Mailing-List: contact cygwin-developers-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-developers-subscribe AT sources DOT redhat DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin-developers/>
List-Post: <mailto:cygwin-developers AT sources DOT redhat DOT com>
List-Help: <mailto:cygwin-developers-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-developers-owner AT sources DOT redhat DOT com
Delivered-To: mailing list cygwin-developers AT sources DOT redhat DOT com
Date: Thu, 31 May 2001 21:59:39 +0200
From: Corinna Vinschen <vinschen AT redhat DOT com>
To: cygwin-developers AT cygwin DOT com
Subject: Re: [RFD]: Egor's proposal for a Cygwin server process
Message-ID: <20010531215939.Y1870@cygbert.vinschen.de>
Reply-To: cygdev <cygwin-developers AT cygwin DOT com>
Mail-Followup-To: cygwin-developers AT cygwin DOT com
References: <20010531124452 DOT G1870 AT cygbert DOT vinschen DOT de> <48146951254 DOT 20010531164356 AT logos-m DOT ru> <20010531151226 DOT I1870 AT cygbert DOT vinschen DOT de> <192151639946 DOT 20010531180204 AT logos-m DOT ru> <055e01c0e9da$dc2812d0$0200a8c0 AT lifelesswks> <20010531170820 DOT N1870 AT cygbert DOT vinschen DOT de> <20010531140932 DOT F23914 AT redhat DOT com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.2.5i
In-Reply-To: <20010531140932.F23914@redhat.com>; from cgf@redhat.com on Thu, May 31, 2001 at 02:09:32PM -0400

On Thu, May 31, 2001 at 02:09:32PM -0400, Christopher Faylor wrote:
> >Hmm, personally I would prefer using a Win32 interface like the
> >above (named pipes on NT, shared memory on 9x). Security doesn't
> >matter on 9x and the low level interfaces are typically faster
> >and (from my point of view) easier to understand. Anyway, if we
> >agree to use COM I wouldn't step back.
> 
> Are we thinking about making this server able to maintain multiple
> platforms?  I think it could possibly be useful to have a single
> setuid server running in an NT domain.
> 
> Or maybe that's far into the future...

Far, very far in the future...

> I have a book on COM sitting around somewhere.

It's just sitting around? What is it doing in all it's spare time???

> I wonder how much
> overhead that would add to the Cygwin DLL, though.

We could use the plain C interface as in shortcut.c probably.
The C++ interface is not only more overhead but raises some
new problems. You remember the compiler switch needed? Wasn't it
`-fvtable-thunks'?

> Also, I don't know if this has already been mentioned or not but
> we also have to make sure that cygwin works ok without the server.
> Maybe that means that ttys are not a good candidate for the server
> because they are a heavily used feature and we don't want to limit
> functionality.

I think everything is a good candidate. From my point of view the
real base functionality should still be implemented in the DLL.
The server just calls that functionality in turn. This enables
us to short-circuit the server if it isn't available. We then
just drop some features (IPC functions return ENOSYS) or some
security (tty handles are unsecure again) whatever is concerned.

Corinna

-- 
Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Developer                                mailto:cygwin AT cygwin DOT com
Red Hat, Inc.