I have skimmed through the current r7 working paper and I was expecting a layered approach to the API. Namely that at the bottom would be a pixel surface structure with simple get/put pixel facilities and on top would be a rendering layer that provided the facilities for brushes, pens, stroke text, raster text, etc.
This layered approach has been quite common for graphics APIs for some time now, allowing you to use a reference (correct, but slow) implementation for the advanced rendering features and implement get/put pixel for your device.
However, I seem unable to find get/put pixel functionality in this proposal. Is that deliberately left out? If so, where is the declared rationale for excluding this?
That functionality is part of the GraphicsSurface implementation rather than part of this specification. For example, it would be meaningless on a plotter.
I have skimmed through the current r7 working paper and I was expecting a layered approach to the API. Namely that at the bottom would be a pixel surface structure with simple get/put pixel facilities and on top would be a rendering layer that provided the facilities for brushes, pens, stroke text, raster text, etc.
This layered approach has been quite common for graphics APIs for some time now, allowing you to use a reference (correct, but slow) implementation for the advanced rendering features and implement get/put pixel for your device.
However, I seem unable to find get/put pixel functionality in this proposal. Is that deliberately left out? If so, where is the declared rationale for excluding this?