Re: [css3-transitions] transitioning to/from auto value

On Apr 28, 2010, at 2:54 AM, Boris Zbarsky wrote:

> On 4/28/10 2:29 AM, Brad Kemper wrote:
>> I don't understand. The transition has to start sometime.
> 
> Yes.
> 
>> So the first thing it does is get the offsetHeight to transition from.
> 
> Which might well give different results in different browsers, or even in the same browser depending on memory layout.  At least if implemented the way you just described.
> 
> For example, if a browser batches up style changes and processes the batch all at once, the transition will start partway through the processing of the batch.  At this point, some of the pending style changes have been processed and some haven't; which ones have or haven't may well be effectively random.  

In this scenario, even without transitions wouldn't there be the possibility that the attribute that was just set to auto to bounce around?  It seems like users would not like that either.

> Is the idea that all change processing needs to complete before sampling the "auto" height the transition should run to?  Something else?
> 
>> It may not be perfect, but it would be better than me trying to set a fixed height on everything I wanted to transition this way (and I could still do that anyway, if I liked it better).
> 
> Honestly, my feeling on this is the same as the webkit developer who commented earlier in this thread.  The moment such a thing is implemented, people will start demanding that various corner cases "work correctly", and work the same way in different browsers, etc.  It sounds like a huge time-and-effort-sink...

Seems like there is a middle ground somewhere.  Define the simple case where only one attribute is in flux at the time and it is going from fixed to auto or auto to fixed and declare that anything more complex the author will get a strongly worded letter from the UN :-)

Received on Wednesday, 28 April 2010 13:35:42 UTC