Meeting minutes
Agenda
<kaz> agenda for today
Minutes Review
<kaz> July-13
Sebastian: any objections?
Sebastian: hearing none, let's publish
TD
Cancellation
Sebastian: no TD call next week due to the Testfest
Implementation status
TD Editor's draft (with yellow highlights for unimplemented features)
Sebastian: Let's see implementation amount
Sebastian: the focus is profile but we should fill the gaps for the TD as well
Daniel: we should regenerate it more often
Kaz: mccool will inform everyone about the instructions for the Testfest, so please give your comments and then let's finalize the instructions on Monday.
TD schedule update
<kaz> wot PR 1027 - Update wg-2021-extension-plan.md - TD schedule
Sebastian: I will merge it
4th WD
Sebastian: we can publish it next week
Kaz: we need to add the change log since the March WD.
dape: will work on a PR.
IANA Registration
Sebastian: I have also asked coap id number
Sebastian: it will be good to have that when we go for CR transition
PRs
PR 1587
<kaz> PR 1587 - Add hreflang to schema
Sebastian: ege worked to validate the hreflang attribute
Ege: the ietf spec does not provide the pattern but the one I found validates against their examples
Sebastian: I hear no objections, merging
PR 1593
<kaz> PR 1593 - Make additionalProperties explicit in JSON Schema
Ege: this makes it more explicit to define new new security terms
Sebastian: we like this in the TD to put new metadata
PR 1564
<kaz> PR 1564 - explain contentType usage
Sebastian: I will convert to a draft
Oracle Review
<kaz> Issue 1548 - Oracle review comments on CR draft - Part 1
Daniel: Should we move forward?
Sebastian: I have sent him an email, waiting for reply
… let's wait and give him more time
tm:required default value
<kaz> Issue 1581 - Should tm:required have a default value?
Sebastian: this is an interesting topic
Sebastian: if there are many properties in a TD, the tm:required term might become huge
Sebastian: we cannot use a wildcard since it is not a valid json pointer
Daniel: this solves one use case but there can be other use cases that want the opposite behavior
Cristiano: we should think of how people use TMs
Cristiano: why would I make a template if everything is optional
mjk: similar thoughts came during the sdf discussions
mjk: I agree with Cristiano, this is not about validation
<kaz> (some more discussion around SDF)
<kaz> ca: question on composability
<kaz> ... regarding the common use cases
<kaz> ... by default it's optional
Kaz: should we not get feedback from implementors?
Kaz: who is implementing this now?
Sebastian: editdor is implementing this
Cristiano: node-wot too
<Zakim> dape, you wanted to consequences of removing tm:required
https://
Ege: ditto follows the default assumption and every affordance is required
Cristiano: I agree with daniel that it is weird that if it is not present, it is full of values
Cristiano: if there is no tm:required, all TDs can claim to implement that TM
Ege: yes a bit like json schema. If there is no required, all json objects validate the schema
mjk: it also depends on the tool
Sebastian: maybe we should make it mandatorz
Cristiano: it might be dangerous since tools would show different behavior
Sebastian: my default assumption is to take everything from it
Cristiano: maybe lets see the options and vote
<mjk_> 1. no default
<mjk_> 2. default is "all required"
<mjk_> 3. default is "none required"
Sebastian: I write all the options to github issue 1581
Issue 1594
<kaz> Issue 1594 - Using tm:required for non-affordance members
Ege: this is more of a question, what if we want to say your TD must have basic security etc.
Cristiano: derivation algorithm copies all the terms no?
Ege: in that case, we require all affordances?
Daniel: it seems more and more clear that we need tm:optional instead of tm:required
Daniel: what does this imply schedule-wise
Cristiano: I agree with daniel
Kaz: that is why I proposed to talk with developers
Issue 1214
<kaz> Issue 1214 - TD File Extension Convention
Sebastian: matthias did the registration, I will ping him
<sebastian_> Ege we can hear you but ut seems you are not able to hear us
Binding Templates
Ege: would we publish all of them as separate documents?
Kaz: we can publish them. group notes do not need to be perfect. however, we still need to clarify the potential gaps between the Binding Templates Notes and the existing implementations, e.g., node-wot, too.
<kaz> [adjourned]