W3C

– DRAFT –
WoT Profile

17 July 2024

Attendees

Present
Ben_Francis, Daniel_Peintner, Jan_Romann, Kaz_Ashimura, Luca_Barbato, Tomoaki_Mizushima
Regrets
Ben, Ege, Koster
Chair
Luca
Scribe
kaz

Meeting minutes

Minutes

Luca: (goes through the minutes)

June-11

approved

July-9

approved

Out of box clarification

PR 415 - Out of the box clarification

changes

merged

Profiles for ecosystems

Kaz: any issues and/or documents?

Luca: this is an open discussion for today

Luca: (shows his diagram about profiles vs ecosystems)
… Protocol: HTTP, MQTT, WebTransport, WebSocket
… Data Serialization: JSON, CBOR, BSON, ProtoBuf
… Data Structure: DataSchema, DataMapping
… Meaning of Data: SAREF, SSN
… Behavior: Async Action, Affordance Relationship
… System Architecture: Security, Disovery, Onboarding
… examples of ecosystems: OPC-UA, MATTER
… initial work on OPC-UA example done
… would work on MATTER too

<benfrancis> https://w3c.github.io/wot-binding-templates/#platform-bindings

Ben: sounds to me platform design
… protocol binding and payload binding
… think it's opposite to what we've been doing for Profile
… and don't see any benefit to think about a "MATTER Profile" because MATTER itself already works for its interoperability
… also whey do we need to think about OPC-UA Profile?
… Binding Template should describe the interface for Brown-field devices

Luca: ecosystems do have out-of-box interoperability issues
… more about describing something
… Binding doesn't handle restriction
… Profile would be better fit for that purpose
… if the Consumer is implementing some specific Profile, you can't handle that using Binding
… if we have a Profile exposing a Thing, you for sure know about everything that exposes the Thing
… my Thing supports this, and OPC-UA Thing supports that
… you can use WoT as the bridge for something outside WoT

Ben: that's not we've been handling for Profile...
… there are two possible ways Binding and Profile
… this discussion today is rather opposite to what Profile should do for "interoperability"

Luca: sort of feedback we got was we should handle brown-field devices too
… they can be easily considered as "Vendor Profiles"
… we want to focus more on the interoperability part
… we use Profile to provide a good way to cover everything
… provide a bridge for WoT
… we need something that is a way forward for the bridge
… I can see your point but we need to see a good balance

Ben: don't think it would be helpful to just thinking about supporting subset of WoT

Luca: nothing prevents you to have partial way

Ben: all MATTER devices are already interoperable themselves

Luca: within one specific ecosystem all the devices are interoperable
… but if we want to involve devices from another ecosystem, we need to support interoperability
… WoT is a solution for that purpose

<Zakim> Ben, you wanted to react to benfrancis

Kaz: clarification on what Profile does is good
… but we should clarify what Profile does and Binding does based on some specific use case and component setting. For that purpose, probably we should start with our experienced examples.

Luca: agree most parts can be handled by Binding Template
… on the other hand, we have to make sure what to be done for ecosystems
… regarding the part about MATTER
… Koster mentioned that
… regarding OPC UA, they're ready for discussion

Ben: but can be handled by Binding Templates

Luca: right
… but Binding Templates can't handle everything

Ben: let's describe based on a concrete use case

Luca: can't say much about OPC UA itself since I'm not an expert of OPC UA

Mizushima: agree with Kaz
… but what is the purpose of standardization for Profile here?

Luca: expected it would good to think about mapping between Profile and Ecosystem
… but it seems that's not a good idea
… if we can find a good reason for this direction, we can get a good starting point
… need out of box compatibility

Kaz: Basically, interoperability with other ecosystems can be handled by Binding Templates, but if there are any pain points with that, we need to clarify the pain points for interoperability with external ecosystems a bit more

Ben: can see the argument for OPC UA Profile

<Tomo> +1 for kaz

Ben: maybe OPCF might want to use a WoT Profile but don't see benefit for a possible MATTER Profile

Luca: wanted to see what can and can't be done for MATTER too based on Koster's feedback
… during the next meeting, we can concentrate on the OPC UA part
… next call will be held in 2 weeks

[adjourned]

Minutes manually created (not a transcript), formatted by scribe.perl version 221 (Fri Jul 21 14:01:30 2023 UTC).