#clojure log - Apr 02 2016

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

0:07 deepakprasanna: How can I create a lazy sequence containing all the combinations of 0, 1 and 2?

0:07 Like this

0:07 (0 1 2 00 10 20 01 11 21 02 12 22 000 100 200 010 110 210 020 120 220 001 101 201 011 111 211 021 121 221 002 102 202 012 112 212 022 122 222 0000 1000 2000

0:07 .....

0:07 Z3nnY: in wich language ?

0:07 and for wht

0:43 troydm: is it faster to conj! each element of vector or list performance wise into transient vector?

0:45 also is there a macro that will expand (into! trans [a b c]) into (conj! ((conj! (conj! trans a) b) c) ?

1:28 backnforth: Hi

1:28 How do I use the repl to run a project?

1:31 i thought it was simply to use the command "lein repl" inside a projects directory

1:35 TEttinger: backnforth: to run an application from source you can use lein run, which needs a -main fn defined and specified in project.clj

1:36 lein repl lets you interactively run fns

1:36 backnforth: I've used lein repl before to act as lein run but removing the main fn from the project.clj

1:36 i just cant remember how lol

1:37 TEttinger: I mean, at the repl you can call (-main)

1:37 it's just a fn (Clojure catchphrase)

1:37 backnforth: "run -main" ?

1:37 TEttinger: hm?

1:37 backnforth: how so?

1:37 TEttinger: no, I mean if you have a repl running, you can run main since it's a normal fn

1:37 have you used the repl before?

1:37 backnforth: very little

1:38 i understand that its great for fn functions

1:38 but i dont know how to call a projects function with it

1:39 I know I would have to start off with the command "lein repl"

1:39 TEttinger: do you mean at the command line, before you get into the repl?

1:39 backnforth: and thats while im inside the projects directory

1:39 but i dont know what command to use in the repl

1:39 TEttinger: oh, the repl is all clojure code

1:40 there aren't special commands, really.

1:40 try this in the repl

1:40 (println "Hello!")

1:40 should print the string you gave it and also print nil at the end, where nil is just what println returns

1:42 backnforth: TEttinger, yes i understand that

1:42 i love that about the repl

1:42 TEttinger: since -main is a fn, if the repl loaded the same ns as the ns main is in, you can just type (-main) to call it with no args. if you aren't in the same ns, you can use or require the ns that has -main

1:42 backnforth: ,(require ns.core)

1:42 like that?

1:42 clojurebot: #error {\n :cause "ns.core"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.ClassNotFoundException: ns.core, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6688]}\n {:type java.lang.ClassNotFoundException\n :message "ns.core"\n :at [java.net.URLClassLoader$1 run "URLClassLoader.java" 366]}]\n :trace\n [[java.net.URLClass...

1:42 TEttinger: require allows you to add :reload at the end of the args, too

1:43 heh, clojurebot doesn't have your code, but almost

1:44 (require '[ns.core]) is how I think it expects it, but there's also (require '[ns.core :reload]) for when you changed the code and want to update the code in the running repl

1:44 I might have the syntax wrong

1:44 justin_smith!

1:45 backnforth: (require 'ns.core) worked for me :D

1:45 this is excellent

1:45 Thank you TEttinger

1:46 TEttinger: woo, glad it works!

1:46 backnforth: i feel "repl"ed

1:46 its great

1:56 Hey TEttinger

1:56 How do I go about running the main function afterwards?

2:00 TEttinger: is it called -main in ns.core?

2:00 if you required it, I think (core/-main)

2:03 backnforth: I get a file not found exception with (require 'ns.core/-main)

2:04 and im using the right ns lol

2:06 FileNotFoundException Could not locate _main__init.class or _main.clj on classpath. Please check that namespaces with dashes use underscores in the Clojure file name. clojure.lang.RT.load (RT.java:456)

2:07 but it works when i remove /-main

2:07 TEttinger: eh?

2:07 no, you don't require a fn

2:08 you require the whole ns, it has the fn in it

2:08 that should let you call -main with either (ns.core/-main) or something else if yous used require with :as

2:08 backnforth: oh lol

2:08 ohh ok

2:08 TEttinger: I forgot the ns. part

2:09 ,(require '[clojure.string :as st :reload])

2:09 clojurebot: #error {\n :cause "No value supplied for key: true"\n :via\n [{:type java.lang.IllegalArgumentException\n :message "No value supplied for key: true"\n :at [clojure.lang.PersistentHashMap create "PersistentHashMap.java" 77]}]\n :trace\n [[clojure.lang.PersistentHashMap create "PersistentHashMap.java" 77]\n [clojure.core$hash_map invokeStatic "core.clj" 387]\n [clojure.core$hash_map doInvoke "...

2:09 TEttinger: hm.

2:09 ,(require '[clojure.string :as st] :reload)

2:09 clojurebot: nil

2:10 TEttinger: ,(st/join ", " (range 5))

2:10 clojurebot: "0, 1, 2, 3, 4"

2:11 TEttinger: ,(require '[clojure.set])

2:11 clojurebot: nil

2:11 TEttinger: ,(clojure.set/union #{:a} #{:b})

2:11 clojurebot: #{:b :a}

2:13 backnforth: aha

2:13 it works

2:14 Thanks TEttinger

2:14 i'm logging off now.

2:14 TEttinger: great glad it works

2:14 backnforth: gn

3:43 deepakprasanna: I have a function which spits a very long output on the repl.

3:43 How do I silence the output?

3:43 I want to measure the time, using (time expr).

3:43 luma: (time (f) nil)

3:43 deepakprasanna: I tried (time (do (expr) nil))

3:44 luma: Thanks mate.

3:44 luma: right, it needs the (do ..)

3:44 time only takes one expression

3:44 so (time (do (f) nil))

3:44 so it returns nil

3:45 deepakprasanna: luma: ArityException Wrong number of args (2) passed to: core/time clojure.lang.Compiler.macroexpand1 (Compiler.java:6636)

3:45 luma: If I wrap it with (do ...), the expr is not getting evaluated.

3:46 luma: do evaluates all the expressions inside it

3:46 is your function returning a lazy sequence?

3:46 if it is, try (time (do (doall (expr)) nil))

3:50 deepakprasanna: luma: Worked, thanks,

4:16 narwhal01: guys what is the best way to remove maximum key from map?

4:19 apply max on keys

4:19 got it

4:19 luma: why do you want to remove the maximum key?

4:20 dysfun: what you probably want is an ordered map and to remove the *last* key

4:20 ridcully: find it and dissoc it

4:24 ane: ,(let [m {12 :a 99 :b 3 :c}] (dissoc m (apply max (keys m))))

4:24 clojurebot: {12 :a, 3 :c}

4:30 kwladyka: Did you do OAuth 2 server / client / resource app and can share some experience with me about solutions?

4:37 Is any good quality solution in Clojure or i should looking in Java?

4:38 I found mainly 2 solutions https://github.com/pelle/clauth and https://github.com/ddellacosta/friend-oauth2 but i feel something is wrong with them. Just some intuition, but maybe i am in mistake :)

4:39 dysfun: i've used friend, though not with that plugin and it's alright

4:40 kwladyka: dysfun you use https://github.com/cemerick/friend without https://github.com/ddellacosta/friend-oauth2 ?

4:41 dysfun: yes

4:41 kwladyka: and works like that: client <-> authorisation app ; client <-> resource app without any problem?

4:41 do you this on github?

4:42 dysfun: no, i'm merely saying 'friend is alright' and that i can't comment on that friend-oauth2 extension

4:42 kwladyka: *have

4:42 oh

4:42 last commit of this extension is 1 year ago. It is so good or so bad :)

4:42 but i see issues from 2015 didn't touched

4:43 it alerts me a little

4:43 dysfun: have a read of the code and satisfy yourself. if it doesn't work, are you willing to fix it?

4:43 kwladyka: oh even from 2014

4:43 yeah but first i want be sure it is the right solution. Maybe exist something simpler.

