#clojure log - Dec 29 2015

The Joy of Clojure
Main Clojure site
Google Group
List of all logged dates

1:28 felixn_: https://gist.github.com/munro/88c578c7eb0551daacf9 <-- hey any feedback for this function that conditionally zips against a second sequence? I'm thinking I could make it a transducer by passing in a 3rd argument as the reduced amount. only problem is the entire operation should fail if it's not complete

1:29 actually, that last condition can just be checked at the end, and not have anything to do with this function

1:32 http://ignaciothayer.com/post/Transducers-Are-Fundamental/ <-- also, I was reading this article about transducer. when would the signature "[result input & inputs]" for a map transducer ever apply?

4:03 Trioxin: I would love to see an example of convoluted artificial neural networks in clojure

4:04 convolutional*

4:05 something that isn't ages old

4:05 because state of the art changes there every so often

4:05 there's synaptic modified 8 months ago. not so bad I suppose

4:06 looks like the last thing they fixed was convolution

4:16 I'm surprised because I've brought things like this up here before and it seems like a lot more clojure devs would be working in machine learning.

4:17 given the nature of the language

4:17 why should c++ have all the fun?

4:18 oh sweet. I see there's a matlab interface for clojure

4:50 underplank: Hi all. Any idea why when running tests using ring-mock the :body of the PUT request is an InputStream whereas when running the service using jetty its just a string?

6:55 engblom: Does clojure have anything ready allowing you to have a circular double linked list?

6:57 I have ([1,0,0,1], [1,0,0,0], [1,1,0,0], [0,1,0,0], [0,1,1,0], [0,0,1,0], [0,0,1,1], [0,0,0,1]) where the last element should connect with the first, so that when I am at the last element and take 'next' I get the first, in the same way when I am at the first element and asks for 'previous' I get the last from the list

7:04 justin_smith: engblom: cycle

7:04 engblom: and there is no "previous" in clojure really

7:04 engblom: justin_smith: I already used cycle for forward... but the backward was the thing I was more wondering about

7:05 justin_smith: engblom: you need mutable data for the whole next/previous based on a single element thing

7:05 though I guess there are zippers...

7:16 engblom: actually, there is a way to do it with promises I think... one moment

7:22 engblom: justin_smith: I realized I could quite easily do it by using 'mod'. Then I need to just pass around the last position and going forward or backward is just a modulo to get the position in the initial vector

7:22 justin_smith: oh, there is that

7:23 engblom: I made something actually circularly linked - it works but printing it is a very bad idea :)

7:23 ,(defn make-node [v] {:prev (promise) :v v :next (promise)})

7:23 clojurebot: #'sandbox/make-node

7:23 justin_smith: ,(defn link-nodes [& nodes] (deliver (:next (last nodes)) (first nodes)) (deliver (:prev (first nodes)) (last nodes)) (reduce (fn [a b] (deliver (:next a) b) (deliver (:prev b) a) b) nodes) nil)

7:23 clojurebot: #'sandbox/link-nodes

7:24 justin_smith: ,(def foo (let [f (make-node :foo) b (make-node :bar) q (make-node :quux)] (link-nodes f b q) f))

7:24 clojurebot: #'sandbox/foo

7:24 justin_smith: ,(-> foo :v)

7:24 clojurebot: :foo

7:24 justin_smith: ,(-> foo :next deref :v)

7:24 clojurebot: :bar

7:24 justin_smith: ,(-> foo :next deref :next deref :v)

7:24 clojurebot: :quux

7:24 justin_smith: ,(-> foo :next deref :next deref :next deref :v)

7:24 clojurebot: :foo

7:24 justin_smith: ,(-> foo :next deref :next deref :next deref :next deref :v)

7:24 clojurebot: :bar

7:25 justin_smith: ,(-> foo :next deref :next deref :next deref :next deref :next :deref :v)

7:25 clojurebot: nil

7:25 justin_smith: ,(-> foo :next deref :next deref :next deref :next deref :next deref :v)

7:25 clojurebot: :quux

7:25 justin_smith: ,(-> foo :next deref :next deref :next deref :next deref :next deref :prev deref :v)

7:25 clojurebot: :bar

7:25 justin_smith: ,(-> foo :next deref :next deref :next deref :next deref :next deref :prev deref :prev deref :v)

7:25 clojurebot: :foo

7:25 justin_smith: etc.

7:25 engblom: Thanks!

7:26 justin_smith: engblom: that "works" but is very clumsy, and can't be edited - you pretty much have to make a whole new set of nodes from scratch and re-link

7:28 engblom: justin_smith: In my case, a simple 'if' or 'mod' will be easier. The sequence I gave first was for driving a stepper motor. If that one would be circular, it would be easy to take one step forward or one step backward.

7:29 justin_smith: engblom: yeah, that is definitely simpler

7:29 lokien: Hey, is there a way of preventing my stupidness? Like declaring a function with too much params - (defn wrong [x y] (identity x y))

7:30 justin_smith: lokien: the eastwood plugin for lein will catch that

7:30 lokien: I mean, this function happily doesn't throw an error, I'd like it to throw one

7:30 justin_smith: sweet name

7:30 justin_smith: lokien: because it's a linter

7:30 lint eastwood, lol

7:31 lokien: why people are so funny in functional programming, I don't know

7:31 imperative dudes are all like "we're engineers"

7:31 but our entire environment is puns and jokes

7:32 justin_smith: and fp guys are like "hell yeah, thomas the tank engine, *toot* *toot*"

7:32 lokien: it's true

7:32 lokien: that's why I love you

8:10 * TimMc grumbles something about the imperative folks driving toy trains just like everyone else

8:10 TimMc: justin_smith: I'm using that. :-P

8:23 jink167: Anybody know why grimoire at conj.io has been down for days?

8:25 justin_smith: jink167: at first arrdem wasn't aware it was down - but supposedly he's watching it now so... dunno

8:26 jink167: thx

8:48 ksaua: Hi. Is there a function which given a seq [1 2 3 4] returns [[1 2] [2 3] [3 4]] ?

8:48 j-pb: partition

8:49 ksaua: ty

8:50 BRODUS: ,(partition 2 (butfirst (butlast (interleave [1 2 3 4] [1 2 3 4]))))

8:51 clojurebot: #error {\n :cause "Unable to resolve symbol: butfirst in this context"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Unable to resolve symbol: butfirst in this context, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6704]}\n {:type java.lang.RuntimeException\n :message "Unable to resolve symbol: butfirs...

