Re: [Groff] space width

[hohe72]

Peter Schaffter <address@hidden> wrote (Tue, 4 Feb 2014 11:32:00
-0500):
> On Tue, Feb 04, 2014, Dave Kemper wrote:
> > I understand the need for backwards compatibility, but I more and
> > more find myself wishing groff had a global option to choose
> > between "follow historical usage" and "be sane."  For someone in
> > 2014 writing a new groff document, there is zero advantage to, for
> > example, having calculations treat addition and multiplication as
> > the same precedence, or any of a dozen other little pitfalls that
> > exist solely for compatibility with a 1970s back end.
> 
> 
> I have to say I completely agree.  Backward compatibility is
> essential, but more and more, I wonder about future compatibility.
> As far as I know, I'm the only person actively developing
> a macro set for groff.  I can tolerate--just barely--the
> contortions I have to go through sometimes in order to work with
> a backwardly-compatible groff, but I have the advantage of my age
> and an appreciation for the long history of (n/t/g)roff.  What of
> future macro programmers, though?  How many who might contribute to
> groff are going to shake their heads over what to them will seem
> absurd anachronisms and simply move onto programming for something
> unemcumbered by what are rapidly becoming absurd historical
> idiosyncracies?  Is the absolute purity of backward compatibility
> worth relegating a powerful and useful program to the museum?

To be able to compile old groff sources, a backward compatible groff
needs to maintained -- and will. Branching a fourth groff will
also spread development resources. Isn't it? Having plenty?

I also don't agree to the term "idiosyncracies". AFAIK, a concept like
the vaporized AT&T's will never be accomplished by a community. Even
the Documentation (Troff User's Manual) is 34 pages where the TeX's
(TEX for the Impatient) is 300. This being quality.

However, the pitfalls .. should have been assembled first to make such
a decision as to branch. Addition and multiplication having the same
precedence really is a pita.

  Holger