4:44 For example i can use Java library

4:44 or maybe write my own part for OAuth 2 is simpler then i think

4:44 i don't have experience with OAuth 2

4:44 so i am not sure what i am doing now :)

4:44 that the reason why i am looking something ready to learn from that also.

5:56 lq: http://shrinkmy.com/lJn7D5Tq

6:05 kwladyka: lg spamming bot?

7:51 irctc: join

8:01 Guest64304: http://oortr.com/YmU5NT

9:18 Guest42377: http://oortr.com/YmU5NT

9:42 justin_smith: (time (do (dorun (f)) nil))

9:43 oh man I was scrolled back

9:50 lqlq: http://oortr.com/YmU5NT

10:19 sdegutis: What non-alphanumeric symbols are available to use in Clojure binding names?

10:20 Thanks in advance. Regards, me.

10:20 justin_smith: sdegutis: officially, or in practice?

10:20 sdegutis: In practice.

10:21 justin_smith: everything but #, {, }, [, ], (, ), @ should work

10:21 oh, and .

10:21 sdegutis: I have models and presenters for the same thing, so I often create :require-based aliases such as :user-presenter or :user-view, and I'd like to shorten that up a bit with a convention. Perhaps something like user$ for view and user% for presenter, I dunno.

10:22 justin_smith: my bad, # is valid (just not in the beginning)

10:22 same as '

10:22 sdegutis: ,(let [x$ 1, x% 2, x* 3, x# 4] [x$ x% x* x#])

10:22 clojurebot: [1 2 3 4]

10:22 justin_smith: sdegutis: those are both valid - as would be user# and user'

10:25 sdegutis: ,(let [x$ 1, x% 2, x* 3, x# 4, x' 5, x! 6, x& 7, x+ 8, x- 9, x= 10, x| 11, x? 12] [x$ x% x* x# x'x! x& x+ x- x- x| x?])

10:25 clojurebot: #error {\n :cause "Unable to resolve symbol: x'x! in this context"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Unable to resolve symbol: x'x! in this context, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6688]}\n {:type java.lang.RuntimeException\n :message "Unable to resolve symbol: x'x! in this co...

10:26 justin_smith: sdegutis: you missed whitespace

10:26 sdegutis: ,(let [x$ 1, x% 2, x* 3, x# 4, x' 5, x! 6, x& 7, x+ 8, x- 9, x= 10, x| 11, x? 12] [x$ x% x* x# x' x! x& x+ x- x- x| x?])

10:26 clojurebot: [1 2 3 4 5 ...]

10:26 sdegutis: Now it's just a matter of picking three that seem reasonable for presenter, view, and model.

10:30 Presenter should be $ since it usually involves formatting, similar to turning 1234.56 into "$1,234.56"

10:32 View should be x# since it's just Hiccup which often involves HTML-based #id tags.

10:33 Hmm, maybe model should be x! since it's just way so exciting?

10:33 Not x= because that implies some sort of equality checking.

10:34 Maybe model could be x+ because it represents the superior form of x, in terms of purity.

10:40 https://gist.github.com/sdegutis/dad5e45b905c5161baaf0d96e37a6d63

10:47 Done, x* is chosen for no real reason.

10:55 justin_smith: thanks

10:56 Oh dang, there's also services.

10:56 But that one makes sense as !

10:57 (email!/send-email ...)

10:57 etc

10:59 justin_smith: ,(defn *&!%?$#! [] (throw (Exception. "OH FUCK!")))

10:59 clojurebot: #'sandbox/*&!%?$#!

11:03 sdegutis: ,(try (*&!%?$#!) (catch Exception e))

11:03 clojurebot: sdegutis: Gabh mo leithscéal?

11:03 sdegutis: wha??

11:04 ,(*&!%?$#!)

11:04 clojurebot: #error {\n :cause "OH FUCK!"\n :via\n [{:type java.lang.Exception\n :message "OH FUCK!"\n :at [sandbox$_STAR__AMPERSAND__BANG__PERCENT__QMARK_$_SHARP__BANG_ invokeStatic "NO_SOURCE_FILE" -1]}]\n :trace\n [[sandbox$_STAR__AMPERSAND__BANG__PERCENT__QMARK_$_SHARP__BANG_ invokeStatic "NO_SOURCE_FILE" -1]\n [sandbox$_STAR__AMPERSAND__BANG__PERCENT__QMARK_$_SHARP__BANG_ invoke "NO_SOURCE_FILE" -1]\...

11:04 justin_smith: sdegutis: clojurebot doesn't allow try/catch

11:04 sdegutis: haha stupid cljdbot

11:04 justin_smith: oh man, I like the munged name there

11:05 sdegutis: ,sandbox$_STAR__AMPERSAND__BANG__PERCENT__QMARK_$_SHARP__BANG_

11:05 clojurebot: #error {\n :cause "Unable to resolve symbol: sandbox$_STAR__AMPERSAND__BANG__PERCENT__QMARK_$_SHARP__BANG_ in this context"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Unable to resolve symbol: sandbox$_STAR__AMPERSAND__BANG__PERCENT__QMARK_$_SHARP__BANG_ in this context, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "...

11:05 sdegutis: justin_smith: is it possible to access it manually?

11:07 justin_smith: sdegutis: wow, I don't know - maybe it's private or protected or something?

11:07 sdegutis: Like, is it ever possible to do that?

11:07 * sdegutis tries locally

11:08 sdegutis: dunno

11:38 kwladyka: I send data between APP1 <-> APP2 - what is the best solution in Clojure (or generally) to secure that connection to prevent from read this data?

11:41 sdegutis: dunno

11:42 TimMc: sdegutis: Clojurebot forbids try/catch but allows eval... so it also allow try/catch. :-P

11:43 ,(eval '(try (/ 1 0) (catch Throwable t :oops)))

11:43 clojurebot: TimMc: Gabh mo leithscéal?

11:43 TimMc: ,(eval '(~(symbol "try") (/ 1 0) (catch Throwable t :oops)))

11:43 clojurebot: TimMc: Excuse me?

11:43 TimMc: ,(eval '(~(symbol (str "t" "ry")) (/ 1 0) (catch Throwable t :oops)))

11:43 clojurebot: TimMc: excusez-moi

11:43 TimMc: Ooooh, is it the catch?

11:44 Oh that was silly, I would need ` anyway.

11:46 ,(eval `(~(symbol (str "t" "ry")) (/ 1 0) (~(symbol (str "c" "atch")) Throwable ~'t :oops)))

11:46 clojurebot: :oops

11:46 TimMc: So tedious.

11:46 justin_smith: lol

11:49 kwladyka: Do you know channel where people talk about design software non depended with any language? Just channel for software designer / architects ?

11:50 justin_smith: #programming or maybe it was ##programming

11:51 kwladyka: thx i will try :)

12:50 sdegutis: Never mind.

12:50 My idea was terrible.

12:58 Good day.

12:58 How are you amalloy?

12:58 user_: Hi

13:04 backnforth: Hi

13:14 Kototama: hi, i'm updating some of my old clojure libraries and leiningen now complains that I have a version range for the clojure in the :deps

13:14 what are the current best practice to support different version of clojure in libs?

13:15 justin_smith: Kototama: it's a warning, ranges are discouraged

13:15 Kototama: don't

13:15 Kototama: I should force my users to use clojure 1.8 ?

13:15 justin_smith: Kototama: just write your code, require whichever version of clojure makes sense, and a user is free to override that decision

13:16 Kototama: it's impossible to force someone to use a specific clojure version unless you aot compile, and you shouldn't ever AOT compile libs

13:16 Kototama: with :exclude ?

13:16 justin_smith: Kototama: or just by specifying the version of clojure they want

13:16 Kototama: ok

13:16 justin_smith: that implicitly excludes the versions all libs ask for

13:16 Kototama: i though it would give a conflict

13:17 was it always like this?

13:17 justin_smith: Kototama: if you go through the libs you use, one or two might specify a range, the rest all ask for different clojure versions, the conflict is resolved by using the clojure version explicitly provided in the project version (if any)

13:17 Kototama: always

13:18 Kototama: does it apply for all libs or is it specific to clojure lib? and is it how maven works?