8:51 BRODUS: ,(partition 2 (rest (butlast (interleave [1 2 3 4] [1 2 3 4]))))

8:51 clojurebot: ((1 2) (2 3) (3 4))

8:52 j-pb: BRODUS: works but is ugly

8:52 ,(partition 2 1 [1 2 3 4])

8:52 clojurebot: ((1 2) (2 3) (3 4))

8:52 BRODUS: aaa i didn't know partition could do that

9:08 TimMc: (def rest butfirst) :-P

9:10 BRODUS: TimMc: i need to go to bed :)

9:12 TimMc: BRODUS: partition has even more tricks than that

9:13 ,(partition 5 3 "hi" (range 8))

9:13 clojurebot: ((0 1 2 3 4) (3 4 5 6 7) (6 7 \h \i))

9:23 justin_smith: new thing added to my list of evil - ansi escape sequences inside files so that git diff breaks in the terminal

9:24 breaks / shows misleading output

10:08 anti-freeze: Hi everyone, so I'm using clojurescript with boot and everything was working just dandy until about 2 hours ago, then figwheel shat itself. Has anyone seen this error before? https://www.refheap.com/113206

10:12 TimMc: justin_smith: font-lock-fontify-buffer doesn't do it. :-/

11:26 justin_smith: TimMc: odd...

11:26 anti-freeze left, but when that happens you need to clean your cljs and rebuild

11:34 ,(partition 5 4 (cycle "hi") (range 9))

11:34 clojurebot: ((0 1 2 3 4) (4 5 6 7 8) (8 \h \i \h \i))

11:34 justin_smith: fancy

11:46 ,(partition 4 3 (cycle "lo") [4 2 0 6 6 6 69])

11:46 clojurebot: ((4 2 0 6) (6 6 6 69) (69 \l \o \l))

11:58 devth: sadness = the inconsistency between def* forms and whether they allow docstrings

11:59 justin_smith: devth: would the fact that defonce rhymes with beyonce cheer you up?

11:59 devth: nice try!

12:00 i want to hear someone say it like that IRL

12:00 justin_smith: I nominate gfredericks

12:01 devth: justin_smith: btw, i switched from bindings to passing state around explicitly. only took about 15 min per ns and my sanity is restored! good call.

12:01 justin_smith: glad to hear it

12:08 gfredericks: devth: justin_smith: http://tmp.gfredericks.com/defonce.m4a

12:09 devth: http://www.reactiongifs.us/wp-content/uploads/2013/08/picard_clapping.gif

12:09 justin_smith: gfredericks: you're the best

12:10 gfredericks: mind if I tweet that link? I guess it has tmp in it so it isn't meant to be a long term thing

12:11 gfredericks: SEMANTIC SUBDOMAINS!

12:12 um here I'll put it elsewheres

12:12 http://upload.gfredericks.com/defonce.m4a

12:12 justin_smith: OK, tweeting that now

12:14 TimMc: leenucks

12:15 justin_smith: yeah, it would be cool to get "this is Rich Hickey and I pronounce defonce as defonce" - with the gfredericks approved pronunciation of course

12:15 TimMc: actually hearing that recording, after fighting with kernel configs and drivers and such, was a moment of pure bliss in an otherwise bleak existence.

12:17 TimMc: :-)

12:18 $findfn nil ()

12:19 Oh right, no lazybot. :-(

12:23 ,(doseq [v (remove (comp :macro meta) (vals (ns-publics 'clojure.core)))] (try (when (= (v nil) ()) (println v)) (catch Throwable t nil)))

12:23 clojurebot: TimMc: Gabh mo leithscéal?

12:23 justin_smith: TimMc: concat

12:23 devth: what happened to lazybot?

12:23 TimMc: rest, reverse, sort, interleave, cycle, seque, vec, concat, drop-last, distinct, flatten, sequence

12:23 justin_smith: entropy

12:23 TimMc: oh yeah, those would do it too

12:24 I'd venture that (distinct nil) is the classiest way to get an empty list from nil

12:28 TimMc: What if I don't want to modify the list when it's not empty?

12:28 justin_smith: concat

12:28 that's the only one that will act like identity if a single list is arg I think

12:28 ?

12:28 oh, or sequence

12:29 TimMc: or vec

12:29 justin_smith: right

12:29 TimMc: wtf is sequence

12:29 flatten ;-)

12:29 justin_smith: TimMc: a transducer that doesn't reverse the input like (into () ...) does

12:29 very useful

12:29 TimMc: ooh, that's actually maybe what I want

12:30 ,(-> sequence var meta :added)

12:30 clojurebot: "1.0"

12:30 TimMc: Fascinating.

12:30 justin_smith: oh!

12:30 well, it became more useful when it became a transducer, to be sure

12:30 TimMc: No, concat is right, but this is useful to know I guess.

12:31 I have a fn that yields a sequence of stuff, and it is currently implemented entirely by another fn that yields either a vector or nil, and I don't want the test to distinguish between nil and [].

12:32 I could also just wrap the "actual" side of the test in a seq

12:32 Oh right this is a midje (provided ...) block so I need the actual arg.

12:32 justin_smith: if it's a vector or nil, what about vec?

12:33 ,(vec nil)

12:33 clojurebot: []

12:33 TimMc: concat it is, since that conveys that there could, in the future, be more data sources.

12:34 ,(->> 'clojure.core ns-publics vals (filter #(and (not (:macro (meta %))) (try (= (% nil) ()) (catch Throwable t nil)))) (map #(.sym %)))

12:34 clojurebot: TimMc: No entiendo

12:34 TimMc: I wonder what its beef is.

12:34 justin_smith: catch

12:35 TimMc: fine

12:35 justin_smith: or more specifically I think it is try that is forbidden

12:40 TimMc: ,(eval '(+ 1 2))

12:40 clojurebot: 3

12:42 TimMc: ,(->> 'clojure.core ns-publics (filter #(and (not (:macro (meta (val %)))) (eval `(~(symbol (str "t" "ry")) (= ((identity ~(key %)) nil) ()) (~(symbol (str "cat" "ch")) Throwable ~'t nil))))))

12:42 clojurebot: ([eduction #'clojure.core/eduction] nil[dedupe #'clojure.core/dedupe] [rest #'clojure.core/rest] [reverse #'clojure.core/reverse] [sort #'clojure.core/sort] ...)

12:42 justin_smith: TimMc: SNEAKY

12:42 TimMc: WAS THAT SO HARD?

12:42 that bot, seriously

12:48 ,(->> 'clojure.core ns-publics (filter #(and (not (:macro (meta (val %)))) (eval `(try (= ((identity ~(key %)) nil) ()) (~(symbol (str "cat" "ch")) Throwable ~'t nil))))) keys)

12:48 clojurebot: (eduction nildedupe rest reverse sort ...)

12:48 TimMc: bah

12:48 ,(+ 1 2)

12:48 clojurebot: 3

12:48 TimMc: Interesting, my REPL bombs out after I run that.

12:50 "SocketException The transport's socket appears to have lost its connection to the nREPL server"

12:55 justin_smith: TimMc: yeah, weird...

12:58 TimMc: Using binary search to find it...

12:59 set-agent-send-off-executor!

