Trials & Tribulations on the Road to Digital Sovereignty

by Chris McGivern & Jason Hudson, 20 December 2021

To paraphrase the poet Robert Burns, the best-laid plans of mice, men, and women often go awry. It is with huge regret and disappointment that our move to open source our operational technology has been beset by too many problems. So, for now, we are pressing pause…

Earlier this year, we set out our intention to reclaim our digital sovereignty from the many proprietary services we use to organise the Foundation’s day-to-day business. Moving to open source alternatives is not just the right thing to do but an imperative in terms of privacy, security and ownership; for us, our Fellows, and our colleagues.

We stand by that decision and still intend to disentangle ourselves from the clutches of Big Tech. Monopolism and surveillance capitalism are antagonists to open. There must be an alternative. But if we are to fulfil our roles as principal actors in the story of openness, we must - at the very least - be given space to say our lines. Our experience so far has been one of backstage drama, all detracting from the main performance.

We missed important meetings due to problems with sharing calendars and sending or accepting events. Sometimes it worked; others, it didn’t, and this chaotic randomness was not self-contained within the team. It caused a chain reaction and had real-world impacts on others.

Our efforts to seek out fully-encrypted email were successful, but we didn’t fully realise that our newly-found water-tight security came with a significant trade-off. We could no longer search the body text of our emails, only the sender or email header. It was a huge sacrifice in terms of usability, particularly when those emails contain your entire institutional memory dating back to the mid-2000s. Ultimately, the features and usability of the tools we rejected became a seductive and alluring Siren’s song.

This is especially true of the open alternatives we used to replace Google Suite. We struggled to collaborate on important documents and lost a lot of work in the process. A lack of real-time collaboration meant our productivity took a nosedive. And if you are constantly second-guessing a system, you will stop using it altogether, which is hugely problematic. Furthermore, if people connected to our organisation work on company files in their personal Google accounts, it has significant implications around trust, data integrity, and GDPR compliance.

Also, open source could - and should - be a lot more inclusive. For example, our instance of open source video conferencing required hyper-stable connections., and the slightest dip in your Wi-Fi meant significant judder and delay - if not complete breakdown. Yet stable connections are still a pipedream for many. Not everyone lives in California, with all that sun and all that bandwidth.

We understand the counterarguments. Maybe it is our problem and demonstrates a catastrophic failure to adapt. Perhaps we have grown too accustomed to the usability of pervasive (and invasive) proprietary tools. Yes, of course, we want to switch away from technology companies that spy on people and build trillion-dollar industries by selling their data. But we genuinely feel it shouldn’t be this hard to ‘eat healthy food’ and participate in secure, private, open technology that works.

Workarounds exist for all the issues we encountered, and we spent hours on each one to try and make this experiment a success. But ultimately, we don’t want to have to think about email, or calendars, or video conferencing. It’s the email’s message, or the meeting arranged on our calendar, or the people we meet online that are important to us, not the technology underneath it.

So, for now, we reluctantly go back to the ‘other side’ for the sake of productivity and - admittedly - restoring some sanity. We do feel guilty, but this is not a backtrack; instead, it’s a necessary pause in our journey while we map out a new route to our destination. And we are not abandoning all of our newly acquired open source tools. We are still using Cloud68 for hosting our Gitlab instance and open source web stacks. The backend of our technology is brilliant, and our infrastructure remains open. It is the front-facing services for communicating with others that are problematic, and that we - and the open source community - still need to work out.

Posted in announcements