W3C

– DRAFT –
Data Shapes Working Group teleconference

16 February 2026

Attendees

Present
AndyS, bergos, caribou, edmond, elianaP, HolgerK, HolgerK simonstey bergos, ieben, mgberg, nicholascar, simonstey, YoucTagh
Regrets
-
Chair
nicholascar
Scribe
simonstey

Meeting minutes

phase 1 deliverables

nicholascar: anything to know about the phase 1 deliverables?
… haven't heard anything, I'll follow up for feedback

HolgerK: I realized that some builds were failing for the node-expr documents (seems as if the HTML rules were made stricter)

nicholascar: seems to be resolved now

phase 2

nicholascar: any reports on the TF reports?

<HolgerK> https://github.com/w3c/data-shapes/pull/777/changes

HolgerK: PR 777 (please approve; it fixes the pixel thingy and also some bugs in the turtle file)
… another topic is about a new feature which is somewhat of a property pair constraint
… values of one property must be a subset of the values of another property
… kinda like subpropertyof
… it's subsetequal
… sparql part is basically dormant currently

<HolgerK> w3c/data-shapes#786

nicholascar: TF profiling had a meeting 1 1/2 weeks ago
… we added use cases and set a timeline for the completion of the document in a month and a half from now.
… 7 use cases were submitted over time
… there were 2 sections added, one is about: if you do validation and validate your data graph against a data graph
… if you want to indicate that your data graph conforms to a specific shapes graph
… how do you do that?
… so we defined that
… we are meeting again this thursday

mgberg: no additional topics, but andy brought up the possib

<HolgerK> https://www.w3.org/TR/shacl12-core/#explicit-shape-target

mgberg: possibility of the interaction between profiles and rules

<HolgerK> https://www.w3.org/TR/shacl12-core/#sh-shapes-graph

HolgerK: just wanted to mention sh:shape, sh:shapesGraph
… both can be used to link shapes and data graphs in one form or the other

nicholascar: yeah I was looking at that definition today when I was writing a little section there
… basically conventions for claiming and recording that data has been validated against SHACL shapes, so systems don't need to repeatedly revalidate data that's already been checked.
… created a profiling mechanism that establishes hierarchical relationships between validators -> if data is valid according to a specific profile, it's also valid according to broader standards it profiles. e.g. built validators for standards like DCAT and GeoSPARQL, where the GeoSPARQL validator was published by the same working group as the

spec itself, making it the authoritative validator for determining GeoSPARQL compliance.

AndyS: I'm wondering whether that's more of a workflow question rather than SHACL spec

edmond: we are triaging UI issues
… categorizing them into two groups (high/low priority)

nicholascar: (for the previous part) creating a standardized way to persistently store validation results rather than re-running validations repeatedly. while validating data (like graphs) through workflows, we need a conventional mechanism to record which data chunks have passed validation without having to re-validate everything each time
… e.g., if 10% of 1,000 graphs fail, they want to store the results for the 900 that passed rather than re-checking all 1,000. i.e. using profiling specifications to make claims about validated data objects by treating a package of SHACL shapes as a validator object that data can conform to, rather than just storing instantaneous validation

reports that lack persistent semantic meaning.

AndyS: update on rules document

simonstey: apologies for not reviewing PRs, will do better this week!

<AndyS> w3c/data-shapes#789

nicholascar: we don't have any representatives of compact syntax -> ashley summer will rejoin as invited expert, Simon Werner has left the group due to other obligations

open PRs, robert went through a lot of those before his leave

,,. 2 by alex and 3 by vladimir
… are still open

<bergos> branches to delete: https://gist.github.com/bergos/10e05968af10d7c9ed7f90d57e18af6b

bergos: I forgot to add the branches to delete onto the agenda
… in this link are 8 total branches that can be deleted

<bergos> branch view: https://github.com/w3c/data-shapes/branches/all

bergos: I don't have the permission to delete those

nicholascar: [deleting stale branches]

nicholascar: trying to see who we can name and shame quickly
… 4x HolgerK :)
… rest alex
… UI meeting tmrw, profiling in a few

Minutes manually created (not a transcript), formatted by scribe.perl version 248 (Mon Oct 27 20:04:16 2025 UTC).

Diagnostics

Succeeded: s/andy/AndyS

All speakers: AndyS, bergos, edmond, HolgerK, mgberg, nicholascar, simonstey

Active on IRC: AndyS, bergos, caribou, edmond, elianaP, HolgerK, ieben, mgberg, nicholascar, simonstey, YoucTagh