On 2025-02-11 at 08:50 -0800, Junio C Hamano wrote: > Ivan Shapovalov <intelfx@xxxxxxxxxxxx> writes: > > > > > --update-refs:: > > > > --no-update-refs:: > > > > +--update-refs=interactive:: > > > > > > Based on `git grep -e '--.*\[=' Documentation/git-*.txt`, I think this > > > should be more like > > > > > > --update-refs[=interactive]:: > > > --no-update-refs:: > > > > > > But maybe that unintentionally suggests that `=interactive` is the default? > > > > Perhaps --update-refs[=(yes|no|interactive)] then? Or is that too > > verbose? > > If `--update-refs` does take values that the git_parse_maybe_bool() > helper parses as a Boolean value, I do not think the above is > verbose at all. Rather, it is a disservice to the users if the > documentation does not mention yes/no in such a case. I'd say > listing other Boolean synonyms like yes/true/on/no/false/off is > too verbose, though ;-). > > > Anyway, I don't have a preference, I'll just do what I'm told > > That is not quite in line with how we'd like to operate. > > It is your itch. Others may give suggestions to help you polish it, > but ultimately, we would not want to accept a patch that the author > does not agree with. Of course, I care about the patch and the feature; what I wanted to say is that I do not care (comparatively) about the formatting of the help text: I couldn't figure it out on my own, so whatever you tell me is the proper way of formatting it, I'll do. -- Ivan Shapovalov / intelfx /