Re: [E-devel] death to xlib...? :)

[Wesley L. <wle...@ga...>]

On Tue, 2003-05-06 at 20:21, Carsten Haitzler wrote:
> On 06 May 2003 19:36:50 -0500 Wesley Leggette <wle...@ga...> babbled:
> 
> > Yeah, I know XCB is only a replacement library. The question I'm asking
> > is would it be faster to discard X all together. As far as I can see,
> > the advantages would be greater speed and better transparency support
> > (if X was abadoned). Not that anyone should actually do it. I just
> > thought it was an interesting point to bring up. I've heard a lot of
> > talk among different circles about just scraping X, and I wonder if
> > performance wise it would really be worth it.
> 
> I've actually said this in several of my talks before. x's biggest problem is 1.
> it doesn't keep up with the times in terms of features (alpha channels etc.) 2.
> is hard to get into (code-wise) and 3. is fucking huge.
> 
> it has very few problems performance-wise. most of the performance issues are
> due to badly written apps by people who don't understand what they are doing
> with x, and thus dont understand the performance implications of some of the
> things they may do in their code. also the same problems in widget sets etc.
> exacerbate this. xlib also sometimes puts an app in a position where it runs
> into an very non-optimal path and the app has no choice. xcb lets us solve that
> (though now puts the burden on the programme to know more about x and the
> protocol). and the rest is really a lack of features and extensions to provide
> those features.

Can you expound on this a little bit? Are you talking about something
like refreshing too much before redraws are done or something more
complicated? And is this the same problem faced in widget sets, or some
other flaw?

Also, what do you mean by "non-optimal path"? I'm not as familiar with
xlib as I am with the X protocol in general (When I started programming
low level graphics, using xlib was too complicated for what I wanted to
do, so I used OpenPTC instead).


> 
> > > > > -------------------------------------------------------
> > > > > This sf.net email is sponsored by:ThinkGeek
> > > > > Welcome to geek heaven.
> > > > > http://thinkgeek.com/sf
> > > > > _______________________________________________
> > > > > enlightenment-devel mailing list
> > > > > enl...@li...
> > > > > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
> > > 
> > > 
> > > 
> > > -------------------------------------------------------
> > > Enterprise Linux Forum Conference & Expo, June 4-6, 2003, Santa Clara
> > > The only event dedicated to issues related to Linux enterprise solutions
> > > www.enterpriselinuxforum.com
> > > 
> > > _______________________________________________
> > > enlightenment-devel mailing list
> > > enl...@li...
> > > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
> > -- 
> > Wesley Leggette <wle...@ga...>
> > 
> > 
> > 
> > -------------------------------------------------------
> > Enterprise Linux Forum Conference & Expo, June 4-6, 2003, Santa Clara
> > The only event dedicated to issues related to Linux enterprise solutions
> > www.enterpriselinuxforum.com
> > 
> > _______________________________________________
> > enlightenment-devel mailing list
> > enl...@li...
> > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
-- 
Wesley Leggette <wle...@ga...>