Meeting minutes
Introductions & Announcements
laurens: we will discuss several open PRs
… update on the notification PR
… and the storage description resource
Auxiliary Resources (PR https://github.com/w3c/lws-protocol/pull/165 )
<laurens> w3c/
<gb> Pull Request 180 Link set profile (by termontwouter)
laurens: a new PR has been opened #180
<gb> Pull Request 180 Link set profile (by termontwouter)
<rbreitman> #present
termontwouter: we merged PA's PR from last week
… this is a minor change, introducing a profile parameter
… this distinguishes a linkset resource from a dataset resource
laurens: this is a draft PR, so folks are invited to add comments there
<Zakim> elf-pavlik, you wanted to ask what were the CORS issues with profile param
elf-pavlik: are the CORS issues still relevant in this case?
termontwouter: we use a non-URI
acoburn: the string selected will not run afoul of CORS
TallTed: the new PR was just opened, what are we reviewing right now?
laurens: the PR from pchampin made #165 obsolete
<gb> CLOSED Pull Request 165 Define aux & meta in terms of representations (by termontwouter)
laurens: the goal is to introduce this pr this week and discuss further next week
termontwouter: I realize this was last-minute, but left in draft mode before moving to a vote
laurens: next week we can proceed to a vote if there are no significant changes
Replace GET/POST with new HTTP QUERY RFC10008 (PR https://github.com/w3c/lws-protocol/pull/179 ) agendabot]
laurens: this PR was opened by eBremer to replace GET/POST with the QUERY verb
eBremer: the QUERY RFC was advanced as an IETF proposed standard only a few days ago
… this PR adjusts the type index from GET/POST to only the QUERY operation
… this RFC (QUERY: RFC 10008) seems to be gaining a lot of support very quickly
laurens: one question is whether we want to give implementers the option of GET/POST *and* QUERY
<pchampin> +1 to allow a fallback to POST
laurens: would like to hear from the rest of the working group on this
TallTed: clients can ask servers what methods they support
… I don't expect QUERY will have full support in the near future, but we should be able to talk about it
… I don't think we should *require* QUERY at this point
laurens: at present, the type index is optional, so we would not be imposing QUERY on anyone
… please continue discussion on the GH pull request
Notifications (PR https://github.com/w3c/lws-protocol/pull/162 ) and integration into the core specification
laurens: I hope to open a PR this week to integrate notifications into the core protocol
… in addition, we still have PR #162 open (additional notification channels)
<gb> Pull Request 162 feat(notifications): Add WebSocket and Server-Sent Event channels (by laurensdeb)
laurens: at present, only webhooks are defined
… over the next few days, a PR will follow
Storage Description resource as a CID (PR https://github.com/w3c/lws-protocol/pull/183 )
acoburn: This is the result of conversation at the F2F meeting
… The storage description resource looks a lot like a CID.
… So the question is, should this be a CID?
… My intention here is to just introduce this.
… The PR contains three notable changes.
… Firstly, the media type application/lws+cid indicates this is a specialization of a CID.
… Secondly, there are now two required contexts.
… The core CID context, and a specialized LWS context.
… This avoids us copying things in the LWS context.
… The JSON itself however looks exactly the same.
… We can continue the discussion now or in the PR.
<Zakim> ericP, you wanted to ask about tools that can work with CIDs
ericP: Are there tooling and libraries that work generically with CID that we can re-use?
acoburn: I don't know in particular.
… I would imagine a lot of libraries rely on JSON parsing.
… As a CID is a JSON document.
… pchampin might know more.
<pchampin> I'm afraid I'm not
<pchampin> but will look into it
Review Project Spreadsheet
laurens: (sharing screen)
… related to test suite, do we have updates from ericP?
ericP: no significant updates yet, haven't run the suite against a live environment
laurens: we will have an update from Samu on the third Monday (July 20)
… what is the status of the system diagrams jeremycaine?
jeremycaine: there are some debates on terminology, e.g. virtual resources #150
<gb> Issue 150 Define Virtual Resource (by renyuneyun)
laurens: we discussed virtual resources a few weeks ago
laurens: we could put this item on the agenda for next week
… we need to make decision on what to do with virtual resources
… consensus seemed to be that we shouldn't preclude virtual resources but won't include them in spec
<Zakim> acoburn, you wanted to mention virtual resources
<laurens> acoburn: I also recall it this way, we should not preclude virtual resources.
