> Frank da Cruz wrote on 2001-05-15 23:30 UTC:
> > Not to be a stick-in-the-mud but... Private character sets that do not
> > conform to the minimum international standards of structure (ISO 4873 and
> > ISO 2022) have no place in information interchange, and therefore there
> > is no need to register them.
>
> Come on, the ISO-IR is already full of dozens of funny and obscure
> private character sets nobody has ever seen in use. We wouldn't reduce
> the aesthetics of the collection in any measurable way by adding the DOS/
> Windows code pages. In addition, ISO 2022 has perfectly good mechanisms
> to access the C1 control characters via the CL code value range
> (0x00-0x1f). This is not only needed for when CR+GR is filled with
> graphical characters (as in Windows code pages), it is also needed when
> CR+GR is filled with wrong-parity-bit versions of CL+GL in 7-bit
> communication systems. 7-bit communication never was a problem with ISO
> 2022, and so shouldn't 128 character sets be.
>
> It would be nicer though to have 128 charcater sets properly defined as
> a class of its own in a new revision of ISO 2022. If there is a separate
> class of ESC sequences for activating them, then software would
> automatically be aware that although C1 is now gone, C0 is still
> available. That is strictly speaking not quite necessarily the case with
> ESC % ... coding systems outside ISO 2022.
>
> Markus
The issue isn't whether or not it is possible for the host to send a
C1 control sequence when an IBM/MS Code Page is in use. We know the
host can always do that. The issue is whether the terminal emulator
knows to ignore C1 controls.
As we all know, ISO2022 compliant terminals have CL,CR,GL,GR which are
mapped to C0,C1,G0,G1,G2,G3. When we use an ISO2022 escape to switch
to UTF8 mode for instance we given up the ability to manipulate the
mappings between GL/GR and G0-G3. I would argue that if we provide
ISO2022 escapes to use IBM/MS code pages that we would in turn be
giving up the ability to perform those manipulations. This is turn
would mean that IBM/MS Code Pages could not be used in conjunction
with the Graphics or Box drawing character sets of VT terminals.
Are you are proposing that an update to ISO2022 be issued that defines
a new mode that combines CR+GR into a single entity for the purposes
of mapping to
C1+G0 C1+G1 C1+G2 C1+G3
but keep the ability to manipulate the mappings of GL/GR? If so, I
do not see how this would work.
The only possible scenario that I can see is the use of ISO2022
escapes which would disable the use of ISO2022 until a suitable return
sequence is issued. This may be suitable for some applications.
However, I believe it will break most terminfo/termcap entries that
assume ISO2022 is used for box drawing.
Jeffrey Altman * Sr.Software Designer C-Kermit 7.1 Alpha available
The Kermit Project @ Columbia University includes Secure Telnet and FTP
http://www.kermit-project.org/ using Kerberos, SRP, and
[EMAIL PROTECTED] OpenSSL. SSH soon to follow.
-
Linux-UTF8: i18n of Linux on all levels
Archive: http://mail.nl.linux.org/lists/
