![](\"/user_uploads/2/7/m4QmUACTijQzr1dEL_6S6RUU/image.png\")
LIFTOFF
", "time": "2025-11-03T14:30:07.000Z"}, {"author": "Richard Barnes", "text": "i believe Gemini will also be taking notes https://datatracker.ietf.org/doc/draft-rescorla-auto-minutes/
", "time": "2025-11-03T14:34:36.000Z"}, {"author": "Jim Fenton", "text": "Pauline, please turn off video
", "time": "2025-11-03T14:36:52.000Z"}, {"author": "Rich Salz", "text": "If you are remote and not talking, make sure to turn your video off.
", "time": "2025-11-03T14:37:14.000Z"}, {"author": "Muhammad Usama Sardar", "text": "Richard Barnes said:
\n\n\ni believe Gemini will also be taking notes https://datatracker.ietf.org/doc/draft-rescorla-auto-minutes/
\n
This looks great. I have surely been in situations where I would have hoped someone else to take notes, or at least help me out. I propose adding the link https://ietfminutes.org/ to the draft for folks to see it in action.
", "time": "2025-11-03T14:42:10.000Z"}, {"author": "Bron Gondwana", "text": "Does this belong in WIT?
", "time": "2025-11-03T14:43:11.000Z"}, {"author": "Martin Thomson", "text": "Why is this not HTTP?
", "time": "2025-11-03T14:44:54.000Z"}, {"author": "Martin Thomson", "text": "I couldn't identify why this was not something that HTTP is capable of doing.
", "time": "2025-11-03T14:45:13.000Z"}, {"author": "Daniel Gillmor", "text": "@meetecho, i think ted hardie is in the room, but the camera doesn't point at him.
", "time": "2025-11-03T14:46:06.000Z"}, {"author": "Lorenzo Miniero", "text": "@Daniel Gillmor you're right, sorry about that!
", "time": "2025-11-03T14:46:53.000Z"}, {"author": "Daniel Gillmor", "text": "all looks good, thanks!
", "time": "2025-11-03T14:47:01.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Martin, I just saw the assertion HTTP doesn't support things but no explanation of why it doesn't.
", "time": "2025-11-03T14:47:30.000Z"}, {"author": "Martin Thomson", "text": "PHB yes, exactly my point
", "time": "2025-11-03T14:47:57.000Z"}, {"author": "Martin Thomson", "text": "The use of GRPC is using HTTP, just with an extra layer on top. So clearly HTTP can do this.
", "time": "2025-11-03T14:48:56.000Z"}, {"author": "Harald Alvestrand", "text": "The protocol as specified repeats some basic errors of HTTP, such as using an URI both to identify a data object and to access it.
", "time": "2025-11-03T14:49:53.000Z"}, {"author": "Harald Alvestrand", "text": "It also squats on a port number.
", "time": "2025-11-03T14:50:00.000Z"}, {"author": "Ben Laurie", "text": "for no obvious reason
", "time": "2025-11-03T14:50:18.000Z"}, {"author": "Martin Thomson", "text": "Port number squatting doesn't bother me Harald, even if there is no obvious reason to do that.
", "time": "2025-11-03T14:50:29.000Z"}, {"author": "Martin Thomson", "text": "The same goes for the URI scheme.
", "time": "2025-11-03T14:50:37.000Z"}, {"author": "Daniel Gillmor", "text": "let the person who has never squatted a port number cast the first stone
", "time": "2025-11-03T14:50:54.000Z"}, {"author": "Martin Thomson", "text": "Port number squatting is only an issue if you think that a port number registry makes sense.
", "time": "2025-11-03T14:50:58.000Z"}, {"author": "Simon Josefsson", "text": "hi! re base64 etc draft, i have no slides but merely was asking for guidance if there is interest in progressing this to Internet Standard
", "time": "2025-11-03T14:51:02.000Z"}, {"author": "Harald Alvestrand", "text": "yeah, it's just untidy. I have never heard of the protocol that 3101 is currently assigned to, but it's registered.
", "time": "2025-11-03T14:51:10.000Z"}, {"author": "Martin Thomson", "text": "URI scheme squatting is not a major concern for a different reason: everyone does it.
", "time": "2025-11-03T14:51:31.000Z"}, {"author": "Eliot Lear", "text": "Acronym award!!
", "time": "2025-11-03T14:51:53.000Z"}, {"author": "Phillip Hallam-Baker", "text": "Well its a registered trademark.
", "time": "2025-11-03T14:52:16.000Z"}, {"author": "Martin Thomson", "text": "That's a very interesting definition of \"need\"
", "time": "2025-11-03T14:52:31.000Z"}, {"author": "Richard Barnes", "text": "To be clear: The Gemini minutes are currently generated off of the Proceedings, so there will be some delay in their publication.
", "time": "2025-11-03T14:52:36.000Z"}, {"author": "Martin Thomson", "text": "They are generated off the transcripts, which are generated from the recordings, which need to be finalized and then uploaded
", "time": "2025-11-03T14:53:09.000Z"}, {"author": "Ted Hardie", "text": "I believe it is trademarked only with the hyphen, Band-Aid.
", "time": "2025-11-03T14:53:15.000Z"}, {"author": "Martin Thomson", "text": "Trademarks only apply within a predefined scope.
", "time": "2025-11-03T14:53:39.000Z"}, {"author": "Thom Wiggers", "text": "Why not http well-known? These agents seem to like http a lot
", "time": "2025-11-03T14:53:45.000Z"}, {"author": "Andrew Campling", "text": "Dispatch to the trademark working group?
", "time": "2025-11-03T14:53:46.000Z"}, {"author": "Daniel Gillmor", "text": "\"grep 3101 /etc/services\" and https://www.iana.org/assignments/service-names-port-numbers/service-names-port-numbers.xhtml show nothing for 3101, fwiw
", "time": "2025-11-03T14:54:03.000Z"}, {"author": "Martin Thomson", "text": "I don't know what scope band-aid has, but I don't know if it extends into AI discovery. I'd be surprised if it did.
", "time": "2025-11-03T14:54:08.000Z"}, {"author": "Martin Thomson", "text": "Thom: well-known on what name?
", "time": "2025-11-03T14:54:34.000Z"}, {"author": "layer8", "text": "My $0.02 is .well-known is a great bootstrap/complement when you already know the right domain; it\u2019s simple to publish/rotate via the web tier and can carry larger, richer JSON documents. But it doesn\u2019t solve \u201cbrand \u2192 domain\u201d by itself
", "time": "2025-11-03T14:54:35.000Z"}, {"author": "Maxwell Gerber", "text": "Does DNS?
", "time": "2025-11-03T14:54:45.000Z"}, {"author": "Rich Salz", "text": "tmsearch.uspto.gov; bandaid is not TM'd
", "time": "2025-11-03T14:54:53.000Z"}, {"author": "Richard Barnes", "text": "the universal impulse to Just Shove It In The DNS
", "time": "2025-11-03T14:54:58.000Z"}, {"author": "Jim Fenton", "text": "Unfortunately DNSOP is also meeting during this timeslot.
", "time": "2025-11-03T14:55:12.000Z"}, {"author": "Thom Wiggers", "text": "I mean I don\u2019t care either way about shoving things in DNS
", "time": "2025-11-03T14:55:29.000Z"}, {"author": "Wendy Seltzer", "text": "Just don't try to make an adhesive bandage of it
", "time": "2025-11-03T14:55:44.000Z"}, {"author": "Martin Thomson", "text": "", "time": "2025-11-03T14:55:44.000Z"}, {"author": "Pete Resnick", "text": "Any of the dns-at-ietf list people in the room?
", "time": "2025-11-03T14:55:47.000Z"}, {"author": "Jonathan Hoyland", "text": "Richard Barnes said:
\n\n\nthe universal impulse to Just Shove It In The DNS
\n
If we don't shove everything in DNS how can it always be DNS?
", "time": "2025-11-03T14:55:52.000Z"}, {"author": "layer8", "text": "the agents yearn for DNS
", "time": "2025-11-03T14:56:00.000Z"}, {"author": "Murray Kucherawy", "text": "DNSOP has a second session usually, don't they?
", "time": "2025-11-03T14:56:04.000Z"}, {"author": "Martin Thomson", "text": "We have this global scale key value store, why not store it there?
", "time": "2025-11-03T14:56:15.000Z"}, {"author": "Daniel Gillmor", "text": "@meetecho please point to the in-person speaker again \u263a
", "time": "2025-11-03T14:56:18.000Z"}, {"author": "Daniel Gillmor", "text": "thanks!
", "time": "2025-11-03T14:56:21.000Z"}, {"author": "Murray Kucherawy", "text": "Yeah DNSOP is also Friday.
", "time": "2025-11-03T14:56:33.000Z"}, {"author": "Ted Hardie", "text": "BoF as outcome?
", "time": "2025-11-03T14:56:50.000Z"}, {"author": "Ted Hardie", "text": "Whether within the AIPROTO or separately?
", "time": "2025-11-03T14:57:00.000Z"}, {"author": "Jim Fenton", "text": "BOF: WG forming or not?
", "time": "2025-11-03T14:57:13.000Z"}, {"author": "Eliot Lear", "text": "A reminder: there is a mailing list to consider how DNS work should be organized at IETF: dns-at-ietf@ietf.org.
", "time": "2025-11-03T14:57:50.000Z"}, {"author": "layer8", "text": "How does the key value store enforce global uniqueness? That's probably showing my ignorance but DNS at least has that solved
", "time": "2025-11-03T14:57:53.000Z"}, {"author": "layer8", "text": "(somewhat)
", "time": "2025-11-03T14:58:00.000Z"}, {"author": "Martin Thomson", "text": "Sorry, that earlier link was dead. This is the wordmark: https://tsdr.uspto.gov/#caseNumber=71199064&caseSearchType=US_APPLICATION&caseType=DEFAULT&searchType=statusSearch
", "time": "2025-11-03T14:58:16.000Z"}, {"author": "Ted Hardie", "text": "@layer8 over what time scale? Because the DNS doesn't guarantee that at all timescales.
