openeeg-spec

Roadmap — v0.1 to v1.0

Last updated: 2026-06-28 (v0.1.0 draft)

This document tracks what we plan to ship, in what order, and what “done” looks like. Dates are targets, not commitments — community feedback may reprioritize.


v0.1 — “Sketch the keystone” (now → 2026-09)

Goal: publish a draft spec, prove the schema is implementable, gather community feedback.

Deliverable Status Target
Spec v0.1 (this document) ✅ Draft published 2026-06
Apache 2.0 LICENSE ✅ Published 2026-06
README + manifesto + example ✅ Published 2026-06
Public spec repo (github.com/ai2nlp/openeeg-spec) ✅ Live 2026-06
openeeg.org site (landing page) 🟡 In progress 2026-07
Recruit 3-5 early reviewers 🟡 In progress 2026-07
Spec v0.1.1 (incorporate early feedback) 🔴 Not started 2026-08
Spec v0.1 lock (community call for breaking changes) 🔴 Not started 2026-09

Done when: at least 3 independent implementers (humans or AIs) can write a parser for v0.1 from the spec alone, without asking me questions.


v0.2 — “Ship the reference implementation” (2026-10 → 2026-12)

Goal: prove the spec is implementable in code, ship the first converters, get the first real users.

Deliverable Status Target
Reference implementation: Python (NumPy-based) 🔴 Not started 2026-10
Validator CLI: openeeg validate file.openeeg.json 🔴 Not started 2026-10
Muse 2 converter: openeeg convert --from muse2 🔴 Not started 2026-11
OpenBCI converter: openeeg convert --from openbci 🔴 Not started 2026-11
Olimex EEG-SMT (P2) reader: openeeg convert --from openeeg-p2 🔴 Not started 2026-12
PyPI package: pip install openeeg 🔴 Not started 2026-12
GitHub Actions CI: lint + validate every PR 🔴 Not started 2026-12
First 10 real users (hobbyists + researchers) 🔴 Not started 2026-12

Done when: a Muse 2 user can record a session, convert it with one command, and load it in MNE-Python with zero manual file editing.


v0.3 — “Reach the indie-research tier” (2027-Q1 → 2027-Q2)

Goal: the format is usable for real research workflows, not just hobbyist demos.

Deliverable Status Target
PiEEG converter 🔴 Not started 2027-Q1
Compressed sample format (gzip wrapper) 🔴 Not started 2027-Q1
Epoched data support (epochs block) 🔴 Not started 2027-Q2
Cross-session dataset description (BIDS-style sidecar) 🔴 Not started 2027-Q2
MNE-Python integration PR: mne.io.read_raw_openeeg 🔴 Not started 2027-Q2
First dataset published in OpenEEG format 🔴 Not started 2027-Q2

Done when: a researcher can publish a 50-session dataset in OpenEEG format that another researcher can load + analyze in MNE-Python without writing a converter.


v0.4 — “Governance + community” (2027-Q3)

Goal: the format is no longer dependent on a single person.

Deliverable Status Target
Working group charter (governance model) 🔴 Not started 2027-Q3
Multi-maintainer repo access 🔴 Not started 2027-Q3
Code of Conduct 🔴 Not started 2027-Q3
First 10 external contributors (non-founder PRs merged) 🔴 Not started 2027-Q3
Discord / Matrix / community channel 🔴 Not started 2027-Q3

Done when: at least 3 active external maintainers, governance is documented, and the project can survive me stepping away for 6 months.


v1.0 — “Lock the spec” (2027-Q4)

Goal: the spec is stable enough to be cited in papers and depended on by tools.

Deliverable Status Target
Spec v1.0.0 (no breaking changes from v0.x) 🔴 Not started 2027-Q4
At least 3 device converters shipping stable releases 🔴 Not started 2027-Q4
At least 1 analysis tool with first-class OpenEEG support 🔴 Not started 2027-Q4
First paper / thesis citing OpenEEG 🔴 Not started 2027-Q4
Migration guide from v0.x to v1.0 🔴 Not started 2027-Q4

Done when: a tool author can write “OpenEEG support” in their release notes without disclaimers.


Beyond v1.0 (candidates)

These are not committed. They’re the v1.x candidates I’d want to explore if there’s community interest:


What is NOT on the roadmap

To keep scope honest:


How to influence the roadmap

The roadmap is a living document. Items get promoted, deferred, or killed based on community signal.


License

This document is licensed under Apache License 2.0.

Copyright 2026 AI2NLP Solutions Inc.