Re: 8859-1, 8859-15, 1252 and Euro

From: Frank da Cruz (fdc@columbia.edu)
Date: Fri Feb 11 2000 - 18:31:26 EST


Doug Ewell wrote:

> "Robert A. Rosenberg" <bob.rosenberg@digitscorp.com> wrote:
>
> > I would have removed the useless junk in the C1 range and put useful
> > glyphs there. What ACTUAL purpose is served by wasting 32 codepoints
> > to duplicate the C0 (x00-x1F) codes? I can see no reason why a
> > character code needs to function as if the high bit is not there. If I
> > strip the high bit off a character ALL the characters in the xA0-xFF
> > range will display wrong so why preserve the C1 codes just so it
> > FORMATS "correctly"? It is junk in any case so just display the
> > formatting incorrectly.
>
> The C1 range contains 32 additional control characters, not just
> duplicates of the C0 range, and it is my understanding that many
> terminals do use these control codes, although PCs and Macs do not.
>
VT220 and above use them, Siemens Nixdorf terminals use them, and many
others. These are directly from ISO 6429 and many of them are used to
implement the character-set designation and invocation rules of ISO
2022:

  SS2 - Single Shift 2
  SS3 - Single Shift 3
  LS1R - Locking Shift 1 Right
  LS2 - Locking Shift 2
  LS2R - Locking Shift 2 Right
  LS3 - Locking Shift 3
  LS3R - Locking Shift 3 Right

These techniques are what allow (for example) a VT320 terminal to
display a screen containing any mixture of ASCII, line/box drawing
characters, math symbols, and accented Roman letters. But this is not
an obscure feature of the VT320, it is an international standard that
is used all over the world.

- Frank



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:59 EDT