Skip to main content

Privacy

Dasher is free, open-source software. We respect your privacy. This page explains exactly what data we do and don't collect.

The short version

  • We never collect what you type. Not your text, not your clipboard, not your Dasher canvas.
  • Analytics are strictly opt-in. If you don't opt in, nothing is sent.
  • No ads, no trackers, no third-party data sharing.
  • Analytics data is anonymised — no name, email, or device ID tied to you.

What Dasher collects today

Nothing. Dasher v6 does not currently send any data to any server. The app runs entirely on your device. Your settings, text output, and training data stay local.

Proposed: optional analytics

We have proposed adding opt-in, anonymised analytics to help us understand how Dasher is used and fix crashes. This is described in detail in RFC 0001 and has not yet been implemented. When it ships, this is how it will work:

If you opt in, we collect:

Data point Example Why
App launched macOS 15.3, Dasher 1.2 Count active users, know which OS versions to support
Input method selected "touch", "switch", "eye-gaze" Prioritise development on the methods people actually use
Alphabet/language chosen "English", "Slovak", "Japanese" Know which languages need support and testing
Settings tab viewed "customization", "speech" Understand which features people look for
Crash report Stack trace, OS, app version Fix bugs that cause crashes

We never collect:

  • Typed text or symbols
  • Clipboard contents
  • Dasher canvas contents or node positions
  • Your name, email, or account information
  • Your IP address (scrubbed at the server)
  • Training text contents
  • Game mode target text
  • Anything from the iOS keyboard extension

How analytics works

  • Opt-in only. On first launch (once analytics ships), you'll see a prompt asking if you'd like to help improve Dasher. If you decline, no data is ever sent. If you accept, you can change your mind at any time in Settings → Privacy.
  • Anonymous. A random ID is generated on your device. It cannot be linked back to you personally. You can reset it in Settings → Privacy.
  • Self-hosted. Data is sent to a PostHog instance controlled by the Dasher Project, not to a third-party analytics company.
  • Offline-friendly. Events are queued locally and sent in batches when you're online. Dasher works fully offline.
  • Open schema. The full list of events and properties is published in the RFC and in each repo. If it's not listed there, we don't collect it.

Your choices

Action How
Opt out of analytics Settings → Privacy → turn off "Help improve Dasher"
Opt in later Same location — it's always your choice
Reset your anonymous ID Settings → Privacy → "Reset analytics ID"
Use Dasher completely offline Just decline analytics. No network activity occurs.

Data retention

If analytics is enabled, anonymised event data is retained for 13 months, after which it is automatically deleted. This aligns with PostHog's default retention policy. Crash reports are retained for the same period.

Third-party services

Dasher does not use any third-party advertising, tracking, or analytics SDKs. The only external service used is the Dasher Project's own self-hosted PostHog instance (and only if you opt in).

Children's privacy

Dasher is used by people of all ages, including children in educational settings. We do not knowingly collect any data from children. The analytics opt-in is a deliberate action — it cannot be triggered accidentally or by a minor without understanding the prompt.

Open source

Dasher is open-source (MIT licensed). The analytics implementation, event schema, and opt-in flow are all visible in the source code. If you don't trust us, you can verify exactly what's sent — or build Dasher from source with analytics removed entirely.

Changes to this policy

If we change what data we collect, we will update this page and notify users in the app. Any expansion of the event schema will go through the RFC process with community discussion.

Last updated:

Contact

Questions about privacy? Open an issue on GitHub Discussions or contact the project lead, Will Wade.