Re: restrictions when defining charsets

[keld@xxxxxxxx (Keld J|rn Simonsen)]

Valdis Kletnieks writes:

> > > -- citing RFC1345 only proves that Keld's interpretation
> > > of ANSI X3.4 is that codepoint 41 octal is an exclamation point.
> > 
> > To be pragmatic, it is the interpretation of almost all people.
> 
> Ahh.. but that's the *current* interpretation.  I've used a lot
> of keyboards that were convinced that a 41 octal was a vertical
> bar.  I've used a lot of keyboards that didn't understand the
> circumflex character (reading the Berkeley 'curses' source, you'll
> come across the phrase "Hazeltine Brain Damage"...).

I have used ANSIs official registration of ASCII, in the ECME
registry according to ISO 2375, registration number 6, dated 1975/12/01.
It says on page 3.10:

2/1 ! Exclamation mark
5/14 ^ Upward arrow head  circumflex accent

No "vertical bar", no "not sign" there. 
Anyway I would not exclude that the official ANSI standard at one time
had provisions for the "not" and "vartical bar". But I would believe
that the current ASCII standard is in conformance with ISO 646,
which would exclude vague wording on the positions used for
! and ^  (well, at least for ! , ^ is at a national use position.)

> Pragmatic decision: To make the PL/I subset fit into the requisite
> number of characters, 2 ASCII codepoints are ambigiously defined,
> so 2 similar-looking glyphs can share a codepoint.
> At the time, "almost all people" interpreted these two characters
> as "similar enough" to let it happen.  Years later, people realize
> that they are different characters and fix the problem.

There are similar examples on 7-bit use in scandinavia, where
{|}[\] are both used for the ASCII characters (eg in C programming)
and as national letters. RFC1345 has som mechanisms to specify this.
This is ugly, but quite useful.

Keld