blis, will look that up right away
05:32	Empperi	CStorm: indeed, that's the right attitude :)
05:35	prohobo	not a CS major?
05:36	ridcully_	knowing SQL and what comes with it is IMHO a valuabe skill for a developer. but maybe i am just old
05:37	ridcully_	also starting with it on most OS is just a `sqlite3` call away
05:49	CStorm	I bought the book now i-blis– looking forward to it.
05:53	i-blis	CStorm: hope you will enjoy it as I did (despite being a long time SQL user, I enjoyed even the chapter on PostgreSQL)
05:57	hyPiRion	I think sqlite3 is easier to get started with, but Postgres is really good once you've managed to get it up and running. That isn't always as straightforward as one would like though.
06:13	Empperi	ridcully_: no, you are not just old. You are correct
06:14	Empperi	that's especially true with Clojure and other languages with powerful set operations
06:15	prohobo	but you are also old
06:18	Empperi	it's all relative :)
06:18	Empperi	34 years old
06:18	Empperi	I personally think I'm still young and in the beginning of my career
06:21	hyPiRion	Yeah, 34 is not much.
07:14	keksi_	I have a problem modifying my data so I can handle it. I want to modify the structure from the from.clj to to.clj kind of data: https://gist.github.com/Keksike/77477abf2bfcbd5754ee , what should I do?
07:14	keksi_	I dont think cheshire had a proper function for this one
07:17	faxmodem	I'm playing with overtone/at-at
07:18	faxmodem	when I'm running a function using (at/every 10000 #(myfunc a b c)) it doesn't behave the same as if run as (myfunc a b c)
07:18	ridcully_	keksi_: pass true as second argument to parse-string. then you get keywords
07:19	keksi_	ridcully_: I tried that too, but it got rid of the [ & ]
07:19	faxmodem	e.g. the side effects of myfunc are not observed when running it inside an every block
07:20	keksi_	ridcully_: (and it put it in parantheses instead)
07:21	ridcully_	yes. its a list then instead of a vector.
07:22	keksi_	how do I convert it to a vector?
07:23	ridcully_	,(vector '({:a :b}{:c :d}))
07:23	clojurebot	[({:a :b} {:c :d})]
07:24	ridcully_	DOH
07:24	ridcully_	s/vector/vec/
07:24	keksi_	wow, thanks :)
07:25	faxmodem	or into
07:25	keksi_	which is preferred?
07:25	faxmodem	good question :)
07:25	faxmodem	in terms of readability I'd vote for vec
07:26	ridcully_	in terms of me remembering vector and vec i vote into ;)
07:27	faxmodem	:-D
08:18	kungi	Is there a difference between (vec) and (into [])
08:31	hyPiRion	kungi: not semantically
08:32	hyPiRion	But vec is likely more performant.
08:33	MJB47	iirc it isnt
08:33	MJB47	but i might be remembering wrong
08:34	MJB47	it was tested by someone here a few weeks back
08:37	Empperi	kinda surprised if vec isn't faster
08:37	Empperi	it uses PersistentVector.create() under the hood
08:37	Empperi	while into uses transducers and conj
08:44	MJB47	http://insideclojure.org/images/20150107/numbers.png
08:44	MJB47	from http://insideclojure.org/2015/01/07/vec-perf/
10:22	digiorgi	Hi, i'm trying to use clojure with maven, to get a good integration with other ides clojure + java. But when i'm starting CIDER on emacs, connecting from maven i get errors.
10:22	digiorgi	WARNING: The following required nREPL ops are not supported:
10:22	digiorgi	apropos classpath complete eldoc format-code format-edn info inspect-pop inspect-push inspect-refresh macroexpand ns-list ns-vars ns-path refresh resource stacktrace toggle-trace-var toggle-trace-ns undef
10:22	digiorgi	Please, install (or update) cider-nrepl 0.12.0-SNAPSHOT and restart CIDER
10:22	digiorgi	but i have installed the required version
10:33	justin_smith	is the cider-nrepl middleware properly loaded? I thought that it wouldn't work without lein.
