Re: Advice on extending CanvasRenderingContext2D to support Depth Streams

On Fri, Oct 31, 2014 at 5:53 AM, Kostiainen, Anssi <
anssi.kostiainen@intel.com> wrote:

> > On 30 Oct 2014, at 23:55, Rik Cabanier <cabanier@gmail.com> wrote:
> >
> > Did you get any feedback from implementor so far?
>
> We have an experimental Chromium implementation. Ningxin can share more
> information on that. Also folks from the Google’s Project Tango are
> informed and have been providing feedback.
>
> That said, this is a FPWD, and thus the expectation is we’ll get more
> feedback and can improve the shape of the API. This is what we state in the
> SoTD section to set the expectations:
>
> [[
>
> This document is not complete and is subject to change. Early
> experimentations are encouraged to allow the Media Capture Task Force to
> evolve the specification based on technical discussions within the Task
> Force, implementation experience gained from early implementations, and
> feedback from other groups and individuals.
>
> ]]
>
> It appears public-canvas-api is a good place to gather such feedback what
> comes to canvas-related extensions, so we’re happy to get your feedback and
> improve the spec accordingly.
>
> > It seems a bit odd to add something like depthData to the canvas 2D
> object which is basically just a holder of pixels.
>
> Do you have a concrete proposal in mind for an improved model? I’d like to
> capture your feedback and open an issue to track this, discuss within the
> TF, but I’d need a bit more details to make the feedback actionable.


Can you explain why you would like to see this integrated with canvas?
It seems that the extension to Canvas 2D could be implemented as a new
standalone object that works in conjunction with a media element (that is
possibly drawn to a canvas)

> It's also unclear from reading the spec what happens when you draw over
> the regions or if this affects canvas tainting.
>
> By drawing over regions, you mean interactions with the drawing model [1]?
> The spec is currently silent on this aspect i.e. this is unspecified.
>
> With respect to canvas tainting, the expectation is the behaviour is
> identical to that of a regular video drawn on a canvas as specified by the
> resource fetch algorithm for a media element (in the HTML spec) and the
> MediaStream-specific modifications/restrictions (spec’d in getUserMedia
> [2]).
>

The model states that this should taint the canvas [1].
Is it possible today to draw a local video stream from your camera to a 2d
context and read the pixels?


> We’ll open new issues for these after getting your clarifications.
>

1:
http://www.w3.org/TR/html5/embedded-content-0.html#concept-media-load-resource