W3C

– DRAFT –
WoT Security

03 April 2023

Attendees

Present
Jan_Romann, Jiye_Park, Kaz_Ashimura, Luca_Barbato, Michael_McCool, Tomoaki_Mizushima
Regrets
-
Chair
McCool
Scribe
luca_barbato

Meeting minutes

Review Minutes

McCool: Some typos to address

<McCool_> regarding minutes, a few typos

<McCool_> "chater" -> "charter", "sigining" -> "signing", "tls" -> "TLS", then in discussion of "onboarding" I did respond to Kaz saying that I felt we should get started on discussing some of the detailed work planning, but that yes, onboarding does not directly impact the charter

McCool: Consensus on the previous minutes edited?

McCool: Published

Review Architecture Assertion Presentation

<McCool_> Slides on Architecture

Segmented network

<kaz> arch-security-consideration-segmented-network

McCool: This assertion is at risk but also easy to implement

McCool: How can test this assertion?

McCool: Access to trusted environment means access to all the devices

McCool: The assertion is about the whole system, since it is in architecture

Jiye: So it applies to the deployer, not the single device

Jiye: The question was about if the device has to be aware it is in a specific segment

Jiye: How do we check it is implemented

McCool: In this case is the deployer has to confirm

Jiye: Then it should not be at risk

Jiye: Would be a good idea to write down a comment in the document

Jan: Would be a good idea to move this to the best practice section?

McCool: The assertions are about the bare minimum for security
… e.g. guest network vs iot network in a hotel deployment

Kaz: I agree with Jan and Luca, but technically we should, but we cannot do that today
… this assertion is not a requirement for the architecture itself
… those assertions are SHOULD
… and they are useful to point best practices

McCool: This section could be downgraded to informative later

Kaz: if we do not have implementations, we can move them to informative
… we can make an editor's note for those best practices

McCool: As long it is not already included in another section

Using PSK

<McCool_> https://w3c.github.io/wot-architecture#arch-security-consideration-use-psk

McCool: Certificates is one way to share pre-shared keys
… there are other ways to share them

McCool: We can add a sentence to say that is not required to use TLS-PSK.

Jiye: Browsers cannot use PSK, if we want to support browsers we have to allow other systems

McCool: <Issue created about it>

McCool: Reword to use Certificate instead of pre-shared-key

<McCool_> w3c/wot-architecture#900

Communication Platform

<McCool_> https://w3c.github.io/wot-architecture#arch-security-consideration-communication-platform

Jiye: Ege is not clear on what this assertion is about and I wrote my understanding of it

McCool: it boils down to the definition of Platform
… a weaker bridge cannot be created if the bridged ecosystem requires to have the same level of security (e.g. OCF)

Luca: This is a specification of the basic compatibility requirement

Luca: How to test it though?

McCool: The test would be about the "bridge" more than to the TD, if the "bridge" is compliant, then the TD has only to faithfully describe it

Kaz: we can open an issue to clarify the relationship between "IoT Ecosystem" and "IoT Platform" here.

<kaz> [adjourned]

Minutes manually created (not a transcript), formatted by scribe.perl version 210 (Wed Jan 11 19:21:32 2023 UTC).