W3C

– DRAFT –
Data Shapes Working Group

21 July 2025

Attendees

Present
ajnelson-nist, AndyS, bergos, caribou, DavidHabgood, HolgerK, nicholascar, TallTed
Regrets
-
Chair
nicholascar
Scribe
nicholascar, AndyS

Meeting minutes

start the meeting

General runtime-configuration for tests

bergos: allow other levels such as Warning not to fail, by configuration

bergos: ...for tests

ajnelson-nist: why are test conformance pass/fail only?

ajnelson-nist: which a validation report, conformance is true/false. Warning could be true or false. These needs to be defined

AndyS: confused as test outcomes contain details

ajnelson-nist: confusion is pass/fail of overall validation report

ajnelson-nist: we should say whether a test run was configured to allow/deny pass on Warning

AndyS: we can't go too far into tooling

nicholascar: will we have a logic of sets of shapes for SHACL Rules

AndyS: packages is (currently) out of scope

AndyS: Currently, rules defined by "data" + "ruleset" => output
… data is one graph (maybe virtual) and how it come into being is not defined by the standard
… ruleset support "IMPORT"

bergos: maybe not all runtime configs should contain this

(Discussion of sh:conforms)

davidr: conforms to a summary so maybe not t/f any more

nicholascar: that's a structured value to sh:conforms

bergos: one triple overall result is useful.
… e.g. data pipeline
… a validation report gets complicated

nicholascar: we do counting of theresults

bergos: example -- subclass warning ... custom results.
… should still have a summary in the validation report

nicholascar: is this part of the "simple usage" story?

holgerk: there is also sh:result
… reports may be combined so leave the summary to the point of use, not in the report itself

nicholascar: provide example queries for report handling?

mgberg: option leave sh:conforms with the assumption it may differ due to external configuration

nicholascar: back to original point of bergos - counts of each level?
… for the WG tests

bergos: use case: running SHACL as a constraint adviser
… I think sh:conforms is quite useful

nicholascar: we should not answer this matter right now- need a GH issue for discussion
… and list alternates to see the overall design

<Zakim> ajnelson-nist, you wanted to talk about configuration-values

ajnelson-nist: if we store a config dictionary - need to say where it goes and the vocabulary inc custom/tools specific settings.

nicholscar: I will create an issue

<bergos> pull request --w3c/data-shapes#409

bergos: for the PR, can we make progress so as to not have it open too long.

nicholascar: status of subgroups
… node expressions
… is there a meeting time?

Robert: many people on holiday
… will send out a doodle for a couple of weeks time

HolgerK: made a first input to node expressions document.

HolgerK: I made PR #432 as a big start on Node Expressions
… changes being made

nicholascar: SHACL-SHACL

ajnelson-nist: 5 mins on Sh-SH and then profiles
… influence of OWL profiles
… other concepts include DX (data exchange WG)
… how do we log SHACL-SHACL in the documents?

nicholascar: we did discuss an overview doc for SHACL.
… we can show the relationship of governance aspects
… both the parts that get fixed by publication and those that don't

nicholascar: SHACL-UI
… start soon when people become available.

Minutes manually created (not a transcript), formatted by scribe.perl version 244 (Thu Feb 27 01:23:09 2025 UTC).

Diagnostics

Succeeded: i/General runtime-configuration/scribe: nicholascar /

Succeeded: s/in the /of the/

Succeeded: s/pr: /pull request --/

Maybe present: davidr, mgberg, nicholscar, Robert

All speakers: ajnelson-nist, AndyS, bergos, davidr, holgerk, mgberg, nicholascar, nicholscar, Robert

Active on IRC: ajnelson-nist, AndyS, bergos, caribou, DavidHabgood, HolgerK, mgberg, nicholascar, Robert, TallTed