12:59 Well, that would do it.

12:59 justin_smith: oh, hahaha

13:00 yeah, I was just gonna say, it breaks a java -jar repl too

13:01 yeah, I think it would be safe for findfn to simply blacklist that

13:01 actually, it could blacklist anything where #(= (last (str %)) \!) safely

13:01 Bronsa: justin_smith: it doesn't for me

13:02 justin_smith: Bronsa: oh, weird - did you try running it a second time?

13:02 it looks OK after the first time

13:02 Bronsa: nope let's try

13:02 justin_smith: but it's secretly broken (here at least)

13:02 running it the second time exposes that brokenness

13:02 Bronsa: oh wow

13:02 I can't even understand what that code does

13:03 justin_smith: tries to run every funciton with nil as an arg and returns the ones that returned ()

13:03 TimMc: Running *anything* after it exposes the breakage. Just try entering 7.

13:03 Bronsa: ah I see

13:03 I guess it's hitting some non terminating function

13:03 TimMc: It's hitting set-agent-send-off-executor!

13:03 Bronsa: or filling up a threadpool or something

13:03 ah

13:03 nice

13:04 justin_smith: Bronsa: well, if it called (set-agent-send-off-executor! nil) - after that things get dicey right?

13:04 Bronsa: yeah

13:04 TimMc: ,(set-agent-send-off-executor! nil)

13:04 Bronsa: :(

13:04 TimMc: ,7

13:04 clojurebot: nil

13:04 7

13:04 TimMc: ^ I wonder why clojurebot is immune.

13:04 justin_smith: I think the sandboxing will help you there, yeah

13:05 Bronsa: TimMc: that works on my repl

13:06 (a java -jar repl)

13:08 TIL there's a private filter-key function in clojure.core

13:08 TimMc: There's also the private fn >1? -- what else is THE MAN keeping secret?

13:09 justin_smith: TimMc: I smell some prs for the 71 lib

13:09 >71? =71? <71? etc.

13:10 TimMc: ooh

13:10 Bronsa: 71->str

13:10 TimMc: but they should be private too, to stay idiomatic

13:10 justin_smith: I mean, given that precedent they should be private of course

13:10 yes

13:10 map->71

13:11 Bronsa: sadly I don't think the reader will accept that one

13:11 Bronsa: yeah you're right :(

13:11 ridcully_: set-agent-send-off-executor-to-71

13:11 Bronsa: ridcully_: missing the bang!

13:11 ridcully_: dang

13:11 TimMc: ,(->> clojure.core quote ns-interns (filter (comp :private meta val)) (map key))

13:11 clojurebot: (prependss is-runtime-annotation? imap-cons add-annotation load-data-reader-file ...)

13:11 TimMc: ,(->> clojure.core quote ns-interns (filter (comp :private meta val)) count)

13:11 clojurebot: 104

13:12 TimMc: (>71? *1) => true

13:12 justin_smith: oh, and let's not forget compare-to-71

13:13 TimMc: win 12

13:13 ugh

13:13 justin_smith: TimMc: ITYM win 71

13:13 TimMc: I do

13:13 ,(>71? 104)

13:13 clojurebot: true

13:13 TimMc: much better

14:32 felixn_: is there a way I can use a clojurescript repl? I'm using cursive ide, so any integration with that would also be nice

14:34 justin_smith: felixn_: yes, there is austin, and also figwheel which wraps up the repl with live code reload etc.

14:35 felixn_: https://github.com/cemerick/austin https://github.com/bhauman/lein-figwheel

14:37 domgetter: felixn_: It's definitely worth watching that 45 minutes talk on Figwheel https://www.youtube.com/watch?v=j-kj2qwJa_E

14:37 or at least the first 15 minutes

14:45 felixn_: justin_smith domgetter: awesome, thanks! dinking around with austin. I think it's not working because not all my code works with clojurescript, is there a way I can just play around with it and ignore my project files?

14:45 musteresel: Hi, about java interop: If I have a (defrecord Foo [bar baz]), how can I access the members of a Foo from java? In general: Are there comprehensive examples for interfacing clojure code from java code?

14:47 felixn_: ahh, I'll just keep dinking with it until it works!

14:58 domgetter: musteresel: is it not yournamespace.core.Foo ?

15:05 lokien: is there a way of configuring vim to align my code as I hit tab? I'm becoming bored manually deleting these unnecessary spaces

15:06 ridcully_: `=-` with the cljfmt plugin, `==` with regular bindings

15:07 lokien: ridcully_: after all this time

15:07 now I can happily sudo apt-get purge intellij*

15:08 musteresel: domgetter: Is this working? Assuming type hints then namespace.core.Foo foo = some_source(); Double d = foo.bar; should work? I currently cannot test this, thus I'm asking ;)

15:09 ridcully_: cljfmt also formats on write (with a repl connection iirc), if that's your thing

15:09 otherwise look into the options vim itself has. e.g. all the `=` stuff

15:09 domgetter: musteresel: I don't know. The docs on it might be of some help? http://clojure.org/datatypes

15:11 lokien: ridcully_: do you use vim-cljfmt? I don't know if it's necessary as == just works

15:12 ridcully_: i use it, as it's a tad more nicer. it also fixes linebreaks etc

15:13 lokien: ridcully_ thanks, I'll use it too

15:13 ridcully_: if you don't like the "auto-format" aspect, i bet it's just a config var away to be turned off and `=-` can still be used (fmt me the current form)

15:14 lokien: any more useful vim plugins? besides fireplace and sexp?

15:14 ridcully_: cljfmt plugin and paredit are the "i insert stuff for you" of the bizarro universe, if you are used to intellij and openoffice "helping"

15:14 devth_: lokien: guns/vim-clojure-highlight

15:14 moderately useful

15:14 lokien: I had used vim for a long time, now, in anger installed intellij

15:15 ridcully_: the sexp-for-regular-ppl or something like that. that puts some of the sexp shortcuts on "vimy" places

15:15 lokien: I'm purging it right now

15:15 devth_: I have it already

15:15 ridcully_: and paredit of course

15:15 lokien: sexp + paredit?

15:16 isn't it sexp or paredit?

15:16 ridcully_: yes, i have disabled some part of sexp

15:16 maybe the bit for the paredit ;)

15:16 lokien: can I have your dotfiles, please?

15:17 ridcully_: i have this set: let g:sexp_enable_insert_mode_mappings = 0

15:17 * lokien mindlessly copying it to his .vimrc

15:18 ridcully_: well that's the bit, that supposedly makes it happy with paredit

15:19 then i only have three other lines. one to map :Figwheel to do some piggiebacking, and map c-c,c-c to eval, since that is what i use also in tslime (i hear some *ugh* from the emacs ranks)

15:21 lokien: I've installed it, and it hadn't crashed, so I think it's okay

15:21 what features of paredit do you use? and what things from sexp are better?

