Canonical Voices

What Mark Shuttleworth talks about

Posts tagged with 'panel'

mark

A few months ago we took on the challenge of building a version of Ubuntu for the dual-boot, instant-on market. We wanted to be surfing the web in under 10 seconds, and give people a fantastic web experience. We also wanted it to be possible to upgrade from that limited usage model to a full desktop.

The fruit of that R&D is both a new desktop experience codebase, called Unity, and a range of Light versions of Ubuntu, both netbook and desktop, that are optimised for dual-boot scenarios.

The dual-boot, web-focused use case is sufficiently different from general-purpose desktop usage to warrant a fresh look at the way the desktop is configured. We spent quite a bit of time analyzing screenshots of a couple of hundred different desktop configurations from the current Ubuntu and Kubuntu user base, to see what people used most. We also identified the things that are NOT needed in lightweight dual-boot instant-on offerings. That provided us both with a list of things to focus on and make rich, and a list of things we could leave out.

Instant-on products are generally used in a stateless fashion. These are “get me to the web asap” environments, with no need of heavy local file management. If there is content there, it would be best to think of it as “cloud like” and synchronize it with the local Windows environment, with cloud services and other devices. They are also not environments where people would naturally expect to use a wide range of applications: the web is the key, and there may be a few complementary capabilities like media playback, messaging, games, and the ability to connect to local devices like printers and cameras and pluggable media.

We also learned something interesting from users. It’s not about how fast you appear to boot. It’s about how fast you actually deliver a working web browser and Internet connection. It’s about how fast you have a running system that is responsive to the needs of the user.

Unity: a lightweight netbook interface

There are several driving forces behind the result.

The desktop screenshots we studied showed that people typically have between 3 and 10 launchers on their panels, for rapid access to key applications. We want to preserve that sense of having a few favorite applications that are instantly accessible. Rather than making it equally easy to access any installed application, we assume that almost everybody will run one of a few apps, and they need to switch between those apps and any others which might be running, very easily.

We focused on maximising screen real estate for content. In particular, we focused on maximising the available vertical pixels for web browsing. Netbooks have screens which are wide, but shallow. Notebooks in general are moving to wide screen formats. So vertical space is more precious than horizontal space.

We also want to embrace touch as a first class input. We want people to be able to launch and switch between applications using touch, so the launcher must be finger friendly.

Those constraints and values lead us to a new shape for the desktop, which we will adopt in Ubuntu’s Netbook Edition for 10.10 and beyond.

First, we want to move the bottom panel to the left of the screen, and devote that to launching and switching between applications. That frees up vertical space for web content, at the cost of horizontal space, which is cheaper in a widescreen world. In Ubuntu today the bottom panel also presents the Trash and Show Desktop options, neither of which is relevant in a stateless instant-on environment.

Second, we’ll expand that left-hand launcher panel so that it is touch-friendly. With relatively few applications required for instant-on environments, we can afford to be more generous with the icon size there. The Unity launcher will show what’s running, and support fast switching and drag-and-drop between applications.

Third, we will make the top panel smarter. We’ve already talked about adopting a single global menu, which would be rendered by the panel in this case. If we can also manage to fit the window title and controls into that panel, we will have achieved very significant space saving for the case where someone is focused on a single application at a time, and especially for a web browser.

We end up with a configuration like this:

Mockup of Unity

Mockup of Unity Launcher and Panel with maximised application

The launcher and panel that we developed in response to this challenge are components of Unity. They are now in a state where they can be tested widely, and where we can use that testing to shape their evolution going forward. A development milestone of Unity is available today in a PPA, with development branches on Launchpad, and I’d very much like to get feedback from people trying it out on a netbook, or even a laptop with a wide screen. Unity is aimed at full screen applications and, as I described above, doesn’t really support traditional file management. But it’s worth a spin, and it’s very easy to try out if you have Ubuntu 10.04 LTS installed already.

Ubuntu Light

Instant-on, dual boot installations are a new frontier for us. Over the past two years we have made great leaps forward as a first class option for PC OEM’s, who today ship millions of PC’s around the world with Ubuntu pre-installed. But traditionally, it’s been an “either/or” proposition – either Windows in markets that prefer it, or Ubuntu in markets that don’t. The dual-boot opportunity gives us the chance to put a free software foot forward even in markets where people use Windows as a matter of course.

And it looks beautiful:

Ubuntu Light

Ubuntu Light, showing the Unity launcher and panel

In those cases, Ubuntu Netbook Light, or Ubuntu Desktop Light, will give OEM’s the ability to differentiate themselves with fast-booting Linux offerings that are familiar to Ubuntu users and easy to use for new users, safe for web browsing in unprotected environments like airports and hotels, focused on doing that job very well, but upgradeable with a huge list of applications, on demand. The Light versions will also benefit from the huge amount of work done on every Ubuntu release to keep it maintained – instant-on environments need just as much protection as everyday desktops, and Ubuntu has a deep commitment to getting that right.

The Ubuntu Light range is available to OEM’s today. Each image will be hand-crafted to boot fastest on that specific hardware, the application load reduced to the minimum, and it comes with tools for Windows which assist in the management of the dual-boot experience. Initially, the focus is on the Netbook Light version based on Unity, but in future we expect to do a Light version of the desktop, too.

Given the requirement to customise the Light versions for specific hardware, there won’t be a general-purpose downloadable image of Ubuntu Light on ubuntu.com.

