Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!nrl-cmf!ames!oliveb!sun!gorodish!guy
From: guy@gorodish.Sun.COM (Guy Harris)
Newsgroups: comp.windows.x
Subject: Re: Questions about kterm
Message-ID: <52853@sun.uucp>
Date: 11 May 88 17:42:11 GMT
References: <8805111155.AA00590@nirvana.cs.titech.junet>
Sender: news@sun.uucp
Lines: 22

> In the (normal?) Unix environment, the tty(pty) device driver strips off
> the MSB's of the characters passed through it.

True of some UNIXes, not of all UNIXes.

4.3BSD can be told not to strip the 8th bit from characters provided as input
for a pseudo-tty by setting PASS8; however, it cannot be told not to do so for
characters written as output to a pseudo-tty unless you set RAW or LITOUT.
Setting either of those bits turns off all special output post-processing,
which is generally more than you want.

System V can be told not to strip input by turning off ISTRIP, and never strips
output (at least not in the tty driver itself; particular hardware can usually
be instructed to do so).  However, vanilla System V doesn't have pseudo-ttys.
System Vs with pseudo-ttys can presumably handle "kterm" in EUC mode correctly.

> They say that SUN OS 4.0 have such capabilities, but I am not sure.

Being the author of the 4.0 tty driver, I am sure.  It never strips the 8th bit
on output (unless you tell the hardware to do so), and won't strip the 8th bit
on input if you set PASS8 (by using the V7/BSD "ioctl"s) or clear ISTRIP (using
the S5/POSIX "ioctl"s).