Minutes of the meeting on 2022-08-31

The minutes are also available here:

https://www.w3.org/2022/08/31-rch-minutes.html



                         RCH WG Kick off meeting

31 August 2022

    [2]IRC log.

       [2]https://www.w3.org/2022/08/31-rch-irc


Attendees

    Present
           dlehn, dlehn1, dlongley, Gregg_Kellogg, ivan, leonardr,
           manu, markus_sabadello, pchampin, phila, TallTed

    Regrets
           -

    Chair
           markus_sabadello, phila

    Scribe
           manu, markus_sabadello, phila

Contents

Meeting minutes

    [3]https://docs.google.com/presentation/d/

    1RtJMebmGYQ3l73ayntcrvxJDqtr3LdlFhM47-lSIZqE/
    edit#slide=id.g142dbe090c2_0_60

       [3]https://docs.google.com/presentation/d/1RtJMebmGYQ3l73ayntcrvxJDqtr3LdlFhM47-lSIZqE/edit#slide=id.g142dbe090c2_0_60


    phila: We will scribe every meeting.

    phila: I'm Phil Archer of GS1, org behind barcodes.

    phila: Been with GS1 for 5 years, before that at W3C working on
    Semantic Web

    phila: I'm here as someone with backgroun in Linked Data,
    vocabularies, ontologies, etc.

    markus_sabadello: Introduces self. From Danube Tech, working on
    decentralized ID technologies
    … member of the DID WG and the VCWG
    … always had a strong interest in digital identities tech,
    JDON-LD, and the proofs that we need.

    Pierre-Antoine: I'm PA, associate professor of computer science
    in Lyon, been participating in RDF-related working groups

    Pierre-Antoine: Took a sabbatical 18 months ago, joined W3C
    team as fellow.

    Pierre-Antoine: I'm team contact of a number of groups
    including this one.

    Pierre-Antoine: I will also be contact in upcoming RDF-Star WG
    and DID WG

    ivan: In "ancient times" I was in the Semantic Web activity in
    W3C, then handed over to phila

    ivan: Now moved on to digital publishing, staff contact of epub
    WG

    ivan: Also staff contact of DID WG, and before involved in
    JSON-LD WG

    ivan: Now I do VC WG

    ivan: I have actually retired a year ago, in the Netherlands. I
    work 50% of my time.

    ivan: I was also part of a group people who got this WG up and
    running

    ivan: I am NOT staff contact in this group, just interested

    Dan Yamamoto: I appreciate the invitation, I work for Internet
    Initiative Japan, not currently a W3C member. I'm an Invited
    Expert.

    Dan Yamamoto: I work on a project for ZKP for Linked Data,
    includes RDF hashing for specialized BBS+ signature schemes

    Dan Yamamoto: Project members were invited by chairs to this WG

    Dan Yamamoto: We have a live demo ZKP-LD playground

    Dan Yamamoto: I look forward to working with you, this is my
    first time in W3C

    leonardr: In this context, I am chair of technical WG of the
    Coalition for Content Provenance, focuses ensuring provenance
    and trust for assets

    leonardr: Been working with DID and VC groups and utilizing
    those technologies

    leonardr: Also project editor+chair for the Working Group 11,
    in an RDF based technology

    gkellogg: Most recently I spent most of my time on JSON-LD, now
    we have a community group working on YAML-LD

    gkellogg: Been involved with most RDF work for 12-13 years

    gkellogg: I've had my hands in almost ever RDF serialization
    format

    Tobias: I'm assistant professor at university in Amsterdam, on
    a part-time basis. I've been working on things related to this
    WG, for the past 10 years, canonicalization and hashing.

    Tobias: I've applied canonicalization and hashing in an
    isolated way, there hasn't been a coordinated effort

    <pchampin> [4]https://nanopub.org/


       [4]https://nanopub.org/


    Tobias: I have some university obligations, but now will also
    be founder of a startup, to create impact

    aalobaid: I'm a researcher at the Ontology Engineering Group in
    Madrid, I work mainly on RDF graph construction and semantic
    labeling. This is my work time participating in a charter

    AndyS: I'm Andy Seaborne. In this group, I'm affiliated with
    Apache Software Foundation, on the Jena RDF toolkit. Also work
    for a startup that uses RDF and has interest in provenance and
    integrity of data, and sharing between organizations.