", "time": "2025-11-03T14:58:38.000Z"}, {"author": "Wendy Seltzer", "text": "Maybe we can help the USPTO with their URL schemes, once our other work is done
", "time": "2025-11-03T14:58:44.000Z"}, {"author": "Thom Wiggers", "text": "A band-aid as the idiom is so ubiquitous and trivialized that I doubt J&J have any leg to stand on
", "time": "2025-11-03T14:59:33.000Z"}, {"author": "Thom Wiggers", "text": "See also: Hoover, Xerox
", "time": "2025-11-03T14:59:43.000Z"}, {"author": "Jay Daley", "text": "For info: The first mention of AI in the RFC Series is (I think) RFC 152 from May 1971 and the first one to include AI output is RFC 439 from January 1973
", "time": "2025-11-03T15:00:05.000Z"}, {"author": "Lucas Pardue", "text": "When I hear band aid, I think 80s haircuts
", "time": "2025-11-03T15:00:43.000Z"}, {"author": "Ted Hardie", "text": "Non WG forming BoF as next step?
", "time": "2025-11-03T15:01:52.000Z"}, {"author": "Andrew Campling", "text": "Non WG BoF seems like a good next step
", "time": "2025-11-03T15:02:20.000Z"}, {"author": "layer8", "text": "Yes, we originally asked for a BOF but Eric recommended dispatch and side meeting first. Which we agreed!
", "time": "2025-11-03T15:02:21.000Z"}, {"author": "layer8", "text": "Please come to the side meeting digitally or in person if you're here and want to talk about some of the mechanics! :)
", "time": "2025-11-03T15:02:40.000Z"}, {"author": "Kathleen Moriarty", "text": "A BoF is a good direction
", "time": "2025-11-03T15:02:43.000Z"}, {"author": "Phillip Hallam-Baker", "text": "It doesn't need to be limited to AI
", "time": "2025-11-03T15:02:46.000Z"}, {"author": "Phillip Hallam-Baker", "text": "Everything that is talked about applies to any service
", "time": "2025-11-03T15:03:00.000Z"}, {"author": "Pete Resnick", "text": "Clapping???
", "time": "2025-11-03T15:03:14.000Z"}, {"author": "Pete Resnick", "text": "Did the plane land successfully or something?
", "time": "2025-11-03T15:03:34.000Z"}, {"author": "Aaron Parecki", "text": "the dispatch was dispatched successfully
", "time": "2025-11-03T15:03:55.000Z"}, {"author": "Kathleen Moriarty", "text": "Yes, Phil, but AI is driving current work due to how prolific it is at the moment as you can see across WGs with scrapers and agents that are accounting for quite a bit of Internet traffic at the moment.
", "time": "2025-11-03T15:04:04.000Z"}, {"author": "Kathleen Moriarty", "text": "Last MAPRG session was quite informative if watching it back is helpful to some.
", "time": "2025-11-03T15:04:35.000Z"}, {"author": "Lucas Pardue", "text": "Is this CFR === NAT?
", "time": "2025-11-03T15:04:56.000Z"}, {"author": "Ted Hardie", "text": "MASQUE enthusiasts to the microphone, please.
", "time": "2025-11-03T15:04:56.000Z"}, {"author": "Thom Wiggers", "text": "This sounds like Apple private relay
", "time": "2025-11-03T15:05:21.000Z"}, {"author": "\u00c9ric Vyncke", "text": "it seems so indeed, OHTTP
", "time": "2025-11-03T15:05:36.000Z"}, {"author": "Christian Ams\u00fcss", "text": "We should just have done proxies right when HTTPS was introduced\u2026
", "time": "2025-11-03T15:06:03.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Kathleen, but the problems with DNS discovery are more general than just AI. Namely, the interaction between SRV records and prefixed TXT records is underspecified, the TLSA record is limited to TLS server with a bunch of peculiar assumptions, etc. etc.
", "time": "2025-11-03T15:06:04.000Z"}, {"author": "Thom Wiggers", "text": "I guess that this layers with ECH instead of having two hops like Apples solution
", "time": "2025-11-03T15:06:32.000Z"}, {"author": "\u00c9ric Vyncke", "text": "Sigh IPv4 address at IETF meeting in 2025...
", "time": "2025-11-03T15:06:32.000Z"}, {"author": "Kathleen Moriarty", "text": "I'm not arguing solutions, but rather justification to explore the problem more in a BoF as is appropriate for dispatch.
", "time": "2025-11-03T15:06:48.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Kathleen this is just layered protocol design, that the code running is an AI agent is irrelevant to a discovery layer
", "time": "2025-11-03T15:06:50.000Z"}, {"author": "Harald Alvestrand", "text": "Actually CFS seems like a perfect place to install destination-policing, end-user-aware censoring firewalls....
", "time": "2025-11-03T15:06:56.000Z"}, {"author": "Murray Kucherawy", "text": "::1 is the only address that understands me.
", "time": "2025-11-03T15:07:09.000Z"}, {"author": "Hannes Tschofenig", "text": "@Thom: It is not a bad thing if a proposal has some resemblance to something that already exists.
", "time": "2025-11-03T15:07:09.000Z"}, {"author": "Thom Wiggers", "text": "Sure
", "time": "2025-11-03T15:07:17.000Z"}, {"author": "Martin Thomson", "text": "0.0.0.0 understands everyone
", "time": "2025-11-03T15:07:31.000Z"}, {"author": "Thom Wiggers", "text": "But acknowledging prior work and what we can learn from that also helps
", "time": "2025-11-03T15:07:37.000Z"}, {"author": "Thom Wiggers", "text": "ETSI has a bunch of standards for middleboxes
", "time": "2025-11-03T15:08:04.000Z"}, {"author": "Thom Wiggers", "text": "But with the opposite goal
", "time": "2025-11-03T15:08:28.000Z"}, {"author": "Martin Thomson", "text": "This could be implemented with CG-NAT. But you would need to stop answering requests from law enforcement at those boxes.
", "time": "2025-11-03T15:08:47.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Kathleen, I was at the AI lab during the last AI winter. I can see the next AI winter starting before IETF125.
", "time": "2025-11-03T15:08:50.000Z"}, {"author": "Phillip Hallam-Baker", "text": "s/wintec/winter/
", "time": "2025-11-03T15:09:13.000Z"}, {"author": "Hannes Tschofenig", "text": "Putting the proposal in context of prior work would indeed be a good idea
", "time": "2025-11-03T15:09:24.000Z"}, {"author": "Ben Laurie", "text": "what is CFR?
", "time": "2025-11-03T15:09:53.000Z"}, {"author": "Eliot Lear", "text": "Customer Facing Relay?
", "time": "2025-11-03T15:10:05.000Z"}, {"author": "Sophie Schmieg", "text": "Isn't this what OHTTP does?
", "time": "2025-11-03T15:10:20.000Z"}, {"author": "Martin Thomson", "text": "CFS = Customer Facing Server
", "time": "2025-11-03T15:10:24.000Z"}, {"author": "Martin Thomson", "text": "This is something that OHTTP does
", "time": "2025-11-03T15:10:32.000Z"}, {"author": "Ben Laurie", "text": "I also don't know what OHTTP is :-)
", "time": "2025-11-03T15:10:37.000Z"}, {"author": "Martin Thomson", "text": "But OHTTP is somewhat specialized.
", "time": "2025-11-03T15:10:40.000Z"}, {"author": "Martin Thomson", "text": "RFC 9458
", "time": "2025-11-03T15:10:44.000Z"}, {"author": "Paul Wouters", "text": "oblivious http
", "time": "2025-11-03T15:10:49.000Z"}, {"author": "Murray Kucherawy", "text": "Oh, HTTP.
", "time": "2025-11-03T15:11:17.000Z"}, {"author": "Deirdre Connolly", "text": "O! HTTP
", "time": "2025-11-03T15:11:29.000Z"}, {"author": "Ben Laurie", "text": "I see - fix your surveillance problem by all using a single choke point?
", "time": "2025-11-03T15:11:33.000Z"}, {"author": "Martin Thomson", "text": "Oy, HTTP
", "time": "2025-11-03T15:11:39.000Z"}, {"author": "Deirdre Connolly", "text": "Oi! HTTP
", "time": "2025-11-03T15:11:47.000Z"}, {"author": "Daniel Gillmor", "text": "i pronounce it \"Oh! TTP\"
", "time": "2025-11-03T15:11:53.000Z"}, {"author": "Deirdre Connolly", "text": "(:flag_united_kingdom:)
", "time": "2025-11-03T15:11:54.000Z"}, {"author": "Paul Wouters", "text": "ben: isnt that what 8.8.8.8 is too? :)
", "time": "2025-11-03T15:12:06.000Z"}, {"author": "Deirdre Connolly", "text": "hehehe
", "time": "2025-11-03T15:12:14.000Z"}, {"author": "Stephen Farrell", "text": "so this is CGNAT then?
", "time": "2025-11-03T15:12:14.000Z"}, {"author": "Ben Laurie", "text": "single choke point? @Paul Wouters
", "time": "2025-11-03T15:12:26.000Z"}, {"author": "Dennis Jackson", "text": "Deirdre Connolly said:
\n\n\n(:flag_united_kingdom:)
\n
Can't use that proxy here mate!
", "time": "2025-11-03T15:12:35.000Z"}, {"author": "Lucas Pardue", "text": "new guidelines for NAT devices? ...
", "time": "2025-11-03T15:12:44.000Z"}, {"author": "Murray Kucherawy", "text": "@Deirdre: O! Canada!
", "time": "2025-11-03T15:12:49.000Z"}, {"author": "Ben Laurie", "text": "if so, then yes indeed...
", "time": "2025-11-03T15:13:03.000Z"}, {"author": "Deirdre Connolly", "text": ":flag_canada::maple_leaf:
", "time": "2025-11-03T15:13:03.000Z"}, {"author": "Rich Salz", "text": "so server-side NAT?
", "time": "2025-11-03T15:13:38.000Z"}, {"author": "Sophie Schmieg", "text": "Ohttp, like an Irish surname
", "time": "2025-11-03T15:13:43.000Z"}, {"author": "\u00c9ric Vyncke", "text": "CG-NAT is not really sustainable though
", "time": "2025-11-03T15:13:46.000Z"}, {"author": "Daniel Gillmor", "text": "NAT guidance would be an interesting transport discussion
", "time": "2025-11-03T15:13:53.000Z"}, {"author": "Eliot Lear", "text": "A few of us suggested this sort of aggregation as a mixer in the SPRINT workshop some time ago.
", "time": "2025-11-03T15:14:13.000Z"}, {"author": "Ben Laurie", "text": "you mean as in a mix network?
", "time": "2025-11-03T15:14:39.000Z"}, {"author": "Rich Salz", "text": "\n\nCG-NAT is not really sustainable though
\n
Isn't this a more theoretical than practial concern so far?
