Digraph proposal from Denmark

[Doug Gwyn (VLD/VMB)]

The problem with upper-case is that it would involve extra keystrokes.
(Shift key.)  Also it does not look as much like part of the language,
which is what these macros are pretending to be (in effect).  (Most of
my applications include a standard configuration file that also
defines a "bool" data type, in lower case since I like to think of it
as on a par with the standard basic types char, int, etc.  Somehow
this seems to matter enough when actually using such facilities on a
routine basis that I made a deliberate decision as described.)

A possibility for an = replacement (do we REALLY need one?  seems the
only problems are with <, >, and perhaps !, not =) would be the _
character.  The S system allows _ in place of its assignment operator
<- and I find it much more convenient.  The nice thing about _ is that
it is a valid C identifier, and furthermore, one reserved for the
implementation, which means that an add-on standard could choose to
give it a standard meaning without affecting existing strictly-
conforming code.

Of course, by now pretty much everyone involved is aware of my opinion
on this whole subject, namely, there is no problem requiring a solution
in standard terms, if properly approached, but that if nevertheless an
intrusive approach is going to be adopted, it should be one that is
entirely compatible with Standard C, which the introduction of new
compiler (preprocessing) tokens would not be.  <iso646.h> is fine but
anthing requiring recognition of %% etc. is horrible.