13:19 justin_smith: Kototama: yes, these are the standard dependency resolution rules, borrowed from maven (but lein doesn't use maven - it uses some of the same underlying libs though)

13:20 you probably know this already, but the reason version ranges are discouraged is they can turn good code into broken code, without you changing anything in your repo

13:21 that, and the end user can easily override any version explicitly, just by asking for that version at the top level

13:21 Kototama: yes

13:21 i guess it should mostly be okay since clojure is most of the time bacwards compatible but i see the point

13:22 justin_smith: yeah - the warning is a generic one, and the risk is much greater if you had a version range in eg. a CIDER dep

13:22 Kototama: actually for having done python in the last year, clojure's story here is much much much better

13:22 justin_smith: Kototama: that's what I hear, I hope I never have to find out for real

13:22 Kototama: :)

13:22 i hope for you too :)

13:22 thanks for the help

13:24 bendavisnc: is there a write up anywhere on the differences between core.async in clojure vs clojurescript?

13:25 am i just crazy or does there seem to be just like <! in a go block and it blocks

13:25 justin_smith: bendavisnc: <! is supposed to park

13:26 it won't resume until something comes in on the channel

13:26 bendavisnc: i don't see a <!! available (the blocking version)

13:26 kwladyka: bendavisnc yes, there is <! and (go) and works as expected

13:26 justin_smith: bendavisnc: this is because js has no threads

13:27 kwladyka: bendavisnc async is especially used when you load another webpage or some data resources from another site

13:27 justin_smith: bendavisnc: <!! would cause deadlocks

13:27 in clj we can deliver to the channel from another thread, there is no such thing in cljs

13:28 sdegutis: Hi.

13:28 kwladyka: bendavisnc this is example of using async https://github.com/r0man/cljs-http

13:28 bendavisnc: ok but these differences, at least api differences, are not anywhere to be found?

13:28 and i think i'm missing the idea of park

13:29 i assumed park was just basically the opposite of block, ie code after will execute without a value existing on the channel

13:29 justin_smith: parking releases your thread of execution to the core.async thread pool

13:29 bendavisnc: parking only resumes when the action that was parked completes (some value is read or written, usually)

13:30 troydm: I have sequence with 5 or more elements and I'm using destructuring using let [[a b c d e] s] but before doing that I'm making sure that destructuring will succed as otherwise I need to handle destructuring using 3 elements instead

13:30 justin_smith: bendavisnc: as opposed to <!! - which does not release your thread, you prevent the thread from doing work until you read, but no other code can use that thread

13:30 troydm: I'm using cond with count , is this correct way?

13:31 bendavisnc: ok @justin_smith, sorry but i'm yet more confused because you keep explaining things with the word "thread" after telling me javascript has no threads

13:31 justin_smith: troydm: what do you mean by "succeed"? are nil values a problem?

13:31 troydm: justin_smith: yes they are

13:31 bendavisnc: so there's something simulated here that you mean when you say thread?

13:31 justin_smith: ,(let [[a b c d e] [1]] [a b c d e])

13:31 clojurebot: [1 nil nil nil nil]

13:31 justin_smith: bendavisnc: no, I am talking about real threads on the OS

13:32 bendavisnc: but js as no threads

13:32 as you said

13:32 justin_smith: bendavisnc: in cljs there is one thread only, and that is why you can use parking ops but not blocking ones - the blocking ones would not return if they blocked

13:32 bendavisnc: you always have a thread

13:32 but maybe only just one

13:32 troydm: justin_smith: anyway my code works just fine, was just curious if I was on right track or not

13:33 bendavisnc: ok ok

13:33 so but in cljs, where we just have one thread

13:34 inside a go block, <! works exactly the same cljs and clj?

13:34 they both prevent execution until something enters the chan?

13:34 justin_smith: yes, both let some other go block use your thread, until some code parks in another go block and you have a value to read

13:34 bendavisnc: ok ok

13:34 kwladyka: bendavisnc bendavisnc http://stackoverflow.com/questions/14130266/is-there-a-difference-between-concurrency-and-parallelism-in-java difference between concurrency and parallelism

13:35 bendavisnc: i think things are starting to clear up

13:35 so in java, you only use the blocking version, <!!, inside a go block, when you want to stop all go block things from happening?

13:35 kwladyka: bendavisnc js is not concurrency but is parallelism

13:36 bendavisnc: i'm having a hard time imagining a case for that right now, using the <!! inside a go block in java

13:36 justin_smith: bendavisnc: you can't stop all go blocks, but you can "steal" a thread (by doing a blocking op in a go block)

13:36 bendavisnc: in java, there's actually a pool of x*cpus+42 threads available to go blocks

13:37 bendavisnc: is there anyone you guys could show me an example inside a go block where <!! makes sense?

13:37 justin_smith: bendavisnc: it's never a good idea - <!! is for usage outside go blocks

13:37 bendavisnc: ok cool

13:37 thanks guys

13:48 Kototama: how can i specify with which to gpg key to sign the JAR when deploying to clojars?

13:51 ah there is a "default-key" option in gpg.conf

13:58 skeuomorf: The grimoire is sad

13:59 kwladyka: Is any other way worth to consider to deploy clojure ring app except nginx? I want run many clojure apps on server with https (probably digitalocean)

14:02 ben_vulpes: kwladyka: jetty does ssl termination

14:03 but determining which app to proxy a request to is probably a job for nginx

14:04 kwladyka: ben_vulpes thx

14:11 sdegutis: Woot.

14:11 justin_smith: never mind my Protocol idea, it was terrible.

14:11 It looked clever at first, but it's just not really helpful.

14:12 justin_smith: sdegutis: the naming thing, or the routing thing?

14:12 sdegutis: The whole (defprotocol web/Feature (routes[_]) (css[_]) (schema[_]) )

14:13 It was interesting, but a bit too convoluted and inflexible.

14:14 ane: to overengineer or not to overengineer, that is the question

14:16 kwladyka: always simplicity :)

14:17 tolstoy: Over-engineering is a great way to learn simplicity, two weeks later.

14:17 Kototama: or two years

14:18 kwladyka: heh

14:18 Kototama: makes it simple and then refactor when the need comes

14:18 sdegutis: yeah for me it was 3.5 years

14:19 tolstoy: Assumes you already know what simple is. That's an ever changing assessment.

14:24 ben_vulpes: put all your db strings in atoms

14:24 slam 'em in there at runtime

14:24 #yolo

14:24 justin_smith: Different amounts of engineering are apropriate in different places. I prefer to do a lot of assertions, data checking, and modularity at subsystem boundaries, and then the deeper I get into one domain, the more ad-hoc and simple I go.

14:25 ane: the nice thing about clojure is that it embodies a certain philosophy of simplicity in all things

14:25 tolstoy: Yeah. Based on a lot of experience, I bet.

14:25 ane: which is why it's such a nice language

14:25 justin_smith: one nice consequence of this is it makes sure I minimize the coupling between components, because all the engineering gets tedious

14:25 ane: writing terribly complicated crap doesn't feel right

14:26 justin_smith: tolstoy: much pain brought me to this approach, yes

14:27 tolstoy: ane But what's "terribly complicated"? For instance, going all in on Component seems to be over-the-top, but, then, it also seems a nice way to not have to understand everything all at once.

14:28 For me, "simple" is on the small level: all functions take the context they need to work, and at the large level, an app is a kernal running nearly independent modules.

14:33 sdegutis: welp

14:33 Now I have a bunch of (defn routes[env] [...])

14:33 which I actually like a ton better than a stupid record

14:36 tolstoy: sdegutis For me, anyway, the advantage of a record with a protocol is for a certain kind of testing. Or if you have lots of "route" things with a core set of functions.

14:37 sdegutis: tolstoy: I have like 20 (defn routes[...]) currently, all in different namespaces

14:37 tolstoy: sdegutis But I think I over use that stuff. Still trying to find the right balance rather than just consider records/protocols an aversion.

