Conversation

Notices

1. Embed this notice
   feld (feld@friedcheese.us)'s status on Friday, 21-Mar-2025 05:48:02 JST feld
   @q66 I wish someone would backdoor the Linux kernel just to add SIGINFO and ctrl+T
   • Embed this notice
     feld (feld@friedcheese.us)'s status on Friday, 21-Mar-2025 06:00:58 JST feld

     in reply to

     • LisPi
     @lispi314 @q66 Linus has always hated it
     
     https://yarchive.net/comp/linux/everything_is_file.html
   • Embed this notice
     LisPi (lispi314@udongein.xyz)'s status on Friday, 21-Mar-2025 06:01:00 JST LisPi

     in reply to
     @feld @q66 Is there any reason it can't simply be contributed upstream?
   • Embed this notice
     feld (feld@friedcheese.us)'s status on Friday, 21-Mar-2025 06:51:07 JST feld

     in reply to

     • LisPi
     • Tim Chase
     @gumnos @q66 @lispi314 makes sense.
     
     Sometimes GNU/Linux tools will use SIGUSR1 or SIGUSR2 for equivalent functionality. I just wish they'd wire up CTRL+T to send one of those signal and then people can actually start building software that makes use of it. Having to open another terminal just to do a kill -SIGUSR1 <pid> is insane
   • Embed this notice
     Tim Chase (gumnos@mastodon.bsd.cafe)'s status on Friday, 21-Mar-2025 06:51:09 JST Tim Chase

     in reply to

     • LisPi

     @feld

     My understanding was that there are a limited number of signals—one for each bit in a 32-bit integer, and I recall reading that Linux chose to use that particular bit for some other purpose (SIGPWR).

     @q66 @lispi314