15:22 ridcully_: i have used paredit before i used sexp and i stayed with them

15:22 lokien: which*, damn

15:22 o-kay, will skim through manuals

15:22 ridcully_: paredit just ballances everything automagically. i just do my leader-i, leader-w most of the time

15:24 i never tried sexp alone and i _guess_ that i have some interference. leader-i is iirc from sexp. maybe paredit is no longer used at all. idk,idc

15:26 * lokien checks leader-i

15:26 lokien: I love you ridcully_

15:28 ridcully_: all the "wrap me this" magic is leader-w-<with what> or leader-i for just ()

15:29 sdegutis: Hello.

15:29 Who is having a great time today?

15:29 ridcully_: depending on your keyboard layout (and even then) mapping the leader to e.g. space makes your life easier

15:29 sdegutis: I bet it's TimMc and TEttinger2.

15:30 lokien: sdegutis: I am!

15:30 sdegutis: Alright! High five lokien!

15:31 lokien: ridcully_: I was going through vim primer from daniel messer, I think I should finally add his bindings (he recommends space as leader too)

15:32 high five, sdegutis! what's up with TimMc and TEttinger2?

15:33 justin_smith: felixn_: usually you have .cljc files for things that work in clojurescript or clojure, cljs for clojurescript files, and clj for clojure files

15:36 sdegutis: lokien: I dunno I just thought maybe they're having a nice day, they're usually so polite and kind in here so they usually seem to be having good days.

15:37 lokien: sdegutis: what about justin_smith?

15:42 justin_smith: I'm always having a terrible day.

15:42 felixn_: justin_smith: played with cljc a bit, it's works really well! I don't need UI for the clojurescript repl, so I ended up finding just piggieback with rhino works. from looking at the austin example, it look like I also had to setup a webserver to script the file, so not sure that's what I wanted for just a cljs repl

15:42 lokien: justin_smith: oh, why? :(

15:43 justin_smith: lokien: just being silly, I should remember that doesn't always translate to IRC

15:44 J_Arcane: so how long until we have cljs running on node so we can write front and back in clojurescript and avoid enormous jvms entirely? ;)

15:44 felixn_: https://github.com/clojure/core.logic/blob/1c4020f63a5b792f08a90434ab6e8be6aa467a82/src/main/clojure/cljs/core/logic.cljs#L210 <-- is there a way I can extract the name from this type? (:name (LVar. "hey" nil)) doesn't work like it does in clojure. also the clojurescript API is slightly different, so I will submit a PR!

15:44 lokien: justin_smith: knew that! you're our wizard here, your days are meant to be awesome

15:45 mgaare: am I wrong in thinking that this should work? (case (class 42) java.lang.Long "yep!" "nope :(")

15:46 sdegutis: lokien: he's always having the best day it would seem

15:46 lokien: on account of how he's always the most helpful even when I'm the most idiotic or rude (sorry about those days justin_smith and everyone else!)

15:47 lokien: and I do believe you're correct, I think justin_smith may have some wizard-blood running through his veins

15:47 amalloy: mgaare: yes, sorry

15:47 sdegutis: Other wise how elsewise would thus he have been able so well and often to contrive such immaculate solutions in pure Clojure for us?

15:47 amalloy: case doesn't evaluate its test clauses, and java.lang.Long is a symbol which *evaluates to* a class, not a class

15:47 mgaare: amalloy: ok, is there a "right" way to do that?

15:48 sdegutis: amalloy: they have to be constants right?

15:48 mgaare: just use cond when you can't use case

15:48 mgaare: righto

15:48 amalloy: or look (class x) up in a map

15:48 sdegutis: mgaare: or possibly condp when the time is right

15:48 but I've never found a time when condp actually makes my code cleaner than like amalloy said just using a map

15:48 Not sure when I'm sposta use condp

15:48 justin_smith: or nested if clauses if you hate people

15:49 sdegutis: I find that lately my route-handling code has a lot of (cond) in it.

15:49 justin_smith: sdegutis: condp is for when your input and the condition are actually two args to a function

15:49 lokien: sdegutis: this clojure family is so great, I was on #haskell for some time and it's nothing like that

15:49 sdegutis: I'm not sure how I like it so far, but it's cleaner than it used to be, so there's that.

15:49 lokien: Oh yeah this channel is great people. Some of the best I've ever met. Not even sarcasm!

15:49 felixn_: ooo, is there a reason (.-name value) works in clojure & clojurescript, but (:name value) only works in clojurescript?

15:49 justin_smith: ,(condp :x = :x "OK" :y "huh?") ; sdegutis - simpler than calling = in every clause

15:50 clojurebot: "OK"

15:50 felixn_: s/only works in clojurescript/only works in clojure/

15:50 sdegutis: lokien: Although you have to mind amalloy he is extremely helpful and knowledgeable but has NO time for putting up with crap, you've gotta be completely serious with him 24/7 but he makes up for it by being excessively helpful and knowledgeable

15:50 justin_smith: but I've rarely had a time when I'm calling = and not just using case

15:50 justin_smith: possibly never actually

15:51 justin_smith: sdegutis: we just discussed a reason to do that just now

15:51 ridcully_: felixn_: (:name somemap) also works ins clojurescript

15:51 mgaare: amalloy: thanks

15:51 lokien: sdegutis: thanks, I'll remember that! can we officially become friends now?

15:51 justin_smith: sdegutis: and sometimes it's not =, what about > for example

15:53 ,(condp > 10 1 "sad" 10 "ho-hum" 100 "just fine") ; sdegutis

15:53 clojurebot: "just fine"

15:53 amalloy: the main person i am not interested in putting up with is sdegutis

15:53 TEttinger: that's sweet

15:53 amalloy: which i would not really bring up except he has just cast me as an unfriendly sort

15:53 justin_smith: TEttinger: useful for scoring and gamey stuff I think

15:54 sdegutis: lokien: sure, in Clojure that's super easy: (swap! *friends* conj 'username)

15:54 felixn_: ,(do (deftype Meow [name]) (let [v (Meow. "foo")] [(:name v) (.-name v)])) ; ridcully_: so weird, this returns ["foo" "foo"] in clojure

15:54 clojurebot: [nil "foo"]

15:54 TEttinger: I do think you don't put up with crap amalloy. that is different from being unfriendly

15:54 amalloy: well, that's fair

15:54 justin_smith: felixn_: deftype does not define keyword access, but defrecord does

15:55 TEttinger: I think I put up with crap too much sometimes. I have a hard time telling people they should learn it themselves

15:55 sdegutis: amalloy: you're certainly *friendly*, no doubt about that! just, got no time for games or jokes.

15:55 justin_smith: felixn_: do you mean "foo" "foo" in cljs?

