On Sun, Nov 13, 2022 at 01:21:41PM -0800, Thiago Macieira wrote:
> I'm assuming that the kernel would provide different error conditions for
> those. But handling those in shell scripting is very difficult: you'd need to
> start a subshell and parse stderr.
You call that "very difficult"?!
> And this is part of my argument of why it's unlikely people will use their
> shells to do this. That shell function is easily another 10 lines of
> scripting, if it's meant to do its job properly. To make that easier, we've
> developed two tools, one of them the OpenDCDiag tool I linked to, but both
> just happen to be written in C and C++ instead of shell.
I've told you already that having a special tool is always more
problematic. I guess you'll make your own experience distributing it but
if you see pushback and people start complaining, you shouldn't wonder
why.
--
Regards/Gruss,
Boris.
https://people.kernel.org/tglx/notes-about-netiquette
