Why most #geekproblem software fails: Trust vs. control

Almost all of our #geekproblem software fails because it’s built with a mindset of control.

Control over users.
Control over systems.
Control over outcomes.

But all good societies, and all durable communities, are based on trust. When we ignore this, we don’t just write bad code, we produce #techshit that nobody uses, that burns out developers, and that confuses users. Then we start over… and call it “innovation.” That’s #techchurn.

Control-driven projects: Examples of failure

Diaspora
Touted as a Facebook alternative, it focused too much on cryptographic control and data silos — and forgot the social UX that makes people actually want to use social media. It never recovered from this early design flaw.

GNOME Online Accounts
Supposed to be a bridge between the desktop and online services. Instead, it became a privacy puzzle with unclear consent and broken trust. Control was enforced without social understanding.

Secure Scuttlebutt (SSB)
A radical peer-to-peer network, very promising. But became increasingly unusable due to overcomplicated trust mechanics and lack of simple social pathways for onboarding new users. The community stalled.

Matrix / Element
Still pushing forward, but has constant friction because it replicates many centralised “control” models in the name of “choice.” Powerful, yes. But still struggles with real decentralised trust outside geek bubbles.

🌱 Trust-Based Systems: What Works?

Fediverse / Mastodon
It works because it’s socially familiar and based on human trust over algorithmic control. You choose who to follow, what server you trust. And it grew because of this — not in spite of it.

Signal (Early Days)
Before turning more into a consumer app, Signal succeeded by focusing on trusted networks — your phonebook — and making end-to-end encryption invisible. It was about trust, not just security.

The real problem is in part to do it money and the funding of the wrong side of tech, in that most funding goes to things that feel safe:

Protocol development

Core backend infrastructure

“Governance” initiatives run by “neutral” NGOs

These are important up to a point, but this “safe” money ONLY reproduces the #geekproblem:

Building tech without communities

Tools without culture

Features without stories

When we do try to fund the social side, the interfaces, user onboarding, documentation, actual relationships, it too often gets handed to parasite #NGOs with no grassroots accountability. Just look at the endless pilot projects by digital rights NGOs that are abandoned 18 months later. Or the “governance frameworks” that never go anywhere. It’s a cycle of buzzwords over boots-on-the-ground.

The people with shovels, in a messy world, the only thing that might work is messy people with shovels, people who compost the shit, clean the broken tools, and patch the networks to keep things going.

These people are rarely funded.
They’re not “scalable.”
They don’t write grant-friendly proposals.
But without them, none of the tools work.

Who funds them?

A call to action: If we want an #openweb that survives the coming waves of #climatechaos and #mainstreaming sellouts… We need to fund trust, not control, to support social infrastructure, not just servers and specs, to back messy doers, not polished whitepapers. We need to talk about this, fund this, and build on this, or we’re just making more compost for the next #dotcons to grow from.

A conversation about money and the #openweb

#NLnet #NGI #NGIzero #EU #funding