[{"author": "Christopher Patton", "text": "<p>hi nerds</p>", "time": "2025-07-24T15:00:41Z"}, {"author": "Thom Wiggers", "text": "<p>*sup nerds</p>", "time": "2025-07-24T15:02:06Z"}, {"author": "Daniel Gillmor", "text": "<p><span aria-label=\"nerd\" class=\"emoji emoji-1f913\" role=\"img\" title=\"nerd\">:nerd:</span><span aria-label=\"nerd\" class=\"emoji emoji-1f913\" role=\"img\" title=\"nerd\">:nerd:</span><span aria-label=\"nerd\" class=\"emoji emoji-1f913\" role=\"img\" title=\"nerd\">:nerd:</span></p>", "time": "2025-07-24T15:02:18Z"}, {"author": "Christopher Patton", "text": "<p>sorry thom, you're so right</p>", "time": "2025-07-24T15:02:23Z"}, {"author": "Mike Ounsworth", "text": "<p>What's Jabber? :P</p>", "time": "2025-07-24T15:02:34Z"}, {"author": "Daniel Gillmor", "text": "<p>it's pronounced Zoo-Lip</p>", "time": "2025-07-24T15:02:50Z"}, {"author": "Uri Blumenthal", "text": "<p>@Daniel, is it a hint on what\u2019s coming off of the Lip(s)? <span aria-label=\"grinning face with smiling eyes\" class=\"emoji emoji-1f601\" role=\"img\" title=\"grinning face with smiling eyes\">:grinning_face_with_smiling_eyes:</span></p>", "time": "2025-07-24T15:04:17Z"}, {"author": "Christopher Patton", "text": "<p>clap clap clap!</p>", "time": "2025-07-24T15:05:51Z"}, {"author": "Dan Harkins", "text": "<p>huzzah!</p>", "time": "2025-07-24T15:06:03Z"}, {"author": "Robert Moskowitz", "text": "<p>@Uri, Just your usual need zoo.</p>", "time": "2025-07-24T15:06:12Z"}, {"author": "Robert Moskowitz", "text": "<p>nerd zoo</p>", "time": "2025-07-24T15:06:20Z"}, {"author": "Christopher Patton", "text": "<p>AEGIS is pronounced \"ee jis\" correct?</p>", "time": "2025-07-24T15:06:34Z"}, {"author": "Thom Wiggers", "text": "<p>@Chris don't say that you're one of those JIF people</p>", "time": "2025-07-24T15:07:06Z"}, {"author": "Dan Harkins", "text": "<p>\"age ist\"</p>", "time": "2025-07-24T15:07:16Z"}, {"author": "Deirdre Connolly", "text": "<p>++ two sessions</p>", "time": "2025-07-24T15:07:57Z"}, {"author": "Frank Denis", "text": "<p>I pronounce it \"age is\".</p>", "time": "2025-07-24T15:08:06Z"}, {"author": "Christopher Wood", "text": "<p>+1 to two sessions</p>", "time": "2025-07-24T15:08:08Z"}, {"author": "Sofia Celi", "text": "<p>lol @chris ahaha</p>", "time": "2025-07-24T15:08:09Z"}, {"author": "Aritra Banerjee", "text": "<p>two sessions all the way</p>", "time": "2025-07-24T15:08:17Z"}, {"author": "Christopher Patton", "text": "<p>+1 two sessions</p>", "time": "2025-07-24T15:08:27Z"}, {"author": "Dennis Jackson", "text": "<p>Two sessions and more selectivity sounds great.</p>", "time": "2025-07-24T15:08:29Z"}, {"author": "Daniel Gillmor", "text": "<p>+1 for two sessions</p>", "time": "2025-07-24T15:08:50Z"}, {"author": "Watson Ladd", "text": "<p>+1 for the split and reserving time for advancing older drafts/status check. we might even want interims if things are kicking around for a while: it's worked for NTP</p>", "time": "2025-07-24T15:09:24Z"}, {"author": "Sofia Celi", "text": "<p>+1 on iterims, work very well on other RG/WG</p>", "time": "2025-07-24T15:09:43Z"}, {"author": "Dennis Jackson", "text": "<p>Thematic interims could work well</p>", "time": "2025-07-24T15:09:57Z"}, {"author": "Christopher Wood", "text": "<p>+1 to Dennis -- focused interims would probably be most effective</p>", "time": "2025-07-24T15:10:20Z"}, {"author": "Daniel Gillmor", "text": "<p>less drive-by tourism at interims</p>", "time": "2025-07-24T15:10:32Z"}, {"author": "Uri Blumenthal", "text": "<p><span aria-label=\"smiley\" class=\"emoji emoji-1f603\" role=\"img\" title=\"smiley\">:smiley:</span></p>", "time": "2025-07-24T15:10:46Z"}, {"author": "Mike Ounsworth", "text": "<p>CFRG sub-groups: privacy preserving, PQC, etc. Makes sense to me.