4 ms·

Overall, I think you're right about Chris's point, but I'm not sure about your comparison of how many gems versus how many manpages. Many gems are libraries ra

Overall, I think you're right about Chris's point, but I'm not sure about your comparison of how many gems versus how many manpages. Many gems are libraries rather than binaries/applications, and as such they don't have user-friendly man pages, but they do have documentation in the form of rdoc/yard/what-have-you. On my system, only gem-man and ronn have manpages, but many other gems have documentation via ri - it's just documentation aimed at other programmers rather than end users. For a library, that makes sense to me. Having said that, it would be nice if Ruby gem authors got into the habit of using ronn and gem-man in the way that Perl CPAN authors use POD. Also, and this is just nitpicky, but your second command shouldn't need awk, sort or uniq: gem man --all | tail -n +3 | wc -l

leif 17y ago
So yes, some of these gems don't need man-pages. I wasn't hoping for 100% man-page coverage (maybe an interesting tool to write, now that I think of it...), but better than one apart from the gem-man quasi-meta-manpage would be a start. In particular, I'm thinking of things like rails, rspec, heroku, and the like, which, while primarily libraries, have cli components that badly need man pages. I can't tell you how many babies I've punched as a result of typing "man rails" or "man spec". It might be just me, but if it can be the first token in a shell command, I expect it to have a manpage, and it is almost (physically) shocking when that is not true. (I do need that awk | sort | uniq, because I have multiple versions of unicorn that all show up with that command)