– DRAFT –
Linked Web Storage

Introduction and announcements

acoburn: this years tpac in japan
… nice to have face to face in Belgium in the fall

<jeswr> brief announcement from my side there will be a presentation on W3C Linked Web Storage (LWS) at https://globaldigitalcollaboration.org/agenda?day=2025-07-02

acoburn: early October...somewhere in Belgium..more details coming

jeswr: there is a presentation of LWS at the Global Digital Collaboration conference, tuesday wednesday this week

acoburn: # of folks including Pierre-Antoine at that conference

Use cases: scope and status

acoburn: fairly significant PR fgrom eBremer
… aggregates the use cases using LLMs
… it would be good to get that moved into the repo
… one of the open questions came from uvdsl whether some things were in-scope or not
… there will be time for us to discuss what is or isn't but right now is not the time top explicitly indicate this
… use cases are all-important, there will be UC that will not be addressed by the protocol...
… somethings are lower priority. Even though they may not be part of the protocol, I would be inclined to be broad in the scope at this time and trim later
… make sure that folks feel the same or different right now

<bendm> +1

cpn: you talked about having a set of use cases that is part of some future version..
… like to see a set of use cases that describes version 1
… and a protocol that addresses them...

acoburn: i agree with you. in a perfect world we would have a set of use cases that maps to what is in the protcol
… but there will be a winnowing process at that point to derive those requirements
… but there will be further winnowing as prioritization is set
… are charter says we will not create new use cases, so we will definitely not have these use cases

cpn: i agree with having a broader scope in the use cases, which features included in a version 1 or deferred to a future version
… if there are features that go in that we dont have use cases for i would be concerned

acoburn: I agree with you that we dont want things in the protocol that dont map to a use case

ericP: try to massage use cases to get only what is in-scope
… lot of things will be out of scope, but we want to be able to add to those use cases with our tech, someone else layer on top of it
… we should be happy to include lots of use cases but make it clear what is actually in scope and which ones are at risk ..

ryey: are we still open to new cases at the moment?

acoburn: still open
… as things become more clear, there may be gaps and we should have use cases that fill those gaps

ryey: right

eBremer: I'd like PR 166 to be ready to merge by end of the day. Christoph made suggestions, I'll work on those

<acoburn> Use cases PR

<gb> Pull Request 166 Update user-stories.md (by ebremer)

eBremer: Some comments about scope, I don't disagree with what was said, but what aspect of the use cases would we address in v1 to address the use case
… Tim had one about interlinkability of medical data. Good use case, but out of scope as is. Elements could be derived from it that could go in the spec
… Pierre commented on the ToC, addressed by Hadrian
… I'll work on it today, but things can still be changed

acoburn: Please shout if you need more time, but want to get this merged

eBremer: There's additional material from ?? that I'll include as well

Deriving requirements from use cases

acoburn: how do people feel about deriving the requirements from the use cases
… we will have a set of requirements for CRUD operations
… for discoverability and capabilities

<bendm> +1 to use LLMs as a first start

eBremer: It's useful for the LLM digest everything then have humans edit. Any interest in seeing that as a starting point?

<Zakim> bendm, you wanted to ask about granularity

<ericP> +1 to asking the LLM for reqs if it seems like the ROI will motivate

<bendm> https://www.w3.org/TR/did-use-cases/#requirements

bendm: Is the level of granularity similar to what the DID use cases and requirements looks like? Those are very high level. A LLM could generate something high level like that

acoburn: What's the right level is a good question. We don't have any requirements right now. Anything will help at this point. I'd be fine starting high level, like the DID requirements
… If we find more granular would be more useful, we can do that. But I want to start, and help move forward quickly with the protocol document editing

bendm: Agree

ericP: get what you can out of the LLM
… any work you are willing to put into it

eBremer: Happy to work on it, then it's up to the group whether to use it or not

<kaefer3000> fyi - here is another use-cases and requirements document that works with more details https://www.w3.org/TR/sdw-ucr/

kaefer3000: that was out blueprint that we tried linking deliverables
… arranged quite differently from DID one
… it did not have use cases and requirements so much in the headlines
…

acoburn: thank you very helpful
… if we can move forward on the use cases...
… dont know have time to have requirements for next week
… to see this momentum

eBremer: I can create a draft PR this week

acoburn: any other thoughts on deriving requirements?

Updates on research teams

acoburn: last week we talked about research teams taking on areas of protocol document
… areas like authorization, discovery....
… wanted to check in on where things are

jeswr: if there are any areas people want to lead on
… or people that you think would make a good invited expert

jeswr: Jackson put his hand up for discovery
… authorizations another big one

<Zakim> bendm, you wanted to put the areas on the mailing list?

bendm: on the mailing list

acoburn: fantastic idea

<kaefer3000> +1 to sending this over the mailing list

acoburn: Jesse put on email

acoburn: any other items to go over?