</p>", "time": "2025-07-24T15:11:06Z"}, {"author": "Deirdre Connolly", "text": "<p>oops typo on the slide</p>", "time": "2025-07-24T15:11:16Z"}, {"author": "Deirdre Connolly", "text": "<p><a href=\"https://datatracker.ietf.org/doc/draft-irtf-cfrg-concrete-hybrid-kems/\">https://datatracker.ietf.org/doc/draft-irtf-cfrg-concrete-hybrid-kems/</a> is concrete</p>", "time": "2025-07-24T15:11:40Z"}, {"author": "Deirdre Connolly", "text": "<p><a href=\"https://datatracker.ietf.org/doc/draft-irtf-cfrg-hybrid-kems/\">https://datatracker.ietf.org/doc/draft-irtf-cfrg-hybrid-kems/</a></p>\n<p>is generic</p>", "time": "2025-07-24T15:11:47Z"}, {"author": "Christopher Patton", "text": "<p><span class=\"user-mention\" data-user-id=\"603\">@Deirdre Connolly</span> what's the difference between PRG and PRNG here?</p>", "time": "2025-07-24T15:12:57Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"3787\">Christopher Patton</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176492\">said</a>:</p>\n<blockquote>\n<p><span class=\"user-mention silent\" data-user-id=\"603\">Deirdre Connolly</span> what's the difference between PRG and PRNG here?</p>\n</blockquote>\n<p>generator</p>", "time": "2025-07-24T15:13:22Z"}, {"author": "Deirdre Connolly", "text": "<p>not an RNG</p>", "time": "2025-07-24T15:13:34Z"}, {"author": "Deirdre Connolly", "text": "<p>one sec</p>", "time": "2025-07-24T15:13:57Z"}, {"author": "Aritra Banerjee", "text": "<p>confused</p>", "time": "2025-07-24T15:14:19Z"}, {"author": "Deirdre Connolly", "text": "<p><a href=\"/user_uploads/2/24/rs8QVj0KkBCfe9iIy5kuuZ-V/image.png\">image.png</a></p>\n<div class=\"message_inline_image\"><a href=\"/user_uploads/2/24/rs8QVj0KkBCfe9iIy5kuuZ-V/image.png\" title=\"image.png\"><img data-original-content-type=\"image/png\" data-original-dimensions=\"1646x618\" src=\"/user_uploads/thumbnail/2/24/rs8QVj0KkBCfe9iIy5kuuZ-V/image.png/840x560.webp\"></a></div>", "time": "2025-07-24T15:14:30Z"}, {"author": "Deirdre Connolly", "text": "<p>This is trying to nail down the requirements of the seed expander function</p>", "time": "2025-07-24T15:14:48Z"}, {"author": "Deirdre Connolly", "text": "<p>This is also used to model ML-KEM expansion of its PKE material from its seed, securely</p>", "time": "2025-07-24T15:15:14Z"}, {"author": "Sofia Celi", "text": "<p>PRG is usually the \"idealised\" abstract definition, while PRNG is the concrete algoithm</p>", "time": "2025-07-24T15:15:18Z"}, {"author": "Watson Ladd", "text": "<p>where do labels come from?</p>", "time": "2025-07-24T15:15:18Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"205\">Sofia Celi</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176508\">said</a>:</p>\n<blockquote>\n<p>PRG is usually the \"idealised\" abstract definition, while PRNG is the concrete algoithm</p>\n</blockquote>\n<p>exactly</p>", "time": "2025-07-24T15:15:30Z"}, {"author": "Aritra Banerjee", "text": "<p><span class=\"user-mention silent\" data-user-id=\"205\">Sofia Celi</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176508\">said</a>:</p>\n<blockquote>\n<p>PRG is usually the \"idealised\" abstract definition, while PRNG is the concrete algoithm</p>\n</blockquote>\n<p>okay</p>", "time": "2025-07-24T15:15:52Z"}, {"author": "Mike Ounsworth", "text": "<p>So a PRNG is an instance of a PRG, where the thing you're generating happens to be a number? PRG covers the space of more abstract group elements that you might want to generate uniformly randomly?</p>", "time": "2025-07-24T15:16:33Z"}, {"author": "Deirdre Connolly", "text": "<p>In these constructions, you source a seed from a real-world RNG/PRNG, pass it though a PRG (SHAKE or something like it) and use the image material to hand to your component keygen schemes</p>", "time": "2025-07-24T15:16:50Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"5873\">Mike Ounsworth</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176517\">said</a>:</p>\n<blockquote>\n<p>So a PRNG is an instance of a PRG, where the thing you're generating happens to be a number? PRG covers the space of more abstract group elements that you might want to generate uniformly randomly?