Re: [Groff] man problem under non-latin1 locale

[Tadziu Hoffmann]

> >  1. There should be only *one* hyphen character in output.
> 
> no there shouldn't. Hyphen != Dash in anyone's book.

I'm quite aware that hyphen .ne. dash, but I said that there
should only be one kind of *hyphen* in output, not one kind
of dash.  Fact is, when I enter my source text in ASCII, the
groff latin-1 device will keep the ASCII hyphen/minus/dash
character (2D) in a word such as "small-scale fluctuations"
(where it is clearly intended as a hyphen) intact in output.
On the other hand, when auto-hyphenating, it will insert
character AD, which comes out as a different glyph.  This is
what I'm saying is bad.  Both should be output as the *same*
character, either both 2D or both AD.  The ps device outputs
both characters as the postscript "hyphen" character.  This
is good.

Concerning minus and dashes, that's why we explicitly enter
these characters as \-, \(en, and \(em  (so groff can
distinguish them from -, whose meaning is "hyphen").  These
should preferrably be mapped to 2D, 2D, and a double 2D.


> Monospaced typography is oxymoronic to me.

Well, to quote from an old USENET article by Dirk van Deun
on MS-DOS batch programming:

  Art lies in self-limitation. Every kind of artistic
  expression is an activity within strict limits.  [...]
  The narrower the limits are, the greater the opportunity
  for the artist to prove his skills.  [...]  To do without
  is the essence of art.



Cheers,
Tadziu