Re: imaging model (was: JPush)

On Wednesday 02 November 2005 08:23 pm, Ralf Juengling wrote:
> On Wed, 2 Nov 2005, Yann LeCun wrote:
> > - a better imaging model for the default graphic driver
> >  (the existing one is based on the old Xlib).
>
> What is a graphics driver's "imaging model"? You might point
> to something I know (e.g., does matlab have an "imaging model")?

An imaging model is the set of graphic primitives that are given to
you for drawing stuff. For example, Xlib, OpenGL, Imlib, PostScript,
PDF, SVG, Quartz, and Cairo all defined an imaging model (some of them
are also APIs).  An imaging model is not exactly an API, because you
could have several APIs for the same imaging model, and a single API
could be compatible (or emulate) several imaging models. It's more
like a set of behaviors.

Xlib, which Lush uses for its default driver, is pixel-map oriented
and doesn't allow such things are resolution independence,
alpha-blending/compositing, antialiasing, 3D, texture mapping, etc. It
was the hottest thing in 1988, but it's badly out of date now.

If we write a new graphic interface for Lush with more functionality,
we might want to consider an imaging model that can be implemented
with standard APIs on various platforms of interest.  We might want to
use Cairo, which has backends on Linux, Windoze and Mac that
can talk to X, OpenGL, Quartz, PDF, PNG, PostScript, SVG, 
frame-buffer,....
See http://cairographics.org/introduction

  -- Yann






-------------------------------------------------------
SF.Net email is sponsored by:
Tame your development challenges with Apache's Geronimo App Server. Download
it for free - -and be entered to win a 42" plasma tv or your very own
Sony(tm)PSP.  Click here to play: http://sourceforge.net/geronimo.php