10:34	justin_smith	well I guess boot maybe uses it too somehow? anyway, you'll need to do something to make it run properly in maven, it won't be automatic.
10:52	X-warrior	how would you guys represent your own tree structure? a map of maps?
10:52	justin_smith	that or a deftype
10:54	X-warrior	I was doing an exercise for an interview and represented the tree as a deftype. Today I received their answer saying that my code is too much OO and they wanted more functional programming
10:54	justin_smith	interesting - did you use mutable fields?
10:55	X-warrior	justin_smith: yes I did
10:55	justin_smith	aha, yeah, I might use deftype, but I would not use mutable fields unless absolutely neccessary
10:56	X-warrior	well, you have 1 node with one child... you add a new node, how can you add it without using mutable fields?
10:57	justin_smith	the same way you do with any other immutable data type - you have an operation that returns a new instance containing the new node
10:57	X-warrior	I guess my problem was that I created some deftype methods to add new node/retrieve specific node and others
10:57	justin_smith	eg (cons 1 ()) returns '(1) but () is not altered
10:58	justin_smith	X-warrior: you can still have methods to insert and retrieve - it's just a different style when working with immutable data
10:59	justin_smith	where of course insert returns a new tree (that can implicitly share structure with the original, since they are immutable, depending on implementation), and the retrieval would work the same as before
11:04	X-warrior	justin_smith: do you mind to take a look at the code to point some improvements that I could do, so it became more functional instead of OO ?
11:04	X-warrior	besides the mutable fields
11:05	X-warrior	now that I already failed at the test and spent so much time doing it, I would like to get better at this
11:06	justin_smith	X-warrior: I can take a look later, but really the mutable fields thing is what will make the biggest difference (I say without even seeing the code)
11:07	Glenjamin	what was the exercise brief?
11:09	X-warrior	Glenjamin: I had to do a reward system where a customer wins points for invitee from invitee on a 0.5^k
11:10	X-warrior	so direct invitee will give me 1 points, 0.5, 0.25...
11:11	ystael	you had to implement a pyramid scheme kickback mechanic as an interview question? o_O
11:11	justin_smith	just watch, they'll use his code to implement uber but for amway
11:11	X-warrior	ystael: well, it wasn't an interview question... it was one part of the hiring process
11:12	Glenjamin	hrm, my instinct is to record the parent of each customer with something like {:parent <parent-id>}
11:12	Glenjamin	and then when adding a new customer, traverse the graph and add points
11:12	X-warrior	https://bitbucket.org/matheusbrat/clojure/overview
11:13	Glenjamin	i guess something like {<id> { :parent <id>, :points n, :id <id> }, ...}
11:13	justin_smith	X-warrior: one issue with using a deftype with mutable fields, is that one assumes you will be using this structure in a webapp or some other process that handles concurrent clients, and mutable fields don't tend to handle concurrent update very well. This is kind of why we use immutable data structures with containers like atoms / agents / refs
11:14	justin_smith	otherwise you end up needing to mess with locking and such
11:14	X-warrior	Glenjamin: that seems as a nice approach
11:16	X-warrior	justin_smith: got it
11:27	ilevd	Hi, clojurebot
11:27	ilevd	,(print "Hi!")
11:28	clojurebot	Hi!
11:39	ilevd	Ask, please
11:41	cortexman_	what's the deal with paredit not allowing you to insert the parens that will fix the mismatch?
11:42	ilevd	I don't know
11:43	MJB47	speaking of paredit
11:44	MJB47	(f\| (f2 v)) -> (f2 v)
11:44	MJB47	whats the command called?
11:44	MJB47	\| being point
11:44	luma	raise
11:44	MJB47	ty
11:56	justin_smith	cortexman_: I've found paredit handles missing ) that you insert by hand nicely, but not other kinds of closing delimiters, which makes me think it's a bug
11:56	justin_smith	cortexman_: the easiest thing is to turn off paredit, fix the matching, then turn it back on again.
