13:37 <stenur> farkuhar: ok, maybe; but the manuals are placed in false paths anyway, which happened during autotools->cmake, and on 3.7 the thing does not build as such

13:43 <stenur> i have no problem with just wrapping manuals the way i do for contrib/ipcalc. it is ugly, and i should check whether the converter is installed and generate the real ones if possible, then. (though i think this particular program did not exist on CRUX at all. i think it needs ruby too, but i have forgotten.)

14:03 <stenur> romster is fit enough for commits!

14:25 <jaeger> What's a "false path"? Does that mean they're in an incorrect location?

14:25 <stenur> yes

14:25 <stenur> it is man/1/ instead of man/man1/ (iirc)

14:26 <stenur> is not found by man anyhow

14:27 <jaeger> ok

15:18 <stenur> can anyone use mpg123 with "-o alsa" (default here, too)? it only works with "-t oss"

15:32 <farkuhar> stenur: https://github.com/xiph/flac/commit/80d064e15d843eb137ff80c9e9797b9c180e1743 is when upstream stopped providing man-pages in groff format; now the distro maintainers are expected to generate them from Markdown. I shared yesterday a Pkgfile that seems to work, but only if you have pandoc-bin installed, otherwise no man-pages are created.

15:33 <stenur> well i will not install pandoc; but why not simply enwrap like i do for contrib/ipcalc when there is no pandoc installed? like this we have the manual

15:33 <stenur> pandoc is that binary blob thing because it is in that strange language... haskell iirc?

15:33 <stenur> functional programming uuuuuuuuuuuuuuh

15:36 <farkuhar> heh, pandoc-bin doesn't require you to install a huge haskell ecosystem, so i don't mind having it installed. But if romster is fit enough for commits, you can just open a ticket and suggest that the groff man-pages be provided directly in the ports tree.

15:37 <stenur> it was beerman who did that in March and September (the 11th iirc)

15:38 <stenur> sox can mp3 and mp2 (thus as i learned)

15:41 <farkuhar> yes, beerman has been bumping romster ports while romster was unavailable. But if romster is (as you say) fit enough for commits, why not relieve beerman of all the extra work he's taken on?

15:41 <stenur> because he messed it again

15:42 <stenur> i mean he is doing a very lot

15:42 <stenur> but the commits in March and September are very strange ones

15:46 <beerman> ¯\_(ツ)_/¯

15:59 <farkuhar> stenur: i just looked at your "enwrapped Markdown" in the ipcalc man-page. it's not a horribly ugly solution, but i would hesitate to use it if the Markdown contains tables or other complicated syntax.

16:08 <farkuhar> in this case, the # SYNOPSIS section of man/flac.md is probably too unreadable as raw Markdown. Anyone wanting a quick reminder of how to invoke flac is better served by running 'flac -h' than by reading an "enwrapped Markdown" man-page like the one you made for contrib/ipcalc.

16:09 <stenur> i wholeheartly disagree and do not even understand, given -h of many programs

16:48 <farkuhar> stenur: for contrib/ipcalc the upstream meson.build already does "check whether the converter is installed and generate the real [man-page] if possible". Your Pkgfile only needs to test whether the conversion happened during the build: http://ix.io/4aWD

16:49 <stenur> farkuhar: i will keep this snippet for the next update, but CRUX does not have ronn anywhere?

16:49 <stenur> farkuhar: thanks

16:52 <farkuhar> stenur: correct, CRUX does not have ronn in its repos. A user would have to run "gem install --user-install ronn" and then modify $PATH in order for meson to find it.