Evolving Unity for Ubuntu Netbook Edition 10.10

Unity exists today, and is great for the minimalist, stateless configurations that suit a dual-boot environment. But in order embrace it for our Netbook UI, we’ll need to design some new capabilities, and implement them during this cycle.

Those design conversations are taking place this week at UDS, just outside Brussels in Belgium. If you can’t be there in person, and are interested in the design challenges Unity presents for the netbook form factor, check out the conference schedule and participate in the discussion virtually.

The two primary pieces we need to put in place are:

  • Support for many more applications, and adding / removing applications. Instant-on environments are locked down, while netbook environments should support anybody’s applications, not just those favored in the Launcher.
  • Support for file management, necessary for an environment that will be the primary working space for the user rather than an occasional web-focused stopover.

We have an initial starting point for the design, called the Dash, which presents files and applications as an overlay. The inspiration for the Dash comes from consoles and devices, which use full-screen, media-rich presentation. We want the Dash to feel device-like, and use the capabilities of modern hardware.

Unity Dash

The Unity Dash, showing the Applications Place

The instant-on requirements and constraints proved very useful in shaping our thinking, but the canvas is still blank for the more general, netbook use case. Unity gives us the chance to do something profoundly new and more useful, taking advantage of ideas that have emerged in computing from the console to the handheld.

Relationship to Gnome Shell

Unity and Gnome Shell are complementary for the Gnome Project. While Gnome Shell presents an expansive view of how people work in complex environments with multiple simultaneous activities, Unity is designed to address the other end of the spectrum, where people are focused on doing one thing at any given time.

Unity does embrace the key technologies of Gnome 3: Mutter, for window management, and Zeitgeist will be an anchor component of our file management approach. The interface itself is built in Clutter.

The design seed of Unity was in place before Gnome Shell, and we decided to build on that for the instant-on work rather than adopt Gnome Shell because most of the devices we expect to ship Ubuntu Light on are netbooks. In any event, Unity represents the next step for the Ubuntu Netbook UI, optimised for small screens.

The Ubuntu Netbook interface is popular with Gnome users and we’re fortunate to be working inside an open ecosystem that encourages that level of diversity. As a result, Gnome has offerings for mobile, netbook and desktop form factors. Gnome is in the lucky position of having multiple vendors participating and solving different challenges independently. That makes Gnome stronger.

Relationship to FreeDesktop and KDE

Unity complies with freedesktop.org standards, and is helping to shape them, too. We would like KDE applications to feel welcome on a Unity-based netbook. We’re using the Ayatana indicators in the panel, so KDE applications which use AppIndicators will Just Work. And to the extent that those applications take advantage of the Messaging Menu, Sound Indicator and Me Menu, they will be fully integrated into the Unity environment. We often get asked by OEM’s how they can integrate KDE applications into their custom builds of Ubuntu, and the common frameworks of freedesktop.org greatly facilitate doing so in a smooth fashion.

Looking forward to the Maverick Meerkat

It will be an intense cycle, if we want to get all of these pieces in line. But we think it’s achievable: the new launcher, the new panel, the new implementation of the global menu and an array of indicators. Things have accelerated greatly during Lucid so if we continue at this pace, it should all come together. Here’s to a great summer of code.

Read more
mark

One of the driving mantras for us is “less is more”. I want us to “clean up, simplify, streamline, focus” the user experience work that we lead. The idea is to recognize the cost of every bit of chrome, every gradient or animation or line or detail or option or gconf setting. It turns out that all of those extras add some value, but they also add clutter. There’s a real cost to them – in attention, in space, in code, in QA. So we’re looking for things to strip out, as much (or more) as things to put in.

I’m not sure we’ll go as far as Microsoft has with their new Windows Phone 7 UI (links to .PPTX), which uses a design language called Metro. It’s radically pared back, and very cool work. It will be interesting to see if they’ve gone too far, or if users take to the more abstract feel of it.

It’s not hard to get people enthusiastic about the idea that less is more. However, it’s quite hard to get people to agree on which bits can be less. It turns out that one person’s clutter is another person’s most useful and valued feature.

Less, it turns out, is still less.

So, for example, consider tooltips on the panel. In bug #527458, there’s some discussion about a decision I made to deprecate tooltips on panel indicators. For quite a lot of people, that’s a little less too far.

On that particular decision, we’ll have to let time tell. For the moment, the decision stands. I’m the first to admit fallibility but I also know that it would be impossible to get consensus around a change like that. If those tooltips are, on balance, really just clutter, then unless someone is willing to take a decision that will be unpopular, they will be clutter forever. And it’s easier for me to make a decision like that in Ubuntu than for virtually anybody else. I apologise in advance for the mistakes that I will certainly make, and which others on the design team may make too, but I think it’s important to defend our willingness to pare things back and let the core, essential goodness shine through. We have to balance innovation and change with clarification and focus. We can’t *stop* innovating and changing, and we have to be willing to remove things that someone will miss.

The bug is a good place to continue the discussion about that particular issue. But I thought it would be useful to issue a call to arms, and invite suggestions from people on the Ayatana list as to what elements of the existing Ubuntu desktop can be trimmed back, on balance making the whole better.

There’s a growing awareness and excitement about the importance of design in free software. A few years ago, folks laughed out loud when it was suggested that design is a good thing for the free software community to build expertise in.  And it’s been slow going, admittedly. It’s hard to bring clarity in a crowd. Or mob. We’ve been doing our part to lead that at Canonical and in the Ubuntu community, both through internal work and through public forums. If you’re interested in design and Free Software, then Ubuntu and Ayatana and related forums are great places to participate. And your participation is welcome!

Read more