Fork me on GitHub
#boot
<
2016-06-25
>
richiardiandrea01:06:26

just a note, I still receive a Metaspace error

maxp07:06:27

how to set different source-paths for the task?

maxp07:06:53

I have garden files in separate tree.

martinklepsch10:06:43

@maxp: You can add multiple paths to source paths. Is there a reason you don't want to do that?

maxp10:06:56

ok, I moved styles to main tree

maxp10:06:26

but could you help me with another problem? -

maxp10:06:06

how yo copy garden-compiled css file to real resources/public directory?

maxp10:06:21

from tmp/fileset

martinklepsch10:06:47

@maxp: you usually want to copy compiled files there but instead serve them from the classpath/fileset. For deployment you might want to look at the target task

juhoteperi13:06:52

@martinklepsch: ^ boot-reload 0.4.10 available

tcrawley14:06:39

@micha: are you calling .close on the shim? That will shutdown agents for you: https://github.com/projectodd/shimdandy/blob/master/api/src/main/java/org/projectodd/shimdandy/ClojureRuntimeShim.java#L35 /cc: @richiardiandrea

tcrawley14:06:26

that appeared in shimdandy 1.2.0

sekao14:06:31

anyone know what to put in set-env! to make the compiler target java 7?

juhoteperi14:06:51

@sekao: I think this option should be passed to javac task: (javac :options ["-target 1.7"])

juhoteperi14:06:15

Or (javac :options ["-target" "1.7"]), not sure how javac handles these options

sekao14:06:43

looks like it’s the latter. thanks

micha15:06:27

@tcrawley: hmm yeah we are calling .close on the shim, so now i am confused

richiardiandrea15:06:23

Yeah confused too here..

danvingo22:06:22

hello, I have a very basic boot question - i have the following boot task: `(deftask dev "Build romr-mvp for local development." [] (comp (serve :port 8000 :handler 'romr-mvp.handler/app :reload true) (watch) ; (speak) (hoplon) (reload) (cljs-repl) (cljs))) ` after starting this task the console output includes the following:

danvingo22:06:26

Started Jetty on http://localhost:8000 Starting file watcher (CTRL-C to quit)...

danvingo22:06:45

I don't see any output from println calls in my clj code though

danvingo22:06:17

so my question is, is stdout redirected somewhere?

danvingo22:06:42

or is there another boot task I can use to get stdout inlined with the watch output?

micha22:06:04

@danvingo: jetty craetes the servlets on demand

micha22:06:25

so you will need to make some http requests for your ring handler code to run

danvingo22:06:45

ah ok, i'll give it a shot, thanks!

micha22:06:11

you'll see the delay when you make the first request as it fires up its threads and whatnot

danvingo22:06:45

got it, i'm seeing the printlns now, thanks

micha22:06:15

sweet! :+1:

richiardiandrea23:06:11

so I was checking again the memory error and while we call .close on a pod's shim, maybe the second shutdown-agents is necessary for the the "common/boot" clojure to shutdown? another experiment/test would be to count the make-pod against the destroy-pod and see if they match for each watch iteration

micha23:06:26

yeah i don't think you want to do shutdown-agents in the main core pod

micha23:06:38

but that's what you did, right?

micha23:06:44

and it took care of the memory issue?

richiardiandrea23:06:24

yes, there is a shutdown-agents executed in the pod but it is actually not targeting the pod

richiardiandrea23:06:35

because .close on the shim does that

richiardiandrea23:06:08

so that is the solution, but I haven't identified the cause