14:37 sdegutis: tolstoy: and a single routes.clj that just concatenates them all

14:38 Guest70632: http://oortr.com/YmU5NT

14:41 tolstoy: Is there an edn pretty formatter for Emacs when you're typing in stuff in, say, Markdown?

14:41 cider-format-edn-region requires an nrepl.

14:52 sdegutis: Hmm. I wonder...

14:52 Is it even possible?

14:53 tolstoy: There's a json prettify.

15:27 OscarZ: if i have (future ...) and there is some exception, will it be swallowed silently ?

15:27 tolstoy: OscarZ Probably. You can register an uncaughtExceptionHandler with the JVM on start-up.

15:29 OscarZ: ok thanks.. maybe i should do that

15:32 tolstoy, do you always do that in your programs?

15:33 tolstoy: OscarZ Yeap.

15:33 It's a good place to, say, intercept an OOM message and just tank the JVM as a result so that a supervisor can restart.

15:34 OscarZ: what do you mean by "tank the JVM" ? is there some way to recover from that ?

15:34 tolstoy: I don't know. Maybe.

15:36 In my use case, I had a process that woke up periodically (scheduledExecutor), gathered data and wrote it down. But, on a 512 Digital Ocean instance, I got an OOM. App kept running, but no data was gathered.

15:37 ben_vulpes: make you some swap

15:37 tolstoy: Much better to tank and restart and loose a few slices of data then go for many hours not noticing. Then we can work on the root cause. ;)

15:37 Like swap. Or whatever.

15:38 Anyway, I use that uncaught thing for surprises, mainly.

15:38 I've done things like (try (do-something-that-eventually-spawns-a-thread) (catch Throwable t (log t)) and wondered why I never got an exception. ;)

15:41 OscarZ: tolstoy, you dont catch the OOM ex with that?

15:41 tolstoy: Exceptions in one thread aren't caught in other threads.

15:42 I just mean I register that handler so that if I make a mistake, I can find it much more quickly.

15:42 OscarZ: sure ok

15:52 sdegutis: tolstoy, justin_smith, ben_vulpes: what do you think of this macro? https://gist.github.com/sdegutis/4112dd4989fe7be17326d57d57b3ca6b

15:58 justin_smith: isn't it more like every-let rather than some-let?

16:00 luma: I named a similar function when-let+, because it's like when-let but allows multiple bindings

16:01 sdegutis: justin_smith: maybe... I was thinking of some-> when I named it

16:02 luma: oh, I like the idea of it being when- rather than if- so that it can have a big body.

16:02 luma: also I was gonna ping you for feedback but I didn't see your name all morning so I thought you were asleep

16:05 luma: It's 11 pm...

16:05 sdegutis: luma: no its 3pm

16:10 hamid: Hi. How can I pop the nth element of a vector?

16:11 Or I'd better ask this: I do have this vector of vectors, [ [x1, y1, w1, z1], [x2, y2, w2, z2],...], what is the shortest/fastest way to drop all the Ws?

16:12 I was using destructing and with a map then I thought maybe I can use "dissoc vec nth", and I tried ... didnt work. :/

16:13 http://stackoverflow.com/questions/1394991/clojure-remove-item-from-vector-at-a-specified-location

16:13 "Removing elements from the middle of a vector isn't something vectors are necessarily good at. If you have to do this often, consider using a hash-map so you can use dissoc." ?

16:15 ridcully: i start the bidding with:

16:15 ,(mapv (fn [[x y _ z]] [x y z]) [[11 12 13 14] [21 22 23 24]])

16:15 clojurebot: [[11 12 14] [21 22 24]]

16:17 justin_smith: ridcully: the problem is that can't share any structure with the input vector - it's a full new vector tree created every time

16:17 ridcully: while a map with numeric indexes using dissoc, you can share structure

16:20 hamid: justin_smith, dissoc on vector? like (dissoc [1 2 3 4] 1); => [1 3 4] ?

16:23 ridcully: ,(map #(dissoc % 3) [{1 1 2 2 3 3 4 4}])

16:23 clojurebot: ({1 1, 2 2, 4 4})

16:24 hamid: ,(hash [1,2,3])

16:24 clojurebot: 736442005

16:24 hamid: ,(into {} [1,2,3])

16:24 clojurebot: #error {\n :cause "Don't know how to create ISeq from: java.lang.Long"\n :via\n [{:type java.lang.IllegalArgumentException\n :message "Don't know how to create ISeq from: java.lang.Long"\n :at [clojure.lang.RT seqFrom "RT.java" 542]}]\n :trace\n [[clojure.lang.RT seqFrom "RT.java" 542]\n [clojure.lang.RT seq "RT.java" 523]\n [clojure.lang.ATransientMap conj "ATransientMap.java" 42]\n [cloju...

16:43 kwladyka: ring-mock return :body as java.io.BufferedInputStream, can i convert it into clojure data using https://clojure.github.io/clojure/clojure.java.io-api.html? How?

17:01 justin_smith: hamid: I mean that you can dissociate indexes from a vector, and translating from vectors to integer keyed hash-maps is simple

17:02 ,(reduce (partial apply assoc) {} (map-indexed list [:a :b :c :d]))

17:02 clojurebot: {0 :a, 1 :b, 2 :c, 3 :d}

17:03 justin_smith: ,(reduce conj [] (map val (sort {0 :a, 1 :b, 2 :c, 3 :d})))

17:03 clojurebot: [:a :b :c :d]

17:03 justin_smith: ,(reduce conj [] (map val (sort (dissoc {0 :a, 1 :b, 2 :c, 3 :d} 2))))

17:03 clojurebot: [:a :b :d]

17:05 justin_smith: oh, those should use into instead of reduce conj

17:07 kwladyka: ha i found.... slurp works

17:22 libertytrader: jh

17:22 Why doesnt this work : (defn dumb [a] (a))

17:22 (dumb [5])

17:27 ridcully: libertytrader: because [5] is no function

17:29 kwladyka: ,(defn dump [a] a)

17:29 ,(dump 3)

17:29 clojurebot: #'sandbox/dump

17:29 3

17:29 kwladyka: ,(defn dump2 [a] (a))

17:29 clojurebot: #'sandbox/dump2

17:29 kwladyka: ,(dump2 3)

17:29 clojurebot: #error {\n :cause "java.lang.Long cannot be cast to clojure.lang.IFn"\n :via\n [{:type java.lang.ClassCastException\n :message "java.lang.Long cannot be cast to clojure.lang.IFn"\n :at [sandbox$dump2 invokeStatic "NO_SOURCE_FILE" 0]}]\n :trace\n [[sandbox$dump2 invokeStatic "NO_SOURCE_FILE" 0]\n [sandbox$dump2 invoke "NO_SOURCE_FILE" 0]\n [sandbox$eval95 invokeStatic "NO_SOURCE_FILE" 0]\n [...

17:30 kwladyka: it is because first value in the () is always treated as a function

17:30 ridcully: ,(def dump identity)

17:30 clojurebot: #'sandbox/dump

17:30 ridcully: ,(dump 3)

17:30 clojurebot: 3

17:30 kwladyka: so:

17:30 ,(1)

17:30 clojurebot: #error {\n :cause "java.lang.Long cannot be cast to clojure.lang.IFn"\n :via\n [{:type java.lang.ClassCastException\n :message "java.lang.Long cannot be cast to clojure.lang.IFn"\n :at [sandbox$eval165 invokeStatic "NO_SOURCE_FILE" 0]}]\n :trace\n [[sandbox$eval165 invokeStatic "NO_SOURCE_FILE" 0]\n [sandbox$eval165 invoke "NO_SOURCE_FILE" -1]\n [clojure.lang.Compiler eval "Compiler.java" 69...

