Repository navigation
Support setting supervisor=start-stop-daemon in service files - #1089
Conversation
|
is there any reason to set supervisor to start-stop-daemon? it's the default (for better or for worse), and well, it's really not a supervisor removing the warning could be okay, but i don't think it should be documented as a supervisor |
|
ive described rationale here: #1064 (comment)
im not sure if its supervisor either 🤔
|
|
in service managers, supervisor is a software that actively supervises a daemon, with things like auto-restart being expected start-stop-daemon is more like, an environment setup toolbox and not much more a service that needs s-s-d is odd though fair, but fwiw end goal is to phase out direct s-s-d usage in favor of supervise-daemon, using things like subreapers on linux to manage double-forking daemons |
and that would be great! but when will it become a reality? If its likely to happen by the end of next year for example, there might be a reason not to fix the non-critical behavior related to S-S-D and instead move toward marking it as deprecated. but for now,
really not sure whether this is a standard definition, rather than just a cognitive distortion shared by many people or maybe a collective image |
|
not to mention that allowing s-s-d to be set would be the first step toward deprecation! |
it really depends on if my life continuing to stabilize, but if so, having all the necessary tools for it end of next year is realistic i think i'm okay merging the code, but i think docs shouldn't bunch it with the other supervisors, instead, should mention it's an option but also say supervise-daemon is preferred |
|
done. word "leagcy" could be added, though: |
|
service-writing guide also needs to be updated to at least mention |
|
service-writing guide needs a full overhaul, and i'm already working on one it'll become a manpage that'll be installed, plus rendered to html for hosting web docs |
...ensuring that it doesnt cause a warning.
closes #1064