11:57	justin_smith	cortexman_: but you can also use commands like kill and yank without being affected by paredit, so that's another approach
11:57	cortexman_	got some good tips from #lisp. i.e., to fix ((( ))) to ((( ])), stick the cursor in the space in the middle, M-z ) C-q ] M-z is zap-to-char and C-q is quoted-insert
11:58	justin_smith	ahh, nice
11:58	justin_smith	so yeah, zap-to-char and quoted-insert must also be outside paredit, like kill-region and yank are
11:58	cortexman_	yeah
11:59	cortexman_	i think it's fewer keystrokes this way, maybe.
11:59	justin_smith	also in evil mode, r is not effected by paredit, so you can insert spaces and use r to change each one to the right kind of delimiter
11:59	cortexman_	i have been thinking about switching to evil mode, but have not taken the dive
12:00	justin_smith	I switched, and I liked it so much that I eventually started using vim
12:00	justin_smith	haha
12:00	cortexman_	lol.
12:00	cortexman_	this is a good reason to use emacs: https://github.com/RallySoftware/rally-emacs
12:01	justin_smith	in terms of being able to extend it to make apps, emacs is at least as good as the browser (though not nearly as pervasively deployed)
12:05	justin_smith	wooh, this guy used my webapp to help write this article! http://www.vox.com/2016/3/15/11225128/donald-trump-gullible-internet
12:05	justin_smith	(webapp is full stack clojure/cljs)
12:06	justin_smith	direct link to the part that came from our app http://www.getlittlebird.com/blog/data-62-of-the-people-donald-trump-rted-this-week-follow-multiple-white-supremacist-accounts
12:27	TimMc	Fun!
12:31	justin_smith	As compared to what my prior work has been used for, it's definitely gratifying to see the product of my sweat used that way.
12:32	TimMc	To the extent to which I can get excited about anything with "social media" in the title, this is neat. :-P
12:32	justin_smith	heh, it's a fun app to work on
12:33	TimMc	(my social media strategy is the LeechBlock browser add-on)
12:59	sfz-	XD
13:43	tolstoy	I use Control-Q (quote) and Control-W (delete) to repair par-edit stuff.
14:09	ben_vulpes	does anyone know where the datomic provisioning clojure lives, if it's even available to the public?
14:13	amalloy	isn't life a lot easier if you just don't break paredit stuff? how does everyone end up with so many unbalanced expressions?
14:17	TimMc	amalloy: It sometimes happens. I have no idea how.
14:17	TimMc	Paredit does have at least one bug around comments, though.
14:36	sdegutis	Welcome.
14:51	sdegutis	Welcome to #clojure.
14:53	TimMc	anything is possible at #clojure
14:53	TimMc	the only limit is yourself
14:55	rcg	,(/ 1 0)
14:55	clojurebot	#error {\n :cause "Divide by zero"\n :via\n [{:type java.lang.ArithmeticException\n :message "Divide by zero"\n :at [clojure.lang.Numbers divide "Numbers.java" 158]}]\n :trace\n [[clojure.lang.Numbers divide "Numbers.java" 158]\n [clojure.lang.Numbers divide "Numbers.java" 3808]\n [sandbox$eval25 invokeStatic "NO_SOURCE_FILE" 0]\n [sandbox$eval25 invoke "NO_SOURCE_FILE" -1]\n [clojure.lang...
14:55	rcg	or the laws of mathematics...
14:55	rcg	scnr
14:56	sdegutis	You can do /anything/ at #clojure.
14:57	TimMc	rcg: http://zombo.com/
14:57	sdegutis	Anything at all.
14:58	sdegutis	This is #clojure, and welcome, to you, who have come to #clojure!
14:59	sdegutis	The infinite is possible at #clojure. The unattainable is unknown at #clojure.
14:59	tolstoy	paredit rarely breaks for me, but sometimes I inadvertently cut/paste something wrong from a non-Emacs buffer.
