While I agree with Deborah that troff is dead-end stuff in that not
everyone is going down that road, nonetheless there continue to be
requirements for delivering information related to this format for various
Linux distros, so the transform itself definitely gets used.

The real question might be "Which flavor of man page?" Years ago, Lars
Kaufman attempted to nail things down with the euphoniously-named
"Committee for the Common Man" which attempted to define a standard of
sorts (see http://xml.coverpages.org/sgml92.html for some interesting
history on that effort). Some of this design effort was reflected in
DocBook's refentry markup design (see
http://www.docbook.org/tdg/en/html/refentry.html). But depending on which
system you are outputting for, the right deliverable might still be either
pre-formatted flat text, man macro marked-up pages, or something like
DocBooks refentry format or DITA's JavaAPIRef format, both of which
expressly mirror a semantic structure rather than a delivery structure. You
could even have on-the-fly output to HTML, skipping the man/troff
formatting step altogether.

So getting back to the DITA OT's troff output, whether it meets your needs
depends on what you expect from it for your product, in your distro, for
your users. Its architecture permits quite a bit of tweaking, as long as
you understand that your tweaks might not necessarily be useful for
everybody's concept of how to deliver a man page. It wasn't really meant
to define low end troff output ;-), but rather to be enough to get the
community engaged in taking it further. Insofar as I understand how it
works, improving it is just a matter of adding to its own pipeline
(basically like adorning generic divs and spans into more sophisticated man
page structures). For example, this code does not handle turning DITA
tables into troff output. You could extend it to generate the tbl macros
and then let the tbl preprocessor generate your result tables, if you wish.
If you were to contribute this back to the Toolkit, we'd be thrilled!

Regards,
--
Don Day
Chair, OASIS DITA Technical Committee
IBM Lead DITA Architect
Email: dond@...
11501 Burnet Rd. MS9033E015, Austin TX 78758
Phone: +1 512-838-8550
T/L: 678-8550

"Where is the wisdom we have lost in knowledge?
Where is the knowledge we have lost in information?"
--T.S. Eliot



"Shawn McKenzie"
<smckenzie23@gmai
l.com> To
Sent by: dita-users@yahoogroups.com
dita-users@yahoog cc
roups.com
Subject
Re: [dita-users] Re: DITA RTF
02/26/2007 01:57 Output - help
PM


Please respond to
dita-users@yahoog
roups.com













Deborah Pickett writes:

>The RTF output in DITA-OT is, like the troff output, something of an
>evolutionary dead-end. Generating RTF output is a Difficult problem.
>I get the feeling that development on the dita2rtf transform has
>halted. It appears that those who care are now looking to a
>DITA-to-ODF or other transformation, though I've never seen actual code.


I haven't even looked at the troff output in the toolkit at all yet, but I've
had
it in the back of my mind to consider a 'man page' specialization, where I use
the
troff output to install man pages on *nix systems.

What's wrong with the troff output?