« Other digests

Redecentralize Digest — November 2020

Digital public infrastructure

In the “Mapping Social Media” project of the Knight First Amendment institute, Ethan Zuckerman analyses the ‘logic’ of several types of social networks (for example ‘chat logic’, ‘civic logic’ or ‘crypto logic’), by comparing networks along five axes: their technology, revenue model, ideology, governance and affordances. As different logics lead to different dynamics and experiences, the goal is to help us imagine other models than those we got used to.

One essay in the series starts off with the question “What if Social Media Worked More Like Email?”; it focusses mainly on Mastodon, and concludes with a narrative and outlook that resonates with us:

“Decentralized platforms are exciting because they hearken back to the original, decentralized vision of the web. Yes, that vision had its flaws—namely the conviction that the web would lead to a digital utopia—but the values at its core are still important today in our world of digital gatekeepers. Unfortunately, decentralized platforms face significant challenges to their widespread adoption. Mainly, addressing their usability difficulties and overcoming the massive network effects of centralized platforms. We are not convinced decentralized platforms can face these challenges alone. Interoperability with existing platforms is likely a key precursor to the widespread adoption of these platforms and may require interventions from governments that require platforms like Facebook and Twitter to adopt open standards or share their APIs and social graphs. Ironically, it seems like for the decentralized movement to succeed, it may need to partner with the governments it once hoped to transcend.”

The current essay series follows on a previous writing where Ethan made the case for a change in economic models for the internet, to regard them as “digital public infrastructure”; which he illustrated with a historical comparison with the different models for radio broadcasting — the hot new medium a century ago.

Much of the internet ecosystem should be considered digital public infrastructure, not just cables, but also search engines and social networks, as made clear in his recent essay that expounds on this topic: “Infrastructures are things we build so we can build other things”. Because of this, infrastructure creates externalities, i.e. benefits (and problems) for other people than the infrastructure’s direct customers, which causes that “markets don’t always provide the infrastructures we need”. Hence he argues we should use public funding to build it, and build it around civic values instead of surveillance capitalist logic; which first of all means making it decentralised and federated.

Radically distributed code collaboration

GitHub has become a nearly inevitable platform for software developers to collaborate on their code, especially for open source projects. While it is built around the git versioning tool, which is elegantly decentralised, the collaboration features GitHub adds (e.g. issue trackers and pull requests) deviate from this design and thereby lock people in to the platform. Various alternatives to GitHub exist that are open source and self-hostable (e.g. GitLab, Gitea), but even those are not (yet) capable of interacting across instances, hence each project relies on a single server where all its (aspiring) contributors will first have to sign up.

Radicle creates an alternative that replaces these collaboration platforms altogether, using a peer-to-peer approach. As they explain:

“Radicle adopts the Scuttlebutt social overlay paradigm by establishing a peer-to-peer replication layer on top of distributed version control systems, starting with git. User accounts and login is replaced by public key cryptography, hosted issue trackers are replaced by local peer replication, and the idea of a single canonical upstream is replaced by a patch-based peer-to-peer or “bazaar” model.”

After first trying to build such a system on top of IPFS, the developers turned to using git’s own data exchange mechanism, by storing its data the git repository itself:

“Storing collaboration artifacts (issues, pull requests, comments, …) in git has been done before and the data structures available in git satisfy all our needs. Paired with a gossip layer, git becomes exactly what’s necessary to store, replicate and distribute code and collaboration artifacts.

By building a peer-to-peer overlay on top of git, we find not only a performant solution, but one that is better adapted for code collaboration. Issues, comments and reviews become local artifacts that are cryptographically signed and interacted with offline.”

For those curious to try, Radicle’s beta version has just been released.

How to fix the internet

The Electronic Frontier Foundation created a podcast mini-series called “How to Fix the Internet”, with each of 6 episodes covering an issue with current (US) tech policy. The fourth episode, on our favourite topic of interoperability, concludes with some motivating words for those who doubt which strategy is the right one. All of them.

Cory Doctorow: “… And the four directions we can go in are: norms, conversations about what’s right and wrong; laws, that tell you what’s legal and not legal; markets, things that are available for sale; and tech, things that are technologically possible. And our listeners … have different skill sets that they can use here, but everyone can do one of these things, right? … So, everybody has a role to play in this fight, but there isn’t a map from A to Z. That kind of plotting is for novelists, not political struggles.”

Cindy Cohn: “… I think it’s true for almost all of the things that we’re talking about fixing, is that the answer to, “Should we do X or Y?” is “yes,” right? … These are multi-strategic questions. Should we break up big tech? Or should we create interoperability? The answer is, yes, we need to aim towards doing a bit of all of these things. There might be times when they conflict, but most of the time, they don’t. And it’s a false choice if somebody is telling you that you have to pick one strategy, and that’s the only thing you can do.”



All are online, unless noted otherwise.

About this digest

The Redecentralize Digest is a monthly publication about internet (re)decentralisation. It covers progress and thoughts relating technology and politics, without ties to a particular project nor to one definition of decentralisation — figuring out its meanings and relations is part of the mission.

This edition was written by Gerben, with thanks to all who contributed.

The digest’s format and content are not set in stone. Feedback, corrections and suggestions for next editions are welcome at hello@redecentralize.org. We don’t spy on our readers, so please do tell us what you think!