Note: This is the continuation of part 1, part 2 and part 3. You can get the full document as a single PDF. Feel free to share this document to anyone that might be interested to work towards the goals outlined.

Conclusion

I’m very excited by the perspective that I outlined in this document. It really resonates with my own mission statement as a Debian developer (written a long time ago):

My main role in Debian is to help Debian to evolve so that it’s always able to face the new challenges that are showing up.

My approach is both corrective and proactive, I work to solve current problems and prepare for tomorrow’s. This requires to remain sufficiently involved to identify new trends, see the deficiencies and be a force of proposition.

Most of the changes require to interact with many people, and problems are often more relational than technical. I will ensure to follow the habits of interdependence (Think Win-Win, Seek First to Understand, Then to be Understood, Synergize) to find a solution acceptable to all and to inspire others to do the same.

The easiest changes to implement are technical (such as improvements to distro-tracker) and require little interaction. This work is used to recharge me by offering me an immediate reward for my efforts.

Finally, and this is a substantive effort, I want to create in the project working conditions that allow all contributors to give their best. It starts with developing a common vision …

But I can’t achieve this alone, I need help from passionate individuals sharing this vision. Let me know if you want to be one of those persons.

Note: This is the continuation of part 1 and part 2.

Going forward: growing Freexian

Part 2: Extending the team

By all accounts, Freexian is still a small company which relies largely on me in many aspects. The growth of its business is however providing enough financial margin to allow looking into ways to recruit external help, be it through direct hiring (for French residents) or  via long term contracting (for people based in other countries). If you believe you could be the right person for one of the roles listed below, or if you know someone that we should contact, please reach out to raphael@freexian.com.

Project manager

I’m looking for someone that cares about Debian and that has the following skills:

• knows how to manage developers and software projects
  • bonus points for any experience in environments mixing volunteers and paid contributors
• is fluent and experienced enough in Python to be able to do software design and code reviews
  • bonus points for experience with: Django, Test Driven Development

That person would handle (some of) the following tasks:

• lead the “Debian project funding” initiative to a success
  • find useful projects to fund, for example by
    • discussing with various Debian teams / contributors (including the DPL)
    • running a survey among Debian developers
    • doing your own analysis
  • help with drafting and specifying the various projects
  • help to find someone to implement and review the projects
  • coordinate with those persons during execution
• manage other free software projects that Freexian would like to pursue
  • debusine: a software factory tailored for Debian packages
    • participate in design discussions, set milestones and goals
      • start with the short term needs of Freexian
      • but take into account the needs of Debian so that it can replace some aging infrastructure within Debian
    • coordinate with contractors, possibly implement some parts
  • infrastructure to run the various Freexian services and automate most of the administrative work (see “Part2: From Debian LTS to Debian for the Enterprise”)
• maybe coordinate the team of paid LTS/ELTS contributors

Debian/Python Developer

While the current priority is on the above role, there could also be room for a “developer” role with the following tasks:

• Creation and maintenance of Debian packages
• Technical support
• Software development in Python (debusine, internal infrastructure)
• Security support (contributor to Debian LTS)

Sales manager / sales representative

Up until now, the growth of Freexian has mostly been organic, through “word of mouth” and increased awareness of Debian LTS within the Debian community. We never spent a single euro on advertising, except for one promotional video and for Debconf sponsorship (with a flyer and stickers).

But if we can manage to make a positive impact on Debian through the funding that Freexian brings, then I’m interested to grow the company so that we can pay more people to work on Debian. That growth likely would have to go through some more active sales work. At the same time, it is an opportunity for me to delegate (some of) the administrative work that lies solely on my shoulders (invoicing, day to day customer relationship, etc.).

I assume it will be hard to find a member of the Debian community that has an interest in those areas, but who knows…

This article is to be continued in an upcoming post. Stay tuned!

Note: This is the continuation of part 1 where I presented Freexian and its purpose.

Going forward: growing Freexian

Part 1: From “Debian LTS” to “Debian for the Enterprise”

Freexian’s “Debian LTS” service has so far been entirely successful, with a steady growth over the years. Thanks to this, and even if there are always new challenges, it is fair to say that the Debian LTS team has met its goal in the last few years.

While this started from the desire to make LTS a reality, many sponsors are only looking for a way to give back to Debian through their company, and to make sure that Debian fits their needs.

But if you look at the bigger picture outside of this small LTS area, you will easily find many issues that need to be addressed if we want Debian to meet the needs of corporate users. Those issues can have widely different types and complexity. They can be as simple as missing the latest upstream version for an important package because the maintainer disappeared and nobody noticed before it was too late (i.e. the release was frozen); or a somewhat basic piece of software not yet packaged at all; or a release critical bug that was left unattended. On the other end of the spectrum, some corporate requirements will prove tougher to solve, for instance for large software suites that are complex to package, or could potentially have an impact elsewhere in Debian.

