Author: Philippe Kalaf

In October 2008, the GNOME UX hackfest at the Boston summit became the place of birth of GNOME 3. The seeds were planted out of a will to induce a more design-centric mentality in the GNOME project. As the GNOME shell was being designed and implemented, the computer world was changing under it’s feet. The touch-screen revolution was under way, starting with smartphones in 2007 followed by the introduction of tablets in 2010. This caused dramatic changes to the way we consume information on our computers. A large chunk of users and use-case that GNOME 3 was trying to cater to started to convert over to iOS and Android running on tablets for most of their tasks, leaving their laptops behind for those times when they needed to produce information. As the whole industry was transforming, discussion amongst GNOME shell designers about addressing the change naturally ensued. But this was not a small change, it meant having to rethink and redesign the shell to cater for touch-screens. In order to keep the momentum going, it was agreed that GNOME 3 is being designed for desktops/laptops but that touch-screens would be kept in mind for later. As we stand today, there is a list of items and planned design specs for making GNOME shell more “touch friendly”, but those are still mostly for the future and application developers are still confused as to what that means exactly for them.

I argued, mostly internally here at Collabora, that GNOME shell needs a clear and definitive answer to this question and that taking a design decision to aim somewhere in the middle between desktops and touchscreens was a recipe to fail at both. Writing a shell that can work for both is one thing, but what about applications? I couldn’t see how one solution could fit both form factors and input methods. Little did I realized that on the other side of the fence, Microsoft was facing the same dilemma with the next release of their famed Windows operating system. Microsoft Windows was king of the desktop; a mouse, a keyboard and a congenitally insensitive screen was their playground. But now, the rules had changed and a new plan of action was required.

Introducing the Microsoft Surface. Hardware was nothing new to Microsoft, but this was a radical step since it could be interpreted as direct infringement on the business of their OEM partners, which historically have been key to Window’s success. So what is this Surface? At first glance, it looks like a tablet, but hook it up to it’s keyboard and it becomes a laptop with a track pad and a touch-screen?! It seems to be both a laptop and a tablet, or at least it’s trying to be. This sounds exactly like what I argued was impossible to get right from a UX point of view. But wait, we are talking about Microsoft here, a company that has been writing and designing UX for as long as I can remember… and they are good at it. So what is this magic UI that is running on these monstrous hybrids? It’s Windows 8, the same UI that will drive your laptop, will also run on your tablet, or will power one of those new convertibles. OK, so maybe they managed to make a shell that works well across the board, but what about the applications? The Surface comes with Office, a primary tool for producing information. Looking at the screenshots, it seems obvious that this was not designed to be driven through the touchscreen but rather the more traditional keyboard/pointer inputs. So Microsoft’s strategy seems to be that some applications will be driven with the keyboard/touchpad, whilst others will be more touch friendly. Basically, there is no reason why each application needs to do both simultaneously. This is reinforced by the fact that Windows 8 runs both traditional desktop applications and applications written using their new UI toolkit named Metro.

I’m willing to acknowledge that Microsoft can pull this off as I personally like the idea of being able to have my tablet and laptop packed into one slim, lightweight device. Nonetheless, my instinct tells me Microsoft is in trouble, that this is Microsoft’s fumbled attempt at making Windows 8 relevant in a world that changed too fast for them. Most critics agree that Windows 8 running on the Surface is either confusing or just plain bad. I guess time will tell, but in today’s world, things happen fast, so I doubt we will have to wait long to see the unravelling of Microsoft’s wager. Either way, I think GNOME shell designers should be watching Windows 8 very closely, as the results of this experiment could be very useful for the future direction of GNOME 3. Should it stick with the desktop and focus on content creation such as suggested by Eitan? Or are some of the currently proposed touch-friendly work items enough to make it a viable UI for tablet based consumption?