From: alawrenc@sobeco.com (a.lawrence) Subject: Re: Towards a better Posix Date: 23 Oct 1992 06:57:42 GMT
In <JEM.92Oct22224058@lk-hp-6.hut.fi> jem@snakemail.hut.fi (Johan Myreen) writes:
>Linux doesn't use the IBM PC character set, it uses the ISO8859-1
>Latin-1 character set. (Well, it tries to.) The high bit is not
>filtered by the kernel. VT100 graphics characters are available by
>writing a Shift Out character (^N) to the console, the normal
>character set is restored with a Shift In character (^O).
There is already support in Linux for the IBM extended ASCII character set.
In a few hours, as soon as it get UUCP back up and running, I will post
too short C programs. One to set "IBM" mode, and another to reset the
display to it's initial state regardless of the current condition.
[... text deleted ...]
>I'll have to admit there is one valid point in remapping the keyboard
>without recompiling: changing the mapping for different national
>keyboards. This would help distributing Linux in binary form, without
>having to provide N different kernel images, each for a different
>keyboard. But I don't think we need the ability to do this on a
>running kernel, a program to change the mapping in a kernel image file
>is sufficient. (A program like this was posted some time ago in this
>newsgroup.) After all, how often do you change keyboards?
When working one a system that supports multiple keyboard definitions,
I will ofter change keyboards several times a day, sometimes per hour.
When editing French text, I use a French keyboard, when programming I
use the standard US, and when creating documentation I often use a
"line draw" keyboard definition.
>PC ANSI graphics? What's that?
>The Linux console does quite a good job emulating a vt100.
>--
>Johan Myreen,
>barbarian
Yes, the Linux console does do a good job of emulating a vt100 which
does a good job of emulating a KSR-35 teletype. Perhaps however, we
should consider the possiblities that the hardware allows us, and
consider making enhancements to the system to exploit those possibilities.
I will be posting some suggestions and an outline of how I propose to
implement them to the kernel channel of the mailing list shortly.
-- If any given program runs correctly, it is obsolete. ========================================================================= Andrew Lawrence, Informaticien Conseil | alawrenc@sobeco.com 3605 St-Urbain, #1605 | uunet!sobeco!alawrenc