Closed Dhul-Husni closed 6 years ago
I've got a Makefile for a FreeBSD Port of Um, but pandoc takes about a day to compile on my computer and I need to compile it six times (twice for each supported release (i386 and x64), this is the only thing holding me off from submitting the port to get it into the FreeBSD repository.
Switching to something like this brings down the compilation time for any FreeBSD users wishing to compile all of their software (which many do).
This is awesome! Hope to test this out myself soon. My only major concern is that md2man
will expect a different flavor of Markdown. Do you know what the major differences might be?
md2man
is a just a wrapper for pandoc, though. This work will be rendered moot when gem that does the translation itself is integrated, and that think that's what ultimately will happen if we're going to fulfill the intent behind dropping the pandoc dependency - namely, more portability as implied by #6 and #4 .
Upon further review: it seems better to use kramdown as mentioned in #10 instead of running with this PR.
@iajrz I think they were referring to the rubygem md2man found here, which is not a wrapper around pandoc. I do agree, however, that kramdown MIGHT be better, as it does not have any additional runtime dependencies.
Um now uses Kramdown to parse Markdown files into roff man pages. Pandoc is no longer a dependency.
Because Kramdown's Markdown is not the same as Pandoc's Markdown, this may break your um pages. See UPGRADE.md for more info.
convert markdown to roff using a gem instead of an external library. This increases the flexibility of um; to run on devices that may not have pandoc installed or supported. requested in #10 ps: I have also made a pull request on the homebrew tap repo to add md2man gem