[ntp:questions] Re: Latest changes toTWikiGettingStarted

Peter Boettcher boettcher at ll.mit.edu
Tue Oct 7 13:54:30 UTC 2003


fredb at immanent.net (Frederick Bruckman) writes:

> In article <blsv8s$snn$1 at dewey.udel.edu>,
> 	stenn at maccarony.ntp.org (Harlan Stenn) writes:
>> 
>> At one time the html2man script worked; I haven't tried it in a while.
>> 
>> The original author said that it needed some tweaks, but I have
>> no record of these things anymore as those notes were in the recent year
>> of email that got lost.
>
> It's terribly out of date. Even after you make the obvious
> corrections for the filename changes, and correct for the extra
> white space -- extra newlines are needed in the output to make
> the "." escapes appear in column one -- the result doesn't look
> at all good in "nroff".

Yes.  The HTML changed considerably (for the better) with the
conversion to GoLive (or whatever it's called).  I think I might just
start again, as the old version guessed at many things and tried to
handle questionable HTML.

The old version of HTML docs converted (IMO) very nicely, with
newlines where appropriate.  All bets are off for the new stuff.

> There seems to me to be a fundamental problem with the concept.
> Man pages, and the SYNOPSIS section in particular, require a very
> rigid structure, which is not evident from the html. How do you know
> if a dash is supposed to be an ".Fl" macro for a command line option,
> or something else? True "nroff" pages contain a lot of information
> that is simply lost on conversion to text, html, or postcript. You
> can't put the toothpaste back into the tube.

Yeah, but you can come close.  There are a number of MAN pages out
there which are not written in nice beautiful nroff, but are legible
and helpful nonetheless.  Again, I agree that HTML is not a good
source format for conversions to other formats.  You'd like something
that logically marks up the text.  But HTML is what we've got, and I
don't want to try to find an alternative that's author-friendly.


-- 
Peter Boettcher
MIT Lincoln Laboratory
boettcher at ll.mit.edu
(781) 981-5275



More information about the questions mailing list