FEP-302f: QuantPub — federated personal metrics for quantified-self sharing

,
8

Thanks @fiddur, this is great progress. The 0.3 direction covers nearly everything we raised, and breaking changes are fine with us at this stage. A solid 1.0 matters more than staying compatible with today’s drafts.

Our plan: VertMatch ships 0.2 in our next release, and we’ll move to 0.3 as soon as the PR lands. We’re happy to be an early implementer and test interop with Aurboda along the way.

We agree with all eight points. A few small requests while the PR is drafted:

  1. Day precision: under timePrecision: "day", could receivers be told to show the date in the offset it was sent with, not convert it to the viewer’s time zone? Otherwise a run sent as local midnight can appear on the previous day for someone further west.
  2. Unitless extension metrics: since unit becomes required on implementation-defined keys, can we confirm that UCUM 1 or an annotation like {TEP} is fine for counts and scores?
  3. Segments: could startOffset/endOffset be optional, so a segment can be described by distance alone? We’d like to publish climbs without timing them.
  4. Transition: a missing unit means different things in 0.2 and 0.3 (for example km vs m for distance). A short note that readers should use the discovery document’s quantpub version to interpret unit-less metrics would prevent some confusing bugs.
  5. Still open from my earlier post: whether in-band quant:metrics stays normative (this matters most to us, so receivers can show a card without a fetch per post), keeping quant:Exercise, and how capability tokens rotate when a follower is removed.

If useful, I’m happy to review the PR.

Thanks again for driving this forward!

New user rules preventing discussion