", "time": "2025-11-03T15:15:01.000Z"}, {"author": "Hannes Tschofenig", "text": "STRINT workshop?
", "time": "2025-11-03T15:15:18.000Z"}, {"author": "Martin Thomson", "text": "CG-NAT is at least as sustainable as any other solution in this space.
", "time": "2025-11-03T15:15:20.000Z"}, {"author": "Deirdre Connolly", "text": "mixers are not great in their privacy properties vs the flavors of onion routing / private relay style
", "time": "2025-11-03T15:15:22.000Z"}, {"author": "Harald Alvestrand", "text": "one issue is what traffic the box would handle. The WebRTC APIs have been abused in order to get at the \"real home\" of the Web page user - because they send traffic that don't fit in HTTP proxies.
", "time": "2025-11-03T15:15:24.000Z"}, {"author": "Eliot Lear", "text": "as in, aggregate (mix) a bunch of traffic together to reduce traffic analysis. OHTTP can now serve that purpose- to a point.
", "time": "2025-11-03T15:15:40.000Z"}, {"author": "Lucas Pardue", "text": "reminds me a bit of SCONE too. So maybe enSCONEced
", "time": "2025-11-03T15:15:42.000Z"}, {"author": "Martin Thomson", "text": "The real challenge with CG-NAT is dealing with the subpoena thing, which means dropping things like deterministic mappings. Which means state management and then less sustainability.
", "time": "2025-11-03T15:16:18.000Z"}, {"author": "Martin Thomson", "text": "Eliot: OHTTP doesn't help here.
", "time": "2025-11-03T15:16:32.000Z"}, {"author": "Daniel Gillmor", "text": "martin: can you be more specific about \"the subpoena thing\" ?
", "time": "2025-11-03T15:16:39.000Z"}, {"author": "Ted Hardie", "text": "Side comment-using draft-companyname is a bit of a conflict with our usually statements about how we participate.
", "time": "2025-11-03T15:16:52.000Z"}, {"author": "Jim Fenton", "text": "Ted: +1
", "time": "2025-11-03T15:17:19.000Z"}, {"author": "Ted Hardie", "text": "(usual statement)
", "time": "2025-11-03T15:17:20.000Z"}, {"author": "Martin Thomson", "text": "dkg: sure. There's a lot of work done in CG-NAT to allow law enforcement to ask for mappings: \"at time T, I saw a flow from IP X and port P, which customer was that?\"
", "time": "2025-11-03T15:17:26.000Z"}, {"author": "Daniel Gillmor", "text": "the goal of nat-as-privacy seems to be \"do not be able to answer those requests\", no?
", "time": "2025-11-03T15:18:11.000Z"}, {"author": "Yoav Nir", "text": "Ted Hardie said:
\n\n\nSide comment-using draft-companyname is a bit of a conflict with our usually statements about how we participate.
\n
The one exception is when your draft documents an existing proprietary protocol. Like \"Cisco's extension for this and that\"
", "time": "2025-11-03T15:18:21.000Z"}, {"author": "Richard Barnes", "text": "Fancy, but not PQ, math
", "time": "2025-11-03T15:18:26.000Z"}, {"author": "Martin Thomson", "text": "dkg: exactly the problem, yes
", "time": "2025-11-03T15:18:28.000Z"}, {"author": "Pete Resnick", "text": "Getting a bit technical there with the \"fancy math\".
", "time": "2025-11-03T15:18:41.000Z"}, {"author": "Deirdre Connolly", "text": "ligero might be
", "time": "2025-11-03T15:18:44.000Z"}, {"author": "Murray Kucherawy", "text": "What was the middle thing?
", "time": "2025-11-03T15:18:51.000Z"}, {"author": "Andrew Campling", "text": "I understand that the Discord problem was caused by a third party storing PII insecurely. So poor implementation practices, not a requirement of any legislation etc.
", "time": "2025-11-03T15:19:08.000Z"}, {"author": "Bas Westerbaan", "text": "Richard Barnes said:
\n\n\nFancy, but not PQ, math
\n
Ligero is PQ AFAIK\u2014it's just that it proves something about P-256, which is not PQ.
", "time": "2025-11-03T15:19:17.000Z"}, {"author": "Ted Hardie", "text": "There's a bunch of work at the W3C that relates to this--why does this not belong there?
", "time": "2025-11-03T15:19:20.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Richard, there is no PQ math that saves the appearances for this
", "time": "2025-11-03T15:19:36.000Z"}, {"author": "Hannes Tschofenig", "text": "How does this relate to the work we are already doing on JSON Web Proofs?
", "time": "2025-11-03T15:19:37.000Z"}, {"author": "Deirdre Connolly", "text": "yep, the anonymity property is actually PQ looks like
", "time": "2025-11-03T15:19:45.000Z"}, {"author": "Murray Kucherawy", "text": "Possibly ignorant, but hey:
\nIsn't this IRTF material?
", "time": "2025-11-03T15:19:53.000Z"}, {"author": "Ben Laurie", "text": "hmmm, this sounds like a highly linkable credential (and hence, not private)
", "time": "2025-11-03T15:19:57.000Z"}, {"author": "Murray Kucherawy", "text": "Oh it's already on CFRG.
", "time": "2025-11-03T15:20:07.000Z"}, {"author": "Martin Thomson", "text": "Only anonymity holds if you have a CRQC. The CRQC can spoof credentials, not break anonymity for existing presentations.
", "time": "2025-11-03T15:20:32.000Z"}, {"author": "Hannes Tschofenig", "text": "From my experience in OAuth, there are unfortunately very few companies who really want to use these algorithms for wallets. This is why the SD-JWT solution is so popular (also in the EU).
", "time": "2025-11-03T15:20:39.000Z"}, {"author": "Richard Barnes", "text": "@Bas - Ah yes, privacy is PQ, but auth is not, right?
", "time": "2025-11-03T15:20:43.000Z"}, {"author": "Martin Thomson", "text": "Ben: This is not linkable.
", "time": "2025-11-03T15:20:44.000Z"}, {"author": "Daniel Gillmor", "text": "Martin: they just said this can be used to sign into Bumble. how is that unlinkable?
", "time": "2025-11-03T15:21:04.000Z"}, {"author": "Ben Laurie", "text": "ah, good - clearly I have to read it more carefully
", "time": "2025-11-03T15:21:12.000Z"}, {"author": "Bas Westerbaan", "text": "Richard Barnes said:
\n\n\n@Bas - Ah yes, privacy is PQ, but auth is not, right?
\n
The whole system they build isn't PQ, because P-256 isn't, but Ligero, the proof system, is PQ AFAIK.
", "time": "2025-11-03T15:21:18.000Z"}, {"author": "Daniel Gillmor", "text": "can i sign into any Bumble account with it?
", "time": "2025-11-03T15:21:26.000Z"}, {"author": "Richard Barnes", "text": "@dkg - I believe what they said is that it can prove your age to Bumble, which is different from sign in
", "time": "2025-11-03T15:21:35.000Z"}, {"author": "Daniel Gillmor", "text": "ah, ok. thx for clarifying.
", "time": "2025-11-03T15:21:47.000Z"}, {"author": "Martin Thomson", "text": "sign-ons are possible with these schemes because you can generate unlinkable context-specific pseudonyms
", "time": "2025-11-03T15:21:48.000Z"}, {"author": "Andrew Campling", "text": "How does this fit with SPICE?
", "time": "2025-11-03T15:22:09.000Z"}, {"author": "Phillip Hallam-Baker", "text": "What is Bumble?
", "time": "2025-11-03T15:22:09.000Z"}, {"author": "Richard Barnes", "text": "@PHB - Dating app
", "time": "2025-11-03T15:22:16.000Z"}, {"author": "Orie Steele", "text": "If this is crypto without container (ASN1, JSON, CBOR, XML, JSON-LD), CFRG seems good.
", "time": "2025-11-03T15:22:19.000Z"}, {"author": "Richard Barnes", "text": "@dkg - still a question of how the ZK proof gets tied to the logged-in identity, but there are non-cryptographic solutions to that
", "time": "2025-11-03T15:22:48.000Z"}, {"author": "Martin Thomson", "text": "This is a way of building proofs that involve SHA-256/P-256 ECDSA...
", "time": "2025-11-03T15:22:51.000Z"}, {"author": "Orie Steele", "text": "If it were JSON or CBOR, spice maybe.
", "time": "2025-11-03T15:22:54.000Z"}, {"author": "Martin Thomson", "text": "This could build on top of virtually anything. But there's a lot of questions.
", "time": "2025-11-03T15:23:27.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Richard, what a strange choice of name.
", "time": "2025-11-03T15:23:32.000Z"}, {"author": "Stephen Farrell", "text": "outcome: BoF, dunno if WG forming or not
", "time": "2025-11-03T15:23:54.000Z"}, {"author": "Richard Barnes", "text": "this seems fine for a focused WG. no BOF needed.
", "time": "2025-11-03T15:24:40.000Z"}, {"author": "Martin Thomson", "text": "I don't know what the scope should be.
", "time": "2025-11-03T15:24:57.000Z"}, {"author": "Dennis Jackson", "text": "What is the problem statement for the WG? 'Proving things in zkp'?
", "time": "2025-11-03T15:25:02.000Z"}, {"author": "Dennis Jackson", "text": "That seems too broad to be helpful. Better as a RG?
", "time": "2025-11-03T15:25:09.000Z"}, {"author": "David Waite", "text": "what they have demonstrated with Longfellow is a zero knowledge release of attributes from an existing credential in a pre-established format like a JWT or MDOC (which is CBOR based). The issued credentials today are not PQ safe (as we know), but the presentation meeting age requirements is PQ safe - in particular that PQ does not leak additional information.
", "time": "2025-11-03T15:25:30.000Z"}, {"author": "Richard Barnes", "text": "Slide says \"people want to interoperate\". I presume on a specific ZK scheme. So I would scope it to that.
", "time": "2025-11-03T15:25:35.000Z"}, {"author": "Murray Kucherawy", "text": "I've never heard \"IETF\" and \"fast\" in the same sentence.
", "time": "2025-11-03T15:25:39.000Z"}, {"author": "Alissa Cooper", "text": "These implementations are moving way, way faster than any of the interoperable messaging stuff.
", "time": "2025-11-03T15:25:46.000Z"}, {"author": "Hannes Tschofenig", "text": "Rifaat and I have regular conference calls with folks from the EC about the progress of the OAuth specifications related to wallets. They need this today. This will come out far too late. Additionally, there are so many cryptographic algorithms that could be used....
