Re: [css-transitions] Normalization of steps()

On Mon, Apr 11, 2016 at 4:39 PM, Brian Birtles <bbirtles@mozilla.com> wrote:
> On 2016/04/08 1:37, Tab Atkins Jr. wrote:
>>
>> On Wed, Apr 6, 2016 at 6:39 PM, Shane Stephens <shans@google.com> wrote:
>>>
>>> In Chrome, Edge and probably Safari, a specified style of
>>> transition-timing-function: steps(3)
>>> normalizes to a computed value of
>>> transition-timing-function: steps(3, end)
>>>
>>> In Firefox, it doesn't. Timing function strings are always maintained
>>> unaltered.
>>>
>>> We should specify something, then conform to it! I don't have a strong
>>> opinion what, but my irony muscle is twitching slightly because Edge
>>> seems
>>> to do the opposite here to what they do with calc :)
>>
>>
>> The standard serialization strategy is to serialize with the minimum
>> amount necessary to accurately represent the value. If omitting some
>> component is the same as specifying a particular keyword, then all
>> values with that keyword serialize with it omitted, etc.
>
>
> Is this something that is defined somewhere in a generic fashion that we can
> assume in other specs, or do we need to update CSS transitions to explicitly
> specify the serialization here?

I thought that CSSOM defined this, but I can't find it now.  zcorpan,
do you think this needs to be added to OM?

~TJ

Received on Monday, 11 April 2016 23:46:05 UTC