17:31 kwladyka: ,1

17:31 clojurebot: 1

17:31 kwladyka: there is no function 1

17:31 but

17:31 ,(defn 1 [] "foo")

17:31 clojurebot: #error {\n :cause "First argument to defn must be a symbol"\n :via\n [{:type java.lang.IllegalArgumentException\n :message "First argument to defn must be a symbol"\n :at [clojure.core$defn__4387 invokeStatic "core.clj" 294]}]\n :trace\n [[clojure.core$defn__4387 invokeStatic "core.clj" 294]\n [clojure.core$defn__4387 doInvoke "core.clj" 292]\n [clojure.lang.RestFn invoke "RestFn.java" 494]\...

17:31 kwladyka: ,(defn '1 [] "foo")

17:31 clojurebot: #error {\n :cause "First argument to defn must be a symbol"\n :via\n [{:type java.lang.IllegalArgumentException\n :message "First argument to defn must be a symbol"\n :at [clojure.core$defn__4387 invokeStatic "core.clj" 294]}]\n :trace\n [[clojure.core$defn__4387 invokeStatic "core.clj" 294]\n [clojure.core$defn__4387 doInvoke "core.clj" 292]\n [clojure.lang.RestFn invoke "RestFn.java" 494]\...

17:31 kwladyka: ech it doesn't work like i wanted to show :)

17:32 justin_smith: ,(type '1)

17:32 clojurebot: java.lang.Long

17:32 kwladyka: ,(defn (symbol 1) [] "foo")

17:32 clojurebot: #error {\n :cause "First argument to defn must be a symbol"\n :via\n [{:type java.lang.IllegalArgumentException\n :message "First argument to defn must be a symbol"\n :at [clojure.core$defn__4387 invokeStatic "core.clj" 294]}]\n :trace\n [[clojure.core$defn__4387 invokeStatic "core.clj" 294]\n [clojure.core$defn__4387 doInvoke "core.clj" 292]\n [clojure.lang.RestFn invoke "RestFn.java" 494]\...

17:32 justin_smith: ,(eval `(defn ~(symbol "1") [] "foo"))

17:32 clojurebot: #'sandbox/1

17:32 justin_smith: :P

17:33 kwladyka: defn is a macro, first arg has to be a literal symbol

17:33 kwladyka: justin_smith like always rules :)

17:33 justin_smith: haha

17:33 kwladyka: so now:

17:33 ,(1)

17:33 clojurebot: #error {\n :cause "java.lang.Long cannot be cast to clojure.lang.IFn"\n :via\n [{:type java.lang.ClassCastException\n :message "java.lang.Long cannot be cast to clojure.lang.IFn"\n :at [sandbox$eval326 invokeStatic "NO_SOURCE_FILE" 0]}]\n :trace\n [[sandbox$eval326 invokeStatic "NO_SOURCE_FILE" 0]\n [sandbox$eval326 invoke "NO_SOURCE_FILE" -1]\n [clojure.lang.Compiler eval "Compiler.java" 69...

17:33 justin_smith: ,((resolve (symbol "1")))

17:33 kwladyka: hmm

17:33 clojurebot: "foo"

17:34 kwladyka: libertytrader ^

17:34 ridcully: i like how this channel turns some kitten questions into a cat butchery

17:34 justin_smith: ridcully: lol

17:35 kwladyka: ;)

17:35 libertytrader: so you cant just return an s-expression in clojure?

17:35 justin_smith: libertytrader: you can, but (f) in clojure is what f() would be in other languages

17:36 libertytrader: and you probably know why 1() would be an error

17:36 libertytrader: if you want to return an unevaluated form, you can use ` or ' or list

17:36 kwladyka: ,(inc 1) ; inc is a function, it is on the first place of the list

17:36 clojurebot: 2

17:36 justin_smith: ,(list 1)

17:36 clojurebot: (1)

17:36 kwladyka: ,'(1) ; now it is a list with first argument 1 because of char '

17:36 clojurebot: (1)

17:37 libertytrader: man... i feel like i understand the ideas behind lisp

17:37 but clojure errors are so bad

17:37 thanks for your help

17:38 justin_smith: libertytrader: they are hard to understand at first, it's true - eventually they become more familiar, and for me learning to interpret them is a small price to pay for the nice parts of the language

17:38 libertytrader: is there a way to know if i have a sequence or collection in lisp?

17:38 i mean clojure

17:38 justin_smith: libertytrader: coll?

17:38 libertytrader: i want to use filter on a something

17:38 ok

17:38 justin_smith: ,(coll? [])

17:38 clojurebot: true

17:38 justin_smith: ,(coll? ())

17:38 clojurebot: true

17:38 justin_smith: ,(coll? nil)

17:39 clojurebot: false

17:39 justin_smith: ,(sequential? [])

17:39 clojurebot: true

17:39 justin_smith: there's a few relevant predicates

17:39 libertytrader: ok so i have a coll... does that mean i have to convert it to a seq to use filter

17:39 justin_smith: libertytrader: filter turns it's arg into a seq (if it knows how)

17:39 libertytrader: most collection functions in clojure call seq on their arg for you

17:39 ,(coll? "hi")

17:39 clojurebot: false

17:40 justin_smith: ,(filter #{\h} "hi")

17:40 clojurebot: (\h)

17:40 justin_smith: ,(filter #{\i \h} "hi")

17:40 clojurebot: (\h \i)

17:41 kwladyka: libertytrader generally most of functions works with everything, but not always and not always in the same way :) But most of cases you don't have to care about converting data to seq or from list to vector etc.

17:41 but sometimes you have to :)

17:42 libertytrader: hmm ok

17:42 kwladyka: for example conj add to vector on last position, but to the list on the first position. So then you have to care about keep in the one form.

17:43 if you care about item orders

17:45 libertytrader: ok i have heard of that one

17:45 is the best way to compare strings compare?

17:46 ridcully: ,(defn dumb [a] (a))

17:46 clojurebot: #'sandbox/dumb

17:46 ridcully: ,(dumb +)

17:46 clojurebot: 0

17:47 ridcully: ,(= "x" "x")

17:47 clojurebot: true

17:47 ridcully: ,(= "x" "y")

17:47 clojurebot: false

17:48 xphil: kinda surprising about conj behavior different for list vs. vector. why is that?

17:49 ridcully: conj does the "optimal" thing

17:50 prepend on a list is trivial to do, so is append on a vector

17:50 sdegutis: I'm back.

17:51 kwladyka: xphil list and vector just have other purpose

17:51 i don't remember technical explanation :)

17:51 xphil: haha ridcully's explanation sounds fine… although appending to a list is pretty easy if you keep track of the last node

17:52 ridcully: you only have the head. so prepending is easier

17:52 kwladyka: xphil append to the list has bad performance

17:52 sdegutis: linked list appends most efficiently at beginning

17:52 memory-sequential vector appends most efficiently at end

17:52 ridcully: "here is the new head"

17:52 kwladyka: http://stackoverflow.com/questions/14657178/clojure-list-vs-vector-vs-set <- this should be helpful

17:53 xphil: cool, thanks

17:56 ridcully: ,(take 5 (conj (range) -1))

17:56 clojurebot: (-1 0 1 2 3)

18:07 user_: Hello, how can I use printf to print multiple arguments included a vector?

18:07 I can't find any good documentation anywhere :(

18:07 TEttinger: ,(doc printf)

18:07 clojurebot: "([fmt & args]); Prints formatted output, as per format"

18:08 user_: What are the formats?

18:08 I seen a %s

18:08 that usually means a string

18:08 justin_smith: ,(apply printf "%s %s %s %s\n" ["same" "as" "any" "function"])

18:08 clojurebot: same as any function\n

18:08 TEttinger: ,(apply printf "%s: %08d" ["TEttinger" 9999])

18:08 clojurebot: TEttinger: 00009999

18:09 TEttinger: formats are a mini-language, check the Formatter docs...

18:09 justin_smith: user_: printf is weak, it uses java.util.Format which is weak. If you want a powerful formatter, there is clojure.pprint/cl-format

18:09 user_: awesome

18:09 will do

18:09 justin_smith: also cl-format is more confusing, and for that check out common lisp docs and tutorials

18:09 TEttinger: https://docs.oracle.com/javase/7/docs/api/java/util/Formatter.html

18:09 user_: Thank you both

18:10 TEttinger: cl-format is a mega-mini-language, it's much deeper

18:10 user_: interesting

18:10 justin_smith: TEttinger: perhaps "language" would suffice!