", "time": "2025-11-03T15:25:51.000Z"}, {"author": "Martin Thomson", "text": "Murray: I have, but I've never taken it as anything other than a joke.
", "time": "2025-11-03T15:25:59.000Z"}, {"author": "David Waite", "text": "There has been work on applying this to credentials protected with post quantum, I can't speak to the results
", "time": "2025-11-03T15:26:00.000Z"}, {"author": "Richard Barnes", "text": "The internet runs on drafts. We could get something started here that could produce interop checkpoints.
", "time": "2025-11-03T15:26:37.000Z"}, {"author": "Richard Barnes", "text": "RFCs are not the only interop tool we have
", "time": "2025-11-03T15:26:57.000Z"}, {"author": "David Waite", "text": "As an author of JWP, it is something I've been tracking and trying to learn more about. I've reached out to the authors of Longfellow to suggest they look at JWP as a \"what if the credential was designed to optimize the ZKP\"
", "time": "2025-11-03T15:26:57.000Z"}, {"author": "Daniel Gillmor", "text": "+1 for interop checkpoints. \"time to RFC\" is not the only measurement
", "time": "2025-11-03T15:27:00.000Z"}, {"author": "Richard Barnes", "text": "yeah, RFCs are not the only interop tool we have
", "time": "2025-11-03T15:27:31.000Z"}, {"author": "Hannes Tschofenig", "text": "There is also a solution being standardized BBS+
", "time": "2025-11-03T15:27:34.000Z"}, {"author": "Richard Barnes", "text": "@Hannes - BBS has the \"reissue passports\" problem, n'est-ce pas?
", "time": "2025-11-03T15:27:52.000Z"}, {"author": "Hannes Tschofenig", "text": "The crypto is in CFRG and then there is the JOSE draft
", "time": "2025-11-03T15:28:09.000Z"}, {"author": "David Waite", "text": "BBS+ is currently missing the predicates, such as \"the person is over 18 at the time you are asking\"
", "time": "2025-11-03T15:28:15.000Z"}, {"author": "Richard Barnes", "text": "but JWP / BBS+ is about new credential formats, not proving things about existing credentials, right?
", "time": "2025-11-03T15:28:36.000Z"}, {"author": "Christian Bormann", "text": "there are different solutions being worked on with different trade-offs and based on different assumptions. It would be extremely helpful to have a space to discuss these things someplace that is spanning the different WGs currently looking at similar solutions imho
", "time": "2025-11-03T15:28:49.000Z"}, {"author": "Hannes Tschofenig", "text": "Not that I am not a proponent of the BBS+ solution. It is what some people believed is a good solution for the requirements in this space.
", "time": "2025-11-03T15:29:03.000Z"}, {"author": "Richard Barnes", "text": "It seems very unlikely that there's one solution that will work for all the use cases here.
", "time": "2025-11-03T15:29:58.000Z"}, {"author": "Richard Barnes", "text": "We can start working on an existing point solution with known value, in parallel with developing the Grand Unified Theory
", "time": "2025-11-03T15:30:19.000Z"}, {"author": "Mark Nottingham", "text": "Richard: that was one of the things that became apparent and widely agreed to at the workshop
", "time": "2025-11-03T15:30:24.000Z"}, {"author": "Hannes Tschofenig", "text": "This is why we are also standardizing the SD-JWT solution as well
", "time": "2025-11-03T15:30:50.000Z"}, {"author": "Christian Bormann", "text": "I wouldn't fully agree with the \"done\" part - afaik it is somewhat unclear how we would verify correctness of the circuits?
", "time": "2025-11-03T15:30:55.000Z"}, {"author": "Martin Thomson", "text": "This is a direct repudiation of BBS.
", "time": "2025-11-03T15:31:05.000Z"}, {"author": "Ben Laurie", "text": "I like @Deidre's suggestion
", "time": "2025-11-03T15:31:05.000Z"}, {"author": "Hannes Tschofenig", "text": "And then there are 100 or so solutions in the W3C
", "time": "2025-11-03T15:31:07.000Z"}, {"author": "Harald Alvestrand", "text": "AD sponsored?
", "time": "2025-11-03T15:31:09.000Z"}, {"author": "Martin Thomson", "text": "Harald: nope, nope , nope
", "time": "2025-11-03T15:31:30.000Z"}, {"author": "Rich Salz", "text": "ISE RFC for nwo.
", "time": "2025-11-03T15:31:31.000Z"}, {"author": "Richard Barnes", "text": "@Martin - How is it a repudiation? seems like a parallel approach
", "time": "2025-11-03T15:31:38.000Z"}, {"author": "Christopher Inacio", "text": "@Deirdre Connolly would that crypto eng WG just standardize algorithms and not protocols?
", "time": "2025-11-03T15:31:43.000Z"}, {"author": "Deirdre Connolly", "text": "Depends where you draw the line at protocols
", "time": "2025-11-03T15:31:55.000Z"}, {"author": "Deirdre Connolly", "text": "HPKE is a protocol, I would include that, if not for the previous history
", "time": "2025-11-03T15:32:06.000Z"}, {"author": "Jim Fenton", "text": "captions are all caps, but people aren't yelling.
", "time": "2025-11-03T15:32:10.000Z"}, {"author": "Richard Barnes", "text": "We could start something on this specifically, and fold it into a bigger crypto eng wg
", "time": "2025-11-03T15:32:17.000Z"}, {"author": "Mark Nottingham", "text": "Defining the building block in a tight scope seems like a good way forward
", "time": "2025-11-03T15:32:26.000Z"}, {"author": "Deirdre Connolly", "text": "Things that come up that look 'bigger' can be spun out
", "time": "2025-11-03T15:32:41.000Z"}, {"author": "Richard Barnes", "text": "right, HPKEWG seems like a good analogue here
", "time": "2025-11-03T15:32:45.000Z"}, {"author": "Kathleen Moriarty", "text": "I like the crypto engineering WG suggestion - a reference group like CFRG makes a lot of sense.
", "time": "2025-11-03T15:32:47.000Z"}, {"author": "Hannes Tschofenig", "text": "You spoke too long
", "time": "2025-11-03T15:32:52.000Z"}, {"author": "Christopher Inacio", "text": "@Rich Salz but ISE is informational, is that appropriate?
", "time": "2025-11-03T15:32:55.000Z"}, {"author": "Mark Nottingham", "text": "\"beeping on site\"
", "time": "2025-11-03T15:32:57.000Z"}, {"author": "Deirdre Connolly", "text": "The ARC stuff in IETF is looking pretty big
", "time": "2025-11-03T15:32:58.000Z"}, {"author": "Yoav Nir", "text": "There's a bunch of things that are common to many sec area working groups, like IPsecME, TLS, SSH. Take some new algorithm, all protocols are going to carry it in the exact same format with just their own little wrapping.
\nMakes sense to have a group that standardizes the encoding and then have the others just assign the identifiers (or leave it to the IANA experts...)
", "time": "2025-11-03T15:32:58.000Z"}, {"author": "Murray Kucherawy", "text": "It's the \"Wrap It Up\" tone
", "time": "2025-11-03T15:33:00.000Z"}, {"author": "Rich Salz", "text": "Not sure the IETF is a good place for a crypto engineering group.
", "time": "2025-11-03T15:33:34.000Z"}, {"author": "Murray Kucherawy", "text": "Edwina... back in bowl.
", "time": "2025-11-03T15:33:54.000Z"}, {"author": "Bas Westerbaan", "text": "@Rich Salz Where else?
", "time": "2025-11-03T15:33:55.000Z"}, {"author": "Hannes Tschofenig", "text": "The security area is a big crypto engineering group
", "time": "2025-11-03T15:34:03.000Z"}, {"author": "Deirdre Connolly", "text": "@hannes exactly
", "time": "2025-11-03T15:34:09.000Z"}, {"author": "Yoav Nir", "text": "With ChaCha20-Poly1305 we did it in CFRG
", "time": "2025-11-03T15:34:10.000Z"}, {"author": "Nick Doty", "text": "is there a link / info on that side meeting?
", "time": "2025-11-03T15:34:19.000Z"}, {"author": "Deirdre Connolly", "text": "@yoav i'd argue that'd be a better fit for CEWG bc it was practically 'done' before it came to IETF
", "time": "2025-11-03T15:34:40.000Z"}, {"author": "Stephen Farrell", "text": "can meetecho fix the alarm? :-)
", "time": "2025-11-03T15:34:40.000Z"}, {"author": "Hannes Tschofenig", "text": "Time for a coffee break?
", "time": "2025-11-03T15:34:43.000Z"}, {"author": "Andrew Campling", "text": "The side meeting schedule is at https://trello.com/b/s1hNprRf/ietf-124-side-meeting-scheduling
", "time": "2025-11-03T15:34:47.000Z"}, {"author": "Lorenzo Miniero", "text": "@Stephen Farrell I wish :D
", "time": "2025-11-03T15:34:51.000Z"}, {"author": "Yoav Nir", "text": "I can't hear it very well. What is the PA saying?
", "time": "2025-11-03T15:34:54.000Z"}, {"author": "Deirdre Connolly", "text": "HPKE grew in CFRG, the PAKEs grew, etc
", "time": "2025-11-03T15:35:02.000Z"}, {"author": "layer8", "text": "it's saying attention, we've received a signal and we are investigating. In french and then english
", "time": "2025-11-03T15:35:10.000Z"}, {"author": "Rich Salz", "text": "\"something is going on, we are investigating\"
", "time": "2025-11-03T15:35:12.000Z"}, {"author": "Murray Kucherawy", "text": "An alert has been received. We're investigating the cause. Stand by for further.
", "time": "2025-11-03T15:35:15.000Z"}, {"author": "Lorenzo Miniero", "text": "@Yoav Nir it's a generic message that says an alert was received and they're investigating, and to stand by
", "time": "2025-11-03T15:35:16.000Z"}, {"author": "Brian Campbell", "text": "there is \"credential\" work happening in OAuth too. And longfellow can be similarly applicable there. Which isn't to say that it should be dispatched there. But rather that SPICE would not be appropriate.
", "time": "2025-11-03T15:35:24.000Z"}, {"author": "Deirdre Connolly", "text": "Good question Martin
", "time": "2025-11-03T15:35:33.000Z"}, {"author": "Daniel Gillmor", "text": "i'd be afraid of the size of a crypto engineering WG. nearly everything the IETF touches these days involves somee amount of cryptographic work.