15:55 felixn_: justin_smith: oohhh, nvm, that doesn't work in clojure. I bet if I check clojure.core.logic it's defrecord XD

15:55 ridcully_: deftype ain't mapish

15:55 sdegutis: amalloy: and that's respectable; anyone who has no time for games or jokes certainly knows the value of his own time

15:55 ridcully_: defrecord is

15:55 lokien: sdegutis: :D

15:56 sdegutis: amalloy: I admire that and it inspires me to waste less time

15:56 amalloy: haha

15:56 sdegutis: So with that.. I'm off to continue working on my /very first legit programming language ever/!

15:56 ridcully_: oO

15:56 sdegutis: :D:D:D

15:56 ridcully_: i said something wrong

15:57 felixn_: justin_smith & ridcully_: disregard, clojure.core.logic sets up a clojure.lang.ILookup for :name on the type, I was just getting really confused

15:59 sdegutis: I was trying to use core.match to help make my route handling functions a little cleaner, but I'm not sure it really helped. What's the intended use-case or best use-case for this lib?

16:00 justin_smith: felixn_: yeah, that's an interface

16:05 felixn_: with the new reader conditionals, will we see more core libs use that instead of duplicating code for clj & cljs?

16:05 justin_smith: one would hope

16:06 dnolen: felixn_: sure if people submit the non-trivial patches needed to push that forward

16:11 lokien: where to add :repl-options in my profiles.clj? I can't find an example

16:11 justin_smith: lokien: either at the top level or inside a specific profile where you want them applied

16:12 TimMc: amalloy is srs bsns

16:13 lokien: justin_smith: http://lpaste.net/148059

16:13 amalloy: TimMc banned for ten hours for poking fun at amalloy

16:13 TimMc: /nick TimMac

16:13 lokien: McTim would be better

16:13 TimMc: MC Tim

16:14 sdegutis: :D

16:14 amalloy: ten more hours for ban evasion. u in real trouble now, mister

16:14 justin_smith: lokien: looks good https://github.com/technomancy/leiningen/blob/master/sample.project.clj#L335

16:14 sdegutis: wow amalloy apparently I had you pegged wrong all along!

16:14 lokien: put him in a monad, amalloy, it always works

16:14 * sdegutis high fives amalloy for having shared interests

16:15 amalloy: lokien: is that why it's called bind?

16:15 TimMc: bonding over bans

16:15 felixn_: https://github.com/clojure/core.logic/blob/master/src/test/cljs/cljs/core/logic/tests.cljs#L7 <-- does a definition starting with a hyphen mean anything? experimental? wondering why the clojurescript version has hyphens in front of some definitions

16:15 lokien: justin_smith: doesn't work though. "vim-cljfmt: Could not locate cljfmt/core__init.class or cljfmt/core.clj on classpath."

16:16 amalloy: I suppose

16:16 justin_smith: lokien: umm - how would you get that error unless your code ran?

16:16 lokien: amalloy: I don't know if it's a serious-enough-topic for you though

16:16 BRODUS: felixn_: i just think thats signalling that they're private

16:16 lokien: :D

16:17 justin_smith: I just do :Cljfmt and this pops out, it was supposed to work

16:17 justin_smith: lokien: a classpath error doesn't mean you have to call require, it means that your deps are wrong

16:17 clojurebot: excusez-moi

16:17 justin_smith: lokien: a require will raise that error if you don't have the right deps for the thing you are requiring

16:18 ,(require 'this-does-not.exist)

16:18 lokien: justin_smith: so, my deps are wrong?

16:18 clojurebot: #error {\n :cause "Could not locate this_does_not/exist__init.class or this_does_not/exist.clj on classpath. Please check that namespaces with dashes use underscores in the Clojure file name."\n :via\n [{:type java.io.FileNotFoundException\n :message "Could not locate this_does_not/exist__init.class or this_does_not/exist.clj on classpath. Please check that namespaces with dashes use underscores...

16:18 justin_smith: lokien: yeah, what project would provide the cljfmt.core namespace?

16:19 lokien: justin_smith: my project! /silently adds deps to his project.clj

16:19 justin_smith: lokien: I'm sure it's something easy to find, and it needs to be in your deps or plugins somewhere

16:19 :)

16:20 lokien: I forgot to add it, sorry for making a trivial problem

16:20 ridcully_: lokien: you just need to add cljfmt to the dependencies in your ~/.lein/profiles.clj

16:21 this and the addon should be enough to make it work with a nrepl connection

16:21 lokien: ridcully_: yeah, I wanted to do that, but I switched tabs and you were talking about funny stuff and I forgot :(

16:21 ridcully_: i certainly was not talking funny stuff ;P

16:22 lokien: "you" was plural here

16:23 ridcully_: by the way, when I type :Console, my fireplace should connect to this repl straight away, yeah?

16:23 ridcully_: if you have salve installed

16:24 lokien: I have it, but it doesn't make a connection

16:24 ridcully_: otherwise either start with a running nrepl in that directory (.nrepl-port file existing) or do a :Connect <theurl>

16:24 lokien: aand I can't remap leader to space, vim is so great

16:25 vim-salve installed, doesn't work. :(

16:25 domgetter: lokien: http://stackoverflow.com/questions/446269/can-i-use-space-as-mapleader-in-vim

16:25 ?

16:26 lokien: domgetter: doesn't work

16:26 <\Space> doesn't work too

16:26 domgetter: lokien: "As spacemanaki comments in the question above nnoremap <SPACE> <Nop> was needed to get this working. "

16:26 Even after that?

16:28 ridcully_: lokien: this is what i have: https://www.refheap.com/205b710b37a6b703cd3d18efb

16:28 lokien: even after that

16:28 ridcully_: lokien: the url there seems rather unrelated - ignore it

16:28 this is on linux + xterm. ymmv

16:29 if \ is _not_ the default leader in vim, then i used that before and kept that around

16:29 lokien: ridcully_: it works! you wonderful santa incarnate

16:31 ridcully_: satan incarnate? thanks!

16:33 lokien: ridcully_ still no live REPL connection, so, yeah, I guess

16:35 ridcully_: :Console needs some setup - things like. iirc i setup tmux to be used. otherwise it starts a new xterm. there should be some "result" from :Console

16:35 s/like/& where to start/

16:35 lokien: :Console fires up another terminal and a repl in it. it's ran from a weird place though, it doesn't even see my namespace

16:36 ridcully_: first step: start your `lein repl` by hand, get the url there and do :Connect <url>

16:36 lokien: (use 'something.core) doesn't work

16:36 works

16:36 I mean, :Connect works

16:37 ridcully_: are starting your vim from the directory the project.clj is in?

16:37 lokien: yeah

16:37 justin_smith: you can double check with (System/getenv "PWD")

16:37 lokien: vim src/something/core.clj

16:38 "/home/lokien/cl/something"

16:40 ridcully_: tried that. works for me. even if there is no tmux a new xterm is opened then. :Console opens a lein repl; then a `cpp` in that opens file works

16:40 lokien: it works, justin_smith is a wizard, confirmed

16:40 ridcully_: s/opens/opened/

16:40 lokien: I don't know, I just typed the same commands, and it wonderfully worked

16:40 thank you all

16:41 now it doesn't work!

