[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: "animations" in oxygen style
From:       Hugo Pereira Da Costa <hugo.pereira () free ! fr>
Date:       2009-10-23 4:37:49
Message-ID: 4AE1331D.7020905 () free ! fr
[Download RAW message or body]

Aaron J. Seigo wrote:
> On October 22, 2009, John Tapsell wrote:
>   
>> I see no reason why the same doesn't apply here.  You need instant
>> response and highlighting.  You can then animate something
>> additionally on top if you wish, but the initial highlighting must be
>> near instant (e.g. around 50ms).
>>     
>
> yes, there's a difference is between visual feedback and result-on-action. and 
> you'll see this is exactly what we do in the plasma tasks widget, btw.
>
> you can also screw it up by starting an animation that is the result of 
> passive focus (e.g. hover highlight) too quickly giving it a feeling of being 
> too jumpy.
>
>   
In fact Lucas (lmurray) and I had this discussion about the windeco 
"shadow" transition. (He complaining that it looked as if the active 
window glow would start with a delay after you actually activate the 
window). The conclusion was that the first "frame" of your animation 
should not be the inactive state, but some (well tuned) fraction of the 
final state. Say that if your animations has four/five frames (which is 
typically what we get for active/inactive oxygen shadow/glow), the first 
one should be about 20% of the final state.
I think the result is quite satisfying IMHO.

To Aaron: the above is actually not implemented in nitrogen style, which 
contributes partly to the slowness you mention.

(and then also you want to optimize your code to minimize the delay 
between the user taking action and the actual animation to start, but 
this is true even without animations).

(btw with QTimeLine and QAnimation, its not you to decide how many 
frames you get. You just set the total animation length and Qt tries to 
cope with it. Big improvement wrt old QTimer).


> ------------------------------------------------------------------------
>
>  
>   
>>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<
>>>       

 
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<
[prev in list] [next in list] [prev in thread] [next in thread]