August 23, 2010

Talking KDE and openSUSE with Jos Poortvliet

Recently I had the pleasure of chatting with newly minted openSUSE community manager, and former KDE marketing lead, Jos Poortvliet about user-visible changes brought in the KDE 4.5 release. We asked him ten questions about changes in KDE 4.5, what problems needed to be overcome, and what the future of KDE looks like — and his thoughts on openSUSE.

Jos Poortvliet was hired by Novell as openSUSE community manager in early August. Before that he was active as volunteer in the international KDE community as marketing team lead. He coordinated and contributed to the work around releases, creation of marketing materials, visiting and organizing conferences, and maintaining contacts with the press. In his "real" life he worked as business consultant at companies like Royal Bank of Scotland, the dutch Governmental Department of Education and KPN, a major dutch Telecom provider.

So, without further ado, let's get on with the interview. My first question is about KDE's migration from 3 to 4. Everyone was quite happy with the state of KDE and 3.5 was about as solid a desktop as you could find. But change was in the air. When KDE 4 first arrived it was hit with a lot of criticism for being unstable and slow. How have you managed to overcome that?

Jos Poortvliet: Absolutely. With the first release of our software on KDE Platform 4 we pushed the boundaries of technology on the Free Desktop. Much of that vision has come to fruition now. We're at the 6th major release since we started on this journey. The lower stack has caught up with us: where especially graphical drivers couldn't cope with what for example the Plasma Workspace did in terms of animations and smooth gradients, it is now mostly capable of sustaining reasonable performance even on low-end hardware. You can see this looking at the move of Plasma to netbooks and even mobile devices. The same goes for Nepomuk, the semantic desktop technology. And of course Nokia's Qt has also been improved in almost every aspect, benefiting us hugely. All these technologies are not perfect (and will never be as there is always something to improve), but in terms of stability and performance we're at least matching Mac OS X or Microsoft Windows.

Meanwhile the advances we have made in terms of functionality and usability are huge. Technically, we're years ahead of Mac OS X and Windows when it comes to our desktop technology. Both Apple and Microsoft require developers to re-write their applications for the mobile space. Plasma offers a real device spectrum. Developers can write an applet which will be aware of its surroundings. It will scale from the desktop and netbook to a mobile phone, up to television screens and media center interfaces. There is NO comparable technology out there!

And you can write these applets in JavaScript, Ruby, Python, C#, C++ — whatever you like. They offer a split between technology and interface: a designer can create the interface in a declarative manner, independent from the developer working on the back-end. It is efficient, synchronizing animations and network activity between applets to minimize power usage and CPU resources. Nokia, Intel, Palm, Apple and Microsoft — none of them have something which comes even close.

Much of the work is now going into finishing and polishing the results, and sharing the technology with other free desktops on A good example of that would be the notification specification, or our work on the system tray.

So, 4.0 was about overcoming the resistance to revolutionary change — dropping the working but evolutionary dead end of the KDE 3 desktop to concentrate on a desktop design for the next ten years, and getting our broad user base to believe in this vision. If you (users) haven't given KDE 4.5 a try yet, you might not realize that there are some big changes that have occurred. Most of those changes are very subtle, but some... not so much. What has been the biggest change from KDE 4.0 to KDE 4.5?

Jos Poortvliet: I would say going from showing a vision to implementing it. It was hard and took time to get the world to acknowledge the vision we had and rallying support for it in the Free Desktop world. The Plasma desktop team had to fight poor performance and a lack of features in the Linux infrastructure and tried to push things which were needed among the Free Desktop. This needed collaboration with for example the GNOME developers but things have been improving on that front.

In terms of technology, well — there have been huge changes. We have had cultural changes in the community, where we increased focus on usability. We've had technical changes, especially when it comes to Plasma. Plasma is, from a technical point of view, a hugely revolutionary product. For the first Plasma releases the developers decided to re-create the traditional desktop (panel on bottom with start button on the left, tasks in the center and systray with clock on the right) to keep things familiar. However, with Plasma you can write completely new desktops, like the Plasma Netbook, Plasma Mediacenter and Plasma Mobile initiatives have shown. Nobody in the industry has something which even aspires to be as powerful as this. most users aren't developers, they don't realize how challenging even a minor upgrade can be for developers. Was their one particular challenge moving from KDE 4.4. to 4.5 that was harder than any other?

