Re: [Groff] inconsistency between .R and \*R in man.tmac

[Tadziu Hoffmann]

> > DEC OSF/1 uses a similar definition:
> > 
> >   .       \" R - Roman font
> >   .de R
> >   .if n .ul 0
> >   .ft R
> >   ..
> 
> And there are man pages that use it on that system?

I haven't found any manpages that use it.  However, it *is*
documented in man(5):

  .R  Sets the text in a roman typeface until another font
      change is encountered. Also ends nroff underline mode
      if it was in effect.

Perhaps the reason they don't actually use it is that the Digital
manpages employ something called "semantic markup language", in
which all text-level markup is done with strings instead of
macros.  (You can imagine that this does not really improve the
legibility of the source files.  Ugh!)  The following logical
styles are available: L - literal, V - variable, A - alphabetic
constant, N - numeric constant, O - ordinary text, C - computer
output, U - user input, E - emphasize.  (The physical styles
were apparently removed in recent versions: R - roman, B - bold,
I - italic, C - computer.)

So instead of

  The file
  .BI /usr/lib/lib name .a
  contains something-or-other.

they have

  The file \*L/usr/lib/lib\*Vname\*L.a\*O contains
  something-or-other.

I find that much harder to read as source.
(Maybe their manpages are generated automatically from
some other format, and nobody writes *roff anymore.)