Auto-launching of applications

Mark Shuttleworth mark at
Wed Feb 25 09:40:04 GMT 2009

Chow Loong Jin wrote:
> If there must be a solution for users who have poor icon observation
> skills, then let it be one that does not irritate those who do have
> proper icon observation skills. 
Well, yes there is :-)

First, we will do *fewer* icons. They will be clearer, bolder, easier to
see in different lighting and sighting circumstances, consistently
positioned, and pixel-perfect. By doing fewer of them, we improve the
quality and the recognisability.

Second, we will strongly deprecate arbitrary additions to the panel.
That means, over time, we will go as far as we can to ensuring that
nothing in main adds icons to the panel, except those that have been
through a rigorous debate and discussion. We know there are several
strong use cases for such icons - you mention email, there are also
network, bluetooth, volume, time and a few others. Each of these will be
"small but perfectly formed".

Third, we will work hard on perfecting the alternatives - non-icon tools
that applications have to draw attention to a decision that the user
has. We have stated that there are four:

  1. ephemeral notifications,
  2. persistent indicators where there has been a design review and
consensus on the need for a category to be created,
  3. opening a decision / action window in the background and calling
for attention via the window manager, and
  4. opening a system modal window in the foreground (obviously, this is
only for total emergencies)

Right now, the implementations of (3) are of varying quality depending
on your WM. So we will be doing substantial amounts of work there. Of
course, WM preferences vary, but we expect people will choose a WM that
does this elegantly *if they care about it*, and therefor we will trust
the WM to be acceptable, especially once we have a good default
behaviour in Ubuntu.

There are some additional twists and innovations that flesh out the
picture, which I won't go into here.

We really need a good list to discuss this on, David Barth is setting
one up and we'll announce it here when it's done. In the meanwhile,
please tag bugs that are related to notifications "notifications" so we
have our arms around all the sharp edges for 9.04 and can address the
most irritating bits, of which update handling is currently the prime
example :-)

I promise we will smooth the experience of update management - it's one
of the very best features of Ubuntu (hey, I invest a lot in giving those
updates away for free to the world, I really want people to love them
not hate them!) and apologise that the initial landing wasn't smooth -
in particular for our core dev team and beta / development release
users, who of course are a key part of the community. Since these
changes had been discussed at UDS, we didn't expect such a big surprise,
but we should have announced the changes here when they landed and
provided a forum for this discussion.

So, thanks to Bruce Cowan for raising the thread.

We had a detailed call last night between Canonical folks in the Ubuntu
Desktop team, the design team and the upstream DX engineering team that
did the notifications work, and MPT will write up that discussion.
Martin Pitt, Sebastien Bacher and Michael Vogt did a great job
representing the development community there, so I am confident we will
take the key rough edges off the new work for 9.04, while recognising
there's room for improvement in 9.10.

Finally (sorry for the long email!), I believe this work is very
important for Ubuntu, and for desktop Linux in general. Like the desktop
work going on in KDE 4 and GNOME 3, we want to be part of the movement
to bring a dramatically better user experience to desktop Linux, so that
tens of millions of people who DON'T know about the GPL today, can learn
about it from the perspective of "wow, this amazing software is ALSO
licensed in this amazing way!".

Being something of a make-the-world-better freak, I'm not afraid of the
idea that for something to get better, it might have to hurt a little
along the way. True of medicine, investing, exercise and probably love,
too! What I'm saying is that we might break things along the way,
sometimes accidentally, sometimes deliberately. I'm asking for the
support of the development community in Ubuntu and upstream to let us
innovate. Some things will work, others won't. We don't experiment
lightly, and we're humble in the face of real evidence that we've made a
mistake. I change my mind, folks who've worked with me can vouch for it.
So should everyone on this list, given evidence. So please continue to
provide patches, bugs, feedback, suggestions, code, drawings,
animations, inspiration, coffee, support, cheers, bouquets and the
occasional brickbat. But deliver the brickbats in the Ubuntu way.

-------------- next part --------------
An HTML attachment was scrubbed...

More information about the ubuntu-devel mailing list