RE: [Groff] Why is it...

[Ted Harding]

On 14-Dec-07 17:25:40, Michael Kerpan wrote:
> ...that groff/troff seems to be written off by so many as
> "obsolete" and "only useful for man pages", despite the
> fact that it can do everything that TeX/LaTeX (seemingly
> the favored non-WYSIWYG document processor) can do but
> while taking up 3 megabytes (as opposed to the 300 or so
> used by the average TeX install) It can't be ease of use,
> as *roff plus -me, -mm or -ms is no harder to use than
> LaTeX or HTML. It can't be availability as the *roff family
> is basically a required component of any Unix-like system.
> It can't even be font support, given that it's MUCH easier
> to install and use random PostScript fonts in groff than
> in TeX... What gives and how can we fix it?

A good question! I recall (many years ago) someone posting
to a Linux mailing list, and in passing saying that groff
was obsolete "and now only used for formatting man pages."

To which the opening words of my response were:

  I respectfully disagree (in other words, bollocks)."

I pointed out the many O'Reilly books typeset with groff,
and the fact that (at least at the time and for some years
later) that the Collins Language Dictionaries were typeset
using troff (and later groff), after translating from XML.

Also, all the reasons you give above.

I think the principal reason why TeX became the norm for
mathematical and much technical writing was that fact that
when Knuth released TeX (around 1989-90) he provided it
with a very rich set of fonts specially designed (by himself)
for good mathematical typesetting; and also used his great
typographical knowledge to design a powerful interface to
layout issues. As a result, TeX (and LaTeX and variants)
became widely and enthusiastically adopted by the academic
community: they could do really weird-looking stuff "straight
out of the box".

This take-up was also stimulated by the fact that Knuth's
3-volume "Art of Computer Programming" was typeset in TeX
(at any rate in the second edition), which again was a
great advertisement for TeX.

Nowadays, it would be difficult to submit an  article to
a mathematical journal electronically if it were not in
TeX; the editors and production team can handle it for
modifications, amendments, and layout issues. This is
not the case with groff these days! If you were to submit
troff source, they would not look at it. If you were to
submit a PDF, they would most probably treat it on the
same basis as a hard-copy submission, and re-writie it
in TeX.

In the early days of TeX, good table layout and good
diagrams were easier to achieve (and usually better)
using groff than TeX; and indeed I sometimes helped
people who were stuck on such things in TeX by doing
them in groff for them! I think TeX has caight up on
these fronts.

Of course, as you point out, it is easy to install extra
PS fonts in groff, and the TeX fonts have been available
as freely downloadable PS fonts for some time now.

So, yes, groff could do everything that TeX can do (with
some extra persuasion for certain things -- TeX's view of
layout is more "global" than groff's). But, unfortunately,
the train left the station a long time ago!

As to what to do about it -- well, I never use TeX, but
I always make a point, when I send anyone a PDF file
(which I will always compose using groff), of ensuring
that the layout is as attractive, clear and readable as
possible. A lot of this has to do with the very fine
control that you can easily get with groff, and I think
the normal TeX user would have a hard time getting anything
as nice. So at least it is a kind of advertisement
for groff!

And, while I have seen some truly dreadful formatting
from TeX in published journals, this is probably due to
failings on the part of publishers' typesetters not
appreciating how to use TeX properly.

Ah well ...

Best wishes to all,
Ted.

--------------------------------------------------------------------
E-Mail: (Ted Harding) <address@hidden>
Fax-to-email: +44 (0)870 094 0861
Date: 14-Dec-07                                       Time: 18:51:31
------------------------------ XFMail ------------------------------