16:42 I type :Console, new terminal is opened with a repl in it. in that repl, I can't use my project's namespace

16:42 pilne: hrm, apply is kinda like fold (but is it foldr, or foldl?)

16:42 ridcully_: it is just a helper to fire up lein repl. there should be no "on-off-behaviour"

16:43 justin_smith: lokien: did you try requiring that namespace?

16:43 lokien: justin_smith: where?

16:43 justin_smith: in the repl

16:43 (require 'my.ns :reload) - you should either be able to access your ns, or get an error message

16:43 ridcully_: and are you typing in the repl or are you sending commands over from vim?

16:43 lokien: FileNotFoundException Could not locate

16:44 justin_smith: :reload is in case there was a prior error loading it that was hidden

16:44 lokien: (System/getenv "PWD") gives me my home directory, strangely

16:44 "/home/lokien"

16:44 justin_smith: that would be your problem right there

16:45 it got launched from the wrong place

16:45 lokien: from dir where I have my project.clj

16:45 justin_smith: it should depend on either what dir you were in when you started vim, or what file you are editing?

16:45 what file were you visiting when you started the repl?

16:45 ridcully_: cd into the dir, where your project.clj is, then just `vim` and :Console

16:46 :Console does not even autocomplete, outside of a project

16:46 lokien: vim src/something/core.clj, yeah?

16:46 ridcully_: try just `vim` first

16:47 i tried with a relative file, and it worked for me

16:47 lokien: Not an editor command: Console

16:47 :(

16:47 kopasetik: Is there an evalbot on this channel?

16:48 justin_smith: ,'yes

16:48 lokien: ,(println "here i am")

16:48 clojurebot: yes

16:48 here i am\n

16:48 ridcully_: are you doing a :cd in vim?

16:48 lokien: ridcully: no

16:48 ridcully_ *

16:49 LauJensen: Gents, datomic question. I have some :days/* which stores stats on a given day, and each day references 24 :day/hours where information specific to each hour. The ID for each day is :day/date (instant). How do I best extract every :hour/whatever value for dates in a given range?

16:51 lokien: ridcully_: vim project.clj and :Console from there doesn't work also. I'll reinstall salve

16:52 ridcully_: lokien: i am out of ideas then. i installed salve for the other commands and console is just a tad convinience for me. starting the repl first and then just starting vim in that dir should work relyable. i also usually have it running in tmux - don't know, if there are differences with the newly spawned xterm

16:52 justin_smith: ,,,,,',,,,,',,,,,',,,,',,,,',,,,'""

16:52 clojurebot: (quote (quote (quote (quote (quote "")))))

16:52 lokien: ridcully_: it had worked for me, but now it's broken, dunno why. thanks for help :)

16:53 ridcully_: lokien: this is the git commitid of salve i run, if it helps: a0dc869

16:55 lokien: ridcully_: I.. I think it's fish

16:56 ridcully_: the protocol or the sea creature?

16:56 lokien: it's not fish, fake info. friendly interactive shell

16:58 going to sleep now, I'll look at that tomorrow. thanks again ridcully_

16:58 ridcully_: yw

17:01 kopasetik: justin_smith and lokien: thanks!

17:10 devth: mistakenly assumed mapcat == (comp flatten map) but flatten goes crazy and flattens everything

17:11 found when trying to switch to pmap

17:11 justin_smith: ~flatten

17:11 devth: since pmapcat doesn't exist

17:11 clojurebot: flatten is rarely the right answer. Suppose you need to use a list as your "base type", for example. Usually you only want to flatten a single level, and in that case you're better off with concat. Or, better still, use mapcat to produce a sequence that's shaped right to begin with.

17:11 justin_smith: devth: flatten is kind of wacky

17:11 devth: oh right

17:11 the cat is for concat

17:11 sdegutis: Yeah I never use flatten.

17:11 justin_smith: ,(flatten {:a 0 :b 1 :c [1 2 3]}) ; user error, but really...

17:12 clojurebot: ()

17:12 sdegutis: I always just do mapcat or for with two pairs.

17:12 That's great enough.

17:12 devth: so: mapcat == (comp concat map) right?

17:12 justin_smith: (source mapcat)

17:12 ,(source mapcat)

17:12 clojurebot: Source not found\n

17:13 devth: yep

17:13 ran it in my repl

17:13 sdegutis: (src mapcat)

17:13 ,(src mapcat)

17:13 clojurebot: #error {\n :cause "Unable to resolve symbol: src in this context"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Unable to resolve symbol: src in this context, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6704]}\n {:type java.lang.RuntimeException\n :message "Unable to resolve symbol: src in this conte...

17:13 sdegutis: clojurebot: come on man why you gotta be that way

17:13 clojurebot: It's greek to me.

17:13 justin_smith: sdegutis: src is not a thing

17:14 devth: (apply concat (apply map f colls))

17:14 ptrx: hello, i am new to clojure and just discovered libGDX. From what i understand i have ideally one clojure codebase and can target libGDX over the RoboVM with it and deploy my app to all kind a platforms. Is this correct? I am interested in gui apps, which arent games. Would i still want to use this combination or are there other tools? Thanks in advance! =)

17:16 justin_smith: ptrx: cljs with one of the react libs (om, reagent, etc.) is good for webapps if that's an option. Otherwise there is seesaw (swing bindings) or javafx - I think there's some clojure binding to javafx out there

17:16 * pilne grumbles about "language" tutorials that "assume" one is going to use "the author's" "editor-of-choice"

17:16 justin_smith: pilne: I'm gonna make a tutorial that assumes one is using ed

17:16 pilne: bwahahahahahaha

17:17 line-only-editing ftw

17:17 justin_smith: I mean it's the most universally installed editor in existence, right?

17:17 ptrx: justin_smith, thank you =)

17:17 sdegutis: What is? vi?

17:17 Oh wait no ed.

17:17 Is sed just like a super ed?

17:18 pilne: i mean.... don't get me wrong.... i have zero hate for emacs... but... when i'm looking for a decent sized project tutorial to grok clojure a bit better... (i learn better that way), i don't want to *have* to use it

17:18 sdegutis: Like, ed++?

17:18 pilne: sed and ed are... really different in use

17:18 justin_smith: ptrx: of all the options I think seesaw is probably easiest to jump into in a clojure way, react webapp is likely the most powerful option right now, and javafx will look and act more like a regular desktop app compared to the other options

17:18 sdegutis: pilne: sure that's fine, no loss.. I can't really use anything except emacs when s-expressions are involved on account of paredit

17:18 WorldsEndless: I tried to index a collection of page objects using zipmap with range, and the order came back shuffled. Should I be using a different function, or should I sort?

