On Thu, 2008-11-27 at 00:41 -0600, Michael Cronenworth wrote: > Before this "manpage-to-HTML" idea goes any further, I'd like to stop > it. > > This is not what I have envisioned. An HTML rendering engine is too > beefy. I'm going on previous Fedora requirements of lighter > documentation engines, particularly the release notes in anaconda. Anything but HTML is insane in this day and age. If a full HTML engine is too heavy, use a simple subset. There's plenty of options, Gecko, WebKit, GtkHTML, Dillo, Lynx, Links... You don't have to write in HTML, you can derive it from just about anything else. Man pages, GNU info, docbook, Doxygen, Javadoc, OpenOffice Writer... No need to force all upstreams to one format. Really the problem here is front end UI. There's yelp, but I wish I could just browse documentation inside Firefox... Hell, carrying around documentation is one of the major disk eaters in a Fedora install. $ du -sh /usr/share/doc/ 377M /usr/share/doc/ What would be neat is if we could transparently grab documentation from online instead of having it eat up disk space even though you probably only ever need a fraction of it... Hmmmm... What say we hack up a script to pull all documentation out of every RPM in a release, convert all man and info pages to HTML, and put it all up on docs.fedoraproject.org? Then embed a Google search on top of it. Ta da, a one stop shop for all documentation.
Attachment:
signature.asc
Description: This is a digitally signed message part
-- fedora-devel-list mailing list fedora-devel-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/fedora-devel-list
