Re: [Gterm-discuss] PTY reading

[Timothy M. <th...@ya...>]

--- Jörg Sonnenberger <joe...@we...> wrote:
> On Fri, 13 Dec 2002 10:59:50 -0800 (PST)
> Timothy Miller <th...@ya...> wrote:
> 
> this. The following leaves the gterm open, but it shouldn't stay (I
> think):
> 	sleep 1000 &  [ keeps pt open]
> 	exec sleep 1000 < /dev/null > /dev/null 2> /dev/null	[closes pt]
> Xterm stays, but I doesn't like the idea of the terminal staying if
> the foreground
> process closed his pty interface. 

I'm not sure what to think about that.  Even if a process closes its
pty interface, wouldn't we still want to be able to ctrl-c it?  Or
background it?

> Attached is a patch using poll instead of select, which explicitly
> checks this,
> doesn't make assumption of read returning 0 and is somewhat simpler
> than
> the select. Since poll is a possible portability issue, I'd like to
> discuss it first.
> 

I suggest we put that in as an #ifdef option.  When someone configures
it for a build, they pick what they want.

But the only thing that must be 100% portable (unix, windows, mac,
etc.) is the library core.  The example doesn't have to be.


__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com