Having Microsoft endorse a GNU/Linux distribution was once not the best advertisement for that distribution. These days, however, that has changed.
Thus when Microsoft recently endorsed Debian GNU/Linux in its Azure marketplace and later announced it would be using the same distribution to launch Linux-based tools for networking, it was taken as a compliment.
New Debian project leader Mehdi Dogguy (seen above) attributes this choice to the fact that Debian is generally a great platform for derivatives.
"It has the largest number of derivatives among popular Linux distributions," he told iTWire during an interview soon after being declared leader. "It is quite stable and integrates a lot of FOSS projects. I believe those arguments speak in favor of their choice."
Dogguy was the lone candidate who stood for the post of leader this year, an unusual situation as there is normally a contest for the post. The last time this happened was a few years ago, when an incumbent, Stefano Zacchiroli, contested.
Dogguy, who hails from Tunisia but now lives in France, is 32 and works for Électricité de France as a technical manager.
{loadposition sam08}The group with which he works is dedicated to HPC-related tasks (High Performance Computing) where he gets Debian ready to use out of the box on some TOP500-ranked machines. "We focus on making sure that Debian is ready for real-world HPC environments and deploy Debian to the HPC clusters," was the way he put it in his election platform.
He was interviewed by email, shortly after taking office on April 17.
iTWire: Having interviewed every DPL for the last 10 years, i have heard everyone talk of improving communication within the project. Yet, during the debates over the change of init systems, there was a lot of bitterness and hysteria. Do you agree? What can be done to improve communication and maintain civility?
Mehdi Dogguy: It is true that the init systems debate (in 2014) was not as calm as we hoped. The discussions on that subject lasted a long time and the project even lost a few contributors in the way. Eventually, the Technical Committee decided on the matter and it was implemented flawlessly (modulo a few reactions). There was some confusion until we identified the main agitators. Many were external to the project.
Project members also called for a vote on init system coupling General Resolution and there they reaffirmed their trust in the procedures for decision-making and conflict resolutions. By voting, project members showed there is a big difference between their views on the matter and the general public perception about Debian!
This was also a key point to understand general project members' position and not mix it up with the observed hysteria (globally due to some external agitators). I am quite proud the project was able to reach a conclusion and implement it in time for Debian 8.0 (Jessie). In some way, it made us stronger as a community!
We can have some heated discussions from time to time. It should not be perceived as a sign of weakness. When needed, the Code of Conduct should help us to keep discussions in a reasonable shape and tone. We've had our share of tough discussions in the past, and the global atmosphere is quite better today than it used to be. So I guess we've learned from our mistakes; we give more attention to others' opinions and are able to debate more serenely today.
iTWire: These are four points you listed in your platform statement: identifying non-trivial bottlenecks; smooth communication between teams; shared goals using a single coherent strategy; and reduction in complexity of our processes. Can you elaborate on each in simple terms, perhaps with examples from the project's past?
MD: Debian is a big project and its complexity has increased over the years. Many volunteers contribute to the project, but we feel always we are understaffed. It is undeniable that we could do more with the help of more hands. But I think that it is also possible to perform some tuning here and there. Some processes may be simplified (like handling of archives content) and other may require more advertisement or simpler documentation.
Since we are many, and organised in many teams, it is not always obvious that everyone shares a global and coherent strategy. I believe a roadmap could help to achieve that. Building a roadmap will require putting together teams' projects and discussing their coherency. It also needs project members to agree with its goal and contribute to it.
iTWire: Debian has generally been focused on releasing when it is ready, not according to some artificial deadline. Why do you need a roadmap for a dynamic project of this kind which creates its own momentum?
MD: This is a very important remark and I thank you for bringing it up! Debian has no need for a roadmap to make its releases indeed. We will continue to do date-based freezes and release when it's ready!
The roadmap described in my platform should not be confused with a release plan.
Many things are done within the project, but not properly advertised. Our new stuff and priorities are not properly communicated. We used to have a list of release goals which somehow served that purpose but the Release Team decided, rightfully, that it is not their job to set technical goals for the project. Now, we have to resume that effort and publish a roadmap that will be useful to our users and upstreams.
Contrary to our historical Release Goals, a roadmap doesn't have to be bound to a release, but should give some idea about when each item will be implemented and where the project is going.
The roadmap will hopefully also help us to attract new contributors. One recurrent issue for them is to identify a subject or an area where to contribute. We already have ways to help them find their way into the project, like using the "how-can-i-help" package, a great tool. But I believe the roadmap can be another tool for newcomers.
Last, but not least, I believe this will greatly help our ecosystem (users, upstreams, downstreams, other fellow f/oss actors, etc) to better understand our vision and priorities.
iTWire: Every DPL I interview has agreed that more interaction with media and more coverage will help the project. Yet few of them — Stefano Zacchiroli and Steve McIntyre are exceptions — have kept their word when the time comes. Will you promise to be accessible to media queries and respond promptly or do you think that the media is just a pain with which you have to put up?
MD: Media is important for obvious reasons. It helps us to reach out more people (not only our regular users) and give us the opportunity to better explain our philosophy and plans. I will count on the media to help convincing more people about the direction the project will be taking in the future! I will try to reply to media queries as much as possible.
iTWire: Why do you think Microsoft chose Debian as the base for its software-defined networking initiative?
MD: Debian is generally a great platform for derivatives (and has the largest number of derivatives among popular Linux distributions). It is quite stable and integrates a lot of FOSS projects. I believe those arguments speak in favor of their choice.
iTWire: You have mentioned that you will maintain a DPL journal listing current subjects and planned actions. Will this be accessible to the public as well?
MD: The journal is described here and is accessible to Debian Developers only. Eventually, all items present in the journal are published in a "Bits from the DPL" mail to debian-devel-announce mailing list. So, it is mainly a way to keep project members up-to-date with the day-to-day tasks. Its content is not meant to be kept private, but it is not written in a proper form to be published as is.
iTWire: Your comments about time commitment in your platform statement are a little confusing. What do you mean when you say that you will not be a full-time DPL?
MD: I have a full-time job. I have a family which requires my presence (and which I need for my own equilibrium). I can dedicate to Debian some amount of my spare time, and my employer allows me to dedicate 20 per cent of my work time to Debian tasks. In that sense, I will not be able to act as a full-time working DPL and people may need to wait sometimes to get a reply.
iTWire: These plans you have outlined - will they be do-able in one term, or will you be nominating again next march?
MD: Before tackling these subjects, there are some subjects that need serious care (like DebConf delegation). Some discussion has to be led with DebConf active members in order to agree on a way Debian and DebConf will be coupled.
The timing of DebConf is also perfect with respect to plans outlined in my platform. It will be a perfect opportunity to meet Debian members and work on those subjects. It is reasonable to expect a first version of the Roadmap published by the end of the year. A new tool for newcomers may also see the light in the same period.
The DPL job is very time- and energy-consuming. Before committing on nominating myself again next march, I'd like to fully enjoy this first experience as a DPL and evaluate my performance in that position. It is not a decision to be taken lightly, IMHO.
iTWire: Anything else that you would like to mention?
MD: I'd like to thank the Debian community! It is great project with fantastic welcoming people! I have the chance to check that fact each time I attend a Debian event or meet Debian friends. I'd recommend anyone interested in the subject to come and join us in Cape Town for our annual conference, DebConf. I've attended DebConf 9 as a newbie and today, I cannot imagine myself out of this extraordinary human project.