A look at Debian's 2015 DPL candidates

By Nathan Willis
March 25, 2015

Debian's technical committee may have attracted quite a bit of attention in recent months, but most of the day-to-day governance tasks in the project are the purview of the Debian Project Leader (DPL). Debian elects a new DPL every year, and the 2015 election is right around the corner. This year, there are three candidates, each of whom has offered a different take on what their term as DPL would mean for Debian and for the world outside the project.

Lucas Nussbaum, who has served as DPL for the past two years, decided not to run again in 2015. Three candidates did step forward: Mehdi Dogguy, Gergely Nagy, and Neil McGovern. As is customary, each candidate has written a campaign platform detailing his background with Debian, his vision for the future of the project, and his goals for the coming year as DPL.

All three candidates have a lengthy history with Debian, having served on a number of the project's teams or in other significant roles. Both McGovern and Nagy ran for DPL in 2014, although Nagy withdrew his nomination before the voting began.

Platforms

Dogguy's platform focuses on "the complexity of collaboration inside Debian;" he notes that the project has been having difficulty making:

solutions that scale to the size of the bigger project. This becomes even a more challenging problem when the number of packages grow more rapidly than we're able to onboard new contributors.

To address this concern, Dogguy says he will conduct a review of Debian's tools, mechanisms, and processes, so that he can identify "non-trivial bottlenecks", smooth communication between teams, reduce the complexity of Debian's processes, and provide a "single coherent strategy" everyone can share.

In more concrete terms, he says that Debian should publish a public roadmap that covers a time frame longer than the goals established for individual releases, and that he as DPL will work to make sure that progress is made. Dogguy also notes several major changes undertaken by Debian (including the init-system change, the Code of Conduct, and the transition away from 1024-bit PGP keys), but says "their implementation was a real pain". Averting such pain is another campaign plank; Dogguy says he will "be present during preparation of important changes (be them technical, social, financial or political) to ensure implementation details have been studied."

In addition, Dogguy says that Debian should start a recruitment and mentoring program designed to familiarize new contributors with Debian's community, processes, and workflow. He also suggested that Debian should devote some effort to making the distribution available through non-traditional installation methods, such as cloud and virtual-machine images.

Nagy's platform makes a point of declining to state a "grand vision". Rather, he says, the DPL's primary purpose is "to be an enabler: the Project Leader is not a front runner to lead the herd to victory, but a gentle shepherd to make them happy."

Doing so, he explains, means the DPL should remove barriers and empower people to pursue their passion. This, he says, means putting other project members first:

That is the vision I have, but I can't do it. At best, I can hope to enable people much better at the these things to do what needs to be done. I wish to take the burden of administration, bureaucracy off their shoulders, so they can focus on what they do best. I feel that this is the most important part of being a Project Leader: to enable the project to grow.

Nagy concludes by saying he wishes to be the DPL that no one remembers. "I'd rather see people remember all the great things the Project - as a whole - accomplished, for there are many."

McGovern's platform also describes a commitment to "support and enable" project members to do their own work by removing "blockers".

He also adds several specific efforts he would undertake as DPL. They include deploying a personal package archive (PPA) system, modernizing Debian's build system and related infrastructure, and promoting the non-packaging portions of contributing to Debian.

In addition, he says he will continue the daily "DPL log" started by former DPL Stefano Zacchiroli, as well as monthly email reports. Finally, he says he will "spend some money we have horded, noting that Debian has several hundred thousand in the bank, and that "we should spend it to make the project more successful."

The questions

Project members can ask all of the candidates questions via the debian-vote mailing list. The 2015 edition of the question-and-answer process elaborates on a few platform points and addresses some other issues.

Dogguy elaborated on his roadmap idea in one response, saying that he does not see it as a plan for the DPL to draw up, but as "a process which will enable us (DDs) to give some visibility to our individual plans". The roadmap would allow Debian developers to find other teams and individuals with similar goals, he said, from which shared priorities would emerge.

McGovern, likewise, was asked to elaborate on the DPL's role in deploying a PPA system. He responded that:

The DPL position holds the ability to influence external parties more than others. The conversations we can have to try and get external interest in getting this (finally) off the ground is much easier as DPL than not.

McGovern was also asked about his "spend some money" statement. He replied that he thinks Debian should spend money on booth paraphernalia, hosting meetings, and "actively recruiting people".

Additional questions about finance occupied much of the discussion thread. Martin Krafft asked the candidates whether they think Debian should offload accounting and finance work to a hired professional. Dogguy said no, while McGovern a relatively straightforward yes and Nagy offered a more tepid yes.

All of the candidates advocated the idea of Debian funding a small number of Outreachy students (from two to four), although McGovern said that Debian should do fundraising specifically to underwrite those students.

In regard to fundraising itself, McGovern later noted that the DPL can act as a project fundraiser. Nagy suggested that Debian should try to minimize the number of fundraising campaigns it runs, due to their unpredictability, and should instead focus on finding long-term sponsors.

The other major topic in the discussion so far was whether or not Debian should relax its acceptance of non-free software. Zacchiroli asked about section 5 of the Debian Social Contract (DSC), which provides for the contrib and non-free package archives, supports their usage by users, and permits their access to the Debian bug tracker and other infrastructure tools. Zacchiroli asked if the candidates thought the time was "ripe" to drop that section from the DSC, either removing contrib and non-free entirely, or simply refraining from publicly sanctioning them.

Nagy replied that dropping section 5 from the DSC but retaining contrib and non-free does not make sense:

So, the only way I could see the drop of SC §5 as a worthwhile goal, is if we also removed non-free (and possibly contrib) too. Unfortunately, I do not think we're quite there yet. But - in the long run - it would be a worthy goal to pursue.

Dogguy seemed generally in favor of retaining contrib and non-free, noting that they increase user awareness of the concept of free software:

As explained elsewhere, this enables some of ours users to use our system based on free software and a little part that is non-free. Which is a good compromise, since the existence of non-free area and having packages there correctly maintained made possible to run Debian at all. It takes users to explicitly install those non-free bits though and it is not automatic (and should remain as such). This increases their level of awareness wrt non-free works... which is a good thing (somehow).

McGovern responded that it was not time to remove section 5, arguing that taking an "ideologically purist view" does a disservice to users:

I would rather Debian is spread, and more people use free software that may require non-free works, than to reject them completely.

This doesn't mean we shouldn't strive to make §5 obsolete! Great work has been done to try and remove non-free blobs from the kernel, for example. I would love to run Debian on all systems without the need for firmware on open hardware, but that day has not yet come. Until it does, we should keep section 5.

To the polls

The question-and-answer period for the candidates is still in full swing. As of today, there are several interesting questions that are still pending responses from some (or all) of the candidates. Nussbaum, for instance, asked where each candidate sees Debian fitting into the free-software ecosystem five years from now. Anthony Towns asked the rather open question "where should the innovation come from?" Users and fans of Debian would be well advised to follow the discussion in earnest over the coming week.

Voting itself will commence April 1, continue through April 14, and the winner will be announced on April 15. The new DPL's term begins on April 17.