Re: [css3-transitions] Interpolating 'other' value types

On Wed, Dec 1, 2010 at 11:43 AM, Chris Marrin <cmarrin@apple.com> wrote:
> On Nov 30, 2010, at 3:25 PM, Simon Fraser wrote:
>> On Nov 30, 2010, at 2:49 PM, Tab Atkins Jr. wrote:
>>
>>> Right now, the Transitions spec still says that only certain types of
>>> values can be transitioned.  Other types of values are ignored (merely
>>> kept in the transition-property list to maintain relative list
>>> positions).
>>>
>>> I thought we'd agreed to let any property without specified transition
>>> rules transition, with the rule that it's equal to the start value at
>>> t=0 and the end value at t>0 (or the opposite, whatever)?  Part of the
>>> reason we introduced the step timing-function was specifically to help
>>> control this case, if I recall correctly.
>>>
>>> I assume this is just a temporary omission, and the necessary edits
>>> will be made in time?
>>
>> David Baron suggested that transition-delay should apply to all properties.
>> I said that I supported that in principle, but was worried about implementation
>> performance implications, given that the default for transition-property is "all".
>
> Maybe we've discussed this in the misty past, but why is the default 'all'? Wouldn't 'none' avoid expensive (in terms of processing) mistakes?

This is noted in the current spec - using transition-duration:0 and
transition-property:all by default is apparently somewhat
aesthetically prettier than transition-duration:250ms (or some
reasonable default) and transition-property:none.

There's no technical reason for it, though.

~TJ