Robert de Bath wrote on 2001-06-12 08:39 UTC:
> > I am slightly concerned though that CSI ... w was apparently already
> > used by DEC, so I guess it will have to be CSI ... !w then.
> Again you pick an uncommon usage that many 'good enough' emulators
> won't ignore. The characters $ > " = are more widely understood and
> skipped.
I couldn't care less. You still haven't got the bigger picture here: SCW
is only of use in UTF-8 terminal emulators, and we have all the relevant
developers of UTF-8 terminal emulators here on this list. DEC VT102
compatibility is *NOT* a design criteria for SCW for this obvious
reason, because a 1970s vitage DEC VT102 does not understand UTF-8 and
never will. DEC is dead. We are working towards the creation of a next
generation terminal standard that will exclusively run under UTF-8 and
therefore will not be VT102 compatible. If that requires a few
developers of terminal emulators to read ECMA-48 more carefully again
and bring their control sequence parsers in line with both ECMA-48 and
it's ANSI predecessor, instead of just following some 1970s user manual,
then that is a GOOD THING[TM]. If a parser chokes on an intermediate
character "!" then that parser has to be fixed.
In all conflicts between ECMA-48 and VT102 behaviour that I have heard
quoted here so far, the ECMA-48 behaviour seemed technically more
desireable to me. Why for example use both "" and "0" as an encoding for
the default value, and therefore loose "0" as an explicit parameter
value? Sounds unpleasant to me. Don't turn VT102 into a religion, if we
anyway use a non-VT102 character encoding already.
Markus
--
Markus G. Kuhn, Computer Laboratory, University of Cambridge, UK
Email: mkuhn at acm.org, WWW: <http://www.cl.cam.ac.uk/~mgk25/>
-
Linux-UTF8: i18n of Linux on all levels
Archive: http://mail.nl.linux.org/linux-utf8/
