From: rchen@fraser.sfu.ca (Robert Chen) Subject: Re: BUG in 0.99-p5 ??? Date: 16 Feb 1993 06:37:37 GMT
In article <1lpg7iINNgq9@senator-bedfellow.MIT.EDU> tytso@athena.mit.edu writes:
> From: torvalds@klaava.Helsinki.FI (Linus Torvalds)
> Date: Sun, 14 Feb 1993 09:56:58 GMT
>
> Arggh. It was me who added the new noctty checks: I also disabled the
> use of /dev/console as a general tty, as that had resulted in some weird
> syslog behaviour. I had gotten a few comments that pty masters weren't
> real tty's, and should also be disabled, and I tend to agree. I'm sorry
> if I broke any binaries, but I don't think I'll change it back unless
> there is any other reason for it than a few weird bsd binaries.
>
>I don't understand why not allowing master pty's to become controlling
>tty's would cause any problems. Why are certain programs relying on
>this behavior at all?
>
> - Ted
There was a big deal made about this a few months ago in the NET
channel. I think that telnet/telnetd relies on some questionable pty
behavior. Does anyone remember the details? (remember new-telnet and
the kernel patch that made it work?)
The newest telnet from the net-bin stuff still works well as of pl5.
I am curious as to what ever became of the master/slave discussion that
finally fixed BSD telnet.
- Rob