Re: PTY feature? or bug?

Craig Milo Rogers (rogers@isi.edu)
Fri, 26 Apr 96 12:09:33 PDT

>NetBSD-current does wait for the master open. So does SunOS. I suspect
>all BSD-derived systems do wait for the master to open. (and remember
>that pty's aren't really defined by any other standard than what BSD
>does, since it was the OS that introduced this innovation).

"the OS that introduced this innovation" -- I don't like that
phrase, it offends my pride as a long-time TOPS10 programmer. TOPS10
is, I think, the OS that introduced the innovative device "pty". BSD
merely incorporated the concept into Unix (and didn't do a very
satisfactory job of it).

So, what historical perspective does TOPS10's implementation
of PTYs have to offer on the issue at hand? Gee, I was afraid you'd
ask that. :-/ TOPS10 didn't have a "fork" operation; essentially, the
PTY master had to be opened before the slave's subprocess could even
exist.

Craig Milo Rogers