Jos Poortvliet: I would say the work on the Kontact suite by the KDE PIM team. They have decided to delay the release of Kontact 4.5 which includes a rewrite of KMail based on the new Akonadi technology. Akonadi offers a unified mechanism for applications to work with personal data like mail, contacts and agenda. It was introduced half a year ago, and with the upcoming KDE Kontact 4.5.1 release it will be built into KMail. This is a huge step. It has become easy to connect to many new online services in a completely transparent way. Akonadi also has a big potential to lower the resource requirements on mobile devices while connecting users to the web through more efficient interfaces than what is offered by a web browser. When it is released, the first visible improvement for users will be that their mails are now properly indexed by Nepomuk, making them available from search interfaces like in Dolphin and the Alt-F2 run dialog. Of course, more will come - there is work on integrating mail with the Plasma desktop, for example.

Of course I am very happy with other changes too — improvements and polish to Plasma, new features in window management like tiling and new effects, and especially the reorganized System Settings. Everyone has their favorite feature on the desktop. But what about you? What feature, in particular, are you most excited about with future releases of KDE?

Jos Poortvliet: Silk and Nepomuk.

Silk isn't really a feature, but more of an idea, a vision. That idea is, in short, taking back the web to our desktop. Integrating the many web technologies and sites out there into our applications and into Plasma, making them available easier and more integrated.

Nepomuk meanwhile is really a technology which is still in its infancy. It is about connecting your data with how you use it, interfacing in a more human way. The normal file system requires the user to adopt the thinking of a computer: a file is in this and this location, placed in a hierarchy. The web doesn't work that way - everything is 'flat', and you use search to find what you are looking for. Google is leading here, making sure that you don't have to literally know what you are looking for: it will suggest related search options and serve pages going that one step further. Look for the name of a restaurant and Google won't just offer the homepage but also a Google map showing the nearest restaurants and a way to contact them. This is what Nepomuk brings to computing devices. Look for the name of a contact and Nepomuk offers you files you received from that person. This of course needs close ties between Nepomuk and other technologies like Akonadi, which will be the upcoming revolution. Is KDE going to do anything differently when GNOME 3 is finally released? Or will KDE continue on as normal?

Well, if you look at what GNOME 3 is attempting — we did that with the KDE Platform 4. Clean up the technology and introduce new interface and infrastructural technologies. They however took a very different approach. We redesigned our infrastructure but tried to build a familiar desktop on top. The current, default Plasma desktop, while fundamentally much more flexible, still follows the traditional panel setup. GNOME introduces GNOME Shell, a much more revolutionary desktop interface, but with far less flexibility. Of course these differences make sense — the GNOME developers had an entirely different vision with their Shell than the Plasma Developers. GNOME wanted to offer the best user experience possible, according to their design. Plasma is a technology to build user experiences... The Plasma Netbook and Plasma Mobile interfaces are examples of those and you could build a GNOME Shell like interface in a reasonably short time. Someone might do that at some point...

And now we've got these huge changes mostly behind us (with Akonadi still lurking around the corner). If you're putting the boot in here you can say that we've got our disruptive changes mostly behind us — completely new Pillars such as Phonon, and Solid are now 3 years old and are mature and reliable and the desktop shell meets KDE 3 users' needs. This frees us up to go innovate with netbook and mobile, instead of trying to retrospectively graft fixes onto a platform which did not change much between major versions.

What we could learn from GNOME is that they worked really hard not to break anything while doing that move. Of course their situation is different and so are their goals. We decided we wanted to push the state of the Free Desktop aggressively and needed to follow the Free Software mantra of 'Release early, release often.' They have decided not do do that and provide a more polished experience from the start. That is better for end users, but less exciting for technology enthusiasts and developers who look for the latest and greatest. We are catering more for end users now we have our platform mostly in order but more work wouldn't hurt. Although this has obviously changed, what was/is your primary involvement with KDE?