", "time": "2025-11-03T15:35:40.000Z"}, {"author": "Yoav Nir", "text": "@Deirdre Connolly At the time, there was ChaCha20 and Poly1305, but no standardized way to connect them.
", "time": "2025-11-03T15:35:57.000Z"}, {"author": "John Klensin", "text": "Puts \"DpS attack\" into a whole new context...
", "time": "2025-11-03T15:36:15.000Z"}, {"author": "Deirdre Connolly", "text": "@daniel this is part of the problem CFRG faces, people want the 'Crypto' group to vet their stuff, but they are a RG and don't necessarily do that
", "time": "2025-11-03T15:36:22.000Z"}, {"author": "Bas Westerbaan", "text": "Daniel Gillmor said:
\n\n\ni'd be afraid of the size of a crypto engineering WG. nearly everything the IETF touches these days involves somee amount of cryptographic work.
\n
It's definitely more involved than \"just create an crypto engineering wg\", also looking at the number of submissions and dispatches to the CFRG every IETF.
", "time": "2025-11-03T15:36:25.000Z"}, {"author": "\u00c9ric Vyncke", "text": "From IETF secretariat: \"A hotel lady came by and told us they're investigating it, we shouldn't worry about it (yet?)
\n\"
", "time": "2025-11-03T15:36:32.000Z"}, {"author": "Deirdre Connolly", "text": "@yoav ok, fair enough
", "time": "2025-11-03T15:36:38.000Z"}, {"author": "Chris Lemmons", "text": "Somebody tell the Fire Alarm that the queue is locked.
", "time": "2025-11-03T15:36:45.000Z"}, {"author": "Murray Kucherawy", "text": "@meetecho: Can we nudge the room volume up a bit to overcome the announcement?
", "time": "2025-11-03T15:37:12.000Z"}, {"author": "Deirdre Connolly", "text": "++ @bas
", "time": "2025-11-03T15:37:18.000Z"}, {"author": "Lorenzo Miniero", "text": "@Murray Kucherawy do you mean the speakers in the room?
", "time": "2025-11-03T15:37:46.000Z"}, {"author": "Murray Kucherawy", "text": "Yes.
", "time": "2025-11-03T15:37:52.000Z"}, {"author": "Hannes Tschofenig", "text": "There are obviously a lot of other use cases beyond age verification for wallets
", "time": "2025-11-03T15:37:52.000Z"}, {"author": "Yoav Nir", "text": "@Deirdre Connolly Don't get me wrong. It a CEWG existed, that's where we would go.
", "time": "2025-11-03T15:37:58.000Z"}, {"author": "Deirdre Connolly", "text": "nod
", "time": "2025-11-03T15:38:06.000Z"}, {"author": "Christopher Wood", "text": "Why do this need an RFC at all?
", "time": "2025-11-03T15:38:36.000Z"}, {"author": "Christopher Wood", "text": "does*
", "time": "2025-11-03T15:38:40.000Z"}, {"author": "Deirdre Connolly", "text": "i'm assuming same reasons why anyone outside IETF wants an RFC
", "time": "2025-11-03T15:39:07.000Z"}, {"author": "Deirdre Connolly", "text": "it's quiet!
", "time": "2025-11-03T15:39:14.000Z"}, {"author": "Christopher Inacio", "text": "@Christopher Wood ++
", "time": "2025-11-03T15:39:14.000Z"}, {"author": "Murray Kucherawy", "text": "ahhhhhh.
", "time": "2025-11-03T15:39:19.000Z"}, {"author": "Murray Kucherawy", "text": "@Martin: OK start over.
", "time": "2025-11-03T15:39:28.000Z"}, {"author": "Martin Thomson", "text": "Finally.
", "time": "2025-11-03T15:39:37.000Z"}, {"author": "Jonathan Hammell", "text": "It is an Informational status alarm? Anyone else think it should be Experimental?
", "time": "2025-11-03T15:39:41.000Z"}, {"author": "Christopher Wood", "text": "It might be useful for the proponents to actually say why they need an RFC before we go down this road
", "time": "2025-11-03T15:39:44.000Z"}, {"author": "Mark Nottingham", "text": "+1 richard
", "time": "2025-11-03T15:39:54.000Z"}, {"author": "Eric Rescorla", "text": "I think this is good work, but I don't think it's appropriate to adopt in IETF. This is plainly quite complicated and needs to be adequately reviewed. Our mechanism for that is the CFRG
", "time": "2025-11-03T15:39:55.000Z"}, {"author": "Richard Barnes", "text": "ATTENTION ATTENTION
", "time": "2025-11-03T15:39:58.000Z"}, {"author": "David Waite", "text": "I'm unsure what they are trying to achieve with this alarm, and think they may be better off taking it someplace else
", "time": "2025-11-03T15:40:11.000Z"}, {"author": "Lorenzo Miniero", "text": "Apparently an emergency button had been pushed accidentally
", "time": "2025-11-03T15:40:14.000Z"}, {"author": "Mark Nottingham", "text": "has another signal been received?
", "time": "2025-11-03T15:40:17.000Z"}, {"author": "Richard Barnes", "text": "@EKR is it research though?
", "time": "2025-11-03T15:40:28.000Z"}, {"author": "Richard Barnes", "text": "it seems like we need a venue for \"engineering that needs sufficient review\" vs. \"research\"
", "time": "2025-11-03T15:40:47.000Z"}, {"author": "Christopher Inacio", "text": "it seems like something that could be included in OAUTH, JOSE, or COSE
", "time": "2025-11-03T15:40:52.000Z"}, {"author": "Eric Rescorla", "text": "No less than Sigma protocols
", "time": "2025-11-03T15:40:53.000Z"}, {"author": "Christopher Patton", "text": "To Martin's point: Longfellow seems like it's pretty tightly coupled to ECDSA: Perhaps the question is what use cases are tightly coupled to reusing credentials already deployed?
", "time": "2025-11-03T15:41:03.000Z"}, {"author": "Andrew Campling", "text": "Unusual to ask Ekr to be quick!
", "time": "2025-11-03T15:41:36.000Z"}, {"author": "Deirdre Connolly", "text": "Just because CFRG is the only mechanism doesn't mean that must remain so
", "time": "2025-11-03T15:41:48.000Z"}, {"author": "Nick Doty", "text": "let's tell the IAB/W3C to hurry up on their report on age-restriction methods, or a joint statement on the harms of age-verification requirements, so that everyone understands the implications of those policies and technologies to satisfy them
", "time": "2025-11-03T15:41:53.000Z"}, {"author": "Christian Bormann", "text": "it's not only about credentials already deployed - it's also about hardware support being available imho (HSMs, SEs, ...)
", "time": "2025-11-03T15:41:56.000Z"}, {"author": "Murray Kucherawy", "text": "+1 @Deirdre
", "time": "2025-11-03T15:41:56.000Z"}, {"author": "Deirdre Connolly", "text": "lololol
", "time": "2025-11-03T15:42:02.000Z"}, {"author": "Eric Rescorla", "text": "Deirdre, it would be fine for you to invent some new forum
", "time": "2025-11-03T15:42:03.000Z"}, {"author": "Christopher Inacio", "text": "it's a mechanism that could be included into one of protocols that are made to exchange identity or signed statements
", "time": "2025-11-03T15:42:04.000Z"}, {"author": "Deirdre Connolly", "text": "Yes a CEWG is one
", "time": "2025-11-03T15:42:12.000Z"}, {"author": "Eric Rescorla", "text": "But until that happens, that's what there is
", "time": "2025-11-03T15:42:13.000Z"}, {"author": "Martin Thomson", "text": "Yes, this ONLY makes sense if it is tied to \"legacy\" credentials.
", "time": "2025-11-03T15:42:19.000Z"}, {"author": "Eric Rescorla", "text": "Sure, and when CEWG exists I'll be in favor of that
", "time": "2025-11-03T15:42:24.000Z"}, {"author": "Deirdre Connolly", "text": "Luckily we are at DISPATCH
", "time": "2025-11-03T15:42:34.000Z"}, {"author": "Richard Barnes", "text": "@ekr - you could imagine a WG here as an experiment towards what Deirdre is proposing
", "time": "2025-11-03T15:42:36.000Z"}, {"author": "Eric Rescorla", "text": "@MT, yes, but I think it's likely there will be a lot of legacy credentials
", "time": "2025-11-03T15:42:38.000Z"}, {"author": "Deirdre Connolly", "text": "++ barnes
", "time": "2025-11-03T15:42:50.000Z"}, {"author": "Eric Rescorla", "text": "@rlb: no, that's something different. That's a specific WG
", "time": "2025-11-03T15:42:57.000Z"}, {"author": "Richard Barnes", "text": "\"Can we make a venue that does eng and gets the right review?\"
", "time": "2025-11-03T15:42:59.000Z"}, {"author": "Wendy Seltzer", "text": "Attention, attention. we must repeat the repeal until work is again impossible
", "time": "2025-11-03T15:43:04.000Z"}, {"author": "Chris Lemmons", "text": "Going to be very interesting to see how the AI note taker copes with our \"emergency\" alerts. :D
", "time": "2025-11-03T15:43:07.000Z"}, {"author": "Eliot Lear", "text": "@richard \"Engineering that needs sufficient review.\" If that's not the IETF, we are doing something quite wrongly.
", "time": "2025-11-03T15:43:07.000Z"}, {"author": "Murray Kucherawy", "text": "Little known fact: Everything in Canada must be repeated six times.
", "time": "2025-11-03T15:43:17.000Z"}, {"author": "Richard Barnes", "text": "ATTENTION ATTENTION
", "time": "2025-11-03T15:43:20.000Z"}, {"author": "Eric Rescorla", "text": "@Eliot: this is some very specialized stuff.
", "time": "2025-11-03T15:43:20.000Z"}, {"author": "Richard Barnes", "text": "THANK YOU FOR YOUR COOPERATION
", "time": "2025-11-03T15:43:26.000Z"}, {"author": "Eric Rescorla", "text": "And the IETF does not in fact have the ability to evaluate it
", "time": "2025-11-03T15:43:32.000Z"}, {"author": "Samuel Schlesinger", "text": "To my knowledge, the requirements for device binding and using existing, deployed cryptographic signatures push this towards a much more inefficient solution with some odd quirks, like using a post-quantum ZKP protocol with a pre-quantum signature.
