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
Subject: Re: Egor's daemon
From: Robert Collins <robert DOT collins AT itdomain DOT com DOT au>
To: cygwin-developers AT cygwin DOT com
In-Reply-To: <20010912123116.A18008@redhat.com>
References: <1000295535 DOT 30404 DOT 67 DOT camel AT lifelesswks>
	<20010912115511 DOT A17668 AT redhat DOT com> <1000310370 DOT 30375 DOT 141 DOT camel AT lifelesswks>
	<20010912121322 DOT A17887 AT redhat DOT com> <6611059312 DOT 20010912202029 AT logos-m DOT ru> 
	<20010912123116 DOT A18008 AT redhat DOT com>
Content-Type: text/plain
Content-Transfer-Encoding: 7bit
X-Mailer: Evolution/0.13 (Preview Release)
Date: 13 Sep 2001 20:27:43 +1000
Message-Id: <1000376864.7335.18.camel@lifelesswks>
Mime-Version: 1.0
X-OriginalArrivalTime: 13 Sep 2001 10:14:34.0314 (UTC) FILETIME=[E2C1CAA0:01C13C3C]

Moved to cygwin-developers.

On Thu, 2001-09-13 at 02:31, Christopher Faylor wrote:
> On Wed, Sep 12, 2001 at 08:20:29PM +0400, egor duda wrote:
> If the client part is linked into the DLL, it should be in the cygwin
> directory, sure.
> 
> I think we are not at the point where we include this in the cygwin
> directory, though.  I'd like to discuss how we are going to be interfacing
> the client code with the cygwin DLL, though.

In what sense? Egor's approach, which I'm planning a minor tweak of to
make more object oriented, is to have a cygserver_request function which
is passed a request_buffer and returns a request_buffer. That structure
is opaque to the transport layer, apart from the size (which is variable
per request type). 
 
So calling it looks like
fill in request_buffer buf
cygserver_request(&buf);
look at results in buf.

> I'm not yet comfortable with making it an integral part of the DLL.

Sure. I'm hoping to get a generic enough proof-of-concept that runs on
NT and 95 equally well, that such decisions can actually be looked at.

As such, I'm going to spend 0 time on rearranging the source, and focus
on functionality. If and when this looks good to you, we can discuss the
directory structure. (I don't actually care where it goes as long as a
make in winsup/cygwin results in it being built. Ie subdirs are fine,
but I don't think it should be in a parallel dir to cygwin.

Rob