1310431 - CSS filter whole property will be considered as invalid if one of the function has no parameter

Status: RESOLVED INVALID

(Core :: CSS Parsing and Computation, defect, P3)

Description

[fireattack [:fireattack]]

User Agent: Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/54.0.2840.59 Safari/537.36

Steps to reproduce:

https://jsfiddle.net/muavjbvs/7/




Actual results:

The whole filter is ignored. It says "invalid property value" in dev tool as well.


Expected results:

According to MDN (https://developer.mozilla.org/en-US/docs/Web/CSS/filter#sepia()_2 ): "If the ‘amount’ parameter is missing, a value of 0 is used." So the opacity part should still take effect.

It's not limited to sepia either. Most of (if not all) the functions without parameter will break filter property.

Comment 1

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

In general, it is expected that invalid syntax makes a property value invalid.

I also don't see anything in https://drafts.fxtf.org/filters/#funcdef-filter-sepia that says that the parameter to the filter function can be omitted, although perhaps I'm missing something.

Comment 2

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

(Do other implementations behave the same?)

Comment 3

[fireattack [:fireattack]]

(In reply to David Baron :dbaron: ⌚️UTC-7 from comment #2)
> (Do other implementations behave the same?)

Chrome is not the same at least

Comment 4

[fireattack [:fireattack]]

If it's not a bug, should I edit MDN to reflect by the way?

Comment 5

[Robert Longson [:longsonr]]

I've fixed MDN to match the CSS filters specification. You're not allowed to miss values out. The defaults are only applicable if we need to infer a missing filter for interpolation purposes.

Comment 6

[fireattack [:fireattack]]

Thanks!

Comment 7

[Xidorn Quan [:xidorn] UTC+10]

Ah, my fault. I thougt [] is optional but apparently it only means grouping.

Comment 8

[Astley Chen (inactive)]

Hi Mike, this is something Blink behaves different from gecko for handling CSS filter values, eg. sepia(). So far there is no end-user facing issue reported here, what's your recommendation to deal with this issue ? Report to webcompat.com or just issue a bug on corresponding browser vendor bug management system ?

Comment 9

[Astley Chen (inactive)]

Based on [1], it's a regression bug from Chrome 53. I believe Chrome team will fix it soon.

[1] https://bugs.chromium.org/p/chromium/issues/detail?id=664866#c1

Comment 10

[Mike Taylor [:miketaylr]]

(sounds good!)