Meeting minutes
<termontwouter> I'm joining on the go today, so apologies in advance for the background noise
Introduction & Announcements
acoburn: welcoming introductions and re-introductions
bigbluehat: co-chair of JSON-LD, verifiable credentials and overlapping involvement
Vote: Terminology PR#164
acoburn: sharing screen to show PR 164 re: Link terminology to definitions
… lots of changed lines, but just mechanical to link terms to defs
… call for vote
<acoburn> PROPOSED: to merge PR w3c/
<gb> Pull Request 164 Link terminology to definitions (by acoburn)
<elf-pavlik> +1
<termontwouter> +1
<eBremer> +1
<jeremycaine> +1
<TallTed> +1
<Luke> +1
<acoburn> +1
RESOLUTION: to merge PR w3c/
acoburn: vote passes; likely more to come, but will keep the text changes moving
Vote: Type Index PR#115
acoburn: sharing screen to show PR 115 re: add Type Index Section
… ask for questions/concerns, and where do people stand on implementation
<Zakim> elf-pavlik, you wanted to mention UCR
elf-pavlik: in review, linked to info about privacy, suggestion about keep connection to UCR for reqs
eBremer: linking the use cases; group had decided to have minimal use cases
… understood to be simple ANDs and ORs to avoid reinventing the wheel
gibsonf1: agree with eBremer about keep spec limited to allow implementation
… search is not just types, for search, just call it "search" so people can just go with it
…
eBremer: agree with gibsonf1 and yield to the group; happy to hear opinions
<Zakim> acoburn, you wanted to talk about naming
acoburn: wants to avoid rabbit hole of naming; search is a little too broad of a name for this
… defining a particular kind of thing; lots of names we could call this, but "search" is too generic
gibsonf1: don't understand acoburn point on the naming; with respect to a given system it can be different
… with LWS define the minimum search; compare to GET method, which the word means a lot of things
… limiting the naming limits implementations to expand upon these minimal definitions
<Zakim> elf-pavlik, you wanted to point out that it may not address some use case requirements from related prior work in Solid CG
elf-pavlik: supports merge for draft to work from; type index work in Solid that it may not address all concerns
… such as "Which container can I create these resources in?"
acoburn: q for group - continue to discuss or vote?
eBremer: proposes deferring naming for later; vote on "is this the kind of functionality we want in the spec?"
jeswr: given this is a feature we have little prior art for, potentially defer on reception of implementations
gibsonf1: agree with eBremer on defer on naming; search impl is done already
<Zakim> acoburn, you wanted to propose options
acoburn: options - vote on merge as-is to spec; jeswr defer or move to ancillary doc;
jeswr: yes, search impls exist; not arguing it doesn't exist as functionality, but capabilities vary enough to want desire convergence
acoburn: consider ordering - first for jeswr to create proposal text to determine top-down or bottom-up model for definition to bring into the spec
… rephrasing - does this need more incubation? Is that the first vote?
<elf-pavlik> +1 to more incubation
gibsonf1: lowest-level search implemented (including type and internal content); works great
… don't have type index, but fairly easy to implement; don't yet do JSON-LD per spec
<Zakim> elf-pavlik, you wanted to propose merging as separate doc
elf-pavlik: consideration for continued incubation based on implementations, but suggest merging PR somewhere to close current discussion
acoburn: first vote for more incubation for this PR
<acoburn> Vote: does the type index PR w3c/
<gb> Pull Request 115 add Type Index Section (by ebremer)
<gibsonf1> -1
<elf-pavlik> +1
<acoburn> +1
<jeremycaine> +1
<TallTed> +0
<termontwouter> +0
<eBremer> +0
<ryey> +0
<Luke> +1
acoburn: clarifying the potential outcomes on the vote
… straw poll is inconclusive; new proposal for merging this PR as-is incoming
<acoburn> PROPOSAL: to amend w3c/
<elf-pavlik> +1
<termontwouter> +1
<jeremycaine> +1
<acoburn> +1
<eBremer> +1
<gibsonf1> +1
<Luke> +1
<ryey> +1
RESOLUTION: to amend w3c/
<gb> Pull Request 115 add Type Index Section (by ebremer)
<Zakim> gibsonf, you wanted to ask how do you define incubation?
acoburn: request to push to separate document
gibsonf1: point of incubation may need discussion about meaning and consequences of anything in incubation being in spec
acoburn: this particular PR has some need of additional feedback due to auth
jeremycaine: when multiple considerations are present it is difficult to vote; awareness of multiple questions would make voting more informed
acoburn: goal is next step to ask question is should it be merged into main or remain separated
<Zakim> elf-pavlik, you wanted to provide another example why I see need for more incubation
elf-pavlik: re: incubation as working further; now we can look for additional art or concerned parties as consideration
TallTed: lots of comments on the PR are unaddressed, maybe because of clarity
<elf-pavlik> Github allows crating issues from comments in PR
TallTed: comments being left open doesn't lend itself to a merge
… prior Solid implementations do not (should not) equate to LWS implementations
… reframing as LWS is a necessity for interoperable spec so implementations are equivalent
… detach implementation details from the spec (SPARQL, SQL, etc.)
<Zakim> gibsonf, you wanted to point out authorization with search
gibsonf1: all comments on PR is just about renaming; all others have been resolved
… re: implementation details - auth has been resolved with solution; auth has to sit between client and implementation
<Zakim> elf-pavlik, you wanted to ask about known implementations besides Samu's sample implementation in C# and possibly Inrupt's server
elf-pavlik: would it help to keep track of implementations? Single place to keep track of them.
acoburn: requests elf-pavlik to keep track of just type search implementations
elf-pavlik: interested in general impl, but will do type search specific also
ACTION: elf-pavlik to open an issue to track implementation feedback on the type search feature
<gb> Created action #170
acoburn: aux resources discussion
Discussion: Auxiliary Resources PR#165 & PR#167 agendabot]
acoburn: competing PRs 165 and 167; requests termontwouter to discuss
termontwouter: update from last week - checked linkset requirement; seems like the PRS are saying the same thing with different words
<Zakim> elf-pavlik, you wanted to ask about examples
elf-pavlik: adding a few examples (e.g. - client discovery); user scenarios to be mapped potentially to test suite
acoburn: both could use examples, even just comments; terminology is covered, but examples would help
jeremycaine: suggestion to address long-running PRs - cancel PRs and work together to create a new, simplified version to avoid duplication of discussion?
<elf-pavlik> +1 to keep two PRs for the group discussion
acoburn: alternative suggestion is that folks weigh-in on proposals to help
<Zakim> termontwouter, you wanted to comment on examples
termontwouter: took some time to write an example out, but it doesn't change anything
acoburn: if the examples end up being the same, great, differences in expression are okay; propose voting next week
elf-pavlik: clarifying which examples are the same
termontwouter: writing a fictional example just re-iterates the spec requests/responses
elf-pavlik: understood this as a different interactions
termontwouter: no functional difference between the two PRs to differentiate with examples
acoburn: continue this discussion at beginning of next meeting
<elf-pavlik> jeremycaine you can add to batch
jeremycaine: see commentary on PR, does applying suggestion accept them?
acoburn: yes
TallTed: options for comment-by-comment or batch commit
acoburn: demonstrating on-screen how to apply
<jeremycaine> quit