RFC: open the spinner directly on touch point

[Thomas Pfeiffer]

>Moreover, I think it would be better (for consistency) to place the spinner in the same place (ie: in the middle of the screen) no matter >where the launcher is. Perhaps we could have a semi transparent overlay (a bit dark) with the spinner in the middle and to the right of it >"Starting: name of app." The overlay would fade away once the application is running. It would also prevent users from tapping more than >once in the same resource/having more than one instance of the same app. As you know, more apps running mean that more system >resources are used, more power consumption, which is translated into less battery.

Spinner in the middle with overlay is exactly what we had and what we abolished because it appeared to block the whole system (but actually didn’t). And I don’t think we should re-introduce it. If a desktop system would prevent any other actions while e.g. launching an application, nobody would want to use it, so why accept it on a tablet? Just because it has a weaker processor? Plasma Active supports multi-tasking, so it should still do so while an application starts.
I think indicating clearly to users that they have to wait for their application to start while allowing them to do something else while waiting is the way to go here. 
One idea that would prevent double-tapping an application on the launcher:
While the application launches, change its icon on the launcher to indicate that it's already launching and disable its tap event. This would also prevent a user from intentionally opening multiple instances of the same application at the same time, but I don't think that is a very common use case anyway.

Cheers,
Thomas

P.S.: I oh so h.a.t.e Outlook! It is completely useless trash when it comes to using mailing lists! That’s the only disadvantage that came with my recent job change: I cannot use a Linux machine with KDE on it anymore and I am forced to use shitty Outlook without a "reply to list" function, without a threaded view and without the possibility to properly format a reply to an HTML message... *grml*