[css-transforms-2] Interpolation of perspective.

[emilio]

I think it's really weird that perspective the transform function gets interpolated differently than perspective the property.

https://drafts.csswg.org/css-transforms-2/#propdef-perspective says:

Animatable: as length

However for the transform function it says:

The transform functions matrix(), matrix3d() and perspective() get converted into 4x4 matrices first and interpolated as defined in section Interpolation of Matrices afterwards.

I just fixed the perspective() function interpolation in Gecko in https://bugzilla.mozilla.org/show_bug.cgi?id=1488414. I haven't tested what other browsers do for the perspective property yet, but I wanted to leave this filed because it seems that if this inconsistency is intended, it should be at least be clarified and explicitly pointed out.

[emilio]

cc @hiikezoe

[emilio]

cc @dirkschulze

[hiikezoe]

https://codepen.io/anon/pen/eLEPWm
Here is an example for 'transform: perspective' and perspective interpolations.

As for the perspective property, Chrome, Edge, Firefox and Safari do interpolate as length (i.e. linear interpolation)

As for the perspective function in transform property, Chrome and Safari do the way what the spec defines (thought I haven't verified the calculation result), Firefox does linear interpolation (but emilio fixed it in the bug). Edge shows somewhat different result.

[css-meeting-bot]

The CSS Working Group just discussed Interpolation of perspective., and agreed to the following:

• RESOLVED: Close https://github.com/w3c/csswg-drafts/issues/3084 no change

The full IRC log of that discussion

<dael> Topic: Interpolation of perspective.
<dael> github: https://github.com//issues/3084
<dael> smfr: emilio noticed perspective property and transform function interpolate different. property is just length value. Function is matrix interpolation. Gives different results. I think he's right they should be same
<dael> smfr: I'm not sure how they differ visually. WOuld be interesting to see if look really different or slightly different
<dael> AmeliaBR: prob different b/c one creates an inverse
<dael> Rossen_: Isn't the codepen in the issue enough?
<dael> smfr: That's just snapshotting one point, not showing animation
<dael> AmeliaBR: Converting perspective funct to matrix isn't consistant with other transforms. rotate we interpolate to a single value
<Rossen_> would this test case be enough? https://bug1488414.bmoattachments.org/attachment.cgi?id=9006258
<dael> AmeliaBR: My preference would be to follow that patter and interpolate the param to the perspective function rather then a matrix. Unless there's an impl reason why that can't work. Apparently FF used to do that and emilio changed to match spec
<dael> AmeliaBR: otoh if we now have conforming impl maybe we stick with that
<dael> Rossen_: Do we have conforming implementations?
<dael> smfr: Chrome and Safari match. Not FF. I think
<dael> mwoodrow: Do you know what Chrome and Safari are doing?
<dael> smfr: Not offhand
<Rossen_> q?
<dael> AmeliaBR: According to issue they do what spec does and generate transform matrix from funt and interpolate. According to linked FF bug which was fixed in FF63 FF is now doing same
<smfr> q-
<heycam> q- smfr:
<heycam> q- Not
<heycam> q- off
<dael> mwoodrow: So we do have matching impl now
<dael> Rossen_: Is proposed to punt on this and resolve no change?
<dael> AmeliaBR: Not used enough to insist on a change
<dael> Rossen_: Objections to resolve close no change?
<dael> smfr: Should we do this without emilio?
<dael> Rossen_: He can re-open if he wants to
<birtles> it seems weird to me
<dael> RESOLVED: Close https://github.com//issues/3084 no change

[smfr]

Animating test: https://codepen.io/smfr/pen/bGbXPgQ

[dbaron]

The group resolved almost a year and a half ago to close this issue with no change unless @emilio objected. It seems to me it's been long enough for him to object, so we should probably close this now.

[emilio]

Yeah, no objection at all, but I wasn't asked directly and I probably either missed or overlooked that sentence in the minutes, so I didn't know closing this was blocked on me... :)

Thanks for closing this @dbaron :)

[smfr]

See also web-platform-tests/wpt#28036