Every newcomers to X think that servers and clients are reversed because they are

Patrick O'Donnell pao at ascent.com
Wed Apr 20 06:49:05 PDT 2011


>Date: Wed, 20 Apr 2011 14:11:47 +1000
>From: Peter Hutterer <peter.hutterer at who-t.net>
>User-Agent: Mutt/1.5.21 (2010-09-15)
>Sender: xorg-bounces+pao=ascent.com at lists.freedesktop.org
>
>On Tue, Apr 19, 2011 at 11:36:17PM -0400, Paul Dufresne wrote:
>> For years I have heard of how we must rethink what is the server and
>> what are the clients when coming to X. ...
>
>...Hardware is not software, and this problem is little more than 
>the English language using the same word for two different things.
>You don't even have to rethink what servers and clients are. ...

I respecfully, partially, disagree with the last.  For me, grasping
the difference between X servers and X clients required fixing a
broken intuition that servers were "there" and clients were "here".
(A slight generalization of your hardware theory.)  In the X case, the
server is closer to the user than the client, offering the resources
the user uses to interact with the machine to other programs to share.
I would call what I had to do, "rethinking".

The other "rethinking" that may be necessary for some is the error
that a program is either a client or a server.  I've used -- and
written -- many programs that act both as clients and servers,
depending on their relationship with other servers and clients.

However, I don't think I ever did think -- or could ever come to think
-- of the X server as a client to the programs that connect to it.  It
was my early intuition that was wrong, not the definitions that the X
Window System used.

		- Patrick



More information about the xorg mailing list