Meeting minutes
minutes review
<kaz> Apr-22
(audio problems)
(closing webex)
(audio restored)
<kaz> Apr-20
sebastian: we need to update the April 20 minutes to change the first topic
Kaz: we can remove that line
Sebastian: looks good to publish, any objections?
… please publish the minutes from April 20
comments from Michael Cooper
<kaz> a11y-request issue 21 - Web of Things (WoT) Thing Description 1.1 2021-01-26 > 2021-04-15
internationalization
<kaz> i18n-request issue 171 - Web of Things (WoT) Thing Description 1.1 2021-02-02 > 2021-03-30
Sebastian: need to wait until (Addison) can join to discuss
TAG
<kaz> design-reviews issue 715 - Web of Things (WoT) Thing Description 1.1: TAG and Security Review
IANA submission
<kaz> wot-thing-description issue 931 - Content Type for Thing Model (IANA registration needed)
Sebastian: sent the form to IANA for proposed registration of TM content type
Sebastian: working with mccool on filling out the questionnaire for security and privacy review
Kaz: regarding IANA registration, have you talked with PLH about this? and got any response?
… need to put the template into the PR document for TM first, then point IANA to the document for review
Sebastian: this is just a request for review, and not an official registration request
… need to get approval from Philip still
pull requests
PR 1452
<kaz> PR 1452 - fix: allow uri value only for in field of APIKeySecurityScheme
Jan: made some changes to the ttl file
… changes to the security scheme, had some issues with the ontology files, hope nothing is broken but looks OK
Sebastian: any comments?
… merged
PR 1466
Sebastian: PR 1466
<kaz> PR 1466 - use "@type" instead of rdf:type
Sebastian: use of @type
Sebastian: this is important for round trip integrity in discovery, so the symbols are the same
… we can't merge this because the author of the PR is not a WoT member
Jan: question about the context being generated from the ontology; are these generated automatically and would it need to be changed in the generator step
Sebastian: will double check where the change needs to be made
Jan: seems like the script needs to be updated
Kaz: we should be strict about the working group procedure, so need to clarify whether this is an editorial change or a not. my understanding is this is an editorial fix for the ontology file.
Sebastian: this is not a feature change, and is editorial in nature
Kaz: we can add the contributor as a participant for the future
… suggest we clarify that this is editorial
Sebastian: will add the contributor as a participant
Kaz: we could accept this PR as editorial but need to create a new PR within the WG to update the script
Sebastian: move on to non-semantic issues
<Ege> it is ready to merge
PR 1470
<kaz> PR 1470 - Create assertion-sync.yml
Sebastian: any more comments before we merge?
Ege: yes, it's ready to merge
… it's more of a utility script to automate changes
Jan: can we use this procedure to automate TD updates also?
cris: we can use it with CI to render all the changes automatically
Sebastian: merged
PR 1472
<kaz> PR 1472 - Reallow context extensions in TD and TM schemas
Jan: schema contains "oneOf" which prevents multiple context definitions, the PR changes this to "anyOf"
Ege: this creates some downstream dependencies that need to be reviewed
… it should not take long
Sebastian: we will return to this later
PR 1473
<kaz> PR 1473 - chore: add npm script for TM schema regeneration
Jan: this is syntactic sugar to make it easier to construct the command
Sebastian: comments?
Ege: good to go
Sebastian: merged
PR 1474
<kaz> PR 1474 - feat: allow definition of additional security schemes
Jan: to allow additional schemas for security schemes
… there was a question about requiring a context extension, but it can't be expressed in JSON schema itself
Sebastian: one problem is that it requires a new pattern that isn't in 1.0, might not be backward compatible
Jan: maybe for now we could allow any value, and add a constraint later
Kaz: concerned about the publication schedule, we are already in wide review
… how and when can we apply fixes?
<Mizushima> +1 kaz
Cristiano: this PR 1474 itself is a king of bug fix, I think
Ege: that's true but not really sure how to fix it
… please see my comment on the PR
Kaz: please remember that our "deliverable" is the Thing Description spec itself
… and schema is an additional mechanism to help us validate the data
… so my question is: Is the TD spec text itself good enough?
Jan: think there is a problem within the spec text as well
Sebastian: the biggest problem is inconsistency between TD 1.0 and TD 1.1
… [[string (e.g., nosec, basic, digest, bearer, psk, oauth2, or apikey)]] within 1.0
… but [[any type (one of nosec, combo, basic, digest, bearer, psk, oauth2, or apikey)]] within 1.1
Kaz: consistency between 1.0 and 1.1 is important, that's true
… but how to specify potential additional security schemes then?
Jan: need some more discussion
Ege: created a new issue for that
https://
PR 1475
PR 1475 - Create Automatic Issue Labeling Action
Sebastian: merged
Kaz: merging this script itself is fine
… but please explain this during the next Editors call
PR 1476
PR 1476 - Add section about IANA Thing Model content type registration
Sebastian: would like to wait for PLH's response
PR 1477
PR 1477 - Add text for composition as alternative
related Issue 1265 - rel:type to multiple TMs
merged
PR 1479
PR 1479 - Clearifications about placeholder map and object values
(conflicts to be resolved)
merged
PR 1480
PR 1480 - default values for read/writeOnly only in PropertyAffordance definitions
preview - 5.4 Default Value Definitions
Jan: the proposed changes make sense
… but kind of concerned about potential impacts for compatibility
Kaz: kind of agree
… would be nicer to see the potential impact of this change with the existing TDs and processors
Sebastian: ok
… can see quickly
Kaz: thanks!
Issues
Issue 1443
Issue 1443 - Can TMs have the V1.0 context URL?
Kaz: TM is a new feature for TD ver. 1.1
… and was not included in TD ver. 1.0 at all
… so we should split TM feature from TD ver. 1.0 in general
… mixing it up would make implementers confused
Sebastian: yeah
… maybe we should leave it as-is
[adjourned]