    AndyS: I have both work and personal interest in this work.

    dlongley: I'm Dave Longley with Digital Bazaar, been working on
    standards for over a decade. Some of the standards are JSON-LD,
    VCs, also some work on input documents of this group.

    manu: Wonderful to see so many face we've known for a very long
    time. This is a great day, this has been 10 years in the
    making. I've worked with many of you including RDFa, JSON-LD,
    VCs, DIDs. Currently I'm editor for Data Integrity
    specification, which hopes to use the output of this group.

    manu: Looking forward to working with all of you.

    dlehn1: Also with Digital Bazaar, been working on JSON-LD,
    usually work on implementation details.

    <TallTed> [5]https://www.w3.org/community/credibility/


       [5]https://www.w3.org/community/credibility/


    TallTed: I'm with OpenLink Software, involved in W3C groups
    related to Linked Data, etc. If you pull a PR, you will find me
    make it more readable. I don't code, but I have a strong
    understanding how things are supposed to fit together. Try to
    liaison with Credibility Community Group.

    [6]https://www.w3.org/2022/07/rch-wg-charter/


       [6]https://www.w3.org/2022/07/rch-wg-charter/


    markus_sabadello: Shares screen to show charter

    markus_sabadello: ... walks through the sections
    … RDF C18N & Hashing is about RDF. VCs have been mentioned,
    there is a link of course
    … VCs can inclde proofs and signatures. For that you need c18n
    and hash
    … we've heard lots of interest in te intros. VCs are one use
    case, but not the only one.
    … Linked Data Spaces will also have interest I think, outside
    VCs
    … c18n ... means creating one default/preferred represetnation
    of something that can have several.
    … we're not canonicalizing a JSON-LD doc, but the abstract data
    model behind it
    … hashing - everyone knows what that means. Can have a large
    number of applications. Not just signing
    … hashlinks construct links that contain a hash of hte expected
    data
    … charter mentions a number of publications as inpur docs

    leonardr: I completely understand the desire to c14n the
    general expression rather than a specific serialization. Does
    that imply that the algos for c18n specific serializations are
    in or out of scope?

    markus_sabadello: We're workng at the abstract data model
    level, not the JSOn-LD

    leonardr: Good. It might be relevant to add to the out of scope
    section that translations of a specific serialization to the
    abstract

    markus_sabadello: Yes, I don't think that's in our scope

    pchampin: We can make some changes to the charter. I think
    liaison is one we can change. Scope inclusing out of scope,
    cannot be changed

    markus_sabadello: Talking of what is out of scope - the
    definition of new crypto algorithms
    … so it's about how to use them in RDF Dataset

    markus_sabadello: Higher level signature schemes are not in
    scope. SCope shoudl be relatively narrow

    dlongley: In response to leonardr - in order to create any test
    factors, we're going to have to define some kind of work
    … or it won't work.

    leonardr: We should talk about that when we et into testing.
    Slippery slope

    markus_sabadello: To create test suites, they can use certain
    tech that is not in scope in order to be able to implement it

    markus_sabadello: That doesn't make it is scope. That's not
    what's being tested

    <dlongley> this group should not be defining new RDF
    serializations (only reusing them as needed)

    <manu> +1

    <AndyS> Existing RDF formats define a mapping to the RDF
    abstract data model as part of their specs.

    <Zakim> gkellogg, you wanted to ask about relationship with
    dataset isomorphism and relationship to RDF-star

    <manu> we also may need to talk about how input/output
    serializations / specs are affected by our work here.

    <leonardr> @manu/@dlongley - though I worry about use perhaps
    showing "preference" to serializations and/or implementations
    of them...

    gkellogg: It seems there is a close relationship between
    isomorphism and the canonicalisation. Might there be somethingn
    to say about that?

    <manu> True, always a concern, leonardr

    gkellogg: ANd the netx thing to think about - if we have quads
    - we are going to take into cnsideration of RDF-star
    … we'll need to discuss that. It needs to be in scope

    ivan: The isoprphism issue came up in the charter discussion.
    There's clearly a close relationship between the two. If wo are
    canonical then they have the smae morphology

    <manu> Yes, we absolutely have to deal w/ graph isomorphism in
    the work we do in this group... it's unavoidable if the goal is
    to create a generalized solution to RDF Dataset
    Canonicalization.

    ivan: We can keep in our minds... but it's not part of any
    recommendation. We can publsh whatever notes we want to do.

    ivan: It's certainly a topic.

    ivan: The topic of RDF* came up late in the discussion. I'm not
    an expert in it, but I thought that builds on the RDF model. An
    RDF* graph can be mapped onto an RDF graph

    ivan: If this is true, then c18n works. But we can't work on it
    normatively as it's not in the charter

    ivan: But we have a liaison and the same staff contact

    markus_sabadello: We should definitiely track RDF* and maybe a
    Note

    pchampin: As the staff contact of both groups, I will ensure
    that communication occurs between this and RDF*. Groups didn't
    exist when we were writing the charter