<laurens> ... If diagrams mention virtual resources, these should be reworked so they are not mentioned.
jeremycaine: my assertion is that virtual resources are not clearly defined
laurens: let's pick up the topic of virtual resources next week
<elf-pavlik> #177
<gb> Issue 177 Create Threat Model Note (by elf-pavlik)
laurens: for existing PRs that mention virtual resources, they will likely not be merged for now
<Zakim> elf-pavlik, you wanted to invite jeremycaine coordinating on diagrams
elf-pavlik: in the issue for the threat model, I would like to invite jeremycaine to work on the diagrams
… I used the threat modeling language from the W3C guide
<jeremycaine> yes elf-pavlik
laurens: add introduction and diagram discussion for next week
laurens: I have a TODO for containment (parent relationship), will address after working on notifications
termontwouter: related to clarifying text for auxiliary resources, this is largely done and can be marked as such
<Zakim> elf-pavlik, you wanted to ask about #89
elf-pavlik: related to aux resources/metadata, I picked up issue #89
… would like to get some feedback, especially on some metadata items, such as size and content type
<termontwouter> will do!
<elf-pavlik> also #92
<gb> Issue 92 Reconsider optional metadata of resources such as media-type and size (by laurensdeb) [needs-discussion]
laurens: issue #116 is still in TODO, any progress termontwouter?
<gb> Issue 116 Abstract 'containers' as Link Set representations of arbitrary relational metadata (by termontwouter)
termontwouter: this is an issue that touches on many aspects of LWS. I think most of this has been addressed, and the issue can be closed
… I will open a more targeted issue for remaining areas
acoburn: Access Request and Notification integration still on my TODO list
laurens: metadata/aux resources status?
termontwouter: this has already been resolved as part of pchampin's PR
laurens: ok, will mark as done
<pchampin> I would consider that there is still work to be done on metadata, but yes, my PR on the terminology touches it
<TallTed> could a link to this spreadsheet be put in the agenda and minutes?
<TallTed> tnx
<Zakim> elf-pavlik, you wanted to comment on grants & indexes
laurens: issue #141 merging DID/CID authn suites
<gb> Issue 141 Does LWS10-AUTHN-SSI-CID already include LWS10-AUTHN-SSI-DID-KEY? (by uvdsl)
acoburn: made a comment a while ago, but no real traction on that direction. Would like to get more feedback
laurens: what about clarifying relationship between Access Requests and type index
elf-pavlik: yes, will look into this with eBremer
laurens: pchampin status of reaching out to IETF OAuth WG?
pchampin: this is in progress
laurens: for threat modeling, Rui was primary, what is the status?
elf-pavlik: I have commented on issue and created an initial diagram as a starting point
<laurens> w3c/
<gb> Issue 177 Create Threat Model Note (by elf-pavlik)
elf-pavlik: would like to coordinate with Rui and will follow-up
… hopefully by next week we will have a PR. If you are interested, please chime in
… related to the RDF WG, we can have a C4 model (higher level) to provide different views
… folks are encouraged to comment on issue. the focus is on end-user being different than the resource owner
laurens: I will look at the issue clarifying semantics of delete
… what is the status of publication of JSON-LD context
pchampin: asked for transition request for vocabularies as a WG note
laurens: with that, we made it through the entire spreadsheet
<Zakim> elf-pavlik, you wanted to ask about query
elf-pavlik: since eBremer made PR for type index, there is also a possibility for using QUERY on notifications or aux resources
… should we look at QUERY more generally?
… should we invite Raul to a meeting to share expertise with IETF?
laurens: let's open a broader issue about the use of QUERY in LWS
elf-pavlik: I will open that issue
<Zakim> jeswr, you wanted to ask what features of HTTP QUERY are of benefit to notifications and aux resources
jeswr: what are the features of QUERY that are of benefit to notifications/aux resources?
elf-pavlik: there are some existing discussions in the various issues
… I will link to the comments
eBremer: I am working on a document that describes "if you are going to implement SPARQL, here's what you should consider"
<Zakim> jeremycaine, you wanted to ask about triage meeting
eBremer: this will inform us about affordances that may be needed now
<laurens> https://
jeremycaine: are the triage meetings continuing?
acoburn: yes, every friday