@borkdude maybe I am late but I think I have often heard "dependency vector"

@lvh https://github.com/jakepearson/quickie

Hi, I have a REST api where one endpoint takes amongst others a time field as input. Two questions: 1. what is considered idiomatic regarding the apis input. I always assumed passing around UNIX timestamps is a good thing to do. Opinions on that? 2. How do you handle these timestamps in clojure, especially if you want to store them in a relational database later. I see myself juggling JodaTime (java 8), java.sql.Time for postgres and clojure.instant right now, having clj-time in the back of my head as a fourth option.

nice puzzle question: - What’s a good way to inline a function call in a clojure file. For example suppose there was something like (+ 1 2 3) among the source, and I wanted to mutate the source file by inlining this to 6.

@borkdude why not let the java compiler do the work?

@sveri I give you this file:

(ns foo)
;; hello world
(println (+ 1 2 3))

@borkdude without some more context I would say I let it do the java compiler later and not care about it at all.

following the golden rule: the best code written is the code that is never written.

@sveri the Java compiler can output this file for me?

nope, but it will inline the code later on compilation part

at least not that I am aware of

@sveri some more context then. suppose I want to give users of this library: https://github.com/borkdude/boot-bundle the possibility to inline the call to expand-keywords, so they get a ‘normal’ build.boot file without a dependency on my library

I could use string search/replace, paren matching, eval, but maybe there is a much better way. tools.analyzer?

@borkdude In that context most probably a simple search/replace will do I think

it seems tools.analyzer also drops comments

@borkdude have you seen https://github.com/xsc/rewrite-clj ? i’ve seen other people suggest it in situations like yours

haven’t used it myself, apologies if it’s irrelevant

i'm trying to learn how to reference local maven builds. So far, I'm following the default build for maven as seen here : https://maven.apache.org/guides/getting-started/maven-in-five-minutes.html

I'm running into trouble in figuring out how to reference it

This is from the pom file, and I have run mvn install

it's correctly in my m2 directory. And I can add it to my project file [com.mycompany.app.my-app "1.0-SNAPSHOT"]

However, I'm not sure how to reference it in my core.clj file

(:require [my-app.app/App :as a]) keeps coming back as missing the init class

Anyone have any ideas?

and other permutations on app.my-app/App, etc

Depends what the namespace of your Clojure code is (this is totally independent from the name of artifact)

That's strange to me that the name of the artifact would vary depending upon the namespace of my code

i'm working on updating some cider documentation when using locally built and installed maven dependencies.

The entirety of the code is as follows

@jrheard not seen it before, taking a look now

Hey, does anyone know any ways of bounding an InputStream so I can reference say the 4th MB of data from a large binary file as an InputStream to clj-http. I’ve used BoundedInputStream in java but wondered whether Clojure offered a better solution? Something like take would only defer realization wouldn’t it?

If efficiency is the primary consideration, I’d probably look at java.nio to mmap the large binary file: https://docs.oracle.com/javase/7/docs/api/java/nio/MappedByteBuffer.html

How does crossclj work? I’m looking at projects that use clojure.tools.reader and reagent is one of them. When I look into the project.clj of reagent, tools.reader is not there. Does crossclj list it because it depends on it indirectly?

is there any examples of using sente with plain javascript?

found something: https://devcenter.heroku.com/articles/using-websockets-on-heroku-with-clojure-and-immutant