This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
- # announcements (2)
- # babashka (31)
- # beginners (608)
- # cider (60)
- # clj-kondo (22)
- # cljsrn (28)
- # clojure (14)
- # clojure-europe (5)
- # clojure-nl (3)
- # clojure-spec (1)
- # clojure-uk (19)
- # clojurescript (38)
- # conjure (20)
- # cursive (9)
- # data-science (26)
- # datascript (4)
- # datomic (19)
- # duct (4)
- # emacs (8)
- # figwheel-main (5)
- # fulcro (7)
- # helix (15)
- # leiningen (12)
- # malli (2)
- # off-topic (20)
- # overtone (3)
- # pathom (14)
- # pedestal (10)
- # re-frame (2)
- # reitit (13)
- # ring (13)
- # shadow-cljs (18)
- # spacemacs (8)
Q: I’m considering re-writing my deploy scripts (currently jvm clj) using bb. my naive design is that it’s just logic wrapped around shell calls to the aws cli. I wonder if there’s a better way than shelling out or some other trick I haven’t thought of. Has anyone done anything like this already?
One alternative would be nodejs for the scripting and using it’s aws client. but bb is so cool, I want to explore the comparison of the two designs
Apparently the aws-api lib doesn't work with graalvm, so I don't have other ideas besides shelling out to aws cli.
@ also had some kind of stripped down AWS clojure lib that worked with babashka or GraalVM I believe
@ For a long term tool I have done a lot of manually work to get part of the aws-api working (https://github.com/AdGoji/aws-api/blob/master/dev/examples.clj), but for more adhoc stuff I had more success wrapping the
aws cli tool https://gist.github.com/jeroenvandijk/e4fca617cf3eea82d58bc87e37e35465#file-spire-clj-L31-L50
@ thanks. that’s really helpful. even just confirming the using shell calls is a valid solution is good
Eshell is a good example of a lisp based shell, where if don't put any parenthesis is just interpreted as something to shell out
It would be a dream to have all the same auto completion I have in fish /bash plus the ability to switch up clojure whenever I want
There is a shell macro here:
Not the same as a REPL, but a custom REPL can be built on top of
clojure.main/repl which is available as a function inside babashka. This allows you to start a REPL inside a script, but can also be used to create custom things.
Alternatively there is https://github.com/dundalek/closh which tries to be a Clojure shell replacement. There is work in progress to leverage sci to be able to run on GraalVM
Babashka.pods also work on Windows! https://twitter.com/borkdude/status/1264483401190473728
Bootleg 0.1.8 released.
the babashka pod release
- add babashka pod support
- YAML upgrade to SnakeYAML 2.0
- Fix Build failed when trying build with graalvm-ce-java11-19.3.0
- Fix the ns/require code from spire in bootleg?
- Fix Running from file in same directory requires leading
- Fix Generating XML preserving the tag name case
- Fix reflection error preventing XML roundtrip
Thanks! Hopefully people put it to some good use. Get their babashka scripts processing heaps of html!
As a large library with lots of namespaces there is bound to be parts that dont work right in the pod form. Remember to open tickets for anything that has issues! Mention in the ticket its as a pod!
bootleg are in scoop now. Scoop users can enjoy the fruits right after update.
I put there a troubleshooting section, because of the vcredist2015 dependency that needs to be installed in addition. But otherwise should be good.
Is someone in here running macOS Catalina? There are some articles on the internet about slowdown of executables due to a notarization check: https://lapcatsoftware.com/articles/catalina-executables.html
immediately after startup it was really slow, but then dropped to instant:
i changed the script and it's name, and got about the same result every time:
.../src on master [✘!?] at dev (dev-view-us) ➜ time /tmp/test.sh && time /tmp/test.sh Hello /tmp/test.sh 0.00s user 0.00s system 0% cpu 10.844 total Hello /tmp/test.sh 0.00s user 0.00s system 0% cpu 10.301 total .../src on master [✘!?] at dev (dev-view-us) took 21s ➜ time /tmp/test.sh && time /tmp/test.sh Hello /tmp/test.sh 0.00s user 0.00s system 83% cpu 0.005 total Hello /tmp/test.sh 0.00s user 0.00s system 87% cpu 0.005 total
.../src on master [✘!?] at dev (dev-view-us) ➜ time /tmp/test-other-one.sh && time /tmp/test-other-one.sh foobar /tmp/test-other-one.sh 0.00s user 0.00s system 0% cpu 1.086 total foobar /tmp/test-other-one.sh 0.00s user 0.00s system 0% cpu 1.012 total .../src on master [✘!?] at dev (dev-view-us) took 3s ➜ time /tmp/test-other-one.sh && time /tmp/test-other-one.sh foobar /tmp/test-other-one.sh 0.00s user 0.00s system 0% cpu 1.015 total foobar /tmp/test-other-one.sh 0.00s user 0.00s system 0% cpu 1.015 total
Yeah, agreed. I’m not sure why it never dipped back down. Definitely feels like there’s something there. Maybe later I’ll install Little Snitch and block some outgoing traffic