General mtg that meets every 4 weeks.
Nick: We like to have anyone who is new introduce themselves. This mtg can be an intro to FDC3.No one new on the call.
Working Group Updates
Updates from all WG. Should Updates should be given from the chair.
API WG
...
...
- The past 4 weeks of August has been a slower cadence than we should see i the coming weeks.
...
- Has been going back and forth with some fundamental questions on the scope and standards.
...
- PR has been approved although the spec hasn’t been fully approved.
...
- Highlights are to try to make clear how one can build a more tighltly coupled RPC system on top of the standards/ Intention.
...
- There is an idea of hte desktop agen that exposes the FDC3 API to the app nd fulfills the interop functionality on behafl of the app. It will connetct ot one or more app directories and multiple agents. Will resolved intents, launch apps, context data. Most of what needs to be standardised are the interfaces.
Context Data
Tim - Had our first context data mtg two weeks ago. Had discussions around how you can use context data, how you can assign particular types. Have some examples but not actual specs. Specifying what an FDC3 vs Other context would be and give some examples.
Riko: Not sure where it fits. What is interesting - is there a way to centralised. Do we need a registry of types? Where you store the types, shape of them, e.g. where does a contact live, what does it look like, how do you suggest a new field?
Nick: It’s a fine line there between context vs object modelling (which is part of FO). We still need to work closely with them
Leslie: One thing I have proposed is that there might be, e.g. 3 different copies of the context data describing the client differently (symphony, proprietary, Tick42) - can be different types or different identifiers (email address, symphony identifier, twitter handle, etc.)
Good topic and there will be continued discussion.
Explore how to assign multple contexts -- mixed types … will it be tied with multiple identifiers
Differentiation what an FDC3 contact might look like (and how that compares with Symphony)
Riko ( JPM): Is there a need to centralize how the context that participate in the interop that participate
Tim: Context is what came up due to the lowest common demoninator -- but as far as finanical onbjects, or the payload and the data, that would be the work of the financial objects working group (sic) [I think they mean the Financial Objects program]
Nick: Can do this through different type, but also through different identifiers -- idea is that you can have any number of identifiers. One thing that would be valuable would
Intents WG Update
Espn could make it. Nick to give update for Espen.
There is a proposal for initial set of intents and schema visible in GitHub. Waiting on some feedback from Use Cases and API WGs.
...
Leslie: The AppDirectory says I can publish these intents and tends to specify context API. General idea is that an intent is invoked with a context and that they work closely together with
Use Cases WG
Kat gave here update
- We’ve been going through both what is a use case as well as the governance of approval
- 3 stage process, everything managed in confluence
- Ad hoc
- Submitted for approval
- Approval / Rejected
- Approval → discussions with relevant use cases; flesh out which WGs need to be involved
- Acceptance guidelines
- How a long discussion of this
- What personas (actors) should we include -- e.g., do we include a CTO in this?
- Decided to do a couple technologists use case to start
- Initial traunch of use cases -- then there will be a vote after the meeting on Thursday; anyone who comments can vote on accept or reject on the initial set.
- We are not going to manage it like this going foward (they will do ad hoc, not by set)
- JT: this is really exciting -- to be asking “is this valid?”, “Is this useful?” -- it is important the people evaluate against a set of criteria -- we are asking participants to review against a set of criteria … vote will happen by email following the call and then close on Mondy