09:09 <merbzt> soxrok2212: yes, try and see if it works

13:13 <zetagon> Hello. I have struggled a little bit with writing a service file for a fairly simple shell script. I have since solved that problem but I figured that it might be nice to document that on the wiki.

13:13 <zetagon> The problem I had was that the script had a loop like this:

13:13 <zetagon> tail -f /tmp/log/messages | while read -r match; do ... done

13:14 <zetagon> procd can't stop this service because that loop lives on after the parent process is terminated (so procd thinks it has stopped the service, but the loop is still running)

13:14 <zetagon> I feel like that is a common gotcha that would be good to document somewhere. Is the OpenWRT wiki a good place to do so?

13:14 <zetagon> I'm thinking this page specifically: https://openwrt.org/docs/guide-developer/procd-init-scripts

13:44 <efahl> zetagon: I just read that wiki page and think it looks like the perfect place to document your warning/findings.

13:46 <robimarko> Please document it, I got bit by procd only tracking the main process recently

13:46 <robimarko> As I had a subprocess in a shell script and it would not get killed on restart or stop

13:51 <schmars[m]> typical gotcha with shells and service supervisors

13:51 <robimarko> Yeah, and easy to forget

13:51 <schmars[m]> yep same :-)

13:55 <zetagon> haha I feel vindicated that I'm not the only one :)

14:45 <zetagon> Follow up question: I solved it by modifying the script I was packaging, but what's a preferred solution if you can't do that?

16:17 <hitech95> who has some experience with netifd call flow for device_type? there are many callbacks/function pointers and I'm a bit lost trying to troubleshoot an issue. Also some tips to setup a easy way to develop/debug it would be welcome :D

21:03 <schmars[m]> blocktrron merge the poemgr version bump? :-) https://github.com/openwrt/packages/pull/25791