Intent to Implement: CanvasPattern.setTransform

[pav...@samsung.com]

Contact emails

pav...@samsung.com, vive...@samsung.com

 

Spec

http://www.whatwg.org/specs/web-apps/current-work/multipage/the-canvas-element.html#dom-canvaspattern-settransform

Summary

pattern . setTransform(transform)

Sets the transformation matrix that will be used when rendering the pattern during a fill or stroke painting operation.

When the setTransform() method is invoked on the pattern, the user agent must replace the pattern's transformation matrix with the one described by the SVGMatrix object provided as an argument to the method.

 

Compatibility Risk

Low.

 

Ongoing technical constraints

None.

 

Will this feature be supported on all five Blink platforms?

Yes.

 

OWP launch tracking bug?

http://code.google.com/p/chromium/issues/detail?id=289572

 

Link to entry on the feature dashboard

http://www.chromestatus.com/features/5100084685438976

 

Requesting approval to ship?

No

- Pavan

[Dirk Schulze]

lgtm. I think this is a sane idea.

On Sep 4, 2014, at 8:42 AM, pav...@samsung.com wrote:

> Contact emails
> pav...@samsung.com, vive...@samsung.com
>
> Spec
> http://www.whatwg.org/specs/web-apps/current-work/multipage/the-canvas-element.html#dom-canvaspattern-settransform
>
> Summary
> pattern . setTransform(transform)
> Sets the transformation matrix that will be used when rendering the pattern during a fill or stroke painting operation.
>
> When the setTransform() method is invoked on the pattern, the user agent must replace the pattern's transformation matrix with the one described by the SVGMatrix object provided as an argument to the method.

What do you plan to do if the passed matrix is singular? In other cases we silently return earlier and the method does nothing (like addPath). It is not specified in the WHATWG spec, but browsers across the board agreed on this behavior for addPath. I think it makes sense to do it for patterns as well.

Greetings,
Dirk

[Rik Cabanier]

I didn't see this in the dashboard. Maybe this feature is too small for its own entry?

  

Requesting approval to ship?

No

This sound good to implement. Firefox implemented this recently: https://bugzilla.mozilla.org/show_bug.cgi?id=1019257

The only strange thing is that you have to create an SVG element to create a matrix, ie:

var svg = document.getElementById("svg1");

var mtx = svg1.createSVGMatrix();

pat.setTransform(mtx.rotate(-45).scale(0.1));

Introducing DOMMatrix will fix this. 

[Rik Cabanier]

On Thu, Sep 4, 2014 at 12:16 AM, Dirk Schulze <dsch...@chromium.org> wrote:

lgtm. I think this is a sane idea.

On Sep 4, 2014, at 8:42 AM, pav...@samsung.com wrote:

> Contact emails
> pav...@samsung.com, vive...@samsung.com
>
> Spec
> http://www.whatwg.org/specs/web-apps/current-work/multipage/the-canvas-element.html#dom-canvaspattern-settransform
>
> Summary
> pattern . setTransform(transform)
> Sets the transformation matrix that will be used when rendering the pattern during a fill or stroke painting operation.
>
> When the setTransform() method is invoked on the pattern, the user agent must replace the pattern's transformation matrix with the one described by the SVGMatrix object provided as an argument to the method.

What do you plan to do if the passed matrix is singular? In other cases we silently return earlier and the method does nothing (like addPath). It is not specified in the WHATWG spec, but browsers across the board agreed on this behavior for addPath. I think it makes sense to do it for patterns as well.

I agree. 

Browsers are currently split on this. 

FF and IE are trying to follow the spec up to a point. WK and blink consistently return early if the matrix is singular. This is a more sane behavior and also faster. 

 

>
>
> Compatibility Risk
> Low.
>
> Ongoing technical constraints
> None.
>
> Will this feature be supported on all five Blink platforms?
> Yes.
>
> OWP launch tracking bug?
> http://code.google.com/p/chromium/issues/detail?id=289572
>
> Link to entry on the feature dashboard
> http://www.chromestatus.com/features/5100084685438976
>
> Requesting approval to ship?
> No
>
> - Pavan

To unsubscribe from this group and stop receiving emails from it, send an email to blink-dev+...@chromium.org.

[ju...@google.com]

I am not an API owner, but speaking as a reviewer and maintainer of 2D canvas, I support this addition. I think it is low risk, lightweight and useful.