Who will replace roff?

I have a vision, … of many courageous and talented experts, coming together, to do this project:

GNU/Linux man pages need to be more easily maintained. They also badly need hyperlinks.


man pages are written in a version of roff using lots of macros, and best edited with emacs, both everyone agrees are ancient.

And software and data languages have come a long ways from those early beginnings. Yet, we still use these antique tools for our modern GNU/Linux documentation. (I imagine that if we could still be documenting in scrolls, there would be those who would staunchly advocate that system.)

What would it take to convert all the man pages to odf (Open Document Format), so they could be editable with LibreOffice Writer, and thus much more richly formatted if desired, but still with a dumbed down text mode output?


It was 34 years ago, that I first used Ventura Publisher, the first graphical WYSIWYG or What-You-See-Is-What-You-Get editor. It blew my mind and opened up a whole new way of writing documentation. I easily wrote a 100+ page, fully typeset manual in it in just a few weeks. That was remarkable back then.

It will take a team of people to get this done, and my 2nd question is, who they will be? I am too old to be any important part of this project. But I can see what needs to be done. I challenge some of you to consider this.

That is not about LibreOffice.

@LeroyG, It looks like you feel that this is not about LibreOffice. Were you thinking that when I said “LibreOffice” in paragraph 5, that I was referring to something else?

Styling, input and output filters, version control, now with LO macro support…

I agree with @LeroyG in that this is not about LO.

You are asking to convert to odf and the just because you COULD edit with LO Writer doesn’t make this a question about something in LibreOffice.

Sorry. You feel this is not a question about something in LibreOffice.

I could have mentioned another editor, but did I mention another editor? Also is there a restriction here that questions must be about LibreOffice, and not what to use it for?

Again it was your question stating:

What would it take to convert all the man pages to odf

Not LO. The rest doesn’t matter.

Personally don’t care about the issue or who if anyone will do it.

Excuse me, but have you tried in askubuntu.com, or the like?

If you want to add more information, you can edit your question, or comment an answer. Do not use Answer to comment.

Mark the circle to the left of the answer that solves your question.

No worries LeroyG. I’m just looking around for a few good people who might be interested. To my knowledge there isn’t an odf forum, but I might be wrong about that. I’m just spending a bunch of time to try and bring up some better documentation tools, and again can see just how urgent it is to get roff replaced, and how really very old the structure of it is. Also I had written a chapter for the Debian manual, and it would be much better IMHO to just use LO rather than troff. It would make it so much easier for new people to come up to speed and help with the docs. Thanks.

@EasyTrieve: I understand the wish (hope?) for a “better” formatting tool for help doc writers. But have you thought of the integration effort of a new solution? It must work both in TUI and in GUI. The present state-of-the-feature is taken as an immutable frozen primitive and many parts of Linux/*NIX eco-systems depend on it. The task is not only about writing/formatting help bits but also about adapting the tools. I think this latter part is much bigger than the former.

So the short answer to your question is: yes, LO can be used for that. And that’s probably all AskLO can answer.

Next, you can start a new project and call for volunteers. Be prepared for an overwhelming amount of work beyond help text itself. Many systems tools must be reviewed and updated for proper integration. If integration is botched, the project will be rejected despite all its qualities.

Yes, many tools must be reviewed and updated, and it will take years. It sounds like you’re just as afraid of this effort as I am. Would things be better if we did nothing to fix this?

No, you can’t just see that the feature could be largely improved and give up. A deep thinking is needed and a tentative specification must be issued. But to whom submit it for peer review?