Minutes actions review from last discussion

Looking back at first principles agreed for AppDirectory the initial goal was to promote strong identity to enable peer-to-peer connections and the ability to verify the veracity of the systems interacting. Given that these are not closed systems AppDirectory is meant to act as an authority for identity. Discovery and the positioning of FDC3 is purposefully lightweight to support peer-to-peer interaction rather than producing a standard that is an “uber container.”

Nick reviewed meeting materials from an early FDC3 meeting, which can be found here. This is publicly available in FDC3 Archive in GitHub so please feel free to go back and review.

Frank pointed out that it’s important to be reminded of the purpose of FDC3 and of keeping the standards as simple as possible to meet the requirements of passing the relevant and essential information between the applications. Tick42 has been vocal in wanting to pass more detailed information, which makes sense in many cases but could also potentially limit uptake.

Fundamental question for the group is to agree on how much to put into the first version of the specification.

Following some continued discussion it was agreed that Frank would put in a PR to remove appType. It is important for the group to review this and comment on it.

Riko/All

There was discussion via email regarding whether or not the AppD WG is an appropriate forum to discuss the need for a common registry to facilitate sharing of global intents and context type definitions (in addition to an app directory for applications, necessary to be able to deliver functionality required by the API, Intent and Context Data Working Groups).

This was also discussed during the meeting including these points:

• It is not the right time to add this as a requirement given the focus on completing a first version of the specification.
• There is currently discussion in the FDC3 PMC and in the Financial Objects Program about where ratified standards and the future catalogue of models should live and also how this gets disseminated. To date, the thought has been that a spec would live on Confluence or GitHub and for APIs, Intents and Context Data there would be a typsecript definition in GitHub and npm. This may be a suboptimal way to consume the info if you are building an app directory.
• There are still a number of open questions: do we need to catalogue types, namespace the types, do we need a mapping of intents in the context of an app vs the global list of intents, etc.
• There will be need for a repository and guidance as people start to develop against the specs in earnest and there are a number of approaches from as simple as a consistent URL or more complicated like a production web service, or an npm module that publishes a package.
• This is a good topic for the General Meeting as it stands across multiple working groups and gets to the heart of what we want FDC3 to do for us vs. what we want to do on our own.

In wrapping this conversation up (for the meeting), Tim pointed out that getting an MVP out is most important right now and then extending this as the group gets more information, additional use cases and feedback from initial implementations.

Changing the scope of WG or consolidating WGs is a decision of the PMC, requiring input and advice from all of the WGs, which is part of the reason for the general meeting.