W3C

- DRAFT -

WoT Security

18 Mar 2019

Agenda

Attendees

Present
Kaz_Ashimura, Mike_Bergman(guest_from_CTA), Michael_McCool, Tomoaki_Mizushima, Elena_Reshetova
Regrets
Chair
McCool
Scribe
kaz

Contents


Cross-alliance IoT Baseline Security Workshop

McCool: we're interested in the interfaces
... how to make the IoT devices accessible
... also working with the HTTPS local CG within W3C
... HTTPS in browsers expect global visibility

Bergman: issue with local hosts

McCool: yes
... the other group interested is IIC
... CTA is related to consumer devices?

Bergman: we're also reaching out IIC
... we're trying to talk with multiple sectors

McCool: (starts to share his screen for the questionnaire from CTA)
... Elena Reshetova, one of the active members, is not available today but would like to review this spreadsheet
... we're working for official transition for our spec now
... but happy to go over the questionnaire
... activities we're working with

Bergman: the genesis is US regulatory activity
... but people are not only interested in US regulatory
... but also global activities
... the concern is many organizations creating many Best Practices, etc.
... many philosophies, etc.
... too many fragmentations is not good
... general principle and requirements needed
... we got over 70 organizations
... we want de-fragmentation
... medical, automotive, other industries, ...
... some requirements specific to industries

McCool: whole a lot of good topics so far
... OCF, IIC, etc.
... but each work covers some specific areas
... IETF's best practice work as well
... the other issue is conformance
... baseline for ecosystems
... however, big problem with IoT
... so many people are building stuff

Bergman: we have similar discussion
... suitable set of specs for conformity
... some kind of label structure
... general baseline for reasonable version

(some more discussion on IoT fragmentations)

Bergman: what we did was sent out invitations
... all the government agencies
... also internet societies
... baseline security
... EU, UK, etc.
... not only US government
... have spoken with METI from Japan, Canada, etc.
... have you seen the attendees list?
... partners involved
... good time to go through the spreadsheet

McCool: how to deal with the fragmentation problem
... this is "narrow waist" idea by WoT
... deep history of multiple industries
... like automotive, factory automation
... (shows the spreadsheet)
... suspect your documents are more than this
... but good questions listed here
... many arguments
... we have been having many discussions
... the issue is there is too much work for a small group like us

Bergman: here is 15 organizations here
... if some of them are out of scope of your work, that's also useful input

McCool: should put that kind of comment here?

Bergman: column G to be used

McCool: ok
... most of the things are reasonable
... there are common references
... but no definition about lifecycle so far
... standard terminology for lifecycle, etc.

Bergman: the idea of device caries out some XML data
... what the intent is
... what if IoT attempt to contact Facebook, etc.?
... should be denied

McCool: very interesting
... could be supporeted with links in TD
... general mechanism for extension using hyperlink
... relation with MUD things here
... a lot of topics here are interesting

Bergman: in the initial instruction about the spreadsheet
... you can't change the order of the lines
... but can add additional lines

McCool: ok
... unfortunately, we don't have Elena today
... but our main spec from W3C WoT is Thing Description
... might be various ways to handle metadata for TD
... change the ID, track ID, etc.
... various risks associated to IDs
... probably two sub categories here
... one is confidentiality
... another is modification protection

Bergman: metadata confidentiality and arguments on browser fingerprinting
... on the other hand, interoperability of systems
... my work related to internet video distribution
... can explain both sides of the picture
... in terms of integrity of the data
... can see link for encryption

McCool: thinking more about end-to-end connection
... link protection is given
... might want to cash intermediate data?
... may want to have a trusted gateway

Bergman: there is a section about protection and trust
... requirements to go for E and comments for G

McCool: one of the fundamental things is end-to-end security
... (7E) Yes, if includes COSE (object security), ENd-to-end security is a necessity. Link security is inadequate if intermediaries
... one thing people are concerned is privacy information
... I think this list is for the first level requirements

Kaz: note that "security and privacy" has two target areas
... device side and user side
... WoT related to the device side
... and Verifiable Claims from W3C related to the user side
... so would make sense to talk with the Verifiable Claims WG as well

(Elena joins)

McCool: mentions several organizations/events

Bergman: why don't you send a message on the collocations

McCool: ok
... e.g., IIC
... CTA's workshop is occurring soon
... we should include summaries of those related specs

Elena: is the spreadsheet sent to me as well?

McCool: yes

Elena: will check it

Bergman: we've not done our final requirements yet
... discussion on Thursday
... sorry not to provide remote connection for the workshop

McCool: np
... Elena, can you look into the spreadsheet?

Elena: will do

McCool: double check the basic requirements, check the scope, give comments, etc.

Bergman: when we look at 6-10 equivalent spreadsheets
... comments on E and G are also useful as well as the questionnaire at F
... probably other people also have similar situation if you have specific notes at G

McCool: ok
... (adds comments to G26)

Bergman: can you check the data in column F?
... please make sure the column F would not effect other lines after you copied it from line 28

McCool: ok
... (adds comments on G29)
... baseline requirements

WoT explainer documents

McCool: have updated the explainer documents

https://github.com/w3c/wot-thing-description/blob/b1c5afe239af2bca8cb5b711c4d785a8fb6777d0/explainer/Explainer.md

TD explainer above

McCool: listed features at risk
... based on the test results
... we need 2 implementations for each feature
... TD security scheme to be included in the results soon

[[

In summary, the features currently at risk are:

The APIKey, Cert, PSK, OAuth2, PoP, and Public security schemes.

The scopes field in forms since it is only used with the OAuth2 security scheme.

Certain options for security, such as proxy fields and some specialized options for some security schemes.

]]

McCool: (explains the features at risk)
... would clarify what "runtime" is for the Architecture document
... management of private information, etc., as well
... clear separation is the direction to go
... better to have a separate container
... get rid of confusion
... please continue to work on the CTA spreadsheet

Elena: ok

[adjourned]

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version 1.154 (CVS log)
$Date: 2019/04/16 20:52:00 $