\nA working group would be nice, because there are probably many deployments where a more efficient solution is possible.
", "time": "2025-11-03T15:43:34.000Z"}, {"author": "Martin Thomson", "text": "I worked out a Meetecho bug. The chat stops scrolling with new messages at the point that you type enough to have the message take a second line.
", "time": "2025-11-03T15:43:35.000Z"}, {"author": "Deirdre Connolly", "text": "If The Protocol looks big enough (this current I-D is quite narrow) that should/could become its own WG, like TLS SSH etc do
", "time": "2025-11-03T15:43:56.000Z"}, {"author": "Vanshaj Singhania", "text": "\n\nLittle known fact: Everything in Canada must be repeated six times.
\n
is that per language or total?
", "time": "2025-11-03T15:44:18.000Z"}, {"author": "Eric Rescorla", "text": "@Deirdre: yes, once the underlying crypto primitive is nailed down then a protocol that used it would make a lot of sense
", "time": "2025-11-03T15:44:20.000Z"}, {"author": "Murray Kucherawy", "text": "Attention Attention: The IETF has too much PQ stuff going on all over the place. Please stand by while we investigate.
", "time": "2025-11-03T15:44:34.000Z"}, {"author": "Bas Westerbaan", "text": "\"try not to and fail\" <- this
", "time": "2025-11-03T15:44:39.000Z"}, {"author": "Lorenzo Miniero", "text": "@Martin Thomson I'll pass the info along so that we can check if there's a bug there, thanks for the heads up!
", "time": "2025-11-03T15:44:55.000Z"}, {"author": "Deirdre Connolly", "text": "Ligero is pretty solid/concrete, the application in the I-D is basically just writing that down
", "time": "2025-11-03T15:45:00.000Z"}, {"author": "Paul Wouters", "text": "let me hand out all the stakes to the stakeholders to use on this vampire
", "time": "2025-11-03T15:45:02.000Z"}, {"author": "Rich Salz", "text": "We fail to prune the lists, so lets provide guidance on how to prune the list?
", "time": "2025-11-03T15:45:05.000Z"}, {"author": "Bas Westerbaan", "text": "I suggested a short-list for that LAMPS draft, and let me tell you: it's not short anymore.
", "time": "2025-11-03T15:45:23.000Z"}, {"author": "Deirdre Connolly", "text": ":vampire:
", "time": "2025-11-03T15:45:35.000Z"}, {"author": "Eric Rescorla", "text": "@Deirdre: well then the CFRG review shouldbe fast
", "time": "2025-11-03T15:45:38.000Z"}, {"author": "Martin Thomson", "text": "Lorenzo: thanks, you just need a busy group, like this. Busy enough that new messages arrive when you are typing.
", "time": "2025-11-03T15:45:53.000Z"}, {"author": "Deirdre Connolly", "text": "review sure, but where does the document get published? CFRG turned it down
", "time": "2025-11-03T15:46:06.000Z"}, {"author": "Ben Laurie", "text": "how about this skepticism: quantum will not get good enough to warrant any of this :-)
", "time": "2025-11-03T15:46:24.000Z"}, {"author": "Eric Rescorla", "text": "@Deirdre: then there is a problem with the interaction between CFRG and IETF
", "time": "2025-11-03T15:46:26.000Z"}, {"author": "Deirdre Connolly", "text": "There is, hence the suggestion for a WG that lets you publish crypto like this
", "time": "2025-11-03T15:46:44.000Z"}, {"author": "Deirdre Connolly", "text": "Because it's a RG, not a WG
", "time": "2025-11-03T15:46:54.000Z"}, {"author": "Ted Hardie", "text": "Hmm, the queue closing before the speaker finishes is a bit pessimal
", "time": "2025-11-03T15:46:55.000Z"}, {"author": "Eric Rescorla", "text": "I think we've got as far as we can get in this forum
", "time": "2025-11-03T15:46:56.000Z"}, {"author": "Richard Barnes", "text": "The level of this doc is just badly wrong
", "time": "2025-11-03T15:47:02.000Z"}, {"author": "Eric Rescorla", "text": "yes, chairs please unlock the queue
", "time": "2025-11-03T15:47:25.000Z"}, {"author": "Bas Westerbaan", "text": "I agree with the existence of Stephen's document (although I don't agree with everything he'd put in it.)
", "time": "2025-11-03T15:47:38.000Z"}, {"author": "Martin Thomson", "text": "This document seems like the wrong thing to be saying, as a whole, even if some of the points are OK.
", "time": "2025-11-03T15:47:39.000Z"}, {"author": "Murray Kucherawy", "text": "holy queue
", "time": "2025-11-03T15:48:08.000Z"}, {"author": "Yoav Nir", "text": "With NIST guidance to more to PQ by 2035 or 2030 or whatever, companies want PQ soon, which means vendors want to provide PQ real soon, which means they're asking the IETF for an RFC on how to put PQ in their IPsec/TLS/SSH/certificates
", "time": "2025-11-03T15:48:14.000Z"}, {"author": "Kathleen Moriarty", "text": "Bring back PKIX - just kidding on the long lived WG.
", "time": "2025-11-03T15:48:15.000Z"}, {"author": "Deirdre Connolly", "text": "++ ekr
", "time": "2025-11-03T15:48:44.000Z"}, {"author": "Richard Barnes", "text": "+1 ekr
", "time": "2025-11-03T15:48:50.000Z"}, {"author": "Britta Hale", "text": "++ekr
", "time": "2025-11-03T15:48:54.000Z"}, {"author": "Ted Hardie", "text": "Given the consensus issue, how about an IESG note? (/s for those unsure)
", "time": "2025-11-03T15:49:39.000Z"}, {"author": "Sophie Schmieg", "text": "How are people queuing with the queue locked?
", "time": "2025-11-03T15:49:40.000Z"}, {"author": "Murray Kucherawy", "text": "The people who were responsible for pruning the lists that have just been pruned, have been pruned.
", "time": "2025-11-03T15:49:43.000Z"}, {"author": "Jim Fenton", "text": "wik
", "time": "2025-11-03T15:49:58.000Z"}, {"author": "Rohan Mahy", "text": "PQUIP: be the sketicism you want to see in the world!
", "time": "2025-11-03T15:50:01.000Z"}, {"author": "Yoav Nir", "text": "General IETF consensus - no. But perhaps coordination between similar working groups? Like TLS/IPsec/SSH - if an algorithm is good for one, it's good for the others; COSE/JOSE; etc.
", "time": "2025-11-03T15:50:02.000Z"}, {"author": "Daniel Gillmor", "text": "@sophie -- you can always join, just after the lock
", "time": "2025-11-03T15:50:06.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@sophie, they are jailed
", "time": "2025-11-03T15:50:11.000Z"}, {"author": "Hannes Tschofenig", "text": "This is a popular discussion topic
", "time": "2025-11-03T15:50:11.000Z"}, {"author": "Martin Thomson", "text": "That was not called for Rich, but I do think that your point about not solving it for single protocols and expecting to solve it for every protocol at once was well made.
", "time": "2025-11-03T15:50:12.000Z"}, {"author": "Yaroslav Rosomakho", "text": "Different protocols have different deployment velocity. Some can upgrade rapidly and can adopt PQ signatures \"when the time is right\". Others take years to deploy and they need concrete recommendations sooner rather than later. I don't think it's right to treat everything equally
", "time": "2025-11-03T15:50:15.000Z"}, {"author": "Rich Salz", "text": "@Murray a deep cut on Python Grail.
", "time": "2025-11-03T15:50:21.000Z"}, {"author": "Deirdre Connolly", "text": "yep re: sean
", "time": "2025-11-03T15:50:27.000Z"}, {"author": "Sophie Schmieg", "text": "+1
", "time": "2025-11-03T15:50:35.000Z"}, {"author": "Murray Kucherawy", "text": "@Rich: I live to serve.
", "time": "2025-11-03T15:50:35.000Z"}, {"author": "Martin Thomson", "text": "Sean++
", "time": "2025-11-03T15:50:35.000Z"}, {"author": "Martin Thomson", "text": "Send it to PQUIP so they can reject it ?
", "time": "2025-11-03T15:50:56.000Z"}, {"author": "Deirdre Connolly", "text": "ha
", "time": "2025-11-03T15:51:00.000Z"}, {"author": "Sophie Schmieg", "text": "Honestly, I don't want to add oxygen to this topic, because I could rant about this for hours
", "time": "2025-11-03T15:51:07.000Z"}, {"author": "Deirdre Connolly", "text": "RoundUp\u2122
", "time": "2025-11-03T15:51:10.000Z"}, {"author": "Deirdre Connolly", "text": "++sophie
", "time": "2025-11-03T15:51:21.000Z"}, {"author": "Christopher Patton", "text": "+1 Thom
", "time": "2025-11-03T15:51:25.000Z"}, {"author": "Sophie Schmieg", "text": "+1
", "time": "2025-11-03T15:51:26.000Z"}, {"author": "Hannes Tschofenig", "text": "We have to also talk about signatures since this topic arises in the context of firmware updates and secure boot.
", "time": "2025-11-03T15:51:35.000Z"}, {"author": "Bas Westerbaan", "text": "If we can't solve the problem, then we kick the can downstream. They won't be any better at solving it.
", "time": "2025-11-03T15:51:43.000Z"}, {"author": "Sophie Schmieg", "text": "Yep, and things like web PKI will easily take a decade
", "time": "2025-11-03T15:52:01.000Z"}, {"author": "Deirdre Connolly", "text": "There may be domain-specific choices like PLANTS, but in plenty of other places just registering a signature codepoint is actually fine
", "time": "2025-11-03T15:52:04.000Z"}, {"author": "Christopher Patton", "text": "The fact that we have so many code points for so many things is just a side-effect of not knowing what to do (quite yet). But we're getting there.
", "time": "2025-11-03T15:52:06.000Z"}, {"author": "Deirdre Connolly", "text": "We do have dozens of elliptic curves registered and only a small subset are used
", "time": "2025-11-03T15:52:30.000Z"}, {"author": "Hannes Tschofenig", "text": "I disagree with Stephen on his signature approach
", "time": "2025-11-03T15:52:33.000Z"}, {"author": "Deirdre Connolly", "text": "I don't find that problematic
", "time": "2025-11-03T15:52:37.000Z"}, {"author": "Christopher Patton", "text": "So let's not panic, but let's keep on chugging
", "time": "2025-11-03T15:52:40.000Z"}, {"author": "Harald Alvestrand", "text": "People feel required to deploy PQC now(ish). Saying \"this is one that isn't known to be stupid - deploy that until we know a good reason to do something else\" would make sense.
