Re: [css3-transitions] starting and reversing animations

On Jan 29, 2010, at 11:01 AM, Tab Atkins Jr. wrote:

> On Fri, Jan 29, 2010 at 12:57 PM, Brad Kemper <brad.kemper@gmail.com> wrote:
>> On Jan 29, 2010, at 9:27 AM, Tab Atkins Jr. wrote:
>>> Yeah, setting transition on "div" still applies to all the subsequent
>>> rules, per normal cascading.
>> 
>> Exactly. So what is unclear? When the state changes and a new rule with its 'transition' property suddenly applies, then that transition begins (or at least its transition-delay begins).
> 
> The issue is that, mentally, transitioning activates *before* the new
> rule takes effect, while the element is in a quasi-state halfway
> between the two.  A new value is coming in, but it has to wait and
> transition gradually instead.
> 
> In my head, then, it makes sense that transitions work based on the
> original value of the transition property, since the new value hasn't
> "arrived" yet.
> 
> Nothing wrong with it using the "destination" state instead, but
> either way seems pretty reasonable to conclude.

If you say so, but I think it would be very unintuitive the other way. When my cursor is hovering over an element, I expect to be in the ':hover' state, not in a ':waiting-to-hover-while-some-other-rule-finishes-up' state. And the 'transition' property could be expected to affect how and when some of the properties in the current rule take effect, but it would be strange indeed for it to affect how and when properties in a different rule are shown (except in the case where it is "undoing" a transition that is interrupted in mid-stream and needs to reverse itself back into the other rule's values).