Re: [Groff] eqn's `...' operator

[Eric S. Raymond]

Tadziu Hoffmann <address@hidden>:
> 
> > you've replaced `cdot' with \(md in eqn -- why not replacing
> > `...' with three \(md characters too?
> 
> Don't!  I want to have the "normal" ellipsis on the baseline.
> (Like in 1, 2, ...;  for special cases it's preferable to have
> a separate "cdots" operator with three \(md characters.)

Actually, Tadziu, I tend to agree with you.  (And I'm not just having
an opinion at random; I used to be a mathematician myself, and have
studied the history of mathematical notation.)

However, the 1976 guide to EQN shows that Kernighan & Cherry intended
... to be set with three medial dots.  So I've checked in this change
in order to make GNU eqn behave like the troff-classic version.  In this
instance I consider compatibility more important than my opinion.

If consensus is for reverting this change and documenting the difference,
I'm willing to do that.  
-- 
                <a href="http://www.catb.org/~esr/";>Eric S. Raymond</a>