On 26 July 2017 at 15:26, Eli Zaretskii wrote: > > From: Reuben Thomas > > Date: Tue, 25 Jul 2017 23:41:03 +0100 > > > > The attached patch against git master adds support for the man-db > > manpath command to woman.el. > > > > For now, this can actually be better than using man, as man -k/apropos > > does not currently support man directories lacking a database or whatis > > file, which is typical for directories derived from PATH. > > Could you please provide a rationale for this addition? > > It sounds strange to me to enhance woman.el, whose main purpose is to > work without ("wo") 'man', including on systems where 'man' is not > easily available, by making it run 'manpath'. If something is wrong > with woman-manpath's built-in database, why cannot we extend it by > adding potential candidates to those already existing there? > ​You're right about woman being designed to run without man, I hadn't thought of that. There's nothing wrong with woman-manpath's built-in database. The idea is simply to use "manpath"'s extra functionality, which looks for man pages in directories formed from the directories on PATH. I guess the logical thing to do would therefore be to replicate this functionality directly, but I don't propose to do it.​ > On top of that, I think we semi-deprecated woman.el because AFAIR it > cannot support newer roff features which are abundant in recent man > pages. So I wonder what kind of use case do you have that on the one > hand needs woman.el, and OTOH finds its manpath emulation lacking. > ​ ​I explained that: man -k does not support man directories lacking a database. I've fixed that for my own purposes by running "mandb" as a per-user cronjob, so I suggest you can ignore this patch (which I agree is ill-conceived), and close the bug report. Thanks as always for your penetrating analysis. -- https://rrt.sc3d.org