</p>\n</blockquote>\n<p>not exactly</p>", "time": "2025-07-24T15:16:58Z"}, {"author": "Deirdre Connolly", "text": "<p>a secure PRG is the security model/property of what SHAKE is doing in X-Wing keygen, for example</p>", "time": "2025-07-24T15:17:27Z"}, {"author": "Deirdre Connolly", "text": "<p>We needed to model this for picking concrete instances of the generic schemes</p>", "time": "2025-07-24T15:17:50Z"}, {"author": "Christopher Wood", "text": "<p>I feel like the distinction you're trying to make is between RNG and _P_R(N)G, where the latter accepts a seed and the former does not. Is that what you're trying to say?</p>", "time": "2025-07-24T15:18:31Z"}, {"author": "Watson Ladd", "text": "<p>my question is clarifying</p>", "time": "2025-07-24T15:18:40Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"128\">Christopher Wood</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176526\">said</a>:</p>\n<blockquote>\n<p>I feel like the distinction you're trying to make is between RNG and _P_R(N)G, where the latter accepts a seed and the former does not. Is that what you're trying to say?</p>\n</blockquote>\n<p>kinda, the PRNG is not abstracted here, you just give the PRG function good randomness from somewhere</p>", "time": "2025-07-24T15:21:13Z"}, {"author": "Deirdre Connolly", "text": "<p>we really just needed to model what we need out of our seed expander function (shake) and this seemed to fit and has been used elsewhere</p>", "time": "2025-07-24T15:21:53Z"}, {"author": "Christopher Wood", "text": "<p>But don't both PRNG and PRGs accept seeds? (I'm confused)</p>", "time": "2025-07-24T15:22:05Z"}, {"author": "Christopher Wood", "text": "<p>(Not important compared to the questions at hand)</p>", "time": "2025-07-24T15:22:14Z"}, {"author": "Sofia Celi", "text": "<p>my guess will be that in the security analysis of these algorithms, the PRGs show up (I haven't checked the proofs): It is not known if cryptographically secure pseudorandom generators exist, as proving that they exist is difficult since their existence implies P \u2260 NP. PRNG is an approximation to a true one, but they are not truly random</p>", "time": "2025-07-24T15:22:18Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"205\">Sofia Celi</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176536\">said</a>:</p>\n<blockquote>\n<p>my guess will be that in the security analysis of these algorithms, the PRGs show up (I haven't checked the proofs): It is not known if cryptographically secure pseudorandom generators exist, as proving that they exist is difficult since their existence implies P \u2260 NP. PRNG is an approximation to a true one, but they are not truly random</p>\n</blockquote>\n<p>exactly</p>", "time": "2025-07-24T15:23:11Z"}, {"author": "Christopher Wood", "text": "<p>OK, so I think we're saying the same thing</p>", "time": "2025-07-24T15:23:31Z"}, {"author": "Alicja Kario", "text": "<p>no way to ensure SHA-3 exists outside the module still is a thing; TLS for example still is SHA-2 for handshake integrity</p>", "time": "2025-07-24T15:24:19Z"}, {"author": "Alicja Kario", "text": "<p>only</p>", "time": "2025-07-24T15:24:27Z"}, {"author": "Sofia Celi", "text": "<p>yes ahaha it matters formally and for security analysis and theoretical complexity. For the engineering perspective we have here, it shouldn't matter</p>", "time": "2025-07-24T15:24:31Z"}, {"author": "Dan Harkins", "text": "<p>if you care what the NSA says, they say that SHA2 must be used for hashing and KDFs</p>", "time": "2025-07-24T15:24:48Z"}, {"author": "Thom Wiggers", "text": "<p>NSA don't allow hybrids anyway though (as was mentioned by Richard)</p>", "time": "2025-07-24T15:24:57Z"}, {"author": "Dan Harkins", "text": "<p>some people need to sell product to people who care what the NSA says</p>", "time": "2025-07-24T15:25:05Z"}, {"author": "Dan Harkins", "text": "<p>true</p>", "time": "2025-07-24T15:25:14Z"}, {"author": "Uri Blumenthal", "text": "<p>\u201cDon\u2019t assume SHA3 exists outside of the tool\u201d is a requirement <em>at this time</em>, if one plans to build a product that US Gov\u2019t (or whoever else that follows CNSA) may buy.