This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2021-04-08
Channels
- # announcements (13)
- # asami (9)
- # aws (1)
- # aws-lambda (1)
- # babashka (37)
- # beginners (70)
- # calva (19)
- # cider (23)
- # clara (10)
- # clj-kondo (15)
- # cljdoc (14)
- # clojure (3)
- # clojure-bay-area (1)
- # clojure-europe (40)
- # clojure-nl (3)
- # clojure-serbia (1)
- # clojure-uk (31)
- # clojuredesign-podcast (1)
- # clojurescript (9)
- # community-development (7)
- # data-oriented-programming (1)
- # datomic (44)
- # emacs (13)
- # figwheel-main (14)
- # fulcro (6)
- # jobs (6)
- # malli (15)
- # meander (7)
- # off-topic (74)
- # other-languages (1)
- # pathom (3)
- # portal (3)
- # re-frame (25)
- # reagent (6)
- # reitit (2)
- # reveal (1)
- # rewrite-clj (6)
- # ring (5)
- # shadow-cljs (11)
- # specter (7)
- # xtdb (7)
I dreamt about babashka last night :face_with_rolling_eyes:
I dream about work all the time, but in weird ways that I work on a problem but it’s so separated from reality that it doesn’t work
in this case I had a dream that I was porting our makefiles to bb tasks
but something about one of the tasks that did something in real life?
If you are using github CI the `setup-babashka` action now supports using CI builds from babashka in addition to normal releases: Example from the tests: https://github.com/turtlequeue/setup-babashka/blob/fff96c154d5aec91d5bbfbb201c5b20f96653084/.github/workflows/check_url.yml#L15-L17

Please note that if you are going to use links from CircleCI directly, these expire after a while, unless github caches these somehow. A workaround for this may be to make your own github project and upload the binaries under a release or something
Question, if I uberjar .bb
scripts and I wanted to embed pod binaries in it as well would adding them to resources
work and pods/load-pod
be able to load from the resources?
@richiardiandrea afaik you can't load binaries from a .jar. most projects I know handle this by writing the binary to a temp directory first and executing it from there
I see, I liked the idea of self contained jars
but I guess that is a bigg-ish feature 😄
yep I am running in a container, I will just expose some sort of PODS_PATH
env var
I’m seeing an issue with Babashka and protocol method implementations when I use the same name arguments in a multi-arity method implementations. Here’s an example:
Note that when some-protocol-method
is implemented with the same named args (`, in this case as I’m ignoring the args) returns
nil`, instead of the constant value passed into the constructor. When they’re named differently (`, and
v`), it works as expected.
Yeah… how did that work? Until I saw NL
I though the envelope must have been dropped off by hand
PostNL (the Dutch mail service) sells stamps that you can activate by hand-written codes they provide online
hah! that is cool. I bet it confused even the american mail carriers
The main open questions for "bb run":
1) Should :depends
be dynamic? e.g. :depends (if (System/getenv "FOO") '[x y] '[x])
2) (How) should (command line) args be passed down to dependent tasks?
E.g.:
{x (println *command-line-args*)
foo {:depends [x]
bb run foo 1 2 3
3) The way it looks right now bb.edn is going to be "code" heavy which is odd a for .edn file. Maybe this belongs in a .clj file somewhere else?
4) Because the above reason, maybe I shouldn't allow any dependencies between tasks and just make this a "shortcut" system and not a half-featured build system?
https://github.com/babashka/babashka/discussions/779...or wait until how we see how the new keyword argument map shakes out and changes programming habits