Task Launcher - mockups

[Thomas Pfeiffer]

On Wednesday 30 January 2013 22:48:52 Marco Martin wrote:
> > Why do you expect integrating the task launcher into the shell to have
> > such
> > a big performance/resource impact? From my technical n00b perspective, it
> > is not that much different from just having more launchers in there,
> > presented in a more sophisticated way...
> > If my specific UI idea could only be implemented in a resource-heavy way,
> > I'd rather try to find a less resource-heavy UI then outsourcing it into a
> > separate application. I'm sure we can find a way :)
> 
> don't know, the application grid is a single view with a single simple
> model, is pretty easy to keep the created objects fairly small, i see the
> task creation ui being a bit more complicated, all of this has a price...
> 
> maybe is possible to keep most of this lazy created enough, don't know

If we agree that we conceptually want a completely task-based Launcher, than 
I'm confident we can find a technically feasible solution together. I know that 
developers see the potential problems first and that's a good thing, but in my 
professional career so far I learned not to be discouraged by initial "That's 
not possible!" or "That would be crazy!" reactions. In the end usually 
together a solution can be found that everyone is at least comfortable with. 
And from my experience with Plasma Active so far, I know that here it's no 
different (and I've seen from things like e.g. the improvements in thumbnailing 
in Reader that you guys can do magic :) )
 
> > > 2-3 horizontal scrolling strips would be already better i think...
> > 
> > I can't picture what you mean by "2-3 horizontal scrolling strips". Do you
> > mean only using 2-3 categories overall?
> 
> yeah, basically trying to arrange it as horizontal scrolling lines of icons

The problem with that is that this would limit us to very few top-level 
verbs/categories, which may result in not finding a suitable category for many 
tasks. That's why I'd like to avoid any hard limits on the number of verbs. 

If all else fails, we can still use the first scrolling row for categories and 
then only show the tasks in the selected category in the second row, but that 
means that users will always have to select a category first, of course.

But well, the central point of my idea was the "Everything is a task" part. 
The rest is design details which we can iterate on in the next step(s) in case 
we agree on it as the general route to go.