</p>", "time": "2025-07-24T15:25:51Z"}, {"author": "Daniel Gillmor", "text": "<p>Chris said YAGNI.  it means \"You Ain't Gonna Need It\"</p>", "time": "2025-07-24T15:26:11Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"1661\">Dan Harkins</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176550\">said</a>:</p>\n<blockquote>\n<p>if you care what the NSA says, they say that SHA2 must be used for hashing and KDFs</p>\n</blockquote>\n<p>So, there's a separate thread being brokered by NIST, and that thread has the flavour of \"Oh Common NSA, a policy of \"submit your commercial hybrid product and we'll decide if we accept it\" is not helpful. Be clearer please\".</p>", "time": "2025-07-24T15:26:56Z"}, {"author": "Martin Thomson", "text": "<p>I'm motivated by ekr's argument to suggest dropping PRE</p>", "time": "2025-07-24T15:27:30Z"}, {"author": "Stephen Farrell", "text": "<p>+1 to ekr, safest + what we're stuck with seems good</p>", "time": "2025-07-24T15:27:30Z"}, {"author": "Watson Ladd", "text": "<p>but you don't get post retirement consulting gigs if you were clear in the requirements in the getgo</p>", "time": "2025-07-24T15:27:38Z"}, {"author": "Martin Thomson", "text": "<p>So GHP + QSF</p>", "time": "2025-07-24T15:27:39Z"}, {"author": "Christopher Patton", "text": "<p>I morally agree with EKR, but it seems like the three options we landed are pretty good. We'll just have to help people choose which option</p>", "time": "2025-07-24T15:27:43Z"}, {"author": "Daniel Gillmor", "text": "<p>Watson <span aria-label=\"rolling on the floor laughing\" class=\"emoji emoji-1f923\" role=\"img\" title=\"rolling on the floor laughing\">:rolling_on_the_floor_laughing:</span></p>", "time": "2025-07-24T15:27:53Z"}, {"author": "Xianhui Lu", "text": "<p>pre-hashing pubic keys need to revise the security proof in quantum random oracle model</p>", "time": "2025-07-24T15:27:54Z"}, {"author": "Alexey Melnikov", "text": "<p>Chairs are giving this discussion 5 more minutes. We will be a bit overtime.</p>", "time": "2025-07-24T15:27:58Z"}, {"author": "Uri Blumenthal", "text": "<p>Since SHA3 <em>does</em> have better properties, better (explicit) design retirements, and better performance profile in hardware (for those who has it <span aria-label=\"wink\" class=\"emoji emoji-1f609\" role=\"img\" title=\"wink\">:wink:</span>), I wouldn\u2019t be surprised if an eventual CNSA-nn includes SHA3. But we aren\u2019t there yet.</p>", "time": "2025-07-24T15:28:48Z"}, {"author": "Simon Josefsson", "text": "<p>What significant non-XWing uses are people planning to use QSF for?</p>", "time": "2025-07-24T15:28:48Z"}, {"author": "Xianhui Lu", "text": "<p>security proof in random oracle is ok for pre-hashing, but quantum random oracle is different</p>", "time": "2025-07-24T15:28:49Z"}, {"author": "Watson Ladd", "text": "<p>right, I need to learn how to use the QROM. which probably means learning more about quantum computers</p>", "time": "2025-07-24T15:28:54Z"}, {"author": "Eric Rescorla", "text": "<p>If we don't <em>drop</em> PRE, we should at minimum strongly recommend either PRE or GHP</p>", "time": "2025-07-24T15:28:57Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"5063\">@Simon Josefsson</span> I think X-Wing-but-P256 has a lot of