17:18 justin_smith: sdegutis: sam is super ed

17:19 sdegutis: justin_smith: I'm still waiting a few years for the react/etc ecosystem to settle down and finally land on something sane before I jump into that + Electron

17:19 pilne: eh, atom, light table, and vim all support paredit from what i've found (and i prefer all three of those (in vim mode for 2 of them) to emacs just for muscle memory.... something tells me i chose the wrong pil years ago

17:19 WorldsEndless: That is, they are assigned the right number, but they come back "1, 7, 13, 3" or some such order

17:19 sdegutis: justin_smith: cuz as it is right now its quite a bit ridiculous to get started with or even use, although reagent looks cool for simple things

17:19 justin_smith: WorldsEndless: hash-maps are not ordered, if you want an ordered map user sorted-map

17:20 sdegutis: justin_smith: wow you're just full of useful Clojure tips/tricks/did-you-know-about

17:20 ridcully_: they appear ordered for some smaller amount of elements. 32?

17:21 pilne: i guess i could just hack and chop my way through some things on rosetta code instead, project euler is fun, but it is all number crunching >.<

17:21 WorldsEndless: ridcully_: I have fewer than 32 and it's bonkers...

17:21 sdegutis: my goal in #clojure is to be helpful often, increasing the chance that someone will compliment me on being helpful, at which point i can reply "Shut up baby I know it B)"

17:21 justin_smith: ridcully_: I think it's 8

17:21 ,(zipmap (range 7) (range))

17:21 clojurebot: {0 0, 1 1, 2 2, 3 3, 4 4, ...}

17:21 justin_smith: ,(zipmap (range 8) (range))

17:21 clojurebot: {0 0, 1 1, 2 2, 3 3, 4 4, ...}

17:21 justin_smith: ,(zipmap (range 12) (range))

17:21 clojurebot: {0 0, 7 7, 1 1, 4 4, 6 6, ...}

17:22 WorldsEndless: Yup, that's my problem....

17:22 So, (sorted-map (zipmap...))

17:22 terom: ,(map-indexed vector (range 10))

17:22 clojurebot: ([0 0] [1 1] [2 2] [3 3] [4 4] ...)

17:22 pilne: hrm... since haskell, lisp, scheme, and clojure all play extraordinarily well with emacs, maybe i should just take the plunge

17:22 justin_smith: ,(into (sorted-map) (map vector (range 8) (range)))

17:22 clojurebot: {0 0, 1 1, 2 2, 3 3, 4 4, ...}

17:22 pilne: (my *favorite* languages)

17:22 justin_smith: ,(into (sorted-map) (zipmap (range 8) (range)))

17:22 clojurebot: {0 0, 1 1, 2 2, 3 3, 4 4, ...}

17:23 ridcully_: yeah it's 8

17:23 WorldsEndless: Ah, into sorted-map did the trick.

17:23 Thanks!

17:24 pilne: now to find the best "default clojure" setup for emacs....

17:26 sdegutis: pilne: just install clojure-mode and paredit

17:26 pilne: also cider

17:26 then bam you're done

17:40 LauJensen: Gents, datomic question. I have some :days/* which stores stats on a given day, and each day references 24 :day/hours where information specific to each hour. The ID for each day is :day/date (instant). How do I best extract every :hour/whatever value for dates in a given range?

17:44 pilne: AAARRGGHHHHHHH!!!! unity's default bindings make using the meta key in frickin emacs impossibru >.<

17:45 * pilne attempts to breathe deeply

17:45 justin_smith: option a: delete emacs, option b: change the default binding key?

17:45 pilne: sadly it isn't just one binding....

17:45 * pilne takes a deep deep breath

17:45 justin_smith: pilne: you can change the meta key it uses

17:45 pilne: (onyxia style)

17:45 or i can take this as my hint to gtfo from unity

17:46 justin_smith: pilne: in either emacs or unity, there are configs that change which key means "meta", one of them can have the windows/apple key

17:46 that works too

17:46 pilne: does gnome 3 do stupid stuff with the meta key by default that anyone knows of?

17:53 justin_smith: is there a browser plugin that makes js code run slower?

17:53 like - intentionally, I want to run this like at 1/1000 speed

17:55 pilne: hrm... maybe i'll just go with mate... i don't remember having these issues with rogue meta key shennanigans back in the gnome2 days >.<

17:55 amalloy: justin_smith: would the js stepping debugger do?

17:55 ridcully_: justin_smith: use IE *ducks and runs*

17:56 Glenjamin: you could use a transpiler to add a while loop between every statement

17:56 justin_smith: amalloy: the problem is that my bug causes the script to redirect before I can intervene - I want to see the steps before that happens

17:57 amalloy: I can't even get at the UI to enable a breakpoint - maybe if I turn js off, enable the breakpoint, then turn it on?

17:57 Glenjamin: at that point i rely on console logging with console persistence-on-page-change

17:57 justin_smith: pilne: http://askubuntu.com/questions/122209/how-do-i-modify-or-disable-the-huds-use-of-the-alt-key

17:58 Glenjamin: how does persistence-on-page-change happen?

17:58 Glenjamin: erm, there's usually a toggle in the browser

17:59 justin_smith: oh... that's what "preserve log" means... I thought it would be like a save-as prompt or something

17:59 Glenjamin: you in chrome?

17:59 ah, yeah - that

17:59 justin_smith: Glenjamin: that will do it, thanks so much

18:00 pilne: ok, before i do anything drastic, emacs default "meta" is alt and not the "super" key yes?

18:01 justin_smith: pilne: yes - but you can pick

18:01 it can use either one

18:01 or both!

18:04 pilne: ok, i can live with this

18:04 justin_smith, saving raging sleep-deprived clojurians one google at a time

18:06 justin_smith: pilne: it's something I've handled myself, back when I tried unity

18:07 pilne: sometimes i hate how much i like it in general....

18:07 i feel like i'm betryaing the FOSS mentality a bit by sticking with ubuntu/unity, but... life does not allow for me to jump ship and get comfy somewhere else right now (too many pressing deadlines overall)

18:07 justin_smith: it's not bad, except the terrible amazon tyin stuff

18:08 pilne: which i nuked

18:08 justin_smith: right, problem solved

18:08 pilne: well, at least disabled, because nuking it makes a lot of stuff... quirky

18:08 although i do use amazon a lot, i'd like to "choose" what they know about me >.<

18:08 otherwise the prime account i share with the gf might have a lot of programming books as reccomendations, instead of things she would actually be interested in lol

18:09 .... which might save some money now that i think of it lol

18:11 Glenjamin: you can have separate accounts and share prime

18:12 pilne: meh, that's true, i have an old account that i rarely use these days that i could tie to unity, but... nah

18:12 ben_vulpes: should i be surprised that function calls in a let don't throw exceptions?