", "time": "2025-11-03T15:53:47.000Z"}, {"author": "Martin Thomson", "text": "One of the biggest problems in this area is the amount of time we spend on discussing meta-level stuff like this. This is even more meta than the endless debate about combiners. Such a waste when the time could be spent fixing and deploying stuff. Thankfully, people are just deploying stuff.
", "time": "2025-11-03T15:54:11.000Z"}, {"author": "Dennis Jackson", "text": "If somewhere then pquip, but the probability of a useful output is near-zero
", "time": "2025-11-03T15:54:11.000Z"}, {"author": "Harald Alvestrand", "text": "PQC for Google Meet is probably on the order of 6 months away from full deployment to production.
", "time": "2025-11-03T15:54:24.000Z"}, {"author": "Andrew Campling", "text": "Is section 3 'we do not recommend use of
\n non-hybrid groups or \"pure\" PQ KEMs', missing the word yet at the end?
@Hannes, so do I. I think we need to have a PQC signature mechanism for doing code updates. I don't see a current need for PQC signatures to achieve non-repudiation and don't regard any mechanisms deployed today as being credible for non-repudiation
", "time": "2025-11-03T15:54:47.000Z"}, {"author": "Deirdre Connolly", "text": "@harald what choices did you make?
", "time": "2025-11-03T15:54:50.000Z"}, {"author": "Deb Cooley", "text": "It would be nice to have a place to discuss some of the more common signature use cases.
", "time": "2025-11-03T15:55:07.000Z"}, {"author": "Hannes Tschofenig", "text": "PHB - we are in agreement
", "time": "2025-11-03T15:55:14.000Z"}, {"author": "Harald Alvestrand", "text": "@Deirdre Connolly I will have to look it up. It's got 559 and KEM in it somewhere, but those are not the most memorable names.
", "time": "2025-11-03T15:55:49.000Z"}, {"author": "Deirdre Connolly", "text": "so mlkemx25519 for TLS , any signature?
", "time": "2025-11-03T15:56:10.000Z"}, {"author": "Sophie Schmieg", "text": "Nobody is currently deploying signatures outside of things like firmware signatures
", "time": "2025-11-03T15:56:29.000Z"}, {"author": "Sophie Schmieg", "text": "This is IMHO addressing a problem that doesn't exist
", "time": "2025-11-03T15:56:41.000Z"}, {"author": "Deirdre Connolly", "text": "ML-DSA is on the roadmaps for several clouds
", "time": "2025-11-03T15:56:46.000Z"}, {"author": "Rich Salz", "text": "@Sophie, you're skeptical of ANSI X.9 ML-DSA specs?
", "time": "2025-11-03T15:56:59.000Z"}, {"author": "Sophie Schmieg", "text": "Oh yeah, GCP KMS is offering ML-DSA
", "time": "2025-11-03T15:57:14.000Z"}, {"author": "Richard Barnes", "text": "wait, wtf was that? SAAG is not a disatch outcome
", "time": "2025-11-03T15:57:14.000Z"}, {"author": "Sophie Schmieg", "text": "And I think that is great
", "time": "2025-11-03T15:57:22.000Z"}, {"author": "John Bradley", "text": "For signature, I am seeing lots of vendors under pressure to show support for PQ signatures, not for wide deployment, but to start doing end to end testing so that something can be ion production in 5 yeares when needed
", "time": "2025-11-03T15:57:24.000Z"}, {"author": "Martin Thomson", "text": "Wow, there's a KMS that does something other than RSA?!?
", "time": "2025-11-03T15:57:45.000Z"}, {"author": "Sophie Schmieg", "text": "Because you need to experiment with these schemes and roll out could take years to decades in quite a few use cases
", "time": "2025-11-03T15:57:47.000Z"}, {"author": "Dennis Jackson", "text": "Richard: eh?
", "time": "2025-11-03T15:57:58.000Z"}, {"author": "Jim Fenton", "text": "@Richard SAAG is a dispatch outcome if the AD says it is.
", "time": "2025-11-03T15:58:13.000Z"}, {"author": "Daniel Gillmor", "text": "Richard: SAAG is \"discuss on a pre-existing mailing list\", which smells like a cross between \"send to a pre-existing WG\" and \"make a new mailing list\"
", "time": "2025-11-03T15:58:23.000Z"}, {"author": "Sophie Schmieg", "text": "But it's not widely deployed to production for actual end user facing systems
", "time": "2025-11-03T15:58:40.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Sophie, agreed, if applications care about non-repudiation, they should be enrolling documents into hash chains. That is more important than the algorithm choice as it binds the signature in time. It also finesses the PQC issue
", "time": "2025-11-03T15:58:43.000Z"}, {"author": "Rich Salz", "text": "@Sophie, tnx. Agree (for what thats worth)
", "time": "2025-11-03T15:58:43.000Z"}, {"author": "Sophie Schmieg", "text": "Which is exactly as it should be
", "time": "2025-11-03T15:58:48.000Z"}, {"author": "Richard Barnes", "text": "@Jim - disagree. it's not a WG / BoF / process next step. if the outcome is \"needs more discussion\" and SAAG is the venue, ok.
", "time": "2025-11-03T15:59:06.000Z"}, {"author": "Bas Westerbaan", "text": "Sophie Schmieg said:
\n\n\nBecause you need to experiment with these schemes and roll out could take years to decades in quite a few use cases
\n
Indeed.
", "time": "2025-11-03T15:59:17.000Z"}, {"author": "John Gray", "text": "We spend 6 years talking about composite signatures in LAMPS. We tried to reduce the number of combinations, every time we tried, we had people request more combinations! I don't think a guidance document will be able to properly scope the industry needs. So we ended up with 18 combinations. Saying hybrid signatures are NOT needed now is not helpful, we have customers that want them. That is how the industry works.
", "time": "2025-11-03T15:59:47.000Z"}, {"author": "Deirdre Connolly", "text": "Apple has rolled out ML-DSA in CryptoKit
\n\nI am not sure what the \"standard\" in the draft name has anything to do with the draft while it is marked as \"informational\".
", "time": "2025-11-03T15:59:53.000Z"}, {"author": "Martin Thomson", "text": "Isn't this trivially falsifiable? https://tee.fail/
", "time": "2025-11-03T15:59:53.000Z"}, {"author": "Deirdre Connolly", "text": "the only PQ sig scheme in there
", "time": "2025-11-03T16:00:00.000Z"}, {"author": "Hannes Tschofenig", "text": "I would dispatch this to RATS since there is work in this area happening already.
", "time": "2025-11-03T16:00:53.000Z"}, {"author": "Dennis Jackson", "text": "It's not clear to me if this draft is specific to remote attestation - or trying to cover a more general note of transparency.
", "time": "2025-11-03T16:00:59.000Z"}, {"author": "Muhammad Usama Sardar", "text": "Martin Thomson said:
\n\n\nIsn't this trivially falsifiable? https://tee.fail/
\n
Exactly.
", "time": "2025-11-03T16:01:01.000Z"}, {"author": "Daniel Gillmor", "text": "MT++
", "time": "2025-11-03T16:01:10.000Z"}, {"author": "Sophie Schmieg", "text": "https://docs.cloud.google.com/kms/docs/digital-signatures here is GCP KMS with ML-DSA
", "time": "2025-11-03T16:01:13.000Z"}, {"author": "John Bradley", "text": "Yubico has alpha test keys with MS-DSA and is testing with Apple and IBM
", "time": "2025-11-03T16:01:26.000Z"}, {"author": "Eric Rescorla", "text": "I'm offline now, but the point MT made I also made in email
", "time": "2025-11-03T16:01:40.000Z"}, {"author": "Martin Thomson", "text": "When we talk about TEEs, we are really just talking about systems that reduce to trust in the operator of the TEE.
", "time": "2025-11-03T16:01:56.000Z"}, {"author": "Eric Rescorla", "text": "We should not be encouraging the idea that you can run stuff in a TEE in the attacker's data center
", "time": "2025-11-03T16:01:59.000Z"}, {"author": "Daniel Gillmor", "text": "in particular, the \"falsifiable\" nature of these claims doesn't seem to distinguish between whether the statement about software is incorrect, or whether there is some sort of vulnerability in the TEE or signing chain.
", "time": "2025-11-03T16:02:09.000Z"}, {"author": "Richard Barnes", "text": "@ekr - unfortunately that is a delusion that underlies several WGs at this point
", "time": "2025-11-03T16:02:24.000Z"}, {"author": "Kathleen Moriarty", "text": "There is overlapping work with RATS.. I need to read the drafts.
", "time": "2025-11-03T16:02:30.000Z"}, {"author": "Eric Rescorla", "text": "@RLB: first rule of holes, etc.
", "time": "2025-11-03T16:02:50.000Z"}, {"author": "Eliot Lear", "text": "is this for software, embedded, both?
", "time": "2025-11-03T16:03:36.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@EKR, depends on the level of assurance you need. Many applications only require lightweight assurance. DRM for example, it is sufficient to ensure 95% of customers pay.
", "time": "2025-11-03T16:04:06.000Z"}, {"author": "Ben Laurie", "text": "Daniel Gillmor said:
\n\n\nin particular, the \"falsifiable\" nature of these claims doesn't seem to distinguish between whether the statement about software is incorrect, or whether there is some sort of vulnerability in the TEE or signing chain.
\n
they would both be in scope IMO
", "time": "2025-11-03T16:04:12.000Z"}, {"author": "Eric Rescorla", "text": "@PHB: sure, but this is explicitly scoped for \"you are running AI workloads in Gemini/OpenAI, etc.\" which is the oppposite of that
", "time": "2025-11-03T16:04:32.000Z"}, {"author": "Nick Doty", "text": "I'm struggling to understand the draft, but I think maybe it's about making claims/promises that don't have additional attestation / proof
", "time": "2025-11-03T16:05:01.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@EKR, I was ignoring the AI garnish.
", "time": "2025-11-03T16:05:34.000Z"}, {"author": "Eric Rescorla", "text": "Eric Rescorla said:
\n\n\n@PHB: sure, but this is explicitly scoped for \"you are running AI workloads in Gemini/OpenAI, etc.\" which is the oppposite of that
\n
If your threat model is \"OpenAI can't afford to pay Daniel Genkin to break the latest version of SGX\" something has gone severely wrong
", "time": "2025-11-03T16:05:36.000Z"}, {"author": "Daniel Gillmor", "text": "this is the problem with a lot of the RATS work, sadly.
