1295404 - remove requiredFeatures from SVGTests

Status: RESOLVED FIXED

(Core :: SVG, defect, P3)

Description

[Robert Longson [:longsonr]]

They've been removed from SVG 2

Comment 1

[Robert Longson [:longsonr]]

Attachment: WIP - still needs various tests to be fixed or removed

Comment 2

[Robert Longson [:longsonr]]

Comment on attachment 8781350 [details] [diff] [review]
WIP - still needs various tests to be fixed or removed

Should I remove hasExtension() from SVGTests here or in another bug?

Comment 3

[Cameron McCormack (:heycam)]

Comment on attachment 8781350 [details] [diff] [review]
WIP - still needs various tests to be fixed or removed

Review of attachment 8781350 [details] [diff] [review]:
-----------------------------------------------------------------

I found bug 1295404 for removing hasExtension(), but I guess I overlooked sending that mail.  In that bug you mentioned we should do this removal behind a pref, presumably because we would be the first ones removing this.  (I just checked, and Chrome/Safari/Edge all still expose SVGTests.requiredFeatures.)

From foolip's comment in https://bugs.chromium.org/p/chromium/issues/detail?id=635420 it sounds like the Blink use counter is sufficiently low for SVGTests in general.  And my gut feeling is that it's safe.  But if you think we should still put this behind a pref for now, then I'm still in favour of making hasFeature() just return true all the time (which would solve bug 1216893's issue).  We should probably still send an intent to unimplement, either way.

Comment 4

[Robert Longson [:longsonr]]

bug 1295404 is this bug.

I think I'll morph this into HasFeature to return true. Then the other bug (whatever it is) can remove it.

Comment 5

[Cameron McCormack (:heycam)]

Oops, bug 1133175 is for hasExtension.

Comment 6

[Robert Longson [:longsonr]]

Attachment: updated patch

Comment 7

[Robert Longson [:longsonr]]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=fc90734bff62182c7e35bd867841ae6abd5669c7

Comment 8

[violet.bugreport]

Chrome has already removed all interfaces in svg/historical.html: https://wpt.fyi/results/svg/historical.html

We probably should also remove the remaining 3 interfaces.

Comment 9

[Robert Longson [:longsonr]]

Well I'm not sure...
useCurrentView is actually pretty useful to us for our own tests. I'd rather leave it as it's hard to test fragment identifiers otherwise.
getTransformToElement is pretty harmless, there's a polyfill somewhere that adds it back.
w3c are still talking about SVGZoomAndPan - https://lists.w3.org/Archives/Public/www-svg/2019Jun/0014.html maybe that will end up being removed altogether since that one really is useless as we don't send any zoom/pan events.

So I'd rather fix other things really.

Comment 10

[Robert Longson [:longsonr]]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=25a475d76f4567c2169fbe9f329755039831cd2b

Comment 11

[Robert Longson [:longsonr]]

Attachment: 1295404.txt

Comment 12

[Robert Longson [:longsonr]]

Attachment: 1295404.txt

Comment 13

[Robert Longson [:longsonr]]

https://treeherder.mozilla.org/#/jobs?repo=try&revision=95145279892d7509ffe91029bef92ca34e7659ab

Comment 14

[Robert Longson [:longsonr]]

Attachment: Bug 1295404 - Remove requiredFeatures from SVGTests r=dholbert r=bzbarsky

Comment 15

[Robert Longson [:longsonr]]

https://www.w3.org/TR/SVG2/struct.html#ConditionalProcessingOverview

Previous versions of SVG included a third conditional processing attribute, requiredFeatures. This was intended to allow authors to provide fallback behavior for user agents that only implemented parts of the SVG specification. Unfortunately, poor specification and implementation of this attribute made it unreliable as a test of feature support.

Chrome removed it some time ago.

bzbarsky for the webidl change

Comment 16

[Boris Zbarsky [:bzbarsky]]

Chrome removed it some time ago.

Does Safari support it?

Comment 17

[Boris Zbarsky [:bzbarsky]]

Nevermind, found https://wpt.fyi/results/svg/historical.html?label=master&label=experimental&q=svg%2Fhistorical.html which shows Safari supports it.

Comment 18

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/ba32151f8fc1
Remove requiredFeatures from SVGTests r=dholbert r=bzbarsky

Comment 19

[Cristina Coroiu [:ccoroiu]]

https://hg.mozilla.org/mozilla-central/rev/ba32151f8fc1

Comment 20

[Kohei Yoshino [:kohei]]

Posted site compatibility note: https://www.fxsitecompat.dev/en-CA/docs/2019/requiredfeatures-has-been-removed-from-some-svg-interfaces/