Bringing those facts together, we would like to have Freexian’s “Debian LTS/ELTS” offering evolve into a more general “Debian Software Assurance” offering, where you commit to a yearly budget for Debian sponsorship in the larger sense. That budget would fund different “projects” and the allocation between those projects would vary over time depending on the desires and needs of the sponsors/customers:

• Technical support: the budget would always ensure that you have a few spare hours of technical support available in case you need them
• Debian LTS: we want this to continue!
• Debian ELTS: when the customer has not managed to migrate their Debian servers in time, they should be able to reallocate their budget towards ELTS and ensure their servers are secure until the migration has taken place.
• Debian for the enterprise
  • Make sure that the packages used by sponsors are in good shape in Debian Testing/Unstable so that they are in the best shape for the next stable release.
  • Package new software that are relevant for corporate users. Offer to pool the maintenance work.
  • Fix bugs that customers are hitting.
  • Etc.
• Debian project funding: that’s the variable part of the budget (and would have a minimum of 10% like we do for Debian LTS right now). When the other projects do not consume the whole budget, we invest the remaining money into generic Debian improvements.

This major shift in our offering would also be an ideal opportunity to build a professional, free-software based infrastructure aimed at sustaining this business, making it easier to administer the various aspects of this work, and easily allowing many more sponsors to join (individuals included!).

On a more pragmatic/operational note, this shift will bring a lot of challenges to the table, and those can hardly be handled with the current resources of Freexian: if we hope to properly implement this new strategy, we’ll need some additional help.

This article is to be continued in an upcoming post. Stay tuned!

TLDR: Freexian’s success means that we have resources to invest into Debian projects. Plainly offering money has not worked so far, so I am looking to hire a “project manager” whose work would be to help spend that money in useful ways. At the same time, Freexian needs to adapt to cope with the growth: with new employees, with new infrastructure and a new offering. I want to give an idea of where we are headed, to try to inspire persons that share our values and our desire to improve Debian. Read on if you are interested.

Note: The original text has been split into 4 blog posts that will be published over a few days.

Introduction

Freexian is an IT service company specialized in Debian. We provide technical support by email on Debian, we create and maintain Debian packages requested by our customers, we also help organizations run an entire Debian derivative (Kali Linux being the most notable one).

On top of this, it runs the commercial part of the Debian LTS service : Freexian invoices many sponsors that need long term support, and uses the money to pay Debian contributors (about 12 currently) to make sure that Debian releases are supported for 5 years instead of 3. With the Extended LTS service, we push that further to 7 years, however only for a smaller subset of packages and in a repository that is hosted outside of debian.org.

Freexian’s purpose

When I created Freexian, it was out of a desire to be paid to work on Debian, and to be able to contribute during work time to the project that was so important to me. That goal has been met a long time ago.

But ultimately what I strive to achieve for Debian is not entirely aligned with the work that Freexian’s customers are requesting. That’s why, in the “long term projects” of Freexian, I always kept “find a business model that can fund the Debian projects that I would like to do”, as well as “if that model works for me, build something so that other can benefit from it too”. The first occasion to experiment something appeared when Debian discussed Long Term Support and when I stepped up to setup a commercial offer to pay Debian contributors.

Step 1: Paying Debian contributors for LTS work

When we started the Debian LTS service, I voluntarily opted to use an hourly rate that was rather high so that any Debian developer regardless of their geographical location,  could participate and not earn (much) less than what they would have from working on other tasks. This choice did imply paying a very high rate for some countries, but I didn’t see that as a problem, quite the contrary: if a Debian developer can earn enough money to cover their cost of living with 15h of Debian LTS, and then spend the rest of their month contributing on Debian, all the better! I’m not sure if anyone made this choice, but that was a dream of my younger self…

From a personal standpoint, the launch of Debian LTS has meant less free time, more administrative work, new duties to coordinate a team of paid contributors, more communication with many Debian-using companies, and many new opportunities too! This ultimately resulted in the launch of Extended LTS and PHP LTS, both of which have been rather successful so far.

Step 2: Funding Debian projects

With the growth of the Debian LTS service, and given that we have reached the required funding level, we decided to put a small share of the revenues aside and use that to fund useful Debian projects, typically in areas that were affected by our Debian LTS work. This effort was fully formalized in the project-funding git repository. We announced this process in November 2020, and we have kept mentioning it in our monthly LTS reports ever since, but so far only a single project has benefited from this. 

This is really the dream offer that I wish had existed when I was younger and was still struggling to get enough customers: hence I don’t really understand this lack of interest. You can find some discussions over the reasons why this offer has not (yet) found its target audience in this debian-vote thread.

I was hoping that spending money would be easy, but I now realize I was wrong! I’m positive that I could find dozens of useful projects to fund, but I just don’t have the time for this extra effort on top of my regular Freexian duties. I still really want to put this money to good use, which is why I’m looking into some solutions.

This article is to be continued in another upcoming post, stay tuned!