Fork me on GitHub
#boot
<
2016-05-26
>
dm307:05:41

what's the proper way to replace clojars/central repositories with mirrors?

dm307:05:15

:repositories is a vector, so updating in merge-env becomes a bit awkward

pesterhazy08:05:16

If I want to fork a library and put up my branch on a public mvn repository, what's the best way/service to do that?

dm309:05:34

@pesterhazy: fork on Github, change the version/group-id, deploy to Clojars (depending on how the project is set up)?

pesterhazy09:05:44

group id in adzerk/boot-cljs would be adzerk?

pesterhazy09:05:57

so if I forked it, it would be pesterhazy/boot-cljs?

pesterhazy09:05:12

(hypothetically, I'm not actually doing that)

dm314:05:46

@micha - could you provide an expert opinion on https://clojurians.slack.com/archives/boot/p1464249161003339 please? 🙂

martinklepsch14:05:01

@dm3: you can use set env with functions as well

martinklepsch14:05:23

@dm3: (set-env! :x (fn [old-val] (conj old-val new)))

dm314:05:43

yeah, just thought strange

dm314:05:00

it's not a map

mynomoto14:05:32

@dm3: order is important

mynomoto14:05:20

You can have artifacts with the same group/artifact on more than repo

tcrawley15:05:21

@micha: are you already aware of boot issues with jdk9?

tcrawley15:05:19

$ java -version
java version "9-ea"
Java(TM) SE Runtime Environment (build 9-ea+119)
Java HotSpot(TM) 64-Bit Server VM (build 9-ea+119, mixed mode)
$ boot
Exception in thread "main" java.lang.ClassCastException: jdk.internal.loader.ClassLoaders$AppClassLoader (in module: java.base) cannot be cast to java.net.URLClassLoader (in module: java.base)
	at Boot.loadJar(Boot.java:228)
	at Boot.main(Boot.java:254)

micha15:05:47

@tcrawley: i haven't seen that yet

micha15:05:16

i compile with java 1.7

tcrawley15:05:42

I'm about to dig in to the classloader tree for the boot classloader anyway, to get dynapath working with jdk9

tcrawley15:05:49

because it breaks lein

micha15:05:52

ah boot uses that

micha15:05:02

but not there

micha15:05:13

that's happening in the shim script

tcrawley15:05:27

and since I can't get lein repl to work under jdk9, I thought I'd use boot :)

tcrawley15:05:34

but I'll let you know what I find

tcrawley15:05:52

it should help with the !dynapath bootloader mucking you are doing as well

tcrawley15:05:05

@micha: it looks like the system classloader is no longer a URLClassLoader. I think the fix for boot would be to have https://github.com/boot-clj/boot-bin/blob/master/src/Boot.java#L227 create a new URLClassLoader with the system CL as the parent, maybe

micha16:05:21

@tcrawley: i bet they made that change because of people who do that kind of thing lol

micha16:05:00

it's probably not good what we do there, somehow

myguidingstar17:05:47

but the file only appears once in my target dir

myguidingstar17:05:23

never again when I re-run the above in repl

myguidingstar17:05:02

it's quite strange because it worked months ago

micha17:05:48

the target dir is in an undefined state while boot is running

myguidingstar17:05:52

I even tried BOOT_EMIT_TARGET=yes but doesn't help

micha17:05:57

especially when you have things going in there other than boot

micha17:05:45

when you say appear only once

micha17:05:49

what do you mean?

myguidingstar17:05:38

I mean the first time I run the above (boot ...), there's a file styles.css in my target dir

myguidingstar17:05:59

from the second time on, there's no such file

micha17:05:21

you can put show -f in between garden and target tasks

myguidingstar17:05:22

the same problem if I put the command in a boot task

micha17:05:28

and see what's up

micha17:05:48

it seems like garden task isn't putting the file in the fileset

myguidingstar17:05:23

in clj syntax, is it (show :fileset)?

micha17:05:40

options are kwargs

micha17:05:47

so (show :fileset true)

myguidingstar17:05:34

hmm, the file is only there the first execution

myguidingstar17:05:34

also, the log Compiling styles.css... (in red) only appears on that first run

myguidingstar17:05:37

also, I tried to set boot ver back to 2.5.5 and clj to 1.7.0 but the problem still remained

micha17:05:07

seems like it could be an issue in the garden task?

micha17:05:20

caching something too aggressively or something like that