order of values in background test

[karlcow]

Should we put the white on line 38 in the first place.

Rationale:
That would test the serialization of multiple background, and not the issue about canonical order in serialization which is being debated.

wpt/css/css-backgrounds/parsing/background-shorthand-serialization.html

Lines 35 to 40 in 1ebef47

	test((t) => {
	element.style = `background: url(/favicon.ico) top left no-repeat,
	url(/favicon.ico) center / 100% 100% no-repeat,
	url(/favicon.ico) white;`;
	assert_equals(element.style.background , 'url("/favicon.ico") left top no-repeat, url("/favicon.ico") center center / 100% 100% no-repeat, white url("/favicon.ico")');
	}, "multiple backgrounds with varying values");

[karlcow]

Context: https://wpt.fyi/results/css/css-backgrounds/parsing/background-shorthand-serialization.html?label=experimental&label=master&aligned

[karlcow]

Ah maybe it was intentional
if I look at the pull request review
https://phabricator.services.mozilla.com/D155077

@emilio would it be better to test those two separately? aka the canonical order and the way background is supposed to work.

[emilio]

That was done in https://bugzilla.mozilla.org/show_bug.cgi?id=743392, but actually given <final-bg-layer> has the <background-color> before rather than after (https://drafts.csswg.org/css-backgrounds/#typedef-final-bg-layer) I think WebKit and Blink follow the spec and Gecko is wrong, so we should probably fix Gecko.

[emilio]

I filed https://bugzilla.mozilla.org/show_bug.cgi?id=1788582 for this.

[emilio]

Actually it's the other way around, Gecko is right. Why testing it separately?

[emilio]

I think the order is clear in the spec, but if you disagree please open a spec issue stating the case?

[karlcow]

Actually it's the other way around, Gecko is right. Why testing it separately?

I think the order is right too.

What I was thinking is isolating tests to make sure to avoid mixing two types of constraints. :) but it's ok as-is too. Just need to improve the serialization of WebKit. Thanks for the comment.