I am currently in the process of interviewing the leaders of every Linux distribution on the planet, with the goal of helping us get to know the people behind the projects better. Having just wrapped up discussions with the heads of both elementary and Fedora (and others in the works) I decided it was time to talk about openSUSE.
This gets a little tricky as, earlier this year, I was elected to a position on the openSUSE board. I thought, for a moment, about either skipping the openSUSE interview or having someone else conduct it – to avoid the perception of bias.
But then I realized something. I am biased. But not just for openSUSE. I, on a personal level, want to see elementary, Fedora and so many other Linux distributions succeed. Add to that the fact that many of these leaders are people that I consider friends... and it becomes crystal clear just how biased I am. But not for any one distro... I am simply biased for Linux as a whole.
So I decided to go forward and interview Richard Brown, the openSUSE chairman of the board. What follows is the full interview, completely unedited.
Bryan: Good morning, Richard! For those who don't know you, could you give a (super short) description of what your role is within openSUSE?
Richard: I'm chairman of the openSUSE board. The board is responsible for leading the openSUSE Project, making sure everyone is talking to each other, resolving conflicts, making decisions when needed, stuff like that. As chairman I have the additional responsibility of making sure there is great communication between openSUSE and its main sponsor, SUSE.
Why do you do it? What is it that makes being the openSUSE chairman worthwhile on a personal level?
I've been contributing to openSUSE since it started 10 years ago. Bug reporting, packaging, advocacy, I've always loved contributing to the project and working with my friends in the community. Being chairman is just another way to scratch that itch of giving something back to the community that's given me so much over the years.
What is it about openSUSE for you? Why not contribute to Debian, Fedora or Arch (etc.)?
I'd say it is something to do with openSUSE's "mindset" or "philosophy"..two words which we don't normally throw around the Project. It's always produced things that suit my needs. Distributions that are technically powerful, but easy to live with. Tools to keep my systems running. Tools to help me learn about and start building software. All with a great bunch of very smart and welcoming contributors. I've spent time using other distributions and their communities, but openSUSE always felt like home for me.
On the distribution side of things... openSUSE has two, very different, distributions - Tumbleweed (rolling) and Leap (not-rolling). Which is running on your systems at home?
Tumbleweed on my Desktop and my Laptop, Leap on my Server.
Over the last year-ish, both openSUSE and Fedora have increased the number of distributions that each project creates (Fedora now has different editions for Cloud, Server and Workstation; openSUSE has Leap and Tumbleweed). Curious what your thoughts on that trend are -- good thing or bad thing?
It's a good thing, mostly. It's important to avoid splitting up or creating new offerings just for the sake of it. In our case with Leap and Tumbleweed we carefully looked at ourselves, our users, and what we were doing in the past and realized that we had two discrete sets of use cases and user bases which we were trying to satisfy with a single distribution. And that ultimately caused more work for us, trying to make one universal distribution that would make everyone happy and doing everything.
Tumbleweed has allowed us to really focus on the needs of our users and contributors who want the latest and greatest on a stable rolling platform, while the availability of the SUSE Linux Enterprise codebase has allowed us to build Leap, an exceptionally stable offering for those users and situations where stability is key, and change is less welcome. Fedora tried to achieve a similar thing with their editions, but in our case we see 'Cloud', 'Server', and 'Workstation' as just simple installation options ontop of whichever openSUSE distribution you choose.
Where do you see openSUSE going in the next few years? At just about every Linux conference I get (half-jokingly) asked about the "SUSE Phone" (in part because it sounds like "sousaphone"). Do you anticipate openSUSE heading into the tablet/phone space?
I certainly think it's possible. We're a contribution led project and our contributors are very often on the more 'technical' side of the spectrum. Professional developers, hackers, sysadmins are audiences where openSUSE really appeals.
If we have contributors who want to get openSUSE tech running on phones we'd certainly like to see that and will support them. But the mobile space is a tricky one, especially as hardware is often not very open or hackable. We're seeing much more interest and activity in other parts of the ARM space, with a huge wide range of support for openSUSE on many ARM development boards. There are even ARMv8 (aka aarch64) hardware vendors like SoftIron even shipping their hardware with Tumbleweed by default.