14:59	sdegutis	Oh man. This site takes me back.
14:59	sdegutis	tolstoy: that's what undo-tree is for
14:59	tolstoy	Missing Plug-in?
14:59	sdegutis	because Emacs without undo-tree is literally just pure unadulterated insanity.
15:00	rcg	TimMc, sorry, due to lack of plugins, i cannot watch that, even got my linux box from the shelf to check. but according to the wikipedia article it sounds quite entertaining :)
15:02	mavbozo	you can get "anything" from infinity at #clojure
15:02	mavbozo	,(get Double/POSITIVE_INFINITY get "anything")
15:02	clojurebot	"anything"
15:06	sdegutis	mavbozo: bravissimo
15:07	sdegutis	TIL you can intensify Italian words like "bravo" by adding "issim" after the final consonant
15:07	sdegutis	Also pianissimo means "very quiet"
15:07	sdegutis	er, very soft
15:17	mavbozo	sdegutis: grazie infinite!
15:22	sdegutis	Today my job is to normalize internal data using a sequence of regex-based migrations.
15:23	sdegutis	What can go wrong?
15:23	ystael	sdegutis: where are you located, roughly? i want to know whether i need to get a car and start driving
15:23	sdegutis	:D
15:37	X-warrior	justin_smith: I changed my solution to use a map as proposed. Did you have time to take a look on the previous version?
15:59	mavbozo	,(:a.b.c/foo {:a.b.c/foo :bar})
15:59	clojurebot	:bar
16:00	mavbozo	is it okay to have keyword which has '.' in its namespace part?
16:01	mavbozo	but this http://clojure.org/reference/reader#_literals says keywords cannot contain '.'
16:02	hiredman	it is refering to the name part
16:03	hiredman	keywords and symbols have name and namespace parts
16:05	hiredman	the set of allowed characters for symbols and namespaces on th reader page don't match the characters the reader actually accepts, and are just for symbols and keywords created via the reader
16:06	mavbozo	got it. thanks hiredman
16:10	hiredman	,:foo.bar
16:10	clojurebot	:foo.bar
16:10	hiredman	,clojure-version
16:10	clojurebot	{:major 1, :minor 8, :incremental 0, :qualifier nil}
16:34	TimMc	,(keyword ":")
16:34	clojurebot	::
16:46	sdegutis	Does something like distinct-by make sense?
16:47	sdegutis	So that (distinct-by first [0 :a] [0 :b] [1 :c]) would -- oh wait
16:47	sdegutis	Hmm. Guess I need to transform my data.
17:29	sdegutis	Bye.
17:38	sdegutis	What's an agreeable way to tell if any item in a collection satisfies a given predicate?
17:38	sdegutis	Something like (any? coll f)
17:39	cfleming	sdegutis: (some f coll)
17:39	sdegutis	Hmm, strange name.
17:39	sdegutis	Thanks Colin.
17:39	sdegutis	I had a feeling that was the function, but it seemed like maybe I was going insane.
17:39	justin_smith	,(some nil? [false nil true])
17:40	clojurebot	true
17:40	sdegutis	Thanks Justin.
17:40	justin_smith	heh, just reminding myself how it behaved
17:41	sdegutis	Hmm. Interesting.
17:42	sdegutis	Our data is just so... actually, no, there aren't words to describe it.
17:48	sdegutis	cfleming: so how's your IDE going?
17:49	cfleming	sdegutis: It's going really well.
17:49	cfleming	Lots of free non-commercial licences gone out, but lots of people paying for it too.
17:50	cfleming	So it's definitely looking sustainable at this point, assuming a reasonable number of people renew and some growth in the community.
17:50	sdegutis	cfleming: So has it become a full time gig now?
17:50	cfleming	Yeah, it has been for a while now.
17:50	sdegutis	Sweet.
17:50	sdegutis	I like your icons of companies who use it on your home page. That's a clever idea.
