Do not hard-fail on process owner check

[zopieux]

ps might be missing in $PATH, or fan2go might be running as a non-root user with write access to the subsystem. I don't think it's reasonable to crash the program for a simple uid check that might actually be wrong. This will eventually crash later when trying to modify paths as non-root anyway.

[markusressel]

Hey, thx for opening the PR! ❤️

There are definitely scenarios in which you do not need to use root permissions for running fan2go, so I can absolutely see why we need to improve the current logic 👍

ps might be missing in $PATH, or fan2go might be running as a non-root user with write access to the subsystem. I don't think it's reasonable to crash the program for a simple uid check that might actually be wrong.

I agree, this code is a remnant from the early days and also my specific usage. However, I think we need to be a bit more specific here, since there is an additional permission related check done while evaluating the configuration file meant to protect users against malicious configuration editors. Did you consider this?

This will eventually crash later when trying to modify paths as non-root anyway.

Did you validate this? Because I am not 100% sure what the current behavior would be 🤔
When a fan cannot be controlled during normal operation, the error should not lead to a crash, but it might do that when the error is already present during startup (of the fan2go service).

[zopieux]

Did you validate this?

I did not, "crash" might be the wrong word here. I meant that I/O while trying to set the PWM will fail in a way or another and hopefully bubble up in the logs, so it'll be pretty obvious something's wrong.

Did you consider this?

I'm not sure what malicious editor protection you're referring to here, sorry, I only skimmed at the surface of fan2go – thanks btw, what a pleasure it was jettisoning that fancontrol garbage!

I don't really mind how this check is done, I just believe depending on a subprocess invocation & a (very common but not necessarily available) Linux util is overkill for a simple user warning message. Maybe Go has a way to report its own uid, like this?

Feel free to close this PR and come up with something better :)

[markusressel]

I don't really mind how this check is done, I just believe depending on a subprocess invocation & a (very common but not necessarily available) Linux util is overkill for a simple user warning message. Maybe Go has a way to report its own uid, like this?

Good find! Seems like a much more reasonable approach 👍

Feel free to close this PR and come up with something better :)

I think we can use it for now, I will think about other special cases in a later PR.

[zopieux]

Nice! Making it non-fatal makes a lot of sense actually.