18:10 ridcully: or universe

18:11 TEttinger: you can do pretty insane formatting with cl-format. I think stuff like parsing a string argument as a number and printing it in roman numerals

18:11 justin_smith: this one is good http://www.gigamonkeys.com/book/a-few-format-recipes.html

18:11 TEttinger: I haven't used cl-format at all

18:11 justin_smith: ,(use 'clojure.pprint)

18:11 clojurebot: nil

18:12 justin_smith: ,(cl-format nil "~r" 123456789)

18:12 clojurebot: "one hundred twenty-three million, four hundred fifty-six thousand, seven hundred eighty-nine"

18:12 justin_smith: ,(cl-format nil "~:r" 123456789)

18:12 clojurebot: "one hundred twenty-three million, four hundred fifty-six thousand, seven hundred eighty-ninth"

18:13 justin_smith: ,(cl-format nil "~@r" 123456789)

18:13 clojurebot: "123,456,789"

18:13 justin_smith: ,(cl-format nil "~@r" 1234)

18:13 clojurebot: "MCCXXXIV"

18:13 ridcully: ah, that what [email protected] meant

18:15 justin_smith: ,(cl-format nil "~{~a, ~}" ["foo" "bar" "baz"]) ; user_ - maybe you wanted this feature

18:15 clojurebot: "foo, bar, baz, "

18:15 TEttinger: why did the last two print different results for "~@r" ?

18:15 justin_smith: TEttinger: some numbers are too big for roman numerals, mentioned in the docs

18:16 also mentioned in that gigamonkeys link I shared above

18:16 format allows things like apply and looping - via string formats

18:16 totally nutso

18:16 sdegutis: whoa

18:16 user_: ,(cl-format nil "~{~a, ~}" ["foo" "bar" [:map1 value1]])

18:16 clojurebot: #error {\n :cause "Unable to resolve symbol: value1 in this context"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Unable to resolve symbol: value1 in this context, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6688]}\n {:type java.lang.RuntimeException\n :message "Unable to resolve symbol: value1 in t...

18:17 justin_smith: (cl-format that is)

18:17 sdegutis: aw, cl-format

18:17 user_: ah oh

18:17 sdegutis: I thought it was format that knew how to do roman numerals

18:17 user_: how do i also print a vector?

18:17 i need the docs

18:17 sdegutis: although, i wouldnt put it past java.util.Formatter

18:17 justin_smith: user_: cl-format comes from common lisp, so the most helpful thing will be the common lisp docs I think

18:18 or books/pages that give good format examples for common lisp

18:18 TEttinger: ,(doc clojure.pprint/cl-format)

18:18 clojurebot: No entiendo

18:18 TEttinger: ,(use 'clojure.pprint)

18:18 clojurebot: nil

18:18 TEttinger: ,(doc cl-format)

18:18 clojurebot: "([writer format-in & args]); An implementation of a Common Lisp compatible format function. cl-format formats its arguments to an output stream or string based on the format control string given. It supports sophisticated formatting of structured data. Writer is an instance of java.io.Writer, true to output to *out* or nil to output to a string, format-in is the format control string and the rema...

18:19 justin_smith: TEttinger: at the end of the string it links to the official common lisp docs for format

18:19 TEttinger: aw

18:20 justin_smith: ,(subs (:doc (meta #'clojure.pprint/cl-format)) 885) ; subs to the rescue!

18:20 clojurebot: "Detailed documentation on format control strings is available in the \"Common Lisp the \nLanguage, 2nd edition\", Chapter 22 (available online at:\nhttp://www.cs.cmu.edu/afs/cs.cmu.edu/project/ai-repository/ai/html/cltl/clm/node200.html#SECTION002633000000000000000) \nand in the Common Lisp HyperSpec at \nhttp://www.lispworks.com/documentation/HyperSpec/Body/22_c.htmn"

18:20 user_: woop

18:21 will do

18:21 yes

18:21 bring on the pprint/cl-format

18:22 justin_smith: if I was feeling more awesome I would have used cl-format to do that substring operation, lol

18:22 user_: lol

18:22 language translation would be a sweet implementation

18:23 so for like those who prefer hexdecimal instead of ascii

18:23 brb

19:02 sdegutis: Good evening.

19:02 ane: hello

19:02 sdegutis: How are?

19:02 ane: hoppy

19:02 justin_smith: sdegutis: should I see the revenant or deadpool?

19:03 sdegutis: justin_smith: I'm confused why you're asking me of all people this.

19:03 ane: though you asked him, BOTH

19:03 (I've seen both)

19:03 justin_smith: sdegutis: I'm amazed you would object to a confusing or off topic question

19:03 sdegutis: justin_smith: Don't mistake me, I certainly object not.

19:03 justin_smith: also, apologies, I blame the vodka

19:04 ane: you inebriated rapscallion

19:04 sdegutis: justin_smith: My wife bought me Merlot today for the first time per my request. I have tried some from both a $35 bottle and a $10 bottle a few minutes ago.

19:04 justin_smith: So far I like the $35 bottle more, it is less, how you say.. "harsh"

19:05 justin_smith: Now please either respond to anything and/or everything I have just said, or answer my frist question.

19:05 justin_smith: ane: thanks

19:06 ane: they are two different beasts entirely

19:06 revevant was intense, deadpool was insane

19:06 sdegutis: justin_smith: why do you ask me?

19:06 justin_smith: indeed, I'm probably in more of a "deadpool" mood

19:07 sdegutis: Also, I've been considering experimenting with root namespaces with unusual names that are unlikely to have conflict, such as _ and !.

19:07 justin_smith: sdegutis: because I figured if I had to ask a silly quesiton, I might as well ask you

19:07 sdegutis: like what vinyasa does?

19:07 werid

19:07 sdegutis: (ns $.home) or (ns !.license)

19:07 ridcully: mhmmm hoppy

19:07 sdegutis: justin_smith: How sweet :)

19:07 libertytrader: what is the clojure replacement for printf

19:07 justin_smith: sdegutis: vinyasa puts things in the . namespace

19:07 ane: ,doc format

19:07 clojurebot: #error {\n :cause "Can't take value of a macro: #'clojure.repl/doc"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Can't take value of a macro: #'clojure.repl/doc, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6688]}\n {:type java.lang.RuntimeException\n :message "Can't take value of a macro: #'clojure....

19:07 ane: eufg

19:08 ridcully: ,(doc format)

19:08 clojurebot: "([fmt & args]); Formats a string using java.lang.String.format, see java.util.Formatter for format string syntax"

19:09 sdegutis: justin_smith: I have looked into both Deadpool and Revenant, and both looked incredibly boring, for different reasons. Deadpool because it was basically an unfunny 110 minute dick joke for 13 year olds, and Revenant because it was basically a 160 minute gory revenge-porn for 25 year olds.

19:09 backnforth: Oh god

19:09 justin_smith,

19:09 sdegutis: justin_smith: and I'm 30 so neither appealed to my sensibilities.

19:09 backnforth: cl-format is kicking my bum

19:10 justin_smith: oh?

19:10 backnforth: i cant find out how to print vectors with it

19:10 sdegutis: ,(do (ns !.foo) (def bar :quux))

19:10 clojurebot: #'!.foo/bar

19:10 justin_smith: what do you want the vector to look like?

19:10 sdegutis: ,!.foo/bar

19:10 clojurebot: :quux

19:10 sdegutis: Nice.

19:10 I may do exactly this, forever.

19:11 ,(do (ns $.foo) (def bar :quux) $.foo/bar)

19:11 clojurebot: :quux

19:11 justin_smith: ,(do (ns .) (def bar :quux))

19:11 clojurebot: #'./bar

19:11 justin_smith: ,(./quux)

19:11 clojurebot: #error {\n :cause "No such var: ./quux"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: No such var: ./quux, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6688]}\n {:type java.lang.RuntimeException\n :message "No such var: ./quux"\n :at [clojure.lang.Util runtimeException "Util.java" 221]}]\n :trace\n ...

