Re: gnuplot vs. emacs' compile command

[Hans-Bernhard Broeker <broeker@physik.rwth-aachen.de>]

On Sat, 24 May 2003, Richard Stallman wrote:

> It is not so bad as all that.  Many things in Emacs can loop, and people
> need to know about C-g.

Right.  But even so, a visible reminder might be helpful.  Since Emacs
won't react to _any_ user input except C-g in this situation, I think it
should announce this fact in a minibuffer message.

> That doesn't mean there is no room for improvement here.  Perhaps Emacs
> should use a different criterion.  But I am not sure what it should be.
> (Currently it waits for all writers to close the pipe, and then for
> termination of the original process.)

The criterion is fine, as far as I'm concerned.  At least gnuplot can be
taught to live with it.

What I'm less sure about is whether the wait loop for this subshell should
affect all the rest of Emacs, too.  At the moment, it blocks all user
interaction until the external process is finished, or C-g is pressed.

-- 
Hans-Bernhard Broeker (broeker@physik.rwth-aachen.de)
Even if all the snow were burnt, ashes would remain.