EGL_MESA_screen_surface proposal

[Jon Smirl]

On Wed, 16 Mar 2005 12:12:31 -0700, Brian Paul
<brian.paul at tungstengraphics.com> wrote:
> Jon Smirl wrote:
> > On Wed, 16 Mar 2005 10:18:26 -0700, Brian Paul
> > <brian.paul at tungstengraphics.com> wrote:
> >
> >>Well, I think that's a pretty esoteric example.  We're starting to get
> >>into things that most people will never come across.  Still, I'm
> >>willing to be flexible.
> >>
> >>Would a function like this suffice:
> >>
> >>const char *eglQueryModeStringMESA(EGLDisplay dpy, EGLModeMESA mode);
> >
> >
> > This solves the problem. You can eliminate the interlace attibute if
> > this call is added.
> 
> I wouldn't want to eliminate the interlaced attribute.  If I did, and
> one wanted an interlaced mode, they'd have to resort to parsing the
> mode name strings.  As an EGL API user, I'd rather use eglChooseMode
> to specify that attribute.  Otherwise, how I am I to know whether I
> should look for "i" or "interlaced" in the mode name strings?

The point here is that you don't want to create an interlaced display.
You instead want the best display at 1920x1080. So if you
glChooseMode(1980, 1o80) and your monitor and hardware both support it
you will get back HDTV 1080p. If either your monitor or hardware
doesn't support 1080p you'll get back 1080i. But both of these scan
out of the same buffer so it doesn't really matter which is returned.
You just want the best display possible.

The only reason you need the interlace attribute is when you have both
1080p and 1080i available and you want to force 1080i. But now that
think about it I don't think you'll ever want to do this so we don't
need the attribute. I can't think of a situation where you want to
force an inferior mode when you have a better one available.

-- 
Jon Smirl
jonsmirl at gmail.com