Conversation

Notices

1. Embed this notice
   Andrew Tropin (abcdw@fosstodon.org)'s status on Wednesday, 14-Sep-2022 20:38:30 JST Andrew Tropin

   in reply to

   • Tom :emacs: :guix:

   @ryuslash It shouldn't be this way, please report it to bug-guix with as much information as possible.

   • PublicLewdness likes this.
   • Embed this notice
     Tom :emacs: :guix: (ryuslash@fosstodon.org)'s status on Wednesday, 14-Sep-2022 20:38:34 JST Tom :emacs: :guix:

     in reply to

     @abcdw I'm guessing that this isn't intended because this sounds like it could be the cause of why my laptop configured with guix home (on arch) doesn't start properly ~50% of the time. I guess one of on-first-login and the other startup script starts and kills shepherd and then the other tries and fails. Now I have no shepherd while still having a pid file for it and the second one won't start shepherd anymore.

   • Embed this notice
     Andrew Tropin (abcdw@fosstodon.org)'s status on Wednesday, 14-Sep-2022 20:38:36 JST Andrew Tropin

     When I stop user's shepherd with kill or herd stop root it doesn't cleanup /run/user/1000/shepherd/socket and to run a new instance I have to remove it manually. Is it intended behavior or it's time to dive into shepherd sources? ?