Brussels / 1 & 2 February 2020


Interview with Thorsten Leemhuis
The Linux Kernel: We have to finish this thing one day ;). Solving big problems in small steps for more than two decades

Photo of Thorsten Leemhuis

Thorsten Leemhuis will give a talk about The Linux Kernel: We have to finish this thing one day ;). Solving big problems in small steps for more than two decades at FOSDEM 2020.

Q: Could you briefly introduce yourself?

I was raised to become a farmer but instead found my way into the world of computers. There I quickly developed a special interest in operating systems and computer hardware. After switching (quite late… :-/ ) from OS/2 to Linux these preferences quickly led to a strong interest in the Linux kernel, as that’s where the drivers are and most of the magic comes from. That’s why I started following the Linux kernel development quite closely.

I make a living by writing about Linux and FOSS for the c’t magazine, which among others is famous for uncovering the SoftRAM scam in 1995. These days it (according to our marketing people) is the biggest magazine dedicated to computers and technology in Europe.

Ten to fifteen years ago I contributed to Fedora quite a lot in my spare time. In 2017 for a while I took care of regression tracking for the Linux kernel (also in my spare time), which got me an invite to both the Kernel and Maintainers Summit.

I also like to cook (preferred cuisine: Indian) and do a lot of sports, if time permits (Cycling, Running, Obstacle Course Racing, Badminton, Inline Skating, Bouldering, …).

And it seems I talk too much, as that wasn’t really brief.

Q: What will your talk be about, exactly? Why this topic?

The talk will look at 20 years of the Linux kernel and Linux kernel development and how that is relevant for Linux today and tomorrow. So stop worrying, it definitely won’t be a history lesson. Sure, I’ll mention some important or funny changes and events from the last two decades. But the focus will be on topics from the past twenty years that are relevant for the present or the future. It’s an ideal point in time to do this, as some recent changes in the Linux kernel might turn out to be the beginning of a small revolution.

I chose this topic because I know a lot about it, regularly give talks about it (albeit normally more technical in-depth on recent changes), and thought Linux kernel topics are a bit underrepresented at FOSDEM.

Q: What do you hope to accomplish by giving this talk? What do you expect?

The plan is to give an educating and entertaining talk, which hopefully gets a few more people interested in the software that drives their systems. I also want the audience to understand better why some things in the land of the Linux kernel are the way they are today, even if they might look odd or stupid to some users — like not integrating ZFS or deliberately avoiding to provide a stable API for kernel drivers.

Q: What has changed over the last 20 years in the Linux kernel and for Linux kernel development?

A lot, come to my talk to learn about this. But to give one thing away: I’ll talk about a few areas where the Linux kernel development did not change while the rest of the world moved on. Some of these differences from typical FOSS projects I will just mention for entertainment, as not using something like GitHub or GitLab looks odd or even antiquated to outsiders.

But I’ll also mention differences because some bear the potential for disrupting the FOSS world. Yeah, I know, that sounds a bit over the top, but remember: twenty years ago a lot of people laughed about the Linux kernel, as it still wasn’t using a Version Control System (VCS) like CVS or SVN as nearly everybody else already did back then. But in the end it led to Linus Torvalds developing this thing called Git you might have heard of.

Q: 2020 marks the 20th anniversary of (F)OSDEM. What contributions has FOSDEM made to the advancement of FOSS, or how did the Linux kernel in particular benefit from FOSDEM?

I’d say: educating the world and bringing people together in real life is how FOSDEM has helped the advancement of FOSS most. There is simply so much you can learn from watching the talks in the devrooms live or via recordings later (thanks, video team and volunteers!). And putting a face to a name you only know from IRC chatrooms or some mailing lists also helps tremendously the next time you interact with that person over a wire.

In regard to the Linux kernel there is no easy answer as to how FOSDEM helped. For some reason unknown to me there hasn’t been a developer room dedicated to the Linux kernel at FOSDEM for years now. But the conference in the end definitely helps the Linux kernel too, as the kernel is the backbone of a lot FOSS. Therefore, in the end it is often somehow involved in lots of talks and discussions happening at FOSDEM. That is especially true in devrooms like Graphics, Virtualization, and Software Defined (Storage|Networking), where the kernel is quite present. The talks there help the world understand new and cool features in the Linux kernel. That helps these features grow and in the end might lead to some developer improving a particular feature even more to build something even better.

Q: Have you enjoyed previous FOSDEM editions?

Definitely! I visit a lot of talks, but also make sure I walk around a lot to see familiar faces again and meet new people. My first FOSDEM IIRC was the last one where the weather served us with loads of snow and a freezing cold at the same time (I think that was 2012). Since then, I came to Brussels every year for the Friday beer event and stay around for the weekend.

Creative Commons License
Creative Commons License

This interview is licensed under a Creative Commons Attribution 2.0 Belgium License.