So I think openSUSE's future will likely be more in that direction, leading-edge technical environments where the whole ecosystem of openSUSE's distributions and tools has the most benefit for those demanding use cases.
Let's pivot away from openSUSE for a moment and talk about the broader Linux world. If you could fix one problem (just one) -- in the Free Software and Linux world -- what would it be?
I want Open Source NVIDIA drivers that are better in every way to the proprietary ones.
That's pretty straight forward. How, in your mind, do we make that happen?
I'm not sure. There is a limit to what can be done via reverse engineering and such, so it probably really depends on NVIDIA realizing they need to do this.
Linus giving them the middle finger has not seemed to work, so I'm betting on either the competitive draw of AMD doing a much better job in this area lately, technologies like Vulkan forcing their hand, or if all else fails an angry mob armed with pitchforks, torches, and stuffed tux toys might be the key.
So we should transition away from Linus giving the finger and move towards a plushie-based strategy. Makes sense to me. If you had to put odds on Nvidia deciding to release completely open source drivers... what would you give it?
Depends on timescale. In a long enough timeframe, I think it's inevitable. More and more of both the software and hardware market is realizing an open model is the only way to remain competitive in the long term. I wouldn't put any money on it happening tomorrow though.
When I asked Matthew Miller (leader of the Fedora Project) what operating system / distro he would be using in the event that Fedora passed a law that forbade him, specifically, from using it... he suggested he might create his own Linux distro. Just because it sounded fun. Same question to you. If the openSUSE board kicked you out and created a rule that forbade you from ever running openSUSE again... what system would you use?
Wow, that's a good question. I'm not sure I'd make my own, there are enough distributions out there. Forking openSUSE would be tempting just so I could keep on using it, but that would be selfish. I think Arch, but I'd implement my own openQA instance between their testing repos and stable and run a custom tested derivative..something in between because I think I'd find their stable repository too slow after running on Tumbleweed for so long”
You do realize that, by law, every Arch user must now post that "openSUSE Chairman would consider using Arch" to reddit no less than three times.
At least it would give them something else to talk about than "Hey did you know, I use Arch Linux?" ;)
One final topic I'd love your thoughts on: Statistics. Getting hard numbers of how many people actively use any given Linux distribution is notoriously difficult -- often resulting in pointing to trends on DistroWatch (which, we can all agree, is not exactly an accurate way to figure out this sort of thing). In the end, it seems like we really don't have any clue what the actual user base is for a large number of Linux distributions. My question for you is two-fold. First: How does openSUSE track that data? Second: How do we get to the point where we have semi-accurate numbers for all of the large distributions?
It's pretty tricky. We have some advantages with the build service and zypper giving us some extra options when it comes to anonymously counting our users, but even then it's still pretty much a dark art with plenty of room for error and interpretation.
As every other distribution has both different tooling and different methods of counting, a meaningful comparison is pretty hard to achieve. I had an interesting chat with Stephen Smoogen from Fedora at FOSDEM about this, and we're going to do our best to collaborate on rationalizing our methodologies going forward. Would be nice if we could come up with a wider solution, but I think with the different technical implementations of every distributions infrastructure, a common solution is a long way off, so distrowatch probably has its market cornered in totally unreliable distribution statistics for some time yet.
If people want to get involved in the openSUSE project... what's a good place for them to start?
Pick something you want to work on. Can be anything, technical, marketing, artwork, find an 'itch' you want to scratch and we'll happily help you scratch it. If you don't know what itch you want to scratch, we have http://101.opensuse.org with a list of ideas as part of our mentoring program. Then join the appropriate openSUSE mailing list or IRC channel and tell people about what you want to do. Make friends, find people to help you, and well, that's it, you've already started, it's all hard work and good fun from there, just like our motto and /etc/motd says..."Have a lot of fun!"