19:11 sdegutis: ,(do (ns #.foo) (def bar :quux) #.foo/bar)

19:11 clojurebot: #<RuntimeException java.lang.RuntimeException: Unmatched delimiter: )>

19:11 justin_smith: ergh

19:11 sdegutis: ,(do (ns %.foo) (def bar :quux) %.foo/bar)

19:11 clojurebot: :quux

19:11 sdegutis: Oh my gosh.!

19:11 I'm going to use these namespaces. I'm going to use them in production.

19:11 justin_smith: ,(./bar)

19:11 clojurebot: #error {\n :cause "Wrong number of args passed to keyword: :quux"\n :via\n [{:type java.lang.IllegalArgumentException\n :message "Wrong number of args passed to keyword: :quux"\n :at [clojure.lang.Keyword throwArity "Keyword.java" 97]}]\n :trace\n [[clojure.lang.Keyword throwArity "Keyword.java" 97]\n [clojure.lang.Keyword invoke "Keyword.java" 110]\n [sandbox$eval264 invokeStatic "NO_SOURCE...

19:11 justin_smith: ahh!

19:11 backnforth: heh

19:11 TEttinger: sdegutis: the pound sign is a legal identifier start in java

19:11 justin_smith: ,./bar

19:11 clojurebot: :quux

19:12 sdegutis: justin_smith: anyway, that was my genuine actual opinion so take it or leave it

19:12 justin_smith: sdegutis: thx

19:12 TEttinger: I think it messes with the clojure reader too

19:12 backnforth: ah

19:12 sdegutis: I'm not opposed to entertainment, though. I just prefer things like Sherlock season 2.

19:12 backnforth: i might just go with the regular (print arg) for now

19:12 libertytrader: In a program i am writing i am getting data from a webserver. How do i save that data to my stak

19:12 stack

19:12 backnforth: or println

19:12 sdegutis: (Which if you haven't seen it is quite worth the watch.)

19:12 libertytrader: so other functions can call it

19:12 backnforth: easier to read anyways

19:13 TEttinger: ,(def ₫ :ding)

19:13 clojurebot: #'sandbox/₫

19:13 TEttinger: ,₫

19:13 clojurebot: :ding

19:14 justin_smith: ,(use 'clojure.pprint)

19:14 clojurebot: nil

19:15 justin_smith: ,(cl-format nil "~{~a ~}" [1 2 3])

19:15 clojurebot: "1 2 3 "

19:15 TEttinger: ,(do (ns ₧) (def ₰ :pfennig))

19:15 clojurebot: #'₧/₰

19:15 justin_smith: ,(cl-format nil "~a" [1 2 3])

19:15 clojurebot: "[1 2 3]"

19:15 sdegutis: Seriously though, are there any drawbacks to using root namespaces like _ and ! in an application's private codebase?

19:15 TEttinger: ,₧/₰

19:15 clojurebot: :pfennig

19:16 justin_smith: sdegutis: like I said, this is what vinyasa does with . as a namespace - the drawback is people think it's weird

19:16 sdegutis: justin_smith: sorry I'll be more specific

19:17 TEttinger: somebody make a dependency insertion lib where the root namespace is ₯

19:17 sdegutis: (1) will it fail to work if the filesystem doesn't support that character?

19:17 (2) will it fail to work in production when turned into an uberjar?

19:18 (3) does it have the potential to clash with other namespaces and is it likely?

19:18 That's all I can think of.

19:20 justin_smith: sdegutis: it will get munged

19:20 sdegutis: every munged thing should work on every fs

19:21 unless you mean to store it in a file with the same name for usage with require etc. .. that might be trickier

19:21 TEttinger: I don't think currency gets munged

19:21 let's try!

19:22 justin_smith: ,(munge "₯"")

19:22 sdegutis: nice!

19:22 clojurebot: #<RuntimeException java.lang.RuntimeException: EOF while reading string>

19:22 justin_smith: um...

19:22 ,(munge "₯")

19:22 clojurebot: "₯"

19:25 TEttinger: it just generated ₧.class

19:25 java is more restrictive than clojure and I made a ₧ class in Java just fine

19:26 this is fun

19:26 you can also do...

19:26 sdegutis: ,(namespace-munge 'foo.bar)

19:26 clojurebot: "foo.bar"

19:26 sdegutis: ,(namespace-munge '$)

19:26 clojurebot: "$"

19:26 sdegutis: ,(namespace-munge '?)

19:26 clojurebot: "?"

19:26 sdegutis: uhh wth

19:27 justin_smith: sdegutis: check out the source for namespace-munge

19:27 sdegutis: ok

19:27 justin_smith: it's super silly

19:27 sdegutis: ,(source namespace-munge)

19:27 clojurebot: Source not found\n

19:27 sdegutis: justin_smith: its not there

19:28 justin_smith: (.replace (str ns) \- \_)

19:28 TEttinger: ,(munge "︴")

19:28 clojurebot: "︴"

19:28 justin_smith: that's all it does

19:28 sdegutis: haha

19:29 justin_smith: ,(ns-munge "-_-_-_-_-")

19:29 clojurebot: #error {\n :cause "Unable to resolve symbol: ns-munge in this context"\n :via\n [{:type clojure.lang.Compiler$CompilerException\n :message "java.lang.RuntimeException: Unable to resolve symbol: ns-munge in this context, compiling:(NO_SOURCE_PATH:0:0)"\n :at [clojure.lang.Compiler analyze "Compiler.java" 6688]}\n {:type java.lang.RuntimeException\n :message "Unable to resolve symbol: ns-mung...

19:29 justin_smith: ,(namespace-munge "-_-_-_-_-")

19:29 clojurebot: "_________"

19:30 TEttinger: ,(munge "_‿⁀⁔︳︴﹍﹎﹏_")

19:30 clojurebot: "_‿⁀⁔︳︴﹍﹎﹏_"

19:30 sdegutis: BLARG.

19:31 I'm not satisfied with any of my code. It's all terrible, no matter how much I try to make it less bad.

19:31 TEttinger: ,(do (ns _‿⁀⁔︳︴﹍﹎﹏_) (def $$$ "MONEY"))

19:31 clojurebot: #'_‿⁀⁔︳︴﹍﹎﹏_/$$$

19:31 TEttinger: ,_‿⁀⁔︳︴﹍﹎﹏_/$$$

19:31 clojurebot: "MONEY"

19:41 tolstoy: If you're developing a data format to feed to something, you would prefer:

19:41 {:prop [:one-of ["a" "b" "c"]]}

19:41 or:

19:41 {:prop {:op :one-of :value ["a'" "b" "c"]}}

19:41 In other words, a map versus a positional tuple?

19:42 A map gives you more flexibility (different ops have different extra keys you can use).

19:43 But a tuple reduces flexibility, thus discouraging you from being too clever.

19:43 And it's easier to read (IMHO).

19:43 justin_smith: maps are much better for manipulating programmatically

19:43 tuples can be clearer if you expect the usage to be literals

19:44 tolstoy: If you want people to type stuff in by hand. On the other hand, said people are Clojure enthusiasts.

19:44 Hm.

19:44 I suppose "both" might work.

19:44 Translate the tuples into the maps.

19:46 justin_smith: right, but accept maps too

19:46 that shouldn't be too hard

19:47 tolstoy: None of it's too hard.

19:47 But, yes, these are meant to be typed in by hand.

19:47 Hm.

19:47 Both might work.

20:01 sdegutis: I wonder..

20:01 So, first of all, why is Hiccup so many files?

20:01 Like, it doesn't seem very hard to translate [:foo "bar"] into "<foo>bar</foo>"

20:03 tolstoy: Don't know, but he has a lot of extra stuff in there, like abstractions for pages, forms, etc.

20:03 sdegutis: True enough. I never used them because they're too limiting.

