[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Can popen2 be avoided?

From: Stef Pillaert (KAHO)
Subject: Re: Can popen2 be avoided?
Date: Thu, 7 Sep 2000 08:46:53 +0200

> On  6-Sep-2000, Stef Pillaert (KAHO) <address@hidden> wrote:
> | I have trouble using popen-close (see bug-octave 2000-276) and
open2( see
> | help-octave 2000-722).
> | For my students, I'm making some simulations, where the calculations are
> | done in octave, but the interface is in Tcl/Tk.
> |
> | The interface (which I start with popen from octave), is perfectly able
> | send strings to octave, but I would like the output of octave to be
shown in
> | the same interface, not to stdout. Since I can't get popen2 to work, is
> | there another way to send the output of octave to the interface? I'm not
> | very familiar with communicating between processes (as you might have
> | noticed...), so...can someone help me?
> There have been some problems with popen2.  I think the current CVS
> sources fix all but one of the problems, which is that it is possible
> for nothing to be available to read from the subprocess (popen2 sets
> that file descriptor up for non-blocking I/O).  In that case, after a
> read that returns nothing, errno should be set to EAGAIN.  If you
> could detect that, then you would want to clear the error on the
> stream, then try again (or better, call select and wait for some input
> or a timeout to happen).  The trouble is that there is currently no
> way in Octave to check errno, no good way to clear the error on the
> stream state, and no interface to the select system call.  :-(
> jwe
Thanks a lot. Is there a patch against 2.1.31 (or will there be a 2.1.32 in
the near future?)


Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:
How to fund new projects:
Subscription information:

reply via email to

[Prev in Thread] Current Thread [Next in Thread]