Prev minutes

<Mizushima> scribenick: Mizushima

<kaz> Aug-12

Taki: approved

Issue and PRs

Daniel: victor take about updating rendering script.

<kaz> PR 951

Kaz: note that this issue 951 has several conflicts, and we need to resolve them before merging it

<kaz> Issue 950

<inserted> Related issue 957

Daniel: This is about unobserveproperty.

Taki: we find this as clarification in TD v1.1.

Issue 952

Taki: This issue is about use case how to describe TD

<kaz> Issue 617

Taki: This issue is about array. And this is continuing to discuss.

Issue 955

Cristiano: This is about validate "oneOf".

Taki: Is this a limitation pattern?

Cristiano: yes
... This is Security Schema.

Daniel: JSON schema can check oneOf constraint.

Taki: we should go on discussion.

<kaz> Issue 956

Daniel: This issue is about JSON schema validation issue.

<kaz> TD draft - 5.3.2.1 DataSchema

Taki: we define JSON schema in TD.

Daniel: we should read iri-referece.
... Ege proposed.

Taki: First session is finished.

<kaz> [break for 5mins]

<kaz> scribenick: kaz

PR 951

PR 951

Taki: Victor raised this issue fixing the rendering mechanism

McCool: rendering may not match up the current index.html
... not sure if it's tested
... would break the content of index.html

Daniel: it's the case here
... we need to update the ttl file as well

McCool: we can once merge this PR and see the difference between the result and the current index.html

Kaz: +1
... during the 1st hour today, I also suggested we remove the conflicts

Taki: (adds comment to PR 951)

Daniel: need to check the security schema too

McCool: mainly OAuth2 changes
... including new flows
... maybe Victor should once pull down the current index.html and see the diff himself

Kaz: yeah, anyway we need to update the index.template.html file as well

McCool: yeah

Taki's comments

(tentatively use "index2.template.html" and "index2.html" to see the rendering results)

<inserted> PR 951; or we can tentatively use "wot-security2.ttl" as well)

(wot-security.ttl to be rebased before merging PR 951)

Daniel: will talk with Victor

Security issues

Security issues

McCool: three security PRs

PR 945

PR 944

Preview - 5.3.1.1 Thing

McCool: "security" term has "CombinationSecurityScheme"
... one more PR to clean up the combination

diff

McCool: Example 12 uses "combo_sc"

Example 12

McCool: another example 16

Example 16

Ege: ok with this change

McCool: functionally not adding anything
... feature we already have on proxying
... but different syntax

Ege: someone can easily test
... but not implemented yet
... so opened an issue

<Ege> https://github.com/w3c/wot-thing-description/issues/901

McCool: that's an issue on implementations

Cristiano: also issue on oneOf scheme

<taki> https://github.com/w3c/wot-thing-description/pull/945

<taki> https://github.com/w3c/wot-thing-description/pull/944

McCool: question of JSON in the proxy
... need an example on proxy
... easy to get multiple producers
... but getting consumers would be problematic
... can add an Editor's note on possible limitation

Ege: why do we need to include features which are not used by any implementations?

Kaz: note that we should think about the need for features based on the use cases and the requirements rather than the implementation status

McCool: yeah
... think there are specific use cases for the combo features
... as mentioned, producer side is easier
... consumer side needs further work

Cristiano: possibly complex boolean as well?

McCool: possible assertion on whether you're allowed to use complicated combination or not
... would be reasonable constraint

Cristiano: what about the backward compatibility?

McCool: let's say 2.0 spec would remove the restriction

Cristiano: so that's more forward compatibility

McCool: yeah, in the future

Taki: (adds comments to PR 944 based on the discussion)

McCool: can add restriction on nested combo setting

Cristiano: ok

Ege: note that node-wot doesn't support multiple security schemes within a TD

McCool: think there is a use case of mashing up multiple micro services

Kaz: yeah, given WoT is expected to be used for multiple-vendor IoT systems, we should consider fusion of multiple micro services
... maybe we need a separate mechanism for that purpose but we could start with this proposal

McCool: I should add an Editor's note on the restriction for now

Kaz: so this is an expected feature for a mashing up server

McCool: mashing up may require basic for a micro service and another scheme for another micro service

Kaz: yeah, not only the security scheme but all the TD capabilities to be handled in parallel for the multiple micro services, e.g., one for TV and another for air conditioner

McCool: right
... let me think about how to modify this PR itself

PR 945

McCool: "security" can just be treated as an array of SecurityScheme objects
... security scheme object vs array
... let me go ahead and think about it
... will update the PR for further discussion next week

Ege: array of names or security scheme object
... or combination of those

McCool: on the consumer side it's easy
... there are different ways to write TDs for the same purpose
... maybe we should capture the point using another issue
... we need different description for assertion tests

Taki: (adds comments)

McCool: OK with holding off this PR 945 until the PlugFest

Taki: let's talk about when to merge later (after the PlugFest)

McCool: we have a preview version of this PR
... and could use it as the basis for PlugFest

Taki's comment

<Zakim> kaz, you wanted to suggest we create a bigger issue on mash up server which integrates multiple micro services

Kaz: issue on mashing up in general?

McCool: yeah
... all the actions/properties to be mashed up for that purpose
... probably we should create an issue

Kaz: yeah, and we need some concrete use case description

Taki: a possible issue for wot-architecture as well?

Kaz: good question

McCool: we already have use cases for digital twin, etc.

Kaz: so we should create a TD issue

McCool: yeah, add mashing up for components

Taki: (creates an issue on that)

Issue 958 (content to be filled in later)

[adjourned]