Meeting minutes
Minutes
<kaz> Nov-24
After going over today's agenda SK starts reviewing the minutes from last call
Sebastian: It probably makes sense to add a pull request topic as PRs have been discussed in the architecture section
As kaz currently away, SK goes to the next agenda point for now
Publication plans
Sebastian: As announced in last week's call we plan to publish a new Working Draft in december
… I would like to do the resolution about new WD in next week
… then we might also have feedback from the W3C regarding a new namespace
Sebastian: Do have any news regarding the namespace, kaz?
Kaz: No, not yet.
Sebastian: I am already wainting for an email response
Kaz: I will ping them on that email thread
Sebastian: I saw that the JSON-LD working group(?) is not using years in their namespace
… we could do the same, as specification is "timeless"
… does not follow the pattern we agreed on before, however
Ege: I would also prefer not to have the year
… would be more compact
… it also wouldn't look outdated once version 2.0 is released
Sebastian: I agree, TD 1.0 looks already kind of outdated due to the 2019 year in the namespace
Kaz: Typically, namespaces are only a URI and not a URL.
… We already have a basic URI and can extend it based on the version number. But we could also change it as well
… we should discuss this with Ralph and Philipp
Sebastian: A year could also be an indicator of how stable a standard is
<kaz> Namespace URI guideline
Kaz: what is important here is what kind of URI would be good to manage multiple versions of TD specs and we should look into existing W3C standards as best practices. Also we can ask Ralph and PLH for advice.
Sebastian: We could invite them to one of our next call to discuss this
Sebastian: Would be cool to have a decision on this by next week for the new WD, which could be released before christmas then
PRs
PR 1264
<kaz> PR 1264 - Fix 948 (add OAuth2 flow examples)
Cristiano: We discussed this in the security call and had some resolution to change this PR
… We are tracking this in the security call and we hope that this PR will be ready next week
McCool: There are still a number of open discussion points
Sebastian: Let's have another look next week again
PR 1283
<kaz> PR 1283 - update webhooks example
Sebastian: There is some discussion going on here
… had no time to look on the latest comment. Ege, do you have some news here?
Ege: I think I added some comments last week
… I checked the RFC spec. The example in the comment is showing the query parameters for uri variables. I think we should add examples for both variants of URI variables.
… I added a final example with both approaches
… I will provide the example in a separate PR
Sebastian: The PR in question will not be merged but Michael Lagally's comments will be taken into account
PR 1301
<kaz> PR 1301 - update Section 8.3
Sebastian: This PR is coming from the architecture call
… addresses the constraint that supported protocols must have an IANA registration
… however, many IoT protocols do not have an official registration like MQTT for example
… therefore I would not add this constraint to the TD specification
… however, I added some information regarding IANA registration to this PR
… making the IANA registration optional for guaranteeing a unique protocol binding
Sebastian: So what I am saying is: If the protocol is IANA registered we are fine. If not we can use the binding templates as best practices
Ege: This is actually already part of the binding templates document
Sebastian: Great, could you add comment on this to the PR later?
Ege: Will do
Sebastian: PR is not going to be merged now, some improvements based on Ben Francis' comments will be incorporated as well as the addition by Ege
PR 1304
<kaz> PR 1304 - Remove TD Canonicalization
McCool: This PR removes canonicalization and updates the implementation report
… it also removes some tooling
… I also removed the comment regarding this addition in the "New features" list
… I updated the list of assertions and moved it to a place where it is easier to find
… there are also some cleanups regarding the JSON schema
… I would propose merging this PR and see if everything renders correctly
McCool: Canonicalization should not go into the profile spec as it is primarily for signing. We should remove it for now and later discuss where should be re-added.
Kaz: I am okay with merging this PR
… we should keep the content somewhere for potential usage in the future
McCool: The content is still on Github. But moving it to its own branch and tagging it might be a good idea
SK creates a new branch as a canonicalization archive ("canonicalization-2021-12-01")
PR 1305
<kaz> PR 1305 - Fix links reported by LinkChecker
Sebastian: Daniel found some broken links
Daniel: Just a quick note: link checker reports some broken fragment identifiers but this seems to be an issue that can't be avoided.
Kaz: note that the link checker should be applied to a static HTML not a ReSpec document
Sebastian: As there are no objections, I will merge this PR
PR 1306
<kaz> PR 1306 - refactor: clarify that property can be retrieved and/or updated
Sebastian: This PR updates an assertion statement that properties can be readable and/or writable
As there are no objections, SK proceeds with merging this PR
Daniel: The are two related issues, one has to be closed automatically
PR 1307
<kaz> PR 1307 - fix observable default assumption in Example 4
Sebastian: This PR adds observable to the list of default values in example 4
… somehow a change from another branch was included initially, I removed it
As there are no objections, SK proceeds with merging this PR
The resolved Issue 1293 is closed manually by SK
PR 1308
<kaz> PR 1308 - Fix roundtripping with language tags
Sebastian: This PR is quite new. I need some more time and would propose to postpone it
PR 1309
<kaz> PR 1309 - New sub-sections to improve reading in Section 6.3
Sebastian: This PR introduces some more subsections under security definitions and forms which makes it easier to jump to these parts of the document
… as this PR is relatively new, should we decide this now or discuss it next week?
… Maybe Ege can have a look on it and then we decide next week
Issues
Issue 977
<kaz> Issue 977 - Introducing assumptions about TD instances that have no IP addresses
Sebastian: This issue deals with devices that have no static IP addresses
… IP addresses of things can change in a network
… also things could have no IP address but the TD is still valid
… similar to webpages where links might be relative to where the document came from
… a similar assumption could be made in TDs
… this was discussed in the issue in an example by Ege
… we could also solve this with TMs which are resolved during runtime
… should we address this in version 1.1 or defer to 2.0?
Kaz: agree we should defer this to 2.0, and even if we defer this to 2.0 we should clarify use cases in more detail, e.g., within a smart home environment with DHCP addressing. And after that kind of clarification, we can discuss the detail again.
Sebastian: Is this already discussed in the discovery taskforce?
Daniel: I guess there is no one left at the moment in the call
Sebastian: We should wait for feedback from the discovery taskforce and discuss this again later
Issue 1280
<kaz> Issue 1280 - UML diagrams contain lines with multiple properties
Sebastian: I think here we need Michael McCool
… to get a status of this issue
Other issues
Sebastian: we need to look into the issues marked as "V1.1"
… most of them are editorial ones
Issue 1300
Issue 1300 - Please check if the rules of chapter 5.2 are valid for a TM
Sebastian: TM relies on TD information model the same rules should be valid also for TM definitions.
Issue 1296
Issue 1296 - Canonicalization - array vs. single value
Sebastian: would suggest we close this issue
Daniel: make sense
Jan: maybe worth thinking about several data types
… maybe for 2.0, though
Sebastian: fundamental discussion has already started
… the mechanism for JSON-LD's @context is not much flexible
… regarding this issue itself, some discussion done in the past
… maybe can be done by alternative compact representation
… anyway this is an issue can't be resolved for TD 1.1
Jan: yeah, I'm also thinking about TD 2.0 for this
AOB?
Sebastian: any specific issue to be discussed today?
(none)
Sebastian: then would close this call
… please review the updated TD draft
… so that we can make a resolution for WD publication
[adjourned]