Date:	Tue, 27 Jan 1998 14:36:29 +0200 (IST)
From:	Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
To:	"Markus F.X.J. Oberhumer" <k3040e4 AT wildsau DOT idv-edu DOT uni-linz DOT ac DOT at>
cc:	dj AT delorie DOT com, djgpp-workers AT delorie DOT com
Subject:	Re: Request for comments: SIGQUIT in DJGPP v2.02
In-Reply-To:	<199801261913.UAA27415@wildsau.idv.uni-linz.ac.at>
Message-ID:	<Pine.SUN.3.91.980127143559.843A-100000@is>
MIME-Version:	1.0

On Mon, 26 Jan 1998, Markus F.X.J. Oberhumer wrote:

> Please excuse me, but the current *design* is definitly "rash".

You will need to explain what aspects of the design deserve this
label.  Otherwise, I won't know what and how to improve.

As I already said, moving SIGQUIT support into termios is IMHO a bad
idea, since signal support has very little to do with termios, and
programs which need signals should not need to link in termios which
affects terminal I/O in fundamental ways.

> I have been asking to find out your motivation for the patch. The
> first time you responded it was emacs and some interpreter, now you
> respond it's not emacs.

My changes include two main features: (1) the ability to define a key
other than Ctrl-C that generates SIGINT, and (2) the generation of
SIGQUIT (whose key can also be user-defined).  Emacs was the main
motivation for the first of these.

> - which programs need it ?
> - what exactly are they needing it for ?

I already answered that: SIGQUIT is required when you need to have two
different actions triggered by a keypress.

But I don't see why should we at all discuss programs which need this.
I didn't invent SIGQUIT itself, it exists both in v2.01 and on other
platforms to which we want to be compatible.  So if the SIGQUIT
feature in its default configuration is harmless (as it is now, since
SIGQUIT is ignored by default), then I really see no reasons to argue
about its inclusion on the basis of its potential users.  For example,
the only user of termios (that I know of) is Bash; does that mean that
we don't need termios?

> - why do you it is neccessary to do B) instead of A) ?

A) is to not change anything.  I'm not willing to flush down the drain
the many hours of work I spent making this happen (including all the
changes that several people requested to make it more transparent)
which result in code that works and is clean enough to satisfy me.
Frankly, I'm surprised you even ask me to consider that possibility.

In summary, I think SIGQUIT support is both important enough and
harmless enough to be included.  If you have specific comments on the
design, I would be glad to hear them.

- Raw text -