Stefan Filipek <srfili...@gmail.com> wrote: > According to this discussion, there is an issue with nice(1) > documentation, which the patch was trying to emulate: > > > EXIT STATUS > > ... > > Otherwise, the exit status of nice shall be that of utility. > > Which is incorrect, since 'nice' doesn't exit in this case at all, > just like 'doas'.
Indeed, the noun is "exit status" But not precisely like doas, as discussed before. doas has cases where it exits and provides an exit status, itself. > Perhaps this line should be omitted, or perhaps removing 'of nice' (or > some other rewording) would make it more accurate while still being > helpful to any user who may not know the full implementation details. Yes, I would agree with removing "of nice". > In any case, doas(1) and nice(1) are documented inconsistently for the > same behavior. Quite a high bar you are setting there. It is possible these wordings could be improved (in particular through vagueness), but it doesn't follow all things must use all the same words. In particular, doas is special since it has two cases: it returns error indication itself for some operations, in other cases it has done an exec and the error comes from the replacement program.
