Re: GTK+ scene graph


On 18 August 2014 19:09, Paul Davis <paul linuxaudiosystems com> wrote:

On Mon, Aug 18, 2014 at 1:55 PM, Emmanuele Bassi <ebassi gmail com> wrote:
[ .. ]

realistically, by comparison with other platforms and with many home-grown
GUI toolkits, there's really only one sane design for any toolkit in 2014:

   * recursively nested objects with a common "draw" virtual method that
renders onto a surface (*) as the only way pixels get drawn
   * access to a GL surface when and if necessary
   * packing and layout orthogonal to drawing from the application code

anything that moves GTK closer to this is good. anything that moves it
further away or impedes motion towards it is bad.

(*) presumably cairo for 2D and GL for 3D but smart people can disagree
about this.

if we ignore the GL side, then good job: GTK+ 3.12 is already pretty
much that. except, you know:

 * the API being weird
 * the lack of a box model for CSS to draw on
 * the lack of support for 3D transforms and animations
 * the lack of an animation API capable of implementing dynamic layout
management policies

and a bunch of other smaller details.

in any case, the end goal is to make GTK+ 4.0 a better API for writing
dynamic and compelling applications for a variety of form factors.
maybe provide enough building blocks for people to write their own
smaller canvases/toy toolkits on top, in order to satisfy ad hoc


[ ] ebassi [ gmail com]

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]