afaik, that's essentially a limitation of sh - inheritting the IO streams should resolve that, but I don't think there's a trivial way to do so with clojure.java.shell/sh - with that said, babashka.process/shell does it by default

and for gum specifically, <https://github.com/lispyclouds/bblgum> wraps the calls if you don't want to deal with the shell layer at all

exactly what I needed ty

You can also change babashka.process/*defaults* and change the :pre-start-fn to print the command, but alter-var-root also works :)

What problem are you solving?

trying to implement function die() with System/exit, but be able to test it in the nREPL session

I think i found an alright workaround for REPL development,

i'll just eval (defn abort [code] (throw (RuntimeException. "abort called"))) when i start a REPL session

@U0J3J79FE You can do (throw (ex-info "..." {:babashka/exit 1}))

the exit code will be picked up by babashka

and in the REPL it's just an exception

my pre-start-fn seems to only be called for p/shell, not p/exec, (either that or it's just not flushing the STDOUT in time for exec)

(alter-var-root #'p/*defaults* assoc :pre-start-fn (fn [{:keys [cmd]}]
                                                     (when *verbose?* (apply println "++" cmd) (flush))))

@U0J3J79FE which version of bb do you have installed?

looks good

i added (flush) after the println, still no good

i looked a little bit at your code for exec, it seems to call it

i can send a repro example if you want

yeah, I can repro it locally now, not sure why this is happening

i don't think its being called

(alter-var-root #'p/*defaults* assoc :pre-start-fn (fn [{:keys [cmd]}]
                                                     (when *verbose?* (apply println "++" cmd)
                                                           (flush)
                                                           (Thread/sleep 10000))))

doesn't hang for exec, but does for shell

I'll debug

I'll add this and then compile:

_ (prn pre-start-fn)

have to go for an appt in 5 mins, but i'll be back in ~2 hours, thanks for debugging this!

and thanks for your project in general

I get what is wrong

I'll fix it

pushed a solution. once master finishes, you should be able to install the update with:

bash <(curl ) --dev-build --dir /tmp

ty ! i’ll try it out when i get back