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

From: Frank da Cruz (fdc@columbia.edu)
Date: Mon Feb 14 2000 - 16:03:11 EST


"Robert A. Rosenberg" <bob.rosenberg@digitscorp.com> wrote:
> ...
> You are not responding to my statement but addressing a different issue. On
> that VT320 what is the ACTUAL character code used on the device?
>
It is determined by the initial Setup mode, as modified by subsequence ISO
2022 character-set designation and invocation sequences from the host.

> We are talking about usage of the ISO-8859-x codes for
> TRANSMISSION/INTERCHANGE of text over the Internet (in Email and HTML) not
> the code used by the computer to talk to the CRT.
>
The Internet also includes Telnet connections. The "thing" in front of the
user's nose on a Telnet connection is, ultimately, a terminal or terminal
emulator that responds to escape sequences, which, for most terminals, respect
the ISO standards for character-set structure and function, and the most
popular ones (VTxxx, xterm, etc) also follow ISO 6429.

This notion is in direct conflict with the notion of graphics in the C1 area.
Terminal/host communication is TRANSMISSION/INTERCHANGE just as much as
email and web browsing. In fact, many people do their email and web browsing
in a terminal-host setting.

Since terminals (and emulators) are an integral part of the Internet, does
it follow that there should be *two* methods of transmission/interchange --
one that follows ISO 2022/4873/6429, and another that violates them?

> At home I use a Macintosh
> (which uses MacRoman [different codes for the x80-xFF range than
> ISO-8859-1] but I have no problems correctly sending/receiving/displaying
> Email or HTML so long as it is ID'ed as ISO-8859-1 (or windows-1252 if
> Windows glyphs in the x80-x9F codepoint range are used)]. The only reason
> for problems is the useless (for interchange) C1 codes.
>
PC Code pages and Apple MacRoman (and Quickdraw, etc) were not designed or
intended for interchange. They are private character sets to be used within
the confines of a particular computing environment.

We would all like it if the Internet were based on the private character set
used on our own desktops. But since there is still a diversity of them, we
need standards, and only standard -- not private -- character sets should be
used on the wire.

- Frank



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