20:04 tolstoy: Me, either. In fact, I almost never use Hiccup. soblano instead (for react-y type stuff).

20:04 sdegutis: Like, I can't give a class to form/text-field as easily.

20:05 Now that I've written a nice CSS lib for myself, I'm thinking we should be able to define components that have both HTML and CSS aspects, and be able to access those aspects of them at compile-time for compiling.

20:05 I'd like to more easily define my CSS alongside my HTML, but Hiccup doesn't allow this.

20:06 tolstoy: Fork and hack! ;)

20:06 sdegutis: That didn't work with Garden.

20:07 I don't understand this compile-time trickery: https://github.com/weavejester/hiccup/blob/master/src/hiccup/core.clj#L7-L14

20:09 tolstoy: I see what you mean. I guess I don't understand macros as much as I thought.

20:09 sdegutis: Me neither.

20:09 I've seen this kind of thing a lot in weavejester's code though.

20:09 Lots of syntax-quoting outside macros.

20:09 justin_smith: do you know what it's doing?

20:10 tolstoy: Maybe you only need the backtick if you're going to interpolate values. But.... huh.

20:12 sdegutis: Hmm.

20:12 Yeah that sounds right tolstoy.

20:12 But.. like.. what?

20:12 I'm confused.

20:12 I'm gonna guess weavejester was a Common Lisp user before Clojure.

20:13 tolstoy: If the macro is executed at eval/compile time, then how is it there are "options" to destructure?

20:13 sdegutis: Oooh, html is a macro.

20:13 I had no idea.

20:13 tolstoy: He gives me the impression of having the luxury of working on a lot of very similar projects.

20:27 I always put ` at the top of my macros, which is why I get confused a bit.

20:31 Is there a seq kinda test for which maps fail, but lists, vectors and sets pass?

20:33 libertytrader: ok when I do (format "%s" x) I get nothing but when I do (prn x) I get a bunch of strings

20:33 this is starting to get very frusterating

20:33 tolstoy: libertytrader (println (format "%s" x))?

20:33 libertytrader: Or, (printf "%s\n" x)?

20:36 libertytrader: tolstoy: that printed out clojure.lang.LazySeq@1e43e368

20:37 tolstoy: ,(format "%s" "hello")

20:37 clojurebot: "hello"

20:37 tolstoy: ,(println (format "%s" "hello"))

20:37 clojurebot: hello\n

20:37 tolstoy: libertytrader: Oh, maybe your x is the lazyseq?

20:38 ,(format "%s" (take 1 (repeat "x")))

20:38 clojurebot: "clojure.lang.LazySeq@97"

20:38 libertytrader: that is probably the case...

20:38 how do i get rid of this

20:38 tolstoy: ,(format "%s" (dorun (take 1 (repeat "x"))))

20:38 clojurebot: "null"

20:39 tolstoy: I forgot the non-do version of that.

20:39 libertytrader: what the heck is a lazy seq and why cant i print it

20:40 hmm

20:41 (printf "%s" (pr-str x))

20:41 works

20:41 except is doesnt do a newline

20:41 major clojure designers were on crack

20:41 tolstoy: Yeah. A lazy seq is a list in which none of the values are produced until you actually referene them.

20:42 libertytrader: i have \n in my string

20:42 tolstoy: So, you can do something like (repeat "x") to produce an infinite list of "x".

20:43 libertytrader: thanks for your help

20:43 i am baffled why it is so difficult to get a newline in clojure

20:44 tolstoy: (printf "%s\n" (pr-str x))?

20:44 (println (format "%s" (pr-str x)))

20:46 libertytrader: I think my data might all be part of one giant string now

20:46 after using pr-str

20:47 tolstoy: Usually, you use doall to realize all the data in a lazy sequence.

20:47 libertytrader: it prints out like "("data" "data" "data")"

20:47 tolstoy: Try just "str".

20:51 libertytrader: with str i get the lazyseq

20:51 tolstoy: ,(format "%s" (seq (take 10 (repeat :word))))

20:51 clojurebot: "(:word :word :word :word :word ...)"

20:51 tolstoy: Seq might be it.

20:56 Or maybe (seq (doall (take 10 (repeat :w))))

20:57 libertytrader: i am trying to do this in a doseq

20:57 basically i have collection of strings and a few functions that will return my desired output based on those strings

20:57 but i cant seem to be able to print out a line

20:58 if you have a better idea than doseq for output based on an array of strings

20:58 im all ears

20:59 tolstoy: When you're going to print out something that's a list, maybe (vec the-list) and see if that helps.

21:00 libertytrader: It seems like clojure is printing out the seq or vector

21:00 not taking them as inputs for a newline

21:02 the "count" of my seq is 19

21:02 maybe a bug in clojure

21:03 tolstoy: I guess I don't understand. Why wouldn't it print out a vector if that's what you're asking?

21:04 libertytrader: I want to print out every element of the vector on a new line

21:04 tolstoy: Or do you want something like (doall (map #(println %) my-strings))?

21:04 libertytrader: The problem with that method is

21:04 I also need to print data from other functions

21:05 tolstoy: (println (string/join "\n" x))?

21:06 Maybe cl-format might work better?

21:06 libertytrader: I want to do something like this for(1 ....10)(printf %d %d %d", i, abcfunction(i), xyzfunction(i))

21:07 in c

21:07 thanks a lot for your help

21:07 i am clueless in clojure

21:07 how would i print out three vectors in paraelle

21:07 parallel

21:07 tolstoy: (dotimes [n 10] (println (format "%d %d %d" n, (abc n) (xyz n))))

21:08 libertytrader: ok except instead of an array of ints

21:08 I have an array of strings

21:09 tolstoy: Like: (doseq [s strings] (println (format "%s" s)))?

21:09 libertytrader: that is what i have been trying to do

21:09 unsuccessfully

21:10 tolstoy: Oh. Maybe your list of strings isn't really a list of strings?

21:10 libertytrader: it is more like (doseq [s [strings]]) though

21:13 tolstoy: (doseq [s1 strings] (doseq [s2 s1] (println s2)))

21:14 Or, just (clojure.pprint/pprint strings). ;)

21:15 libertytrader: ok

21:15 my strings are coming from a function call

21:15 i dont know if that make a difference

21:15 (apply printf (function ...)) wokrs

22:33 Ok I have a bad situation I have no idea how to resolve. In my code I have three S expressions

22:33 They evaluate to the output I want.

22:34 Ideally I want to output my code like "sexpr1.0, sepr2.0,expr3.0" ... etc... until I have gone through all of my lists

22:34 basicially like having three arrays and wanting to printf all three arrays next to each other

22:34 how do i do this in clojure

22:34 i am flabbergasterd

22:43 sdegutis: Hi!

22:45 amalloy: how are you?

22:47 tolstoy: how are you?

23:17 tolstoy: Eating figs.

23:22 Alas libertytrader. I bet it was something simple. refheap sis worth a 1000 misunderstandings.

23:51 backnforth: How I have a vector in a vector, how do I count the amount of the sub vectors there are on.

23:55 tolstoy: ,(count [[1 2] [3 4]]) ?

23:55 clojurebot: 2

23:55 sdegutis: tolstoy: amazing

23:55 tolstoy: Heh. Is that what you mean?

23:55 sdegutis: backnforth: concat count

23:56 try: concat $ count $ vec

23:56 tolstoy: (apply + concat count)? ;)

23:56 backnforth: tolstoy, amazing

23:56 i tried count

23:57 such as (,count (vector :node))

23:57 tolstoy: ,(apply + (mapcat identity [[1 2][3 4]]))

23:57 clojurebot: 10

23:58 tolstoy: Yeah, that's what I was going for, but I bet that's not the problem you described.

23:58 backnforth: tolstoy, it isn't lol, but thanks for trying

23:59 tolstoy, say I have 10 vectors with the namespace :node

23:59 tolstoy: Namespace?

23:59 backnforth: oh

23:59 wrong name?

23:59 should i use variable name?

Logging service provided by n01se.net