Should this be implemented?

This has been in the HTML spec for a while now (see https://html.spec.whatwg.org/multipage/canvas.html#2dcontext),
so we should probably go ahead and implement it.

I personally don't think we should be adding dummy properties to the dictionary that we don't actually support. Safari also only ships a subset of properties.

Backed out for causing wpt failures related to CanvasRenderingContext2D.

• Push with failures

• Failure log

• Backout link

[task 2023-07-26T00:20:11.580Z] 00:20:11     INFO - TEST-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D interface: existence and properties of interface prototype object's @@unscopables property 
[task 2023-07-26T00:20:11.580Z] 00:20:11     INFO - TEST-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D interface: attribute canvas 
[task 2023-07-26T00:20:11.580Z] 00:20:11     INFO - TEST-UNEXPECTED-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D interface: operation getContextAttributes() - expected FAIL
[task 2023-07-26T00:20:11.581Z] 00:20:11     INFO - TEST-INFO | expected FAIL
[task 2023-07-26T00:20:11.611Z] 00:20:11     INFO - ............................................................................
[task 2023-07-26T00:20:11.611Z] 00:20:11     INFO - TEST-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D interface: operation quadraticCurveTo(unrestricted double, unrestricted double, unrestricted double, unrestricted double) 
[task 2023-07-26T00:20:11.611Z] 00:20:11     INFO - TEST-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D interface: operation bezierCurveTo(unrestricted double, unrestricted double, unrestricted double, unrestricted double, unrestricted double, unrestricted double) 
[task 2023-07-26T00:20:11.611Z] 00:20:11     INFO - TEST-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D interface: operation arcTo(unrestricted double, unrestricted double, unrestricted double, unrestricted double, unrestricted double) 
[task 2023-07-26T00:20:11.611Z] 00:20:11     INFO - TEST-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D interface: operation rect(unrestricted double, unrestricted double, unrestricted double, unrestricted double) 
[task 2023-07-26T00:20:11.611Z] 00:20:11     INFO - TEST-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D interface: operation roundRect(unrestricted double, unrestricted double, unrestricted double, unrestricted double, optional (unrestricted double or DOMPointInit or sequence<(unrestricted double or DOMPointInit)>)) 
[task 2023-07-26T00:20:11.611Z] 00:20:11     INFO - TEST-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D interface: operation arc(unrestricted double, unrestricted double, unrestricted double, unrestricted double, unrestricted double, optional boolean) 
[task 2023-07-26T00:20:11.611Z] 00:20:11     INFO - TEST-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D interface: operation ellipse(unrestricted double, unrestricted double, unrestricted double, unrestricted double, unrestricted double, unrestricted double, unrestricted double, optional boolean) 
[task 2023-07-26T00:20:11.611Z] 00:20:11     INFO - TEST-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D must be primary interface of document.createElement("canvas").getContext("2d") 
[task 2023-07-26T00:20:11.612Z] 00:20:11     INFO - TEST-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | Stringification of document.createElement("canvas").getContext("2d") 
[task 2023-07-26T00:20:11.612Z] 00:20:11     INFO - TEST-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D interface: document.createElement("canvas").getContext("2d") must inherit property "canvas" with the proper type 
[task 2023-07-26T00:20:11.612Z] 00:20:11     INFO - TEST-UNEXPECTED-PASS | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | CanvasRenderingContext2D interface: document.createElement("canvas").getContext("2d") must inherit property "getContextAttributes()" with the proper type - expected FAIL
[task 2023-07-26T00:20:11.612Z] 00:20:11     INFO - TEST-INFO | expected FAIL
[task 2023-07-26T00:20:11.918Z] 00:20:11     INFO - ..............................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................
[task 2023-07-26T00:20:11.918Z] 00:20:11     INFO - TEST-OK | /html/dom/idlharness.https.html?exclude=(Document|Window|HTML.*) | took 4467ms

(In reply to Tom S [:evilpie] from comment #8)

I personally don't think we should be adding dummy properties to the dictionary that we don't actually support. Safari also only ships a subset of properties.

I wondered about that, but this seems to parallel how Chrome is behaving with respect to desynchronized, which it always includes in the returned dictionary, even though it doesn't appear to be supported. (Trying to set it to true at context-creation time doesn't make any difference; it remains false in the returned attributes.)

I think desynchronized is a bit of a special case considering that HTML spec doesn't actually require anything to happen when it's enabled and it's more of a hint:

When a CanvasRenderingContext2D object's desynchronized is true, then the user agent may optimize the rendering of the canvas to reduce the latency

https://html.spec.whatwg.org/multipage/canvas.html#concept-canvas-desynchronized

That seems quite different to something like colorSpace, which I presume would actually change how the canvas looks to the user.

https://hg.mozilla.org/mozilla-central/rev/c632111f2a2c

Should we add this to the Fx117 relnotes? Please nominate by setting relnote-firefox? if so.

It seems minor enough that I'm not sure it's relnote-worthy, IMO.

FYI FF117 docs work for this can be tracked in https://github.com/mdn/content/issues/28283

I personally don't think we should be adding dummy properties to the dictionary that we don't actually support. Safari also only ships a subset of properties.

FWIW Does it matter in this case because even though FF doesn't support the attributes being set, the result returned by this method is correct - i.e. colorSpace defaults to srgb, which is what the color space will be, desynchronized is false if you don't support this attribute, and so on?

emphasis mine, mdn doc says

Depending on the attributes supported by the browser, the log below should display a string that looks something like: {alpha: false, colorSpace: 'srgb', desynchronized: false, willReadFrequently: false}

RFP (resistFingerprinting), in Bug 1422237, always returns srgb for color-gamut. So my question is, is the colorSpace here always srgb or does it depend on the device's color-gamut, i.e do we need to take RFP into consideration ?