Re: Groff vs Heirloom troff (was Re: Quick question: how to do .index in groff?)

[Peter Schaffter]

On Fri, Jul 31, 2020, Steve Izma wrote:
> For almost everything I typeset, especially books and
> newsletter-type publications, I always at least a few places
> where I need to use track kerning on a paragraph in order to get
> good word spacing and to shorten or lengthen paragraphs in order
> to avoid widows (the last line of a paragraph starting a column
> of text). When I adjust the kerning (or mortising, if necessary)
> in values of one-hundredth or one-thousandth of a point, it can
> make a difference in whether a word fits on a line or is broken
> or pushed to the next line, thereby making the paragraph too
> long. I avoid trying to adjust only part of a paragraph because
> that can drastically affect the "colour" (i.e., density) of the
> text.

Ditto.  This was "how the pros did it" 40 years ago when I was
apprenticing in a type shop, and it's still the way they do it.  No
programme, application, or algorithm can produce typographic results
as good as those that have been fine-tuned by a human.  The finer
the tuning--i.e. line-by-line--the better the result.

> That said, I've never been convinced that paragraph-at-a-time
> justification makes a difference to the work I need to do for
> getting good word fits and even colour to the page.

Based on my own experience, I agree.  Sure, with the KP algo you
almost never get widows, but that doesn't mean the typographic
strategies used to achieve this golden state are always
satisfactory.

> ...I would like to caution people who think that the
> implementation of that algorithm [Knuth-Plass] in groff is going
> to lessen the effort that goes into high-quality typography.

Again, I agree.

Several years ago, I fielded the idea that, instead of chasing after
the Grail of paragraph-at-once, groff's line-formatting algorithm be
improved instead.  I worked on systems that used the formatting
strategy I proposed

  https://lists.gnu.org/archive/html/groff/2014-03/msg00322.html

and can confirm that it significantly reduced the amount of
intervention required to achieve good grey on a line-by-line basis.
There wasn't much interest in the proposal back then--I felt a bit
like a voice crying in the wilderness--but maybe it's time to try
crying again?

-- 
Peter Schaffter
http://www.schaffter.ca