", "time": "2025-11-03T16:06:16.000Z"}, {"author": "Martin Thomson", "text": "Should the IETF be in the business of telling people that these systems are trustworthy? Especially when the actual trustworthiness is not a whole lot different than an un-backed assertion that they don't do bad stuff.
", "time": "2025-11-03T16:06:35.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@EKR, I am pretty sure that OpenAI has no more than 12 months to run.
", "time": "2025-11-03T16:06:51.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@EKR, but yes, there are limits to this stuff
", "time": "2025-11-03T16:07:05.000Z"}, {"author": "Daniel Gillmor", "text": "but in those 12 months they surely can afford to crack this sort of enclave
", "time": "2025-11-03T16:07:21.000Z"}, {"author": "Nick Doty", "text": "@mt the operators of the systems are going to tell users that they are trustworthy, whether the IETF documents a format or not.
", "time": "2025-11-03T16:07:31.000Z"}, {"author": "Richard Barnes", "text": "@PHB - I have screenshotted that message to revisit in a year
", "time": "2025-11-03T16:07:32.000Z"}, {"author": "Daniel Gillmor", "text": "Nick: so what does this format add?
", "time": "2025-11-03T16:07:43.000Z"}, {"author": "Muhammad Usama Sardar", "text": "We are aiming for a RG in IRTF to deal with confidential computing research questions: https://datatracker.ietf.org/meeting/124/materials/slides-124-hotrfc-sessa-01-gaps-in-confidential-computing-00 interested folks are welcome
", "time": "2025-11-03T16:07:45.000Z"}, {"author": "Murray Kucherawy", "text": "@MT: IMHO, at best, we can present a reliable mechanism for people to make that determination on their own.
", "time": "2025-11-03T16:07:47.000Z"}, {"author": "Martin Thomson", "text": "Nick, the question is whether we intend to be complicit in that
", "time": "2025-11-03T16:07:51.000Z"}, {"author": "Murray Kucherawy", "text": "As in \"Here's what we're able to determine. You decide.\"
", "time": "2025-11-03T16:08:06.000Z"}, {"author": "Martin Thomson", "text": "Murray: we cannot. The threat model for TEEs does not permit that,.
", "time": "2025-11-03T16:08:10.000Z"}, {"author": "Britta Hale", "text": "TESLA makes GPS attacks easier, not harder....
", "time": "2025-11-03T16:08:20.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Richard, when I was at the MIT AI lab, several folk had their Symbolics stock certificates framed on their office wall...
", "time": "2025-11-03T16:08:34.000Z"}, {"author": "Eric Rescorla", "text": "TESLA is back!!!
", "time": "2025-11-03T16:08:38.000Z"}, {"author": "Nick Doty", "text": "@dkg I'm honestly not sure, but I think the question is whether user software could get some benefit from consuming the promises, maybe because the transparency increases the confidence that there's not a targeted attack happening or something
", "time": "2025-11-03T16:08:39.000Z"}, {"author": "Henk Birkholz", "text": "@MT there various types of conceptional messages provided by RATS that constitutes a _basis_ for trustworthiness decision
", "time": "2025-11-03T16:08:49.000Z"}, {"author": "Eric Rescorla", "text": "Nick Doty said:
\n\n\n@dkg I'm honestly not sure, but I think the question is whether user software could get some benefit from consuming the promises, maybe because the transparency increases the confidence that there's not a targeted attack happening or something
\n
But it doesn't if the TEE is broken
", "time": "2025-11-03T16:08:52.000Z"}, {"author": "Hannes Tschofenig", "text": "Ekr: This is a different Tesla
", "time": "2025-11-03T16:08:54.000Z"}, {"author": "Eric Rescorla", "text": "Hannes Tschofenig said:
\n\n\nEkr: This is a different Tesla
\n
@Hannes Tschofenig this is the old timed release of keys stuff, right?
", "time": "2025-11-03T16:09:12.000Z"}, {"author": "Wendy Seltzer", "text": "so many trademarks
", "time": "2025-11-03T16:09:18.000Z"}, {"author": "Henk Birkholz", "text": "You are never mandated to believe what an AR tells you
", "time": "2025-11-03T16:09:33.000Z"}, {"author": "Hannes Tschofenig", "text": "Right. Not the tesla.com
", "time": "2025-11-03T16:09:36.000Z"}, {"author": "Britta Hale", "text": "It does not fix the fundamental design that makes attacks easier
", "time": "2025-11-03T16:09:42.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Richard, I might be out with the 12 months prediction but probably not by much. There is only so long before companies that spend five times their revenues can continue...
", "time": "2025-11-03T16:10:07.000Z"}, {"author": "Martin Thomson", "text": "How do you prevent time-shifting of signals?
", "time": "2025-11-03T16:10:13.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Wendy, this spec predates the car company though
", "time": "2025-11-03T16:10:33.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Wendy, now if only they had trademarked it...
", "time": "2025-11-03T16:10:48.000Z"}, {"author": "Daniel Gillmor", "text": "but it does not predate Nikolai Tesla
", "time": "2025-11-03T16:10:55.000Z"}, {"author": "Wendy Seltzer", "text": "averting more than just name collisions
", "time": "2025-11-03T16:11:09.000Z"}, {"author": "Eric Rescorla", "text": "Daniel Gillmor said:
\n\n\nbut it does not predate Nikolai Tesla
\n
Well it might have predated him but they released it later :)
", "time": "2025-11-03T16:11:14.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Martin, well the time shifting can only happen in one direction...
", "time": "2025-11-03T16:11:28.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Martin, the devices have internal clocks, reject messages too far outside their reference time
", "time": "2025-11-03T16:12:00.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Martin, if they detect they are being spoofed, they stop using the GNS.
", "time": "2025-11-03T16:12:23.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Martin, I suspect you can only spoof effectively for specific locations.
", "time": "2025-11-03T16:13:21.000Z"}, {"author": "Britta Hale", "text": "@Philip: They do not stop using GNS, which then enables further attacks
", "time": "2025-11-03T16:13:30.000Z"}, {"author": "Martin Thomson", "text": "PHB yes, but the primary attack modality for something like GNSS is possible with even a very small delay.
", "time": "2025-11-03T16:13:35.000Z"}, {"author": "Martin Thomson", "text": "And GNSS is your source of time, so I don't see how you can avoid that.
", "time": "2025-11-03T16:13:55.000Z"}, {"author": "\u00c9ric Vyncke", "text": "airspaces around Ukraine/Russia is so GNSS spammed that planes have to reply on good old systems
", "time": "2025-11-03T16:14:20.000Z"}, {"author": "Martin Thomson", "text": "Particularly given that a single GPS message takes 30s.
", "time": "2025-11-03T16:14:38.000Z"}, {"author": "Muhammad Usama Sardar", "text": "Nick Doty said:
\n\n\n@dkg I'm honestly not sure, but I think the question is whether user software could get some benefit from consuming the promises, maybe because the transparency increases the confidence that there's not a targeted attack happening or something
\n
I will talk about my formal analysis tomorrow in UFMRG and lightweight results in SEAT: https://datatracker.ietf.org/meeting/124/materials/slides-124-seat-insights-from-formal-analysis-00 (an initial draft; will add some appendix in slides tonight).
", "time": "2025-11-03T16:14:41.000Z"}, {"author": "Daniel Gillmor", "text": "external standards bodies can't force any group in the IETF to do any specific action.
", "time": "2025-11-03T16:19:27.000Z"}, {"author": "Hannes Tschofenig", "text": "Correct
", "time": "2025-11-03T16:19:34.000Z"}, {"author": "\u00c9ric Vyncke", "text": "Indeed
", "time": "2025-11-03T16:19:35.000Z"}, {"author": "Stephen Farrell", "text": "AD sponsor might be fine here, a small WG would also if there's more than just Bob who'd be involved
", "time": "2025-11-03T16:19:40.000Z"}, {"author": "Hannes Tschofenig", "text": "I guess AD sponsorship is OK but doing more is not useful given that we do not seem to have any change control over it
", "time": "2025-11-03T16:20:08.000Z"}, {"author": "Yoav Nir", "text": "With no change control, I don't see why an AD would sponsor it. Just go to the ISE
", "time": "2025-11-03T16:20:56.000Z"}, {"author": "Daniel Gillmor", "text": "DRIP can also recharter to if they feel it's out of scope
", "time": "2025-11-03T16:20:59.000Z"}, {"author": "Pete Resnick", "text": "To be clear, nothing needs to be DISPATCHed; you could go straight to any of the paths that DISPATCH can recommend. But dkg is exactly right: No external body can force any work to occur in the IETF.
", "time": "2025-11-03T16:21:03.000Z"}, {"author": "Phillip Hallam-Baker", "text": "@Britta, agree importance, agree this is likely insufficient but we don't own enough of this problem to really address it properly
", "time": "2025-11-03T16:21:57.000Z"}, {"author": "Murray Kucherawy", "text": "+1 Pete
", "time": "2025-11-03T16:21:57.000Z"}, {"author": "\u00c9ric Vyncke", "text": "It is also far far away from DRIP
", "time": "2025-11-03T16:22:24.000Z"}, {"author": "Yoav Nir", "text": "AD sponsored is not a compromise between WG document and ISE. I don't see that it makes sense.
", "time": "2025-11-03T16:22:52.000Z"}, {"author": "Martin Thomson", "text": "I'd have to say that the draft is not ready either. It's very, very drafty.
", "time": "2025-11-03T16:22:57.000Z"}, {"author": "\u00c9ric Vyncke", "text": "DING DING Attention Attention Meeting is Closing DING DING
", "time": "2025-11-03T16:23:07.000Z"}, {"author": "Murray Kucherawy", "text": "hah
", "time": "2025-11-03T16:23:13.000Z"}, {"author": "Murray Kucherawy", "text": "Bummer, I was here for the 4648 one
", "time": "2025-11-03T16:23:24.000Z"}, {"author": "Harald Alvestrand", "text": "@Deirdre Connolly the answer was X2559MLKEM768.
", "time": "2025-11-03T16:23:46.000Z"}, {"author": "Daniel Gillmor", "text": "if there's no change control permitted, it does seem like ISE is preferable, if anything. (i'm basically agreeing with Ted, i think)
", "time": "2025-11-03T16:23:46.000Z"}]