W3C

– DRAFT –
RCH WG bi-weekly

23 November 2022

Attendees

Present
aalobaid, AndyS, gkellogg_, ivan, markus_sabadello, pchampin, phila, TallTed, yamdan
Regrets
Kazue
Chair
-
Scribe
aalobaid

Meeting minutes

<phila> FPWD

Phil: the draft has been published

Ivan: It is not yet published

Phil: it is about to be published

<gkellogg_> https://www.w3.org/TR/2022/WD-rdf-canon-20221124/

Phil: it will be published tomorrow

IIW

Phil: Mark, you can explain to everybody about the recent event.

Mark: There were many sessions about harmonisation and different creditional formats.

<phila> IIW website

Mark: another big topic about Web5.

Mark: Another topic is about identify, signatures, and much more. But nothing specific to this working group.

<TallTed> we've skipped web4, and web3 is still being defined as well

Dan: I had to 2 sessions. One is about SPARQL endpoint and signatures.

Dan: and anonymisation about SPARQL results.

Dan: The second is about the canonicalization (but only as a black box). Mainly from the user perspective.

<manu> Plugfest related to RCH work item output -- https://docs.google.com/presentation/d/19GmJ3bLMrbVadesnkmsWaaUr-U71Y9Kr775tZvgs-xI/edit

Manu: Companies participating and using verified signatures.

<manu> Verifiable Credential Data Integrity 1.0 FPWD: https://w3c.github.io/vc-data-integrity/

<manu> https://www.w3.org/TR/vc-data-integrity/

Manu: 17 credentials issuers.

<manu> Transition EdDSA to VCWG in next few weeks: https://w3c-ccg.github.io/di-eddsa-2020/

<manu> Finally, there is a active test suite with multiple implementers: https://w3c-ccg.github.io/di-ed25519-test-suite/#Ed25519Signature2020%20(issuer)

Manu: We will have more test suites from the participants.

Issue

phila: Let us talk about the issues that are recently updated.

phila: Greg. Are there any issues that you would like the group to discuss?

https://github.com/w3c/rdf-canon/pull/40

phila: Any comments on this issue?

ivan: There is a loop in the algorithm that is unecessary.

ivan: We had a discussion about that in issue 23.

ivan: Dave and I agreed that this is unnecessary.

gkellogg_: I took the opportunity to add labels to everystep in the algorithm.

ivan: we need to relabel the steps if the algorithms is changed.

<phila> Proposed Resolution: Accept PR40 that removes the 'simple' processing loop, and close issue 23

<manu> +1

<ivan> +1

<gkellogg_> +1

<pchampin> +1

<phila> Proposed Resolution: Accept PR40 that removes the 'simple' processing loop, and prepare to close issue 23

<manu> +1

gkellogg_: there is some referencing issues that need to be resolved.

ivan: It is a good practice to keep a list of changes to the issue regardless whether the issue is opened or not.

gkellogg_: I will take care of this.

<yamdan> +1

<phila> +1

RESOLUTION: Accept PR40 that removes the 'simple' processing loop, and prepare to close issue 23

phila: any objections?

phila: resolved

https://github.com/w3c/rdf-canon/issues/37

gkellogg_: There is some ambiguity about sorting quads

gkellogg_: there is no formal definition about canonicalised quads

gkellogg_: nquads separaters should be a line terminator or a dot?

<phila> gkellogg_: There's a section in the N triples spec that defines a canonical form. The basic spec allows any amount of white space between components

<phila> ... but you can't have new lines outside the grammar

<phila> ... the canonical form fixes that as a single space character and doesn't discuss newline separator

<phila> ... So erratum would be to include that newline character as part of the canonical form or to describe it only in the canonical form of a document

<TallTed> we might be drilling too far into the details for the moment...

<phila> ... Where a certain no.of quads are used to create a hash - that will help clarify a specific form of these quads without the newline separaor and the terminator

AndyS: I agree with Greg's suggestion.

TallTed: Should we dive into the details about this here.

TallTed: Vocally discussing this here might not be enough.

phila: Understood

https://github.com/w3c/rdf-canon/issues/10

phila: What is the criteria about choosing?

gkellogg_: I don't think it is not up to me to choose the approach.

manu: I agree. We don't have to choose one or the other.

<Zakim> manu, you wanted to note we're not at a fork in the road...

manu: If we decided to do something new, we can do it without disturbing the existing specs

phila: If there is nothing pushing for a change, then we don't have to do it yet.

<Zakim> manu, you wanted to note that it's important to write down what we're optimising for in the current work.

AndyS: Maybe we can add more details about the algorithms.

manu: It is valuable to write down what we optimise (e.g., time complexity, proofs, ...)

<Zakim> gkellogg_, you wanted to discuss a new topic: granularity and approval of future PRs leading to WD publication.

<Zakim> phila, you wanted to support Manu

phila: It is good to include the reasoning for the decisions.

phila: I would like to see a section in the document about the reasons behind the taken decisions.

Granularity of pull requests

gkellogg_: frequency of pull requests and authority for merging pull requests. Maybe we need to discuss the dynamics of this.

<manu> Phil's explanation of the granularity has been my experience.

phila: I think merging several pull requests in the same day will be considered a single change.

manu: I have the same experience and I've never had problems with making several merge requests.

phila: If it is an editorial think, we can go for it greg.

<TallTed> Editors are allowed to edit. :-)

<manu> +1 to depend on Editor's judgement on merging to main. Don't merge unless you're very confident that there won't be objections... make sure you have multiple reviews, etc. :) -- but the Editors already know how to do all of this and have demonstrated to do it well over the past several years. :)

<pchampin> +1

pchampin: I agree that the editors can do changes that is not too large.

phila: Editors should be allowed to edit.

phila: Thanks everyone

Summary of resolutions

  1. Accept PR40 that removes the 'simple' processing loop, and prepare to close issue 23
Minutes manually created (not a transcript), formatted by scribe.perl version 196 (Thu Oct 27 17:06:44 2022 UTC).

Diagnostics

Succeeded: s/(for publication of FPWD)//

Succeeded: s/Thank you Gregg and PA!//

Succeeded: s|-> https://github.com/w3c/rdf-canon/pull/40 PR 40|https://github.com/w3c/rdf-canon/pull|

Succeeded: s|https://github.com/w3c/rdf-canon/pull|Subtopic: https://github.com/w3c/rdf-canon/pull/40|

Maybe present: Dan, Manu, Mark, Phil

All speakers: AndyS, Dan, gkellogg_, Ivan, Manu, Mark, pchampin, Phil, phila, TallTed

Active on IRC: aalobaid, AndyS, gkellogg_, ivan, manu, markus_sabadello, pchampin, phila, TallTed, yamdan