Re: [Ayatana] Idea for improving visibility of running applications

[Evan Huus <eapache@xxxxxxxxx>]

This debate is getting rather heated, and I'm getting quoted and
rephrased a fair bit. Let me take this opportunity to clarify both
what I actually think:

Unity is supposed to be considered 'application-centric'. In a truly
application-centric system (see iPhones, for example) there is
absolutely no immediate indication whether an app is running or not.
If you want to do X, you click on the X launcher, and the app is
presented. The user shouldn't have to care.

This model isn't for everyone. Power-users who tend to have a large
number of the same kind of window open (such as Terminals), and people
coming from older versions of Windows both prefer a window-centric
design. Unity currently has those little white arrows as a bow towards
a window-centric model, but nothing more.

Niklas, and a lot of other people who have made suggestions in other
threads, seem to want to move Unity more towards a window-centric
design, often because they fall into one of the use cases where that
actually makes sense. I disagree with a window-centric model by
default. Although I have no studies to quote here, it seems to be
gospel among interface designers (ie Apple, Canonical Design Team)
that an application-centric model is better for most. I'm fine with
that, and anecdotal evidence supports it fairly strongly. So an
app-centric model by default.

However, many people do need or prefer a window-centric model.
Niklas's current mockup is a good step in that direction (albeit with
some problems) and I would be very interested in trying a working
version of it. It is one potential way towards a window-centric Unity.

In the end, I think we'll end up with a single option for this whole
mess: App or Window model. App model by default, which is Unity
more-or-less as it currently stands. Window model can be selected by
anyone who actually wants/needs it. It still needs to go through a
proper design process but again, Niklas's idea is a big step in the
right direction.

I hope that's a little clearer. It boils down to two things:

1) We need an interface for those who fall into one of the
window-centric use cases. Unity as it stands is not that interface,
leaving a lot of users out in the cold.

2) Moving the default Unity towards that interface is a bad idea,
since that makes it harder for the rest (and majority) of users who
deal better with an application-centric interface.

Cheers,
Evan