Re: termcap ballast

[Frank da Cruz <fdc@watsun.cc.columbia.edu>]

Markus Kuhn wrote on Thu, 04 Nov 1999 14:44:49 UTC:
> Frank da Cruz wrote on 1999-11-04 01:16 UTC:
> > is not true, however, in the reverse direction, where we have applications
> > running on these same platforms that do not use terminfo/termcap, but are
> > hardwired for real physical terminals that are no longer manufactured, and
> > not widely (or well) emulated.  You can laugh at these applications, but
> > that won't make them go away.
> 
> But wouldn't it be much easier to plug a terminal protocol converter on
> to of these applications? Under Unix with the availability of pseudo-tty
> devices, this is very easy to do. A terminal protocol converter running
> on the host could work very similar to say GNU screen.
> 
Conceptually it is possible.  In reality, nobody will do the work.  Each
kind of terminal, except very old and primitive ones like VT52, is actually
quite complicated.  As many of you know, even companies that specialize
in terminal emulation tend not to get them right, even with many years of
trying.

Anyway, 'screen' is not necessarily a solution to all problems, since it
introduces new problems of its own, such as transparency and responsiveness.
And since it is termcap/terminfo driven, it can not mimic a real terminal,
since termcap/terminfo does not fully describe all the behaviors of a real
terminal.

> Did someone already look into UTF-8 support for screen?
> 
Anyone who does this should bear in mind that many terminals, such the VT320,
support full ISO 2022 character-set designation and invocation.  They should
also consider that many terminals support switching into and out of various
"graphics modes" (Tektronix, ReGIS, Sixel, ...).

- Frank
-
Linux-UTF8:   i18n of Linux on all levels
Archive:      http://mail.nl.linux.org/lists/