17:51	cfleming	I'm just getting back to development in the last couple of weeks after the support spike after releasing.
17:51	sdegutis	Yikes, that sounds like it was really boring.
17:51	cfleming	Thanks! I think a lot of companies do that for their products. Lots of companies don't allow use of their logo, sadly.
17:52	cfleming	Yeah, that and licensing, EULAs, payment systems, etc etc - I haven't done much real work since October/November.
17:54	ben_vulpes	'real work' heh
17:54	cfleming	Yeah, none of the rest feels like real work, it's all necessary though.
18:09	cortexman	i'm trying to figure out how to link my cljs buffer with my ClojureScript REPL in emacs
18:13	cortexman	supposedly it's C-c M-J (capital J) but it doesn't seem to hook up correctly
18:15	sdegutis	cfleming: I hear ya, I'm only recently getting back into being really productive.
18:23	rhg135	productivity is a lie
19:19	Kamuela	rhg135: lol
20:02	sdegutis	So I guess ->> can't see on the inside of #() special forms.
20:02	sdegutis	,((->> [1 2 3] (map inc) #(apply +)))
20:03	clojurebot	#error {\n :cause "Wrong number of args (1) passed to: core/apply"\n :via\n [{:type clojure.lang.ArityException\n :message "Wrong number of args (1) passed to: core/apply"\n :at [clojure.lang.AFn throwArity "AFn.java" 429]}]\n :trace\n [[clojure.lang.AFn throwArity "AFn.java" 429]\n [clojure.lang.RestFn invoke "RestFn.java" 412]\n [sandbox$eval25$fn__26 invoke "NO_SOURCE_FILE" 0]\n [sandbox...
20:04	sdegutis	Ever notice that?
20:07	justin_smith	,(macroexpand '((->> [1 2 3] (map inc) #(apply +))))
20:07	clojurebot	((->> [1 2 3] (map inc) (fn* [] (apply +))))
20:09	justin_smith	,(macroexpand '(->> [1 2 3] (map inc) #(apply +))) ; sdegutis - this makes it pretty clear what happens
20:09	clojurebot	(fn* [] (apply +) (map inc [1 2 3]))
20:25	rhg135	,'#()
20:25	clojurebot	(fn* [] ())
20:29	TimMc	,(-> foo #(inc 5) ())
20:29	clojurebot	#error {\n :cause "Can't call nil, form: (nil (fn* foo [] (inc 5)))"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.IllegalArgumentException: Can't call nil, form: (nil (fn* foo [] (inc 5))), compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyzeSeq "Compiler.java" 6875]}\n {:type java.lang.IllegalArgumentException\n :message "Can't call nil, fo...
20:29	TimMc	mmm, right
20:29	TimMc	also, interesting
20:30	TimMc	,(macroexpand-1 '(-> 4 ()))
20:30	clojurebot	(nil 4)
20:33	rhg135	What?
20:33	clojurebot	What is sampling a random integers betwen 2, 12 s..t. P(X = i) = (7-\|i-7\|)/36
20:34	rhg135	Right?
20:39	justin_smith	what a card
20:41	rhg135	I know. clojurebot is so funny and smart and cool
20:42	justin_smith	I just wish clojurebot would stop stealing my girl.
20:43	TimMc	rhg135: The macro sees a list and does (cons (first list-form) (cons inserted-form (rest list-form))) or something
20:43	TimMc	which is (cons nil (cons 4 ())) in this case
20:43	rhg135	Ah
20:44	rhg135	justin_smith: who is your girl?
20:44	TimMc	https://github.com/clojure/clojure/blob/clojure-1.7.0/src/clj/clojure/core.clj#L1564 yep, except it's using unquote-splicing
20:45	justin_smith	rhg135: sadly "my girl" is a rhetorical phrase used for comical effect
20:47	rhg135	Ah, those love triangles
23:48	sdegutis	Hi.
23:48	sdegutis	I forgot why I opened IRC, but here I am.