    AndyS: I think P-A has covered most of it. I'm sure we can do
    isomorphism, even if not c18n

    markus_sabadello: Reviews the the normative deliverables

    markus_sabadello: The hash Rec will specify how to apply a hash
    function to an RDF dataset, and one of the steps will be the
    c18N step

    markus_sabadello: So one spec uses the other

    markus_sabadello: I think there will be some interesting
    variation. We've heard already about zero knowledge proofs

    markus_sabadello: We might talk about VCs and our relationship
    if we have time

    markus_sabadello: We should mention - in some CGs there has
    already been a lot of work on c18n. VCs are already
    canonicalizing and hashing so this group won't start from
    scratch
    … specifically there's a draft from Dave and Manu (and Rachel I
    believe)
    … we want to recognise this existing work
    … Doesn't mean that we have to use it, but we should recognise
    such documents

    Explainer document: [7]https://www.w3.org/2022/07/

    rch-wg-charter/explainer.html

       [7]https://www.w3.org/2022/07/rch-wg-charter/explainer.html


    <gkellogg> The CG also has a testsuite: [8]https://github.com/

    json-ld/rdf-dataset-canonicalization/tree/main/tests

       [8]https://github.com/json-ld/rdf-dataset-canonicalization/tree/main/tests


    <Zakim> AndyS, you wanted to ask about non-normative tests

    markus_sabadello: I think that's it for the scope and
    deliverables. We can cover the time line at a future meeting

    <manu> +1, we do not want (nor expect) a rubber stamp on the
    input documents -- we are all seeking a thorough
    multi-disciplinary expert review and actively analyzed attack
    models against the algorithms presented. :)

    markus_sabadello: we don't have to run for the full 24 months
    :-)

    markus_sabadello: Please review the charter and we'll talk
    about that again next time.

    phila: W3C meeting are expected to finish 5 min early, this is
    good practice

    phila: Will talk about TPAC now

    phila: Does anybody know already that they want to be an
    editor? Requires work, but you will be remembered :)

    phila: We will obviously need editors to drive delivery of the
    documents.

    phila: TPAC is in 2 weeks, we are meeting for 2 hours together
    with VC WG.

    <manu> +1 to meet every other week (alternating w/ the VCWG)

    phila: Our planned call schedule is to meet every 2 weeks at
    this current time, which is aligned with VC WG also meeting at
    that time at the alternate time slot

    phila: Dan, what time is it for you, how is this time slot?

    Dan Yamamoto: It's midnight, but I can accept that

    phila: Anybody else in a difficult time zone?

    gkellogg: It's morning for me

    phila: At the moment we think every 2 weeks is enough, but this
    may change. We may also have task forces that meet separately,
    but we will start with this schedule

    phila: We have the 2 hour meeting at TPAC with VC WG. I will be
    at TPAC in Vancouver

    ivan: I will be there

    <leonardr> I will not be at TPAC - have a conflicting standards
    meeting in Berlin

    Pierre-Antoine: I will be there as well

    <AndyS> I hope to attend remotely.

    phila: We will try very hard to make sure that the people who
    join remotely will not be in any way at a disadvantage

    phila: We will have Dave present his work, and Aidan Hogan will
    present his.

    phila: We will have presentations of 2 input documents, I hope
    to find out the differences between the approaches.

    phila: Does someone else in the room have a 3rd or 4th
    approach?

    <yamdan> I will join remotely

    phila: Any other questions?

    leonardr: I assume the minutes and copies of presentations will
    be made available

    phila: Yes, will be made available to the group

    ivan: Presentation of Dave and Aidan will happen at the TPAC
    meeting? This might be tough on the VC people.

    ivan: This work may be difficult to follow.

    phila: The meeting will also be about the relationship of the
    two groups.

    phila: This group is not just a "special task force" of the VC
    Working Group.

    phila: Thank you everybody, I look forward to the work. See
    some of you in Vancouv

    <AndyS> Thank you!


     Minutes manually created (not a transcript), formatted by
     [9]scribe.perl version 192 (Tue Jun 28 16:55:30 2022 UTC).

       [9]https://w3c.github.io/scribe2/scribedoc.html


Diagnostics

    Succeeded: s/c18n/c14n/

    Maybe present: aalobaid, AndyS, gkellogg, Pierre-Antoine,
    Tobias