Around 21 o'clock on Dec 17, David Dawes wrote:

> Today's many might not be tomorrow's many, and I think it would be
> short-sighted to ignore bandwidth issues that may impact the
> usability of remote X.  I'm not saying that this should stand in
> the way of improvements in the local case, but it shouldn't be
> ignored either.

Render fixes this in the remote case.  The question is whether we should 
move towards making client-side fonts the norm, even if it causes a
performance impact in the remote case for displays without Render.

There are a couple of arguments in favor:

        1)      Existing apps adding Render support generally leave core
                text support in place and make Render text a separate path.

        2)      Apps needing local fonts will only work on Render-enabled
                desktops without this change.

As for the performance impact on remote desktops without Render, I submit
that we've been living with this for a long time; sophisticated
applications have used client-side text for as long as X11 has been around.
As long as you avoid round-trips, performance is acceptable, if not 
stellar.  I've run client-side text over ISDN links, and while not 
stunning, it's not a significant performance problem.  Remember that the 
dominant traffic for any normal X session is image data, even client-side 
text makes little difference here.

For local desktops, we have the ability to add anti-aliasing for static 
visuals; this means that users will gain *additional* capabilities in this 
common environment.

Of course, as desktops advance to include Render support, all of this 
becomes moot.  The question is how to best get from where we are to that 
point.

Keith Packard        XFree86 Core Team        Compaq Cambridge Research Lab


_______________________________________________
Render mailing list
[EMAIL PROTECTED]
http://XFree86.Org/mailman/listinfo/render

Reply via email to