18:13 justin_smith: ben_vulpes: depends, are they lazy?

18:13 pilne: i'm honestly still trying to transition myself to gnome3 or mate, and possibly fedora down the line, but... since the long-term goal is game dev... and steam seems a "decent" way to go... i feel a bit tied to ubuntu/steam (in hopes of minimizing headaches when i get to that point)

18:13 ben_vulpes: justin_smith: might be. the one in question is a my own function call wrapped around a datomic insertion.

18:14 justin_smith: ben_vulpes: datomic stuff is often lazy in my fuzzy recollection

18:14 ben_vulpes: yup.

18:14 looks like datomic transactions return a future

18:17 soooo i should probably deref all datomic-produced futures, neh?

18:17 justin_smith: ben_vulpes: well, if you wanna know how they did, sure

18:19 * ben_vulpes nods

18:19 * ben_vulpes goes off to find more electric fences upon which to piss

19:19 pilne: need moar monitors >.<

19:29 and about them electric fences... ren and stimpy taught me that it is a bad idea to piss on them

19:53 Shayanjm: grats on the new gig arrdem :)

20:05 pilne: hmmm, so by this definition "or returns either the first truthy value or the last value. and returns the first falsey value or, if no values are falsey, the last truthy value." 'and' and 'or' do *not* "short circut"?

20:11 nevermind... i confused myself

20:49 justin_smith: ,(and :pilne false (/ 1 0))

20:49 clojurebot: false

20:51 pilne: yeah... took a bit of re-reading and futzing with the REPL but i got it (:

20:51 and... other than getting used to the keybinds... emacs is rocking my socks

20:52 years ago i went with vim because vi is on every single unix box by default, logically i should have learned enough to just do what i needed to do.. and then learned emacs too

20:52 hindsight is 20/20

21:05 TimMc: pilne: Have you used vim extensively before?

21:08 princeso: how do we avoid in a function getting the args in double parenthesis ((...)) when invoquing like (a-func (cons a b))

21:08 TimMc: I know that plenty of people use vim for Clojure quite happily.

21:08 justin_smith: princeso: this is what the -> and ->> macros are for

21:09 princeso: though usually we use those when there are more than 3 or 4 nestings

21:09 TimMc: Is it? I was having trouble understanding the question.

21:09 justin_smith: (-> (cons a b) a-func)

21:09 TimMc: I figured princeso was talking about argument nesting in parens

21:09 princeso: akey man. let me try :)

21:13 pilne: i've used vim, but the solutions it gives for "pseudo buffers" always didn't feel "right" for some reason

21:13 i'm *probably* going to give spacemacs a shot here after a little while tbh

21:16 TimMc: heh

21:23 princeso: justin_smith it is a macro. A macro that call itself this is the part `(a-macro ~(cons case- (next (next body)))) . Works good exept it get the args in doubles ((args))

21:24 justin_smith: princeso: maybe you want arg splicing?

21:24 `(a ~@[b c d])

21:24 ,`(a ~@[b c d])

21:24 clojurebot: #error {\n :cause "Unable to resolve symbol: b in this context"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Unable to resolve symbol: b in this context, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6704]}\n {:type java.lang.RuntimeException\n :message "Unable to resolve symbol: b in this context"\n ...

21:24 justin_smith: ,`(a ~@'[b c d])

21:24 clojurebot: (sandbox/a b c d)

21:25 princeso: justin_smith thank a lot, let me try

21:38 devth: tfw lein clean takes more than 10 minutes

21:38 i don't get it. isn't it just rm'ing some stuff?

21:39 justin_smith: that's all it should be doing, yeah

21:39 hanging prep task?

21:40 devth: not sure. i don't have any custom tasks on this project

21:40 justin_smith: devth: you can run jstack to see what all the threads in a jvm are doing, if you share a paste link we can probably figure out why it is stuck

21:40 jstack comes with the jdk, it's a command line program 'jstack PID'

21:40 devth: k let me figure out my lein pid

21:40 i have quite a few ...

21:42 hey that kinda looks like a kill -3 on a java process

21:42 https://gist.github.com/devth/13b0731fd7fa2b75d36b

21:42 not totally sure that's the right one

21:43 hm a killall java didn't kill lein

21:44 justin_smith: devth: that was a dump of drip - drip is a tool that keeps jvms ready to deploy to speed up startup

21:44 devth: k

21:44 killing off all my various repls..

21:44 so i can hone in on the correct one

21:45 TimMc: Where's that site that tells you relative usage of different clojure libraries?

21:45 justin_smith: devth: if you use ps ax you can see the command line args

21:45 the one running clean should be distinctive

21:45 devth: that's what i'm using

21:45 they're all huge

21:45 and some are on related projects

21:45 justin_smith: TimMc: crossclj.info

21:46 devth: but they shouldn't all have "clean" as part of their command line

21:46 devth: justin_smith: after i killed off all my repls `lein clean` ran in ~2s

21:46 oh i should have looked for that before i killed it

21:46 ok next time...

21:46 for some reason having a bunch of other `lein repl`s caused lein clean to hang

21:47 TimMc: justin_smith: I'm trying to figure out if there's a more or less canonical fork of clj-aws-s3 that is up to date.

21:47 justin_smith: devth: if you plan to run multiple repls in one project, here's a neat trick: 'lein cp > classpath'; rlwrap java -cp $(cat classpath) clojure.main

21:47 TimMc: There are a bunch of forks.

21:48 justin_smith: devth: after the first time you run cp, you don't need to run it again until you change project.clj

21:48 devth: justin_smith: i usually run 1 per project but that's a cool trick to know

21:48 im gonna put it in my zshrc :D

21:49 justin_smith: TimMc: oh yeah - looks like crossclj.info only tracks one of them

21:49 devth: well it's two parts - you only need to run the lein cp once, then you can run the java -cp part multiple times - starts much faster than lein repl

21:49 devth: nice

21:50 i've been restarting my repl a ton due to poor reloadability

21:50 justin_smith: do you use component or whatever?

21:50 justin_smith: why aren't you coding reloadably?

21:50 devth: yeah, I use stuartsierra's component lib

21:50 devth: legacy code

21:50 on the todo list :)

21:52 that was just due to a huge refactoring though. normally i just reload namespaces.

21:54 time to `lein release` 🎉

21:56 princeso: justin_smith thak you man. yuo are a life saver :) Yes it worked with splicing ~@(cons ....

22:32 justin_smith This was the macro, if yuo have a break. http://pastebin.com/yD9PPXA6

22:40 ben_vulpes: aaaa datomic pull api is teh dopeness

23:06 does anyone have a hot tip for enforcing that *.cljs files get cider-load-file'd into the CLJS repl and that *.clj files get idem'd into the clojure-proper repl?

Logging service provided by n01se.net