@bones You might want to take a look at https://github.com/mikera/core.matrix either to use it or using its source as an example of how to define +, etc operators and protocols to make them work with your custom stuff

Awesome, thanks @jesse.wertheim!

check also https://github.com/ztellman/primitive-math/blob/master/src/primitive_math.clj for method how to actually replace clojure.core/+ to custom one. Additionally there is a machinery to turn replacement on and off (https://github.com/ztellman/primitive-math/blob/master/src/primitive_math.clj#L170 and https://github.com/ztellman/primitive-math/blob/master/src/primitive_math.clj#L182)

I have generated a jar file for my clojure app, and now I want multiple entry points. So basically i want to run different functions with different commandlines. Is this possible?

@hoertlehner does this help? The clojure code

(ns app.core
  (:gen-class))

(defn -main
  [& args]
  (case (first args)
    "a1" (println "hello")
    "a2" (println "hi")))

java -jar app.jar a1
>>> hello
java -jar app.jar a2
>>> hi

@rahul080327 oh yes this helps! Many thanks!

@hoertlehner for more complicated scenarios https://github.com/clojure/tools.cli

@gon thanks! I will study that!

Hello guys, could anyone explain me differences and implications between :keyword and ::keyword?

::keyword will expand to :<namespace>/keyword for the current namespace

to expand a bit, ::foo is just a reader macro to expand into :<current_namespace>/keyword, a keyword in the form of :foo is called a non-namespaced (or simple) keyword, a keyword in the form of :foo/bar is a namespaced keyword there's been a shift in the recent years (mostly thanks to spec) to using namespaced keywords when possible, in order to convey better semantic meaning (e.g. what does :name mean? but :person/name vs :country/name make it clear), while still using unnamespaced keywords mostly for syntax (e.g. in macros like (for [x xs :when x] ..))

so let's say I want to code a library. current best practice guides me to use namespaces keywords, so it is clear that this thing belongs to me and has this exact meaning when used with my library? Which I can perceive as prefixed keywords, right? Is there any effect in term of memory management for example or is it just in the boundaries of semantics?

How do you start Clojure 1.10 (alpha) REPL, please? Do you invoke it as: java -jar ~/Downloads/clojure-1.10.0-alpha5.jar or in what way? I have Sun JDK 10.0 and OpenJDK 11 on a RPM-based Linux (SUSE). I tried both JDKs with clojure-1.10.0-alpha8.jar and two more jars from https://repo1.maven.org/maven2/org/clojure/clojure (and one from http://maven.org). All fail with: Exception in thread "main" java.lang.ExceptionInInitializerError at java.base/java.lang.Class.forName0(Native Method) .... Caused by: java.io.FileNotFoundException: Could not locate clojure/spec/alpha__init.class or clojure/spec/alpha.clj on classpath. at clojure.lang.RT.load(RT.java:463) .... I also tried to build it from source, failed again.

@peter.kehl no, use the clj tool

@ales.najmann yeah that's correct, no performance hit really

the namespace bit is just a string that's set instead of being nil

hey guys. i've just tried to build a new clojurescript project with reagent using lein new figwheel my-app -- --reagent. however on starting with lein figwheel i get a bunch of errors in the console complaining about being unable to find react, e.g.

Uncaught Error: goog.require could not find: create_react_class
    at Object.goog.require [as require_figwheel_backup_] (base.js:681)
    at component.cljs?rel=1538661740772:1

@bhauman may know

I vaguely remember something about create-react-class being deprecated in newer versions of React

@bronsa @danieleneal thank you for your time explaining this to me ❤️

@bronsa Where do you get clj tool for 1.10 aplha?

@peter.kehl https://clojure.org/guides/getting_started

then you can specify 1.10 in .clojure/deps.edn

that is, clj is agnostic to the Clojure version you want to use, so there is no such thing as “clj tool for 1.10 alpha”, there is just “clj”, which can be used with 1.10 alphas: clj -Sdeps '{:deps {org.clojure/clojure {:mvn/version "1.10.0-alpha8"}}}'

@helenelizabethwilliam did you alter the generated project?

(but you can specify any other version of Clojure there too)

@helenelizabethwilliam I just launched it and didn’t get these errors

@bhauman haven't altered the generated project. literally just run those 2 commands

it was working about 2 hours ago and then it suddenly stopped working for me

so you have been working on it

adding code etc

no no

this was in another clojurescript project, creating using the command i was running lein figwheel. lein figwheel stopped working so i tried creating a brand new project, which also didn't work

oh that explains it probably

sounds like caching

interesting

what gets cached?

Try a shift reload

or a hard refresh

well they are both at the same url

localhost:3449

so browsers can be fussy

when you are fetching the same artifact from the same server

:woman-facepalming: thanks - that totally worked

you can avoid things like this if you have a different port for each project

i'll remember that for next time

or just being aware that its an issue 🙂

Thanks @bronsa and @alexmiller.

@peter.kehl How were you running REPLs for Clojure 1.9 etc? You can use the exact same tools -- just specify a different version of Clojure as a dependency. That works for lein, boot, and clj.

I was working on a solution for: http://www.4clojure.com/problem/19#prob-title i know people used

(first (reverse coll))

(fn my-rest [a-seq]
  	(cond
     	(empty? (rest a-seq)) (first a-seq)
     	:else (recur (rest a-seq))))

I think that logically it works! You could use if instead, and avoid calling rest twice with destructuring.

(fn my-rest [[fseq & rseq]]
  (if (empty? rseq)
    fseq
    (recur rseq)))

and because of how destructuring handles empty rest-destructure, (if-not rseq ...) would also work

(or (if rseq ...) and flip the two clauses)

(fn [[f & r]] (if r (recur r) f))

right

oh, sweet, thanks!

the nil-pun of empty collection that seq does can really make things more concise

destructuring seems really nice, saves me from having to use local bindings (let)

i tend to solve some problems using structural recursion on sequences, because seems a lot easier for me, and then i try to simplify the code

I want to study spec lib, any good resources aside clojure website ?

Any free tool like SoapUI to do some load testing on an API?

Are you looking for something like Insomnia? https://insomnia.rest/

Thanks! Ill look into this

hey guys, one quick question. I’m doing my api and something bother me. My handler is calling something like a service “layer” of functions should I validate the params in the handler or into my service ? I mean, should verify the spec on the handler or inside my service layer ( my handlers are super thin , just orchestrating the flow )

one approach I have seen recommended is to put the validation at the system boundary, and I'd argue that the system boundary here is where the client communicates with server, so validate the request rather than the function calls

thanks @noisesmith

Hi, I do a query in datomic with pull syntax and I get this: [#:user{:id #uuid "ef23d1a1-ac61-4a8a-9448-61d14c1ec13d", :first-name "Sozuke", :last-name "aizen", :email "", :password-hash "xxx"}] A simple question: What does #:user signify here?

it's a namespaced map syntax

user=> {:user/foo 1 :user/bar 2}
#:user{:foo 1, :bar 2}

oh, so actual keys are :user/id, :user/first-name, etc?

right

Ok thanks

you can also make map literals that way

user=> (= {::foo 1 ::bar 2} #:user{:foo 1 :bar 2} {:user/foo 1 :user/bar 2})
true