RE: when do transitions occur?

Ok.  For clarification purposes...

If background-image *was* currently animatable, would your answer be A5? 

-----Original Message-----
From: L. David Baron [mailto:dbaron@dbaron.org] 
Sent: Wednesday, September 14, 2011 3:05 PM
To: Brian Manthos
Cc: Tab Atkins Jr.; Jennifer Yu; www-style@w3.org
Subject: Re: when do transitions occur?

On Wednesday 2011-09-14 19:45 +0000, Brian Manthos wrote:
> Wow, example fail.
> 
> I meant to say...
> 
> Example A:
>  from background: blue;
>  to background: none;
> 
> (So that color and image are involved.)

The answer is actually still the same.  background-color still
changes from blue to transparent, and transitions.  background-image
isn't currently an animatable property, so it doesn't transition.

-David

> 
> -----Original Message-----
> From: L. David Baron [mailto:dbaron@dbaron.org] 
> Sent: Wednesday, September 14, 2011 12:06 PM
> To: Brian Manthos
> Cc: Tab Atkins Jr.; Jennifer Yu; www-style@w3.org
> Subject: Re: when do transitions occur?
> 
> On Wednesday 2011-09-14 17:48 +0000, Brian Manthos wrote:
> > 1. Are there any scenarios where an event should fire for the shorthand due to transitioning?
> 
> No (in my opinion).
> 
> > 2. Or are component longhands the only events that ever fire for transitioning?
> 
> Yes.
> 
> > 3. If 2, should longhands that are unchanged in value but members of a shorthand that is explicitly transitioned be excluded from the events?
> 
> Yes.  Transitions are triggered by changes in computed value (see
> section 3, Starting of Transitions).  It doesn't matter whether
> these changes occurred because a shorthand was changed or because
> the individual longhand properties were changed.  Only longhand
> properties have computed values, and only those whose computed
> values changed should have events.
> 
> 
> > Example A:
> >  from background: blue;
> >  to background: transparent;
> > A1: background
> > A2: background-color
> > A3: background-image
> > A4: background-* (all 8)
> > A5: A2 and A3
> > A6: A1, A2, and A3
> > A7: A1 and A4
> > A8: something else
> > 
> > My evaluation of each event set option:
> > A1: it's the author-aligned delta
> > A2: (doesn't make sense in isolation)
> > A3: (doesn't make sense in isolation)
> > A4: all longhands are involved in the delta, no culling of unchanged properties
> > A5: all longhands are involved in the delta, only include changed properties
> > A6: same reasons as A1 and A5, both apply
> > A7: same reasons as A1 and A4, both apply
> > A8: (tbd)
> 
> I think it's clearly A2, since background-color is the only property
> whose computed value changed.  (Also, many of the background-*
> properties aren't animatable at all.)
> 
> -David
> 
> -- 
> 𝄞   L. David Baron                         http://dbaron.org/   𝄂
> 𝄢   Mozilla Corporation               http://www.mozilla.com/   𝄂
> 

-- 
𝄞   L. David Baron                         http://dbaron.org/   𝄂
𝄢   Mozilla Corporation               http://www.mozilla.com/   𝄂

Received on Wednesday, 14 September 2011 22:13:42 UTC