Jos Poortvliet: I am, or rather was, the KDE marketing team lead. Since I have moved to Novell as their new openSUSE community manager, I will become less involved in KDE marketing as I move on to work in openSUSE. This will also mean I will have a closer relationship with GNOME. That is exciting - they have a very different approach to things but are also working more and more with the KDE community. I hope to foster that collaboration. openSUSE is a leader in that area, offering both excellent KDE and GNOME based desktops, as well as MeeGo and other Netbook offerings. This synergy can offer a lot of advantages and I hope to focus on that. Especially now openSUSE offers technologies like the Build Service which allows the building and distribution of cross-distribution software (supporting over 10 major distributions right now) and the SUSE Studio where you can build appliances, small contained operating systems for specific purposes. One of the things you will see from me in this cross-desktop collaboration work is at the upcoming Desktop Summit in Berlin, where I'm one of the organizers. I took care of the marketing for both desktops at the Gran Canaria Desktop Summit already and hope to do that again this year. Many have their own ideas of the future. GNOME, KDE, OS X, Windows...each has a different take on what the desktop should be. Where do you envision the desktop will be in five or ten years?

Jos Poortvliet: I think both Mobile and Cloud technologies are very powerful. GNOME is working hard on mobile technologies, setting up an infrastructure for that. MeeGo builds on that, working with Novell, Intel and Nokia and leveraging the openSUSE Build Service and the openSUSE distribution technology in general. KDE's Plasma technology is also taking advantage of that platform on the MeeGo mobile platform and with a Netbook reference implementation on SUSE Studio. And all those technologies are taking advantage of cloud services. Akonadi does it's part there, as does KDE's ProjectSilk which aims to integrate the web in the desktop. And audio players like Banshee and Amarok are integrating online music stores, podcasts and the like in their interfaces. In your opinion, what kind of influence does the open source community at large have in the decisions made regarding the future of KDE?

Jos Poortvliet: KDE is an open source community, probably the largest really community-controlled organization besides the Linux kernel project. We work with a lot of companies who participate in KDE's community based and meritocratic development model. Each of these companies are community members who have as much influence as others contributors, including volunteers, based on how much effort they prove to development, testing, documentation, etc. We're certainly one of the most open communities in our size class. You'll hear that very often when talking to third parties, be it volunteers of people from companies: they're surprised how easy it is to get KDE contributors excited about and taking advantage of their technologies. Nepomuk is a prime example of this, as it started out primarily as a EU funded university research project where we collaborated with a number of companies and universities. There are so many desktops for the Linux operating system. Each one of them offers something different. From the overly simple Fluxbox to the full fledged desktop environments like KDE and GNOME. If you could adopt one feature from a different desktop (one that KDE does not have), what would it be?

A marketing machine which can force OEM's to use it's technology. Seriously, in terms of features, there are surely things we could add. There are gaps here and there. But that's also the case with Apple's Mac OS X or Microsoft's Windows. We're ahead in terms of technology, but when it comes to general adoption, that's not a real factor. What matters is marketing and market power — that the defining influence. Surely, companies have a bit of a hard time understanding and working with Free Software, that also contributes. But getting Free Desktop technology, be it KDE or GNOME or (more likely) a mix, to the masses - we only lack the power of force. Not KDE-specific. In the eyes of the public, Linux lags behind Microsoft on the desktop. Although this can not be 100% confirmed (due to a lack of an accurate census tool for the Linux operating system), Linux is perceived by the majority as being a non-factor on the desktop. What, in your opinion, does Linux need to help it continue to become more widely adopted on the desktop?

Jos Poortvliet: Well, I just said it. Sure, the general consumer is not familiar with Free Software, that is an issue. But a bigger issue is our own product: if you're a third party selling support contracts on open source solutions you notice it is harder to make money. Because you make money when your customer has issues with the product - and especially on the server, open source is simply better. We need other business models. That's why the mobile space is so interesting, there are other forces at work there.

The desktop in particular suffers from market dominance of two large vendors. Big OEM's like HP, Acer or Samsung might be willing to sell a Linux desktop offering but they still put a 'we recommend MS Windows 7' banner on their site because they're paid to do it. And they will probably lose benefits if they dare to sell open source desktop solutions in serious numbers.

And there you have it. Words of wisdom for any open source programmer in training, a seasoned programmer, or a user of open source software. Whether you are a fan/user of the KDE desktop or not, there is something to glean from Poortvliet's words. And with KDE now at 4.5, the software Poortvliet once worked hard to promote is now primed to make some serious splashes. Will Poortvliet do the same from the GNOME desktop? My instincts say "yes." What do you think? How has the KDE desktop improved with the release of 4.5? Give your fellow readers your experience.

Click Here!