An editor for documents that matter

Why Etherpad exists, what it's for, and what it won't do.

The problem

Most collaborative editors are designed to forget.

They show you the current version of the document. They hide the history behind menus. They obscure who wrote what. They live on servers you don't control, governed by terms you don't write, owned by companies whose business model is the opposite of trust.

For most documents — a shopping list, a meeting note, a draft email — none of that matters.

For some documents, it matters enormously.

The documents that matter

If a country were drafting a new constitution, where would they write it?

If a treaty were being negotiated between governments, where would they draft it?

If a scientific paper were being co-authored across continents, where would the record of authorship live?

If a piece of investigative journalism were being assembled from sources who need to know that their words won't be silently edited by a platform, where would they write?

If a school were teaching children that what they say belongs to them, what tool would they put in front of them?

These documents share something in common. Provenance is not a nice-to-have — it is the entire point. Who said what, when, and why is part of the meaning of the document itself.

What Etherpad is

Etherpad is the editor for those documents.

Every keystroke is attributed. Every revision is preserved. Every change is reversible. The timeslider lets you scrub through the entire history of a document, character by character, watching it come into being. The author colours aren't a hidden feature — they're the product. They make authorship visible, by default, to everyone who reads.

And the whole thing runs on your server, under your governance, with no telemetry, no upsells, no terms of service you didn't agree to. AI is a plugin you install, pointed at the model you choose, running on infrastructure you control — not a feature decided for you in a boardroom you weren't in. The code is Apache 2.0. The data format is open. The history is yours.

Our principles

• Honest. We tell you what the software does and what it doesn't.
• Open. Source, format, governance, roadmap.
• Transparent. No hidden state. No invisible edits. No silent changes.
• Malleable. A plugin system that lets you make Etherpad fit your work, not the other way around.
• Accessible. 105 languages. Runs on a Raspberry Pi or a server farm. Works in any modern browser.
• Truthful. The document tells the truth about itself. Who wrote it, when, and how it got to be what it is.

What we won't do

We won't pander to the trends. We won't add surveillance. We won't pivot to extraction. We won't bury the authorship to make the UI cleaner. We won't give up the values to grow the userbase.

The world is trending the other way. We're holding the line.

A short history of holding the line

Etherpad has been quietly used by Wikimedia, governments, EU public-sector institutions, universities, and self-hosters around the world since 2009. No pivots. No acquisitions. No enshittification. Just sixteen years of doing the same thing, well.

That kind of stability is itself a feature. Institutions that adopt Etherpad in 2026 can reasonably expect it to still be Etherpad in 2036 — still open, still self-hostable, still attributing every keystroke, still owned by no one.

Almost no other software in this category can credibly say the same.

What you can do

If this matters to you:

• Run an Etherpad for your team, your organisation, your school, your community.
• Contribute — code, documentation, translations, plugins, time.
• Tell someone else this exists. A generation of developers and decision-makers grew up after Etherpad's first wave of fame and have never heard of it. Word of mouth is how this kind of project survives.

The documents that matter deserve an editor that takes them seriously.

That's what Etherpad is for.