fans</p>", "time": "2025-07-24T15:29:25Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention\" data-user-id=\"2983\">@Watson Ladd</span> mostly a advantage multiplier</p>", "time": "2025-07-24T15:29:28Z"}, {"author": "Thom Wiggers", "text": "<p>you also can't program the QROM right</p>", "time": "2025-07-24T15:29:55Z"}, {"author": "Watson Ladd", "text": "<p>yeah, I think the logic of QSF is really broken: we want hybrids because we don't trust the the PQ kem, but the PQ Kem might fail to have a property that impacts QSF's security</p>", "time": "2025-07-24T15:30:05Z"}, {"author": "Simon Josefsson", "text": "<p>So revising the question: non-MLKEM uses of QSF?</p>", "time": "2025-07-24T15:30:21Z"}, {"author": "Watson Ladd", "text": "<p>or am I missing something (entirely possible)</p>", "time": "2025-07-24T15:30:22Z"}, {"author": "Mike Ounsworth", "text": "<p>IMO the path-of-least-resistance to aligning the LAMPS and CFRG hybrid KEMs is to:</p>\n<p>1) Have the CFRG thing RECOMMEND a joint-seed-based keygen() but allow separable keygen()s.<br>\n2) Have the CFRG thing add back the HMAC-SHA2 combiner for QSF as an alternative to SHA3.</p>", "time": "2025-07-24T15:30:44Z"}, {"author": "Thom Wiggers", "text": "<p>watson: essentially, ML-KEM's shared secret is already <code>ss = H(ss' || pk || ct)</code> so you end up hashing twice</p>", "time": "2025-07-24T15:31:11Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"2983\">Watson Ladd</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176591\">said</a>:</p>\n<blockquote>\n<p>yeah, I think the logic of QSF is really broken: we want hybrids because we don't trust the the PQ kem, but the PQ Kem might fail to have a property that impacts QSF's security</p>\n</blockquote>\n<p>I'd have to check by details, but I think the X-Wing proof shows that ML-KEM will remain C2PRI, even if broken.</p>", "time": "2025-07-24T15:31:25Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"5063\">Simon Josefsson</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176592\">said</a>:</p>\n<blockquote>\n<p>So revising the question: non-MLKEM uses of QSF?</p>\n</blockquote>\n<p>\"under review\"</p>", "time": "2025-07-24T15:31:26Z"}, {"author": "Richard Barnes", "text": "<p>thanks for the comments at the mic, i'll review the chat after the meeting.</p>", "time": "2025-07-24T15:31:27Z"}, {"author": "Bas Westerbaan", "text": "<p><span class=\"user-mention silent\" data-user-id=\"5873\">Mike Ounsworth</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176595\">said</a>:</p>\n<blockquote>\n<p>IMO the path-of-least-resistance to aligning the LAMPS and CFRG hybrid KEMs is to:</p>\n<p>1) Have the CFRG thing RECOMMEND a joint-seed-based keygen() but allow separable keygen()s.<br>\n2) Have the CFRG thing add back the HMAC-SHA2 combiner for QSF as an alternative to SHA3.</p>\n</blockquote>\n<p>I prefer 1. Phrasing is important. I'd make it SHOULD.</p>", "time": "2025-07-24T15:31:31Z"}, {"author": "Watson Ladd", "text": "<p>right, but the combiner then wouldn't be generic to other schemes</p>", "time": "2025-07-24T15:32:04Z"}, {"author": "Richard Barnes", "text": "<p>\"Mature specification\" is a reason for this group not to adopt a document -- there's nothing for us to do with it!</p>", "time": "2025-07-24T15:32:06Z"}, {"author": "Eric Rescorla", "text": "<p>Parameter flexibility does not seem good to me</p>", "time": "2025-07-24T15:32:18Z"}, {"author": "Richard Barnes", "text": "<p>@Watson i think the idea would be to do SHA-2 in the <em>concrete</em> draft</p>", "time": "2025-07-24T15:32:31Z"}, {"author": "Daniel Gillmor", "text": "<p>sigh, flexibility</p>", "time": "2025-07-24T15:32:32Z"}, {"author": "Alicja Kario", "text": "<p>Parameter flexibility is a disadvantage...</p>", "time": "2025-07-24T15:32:34Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"549\">@Bas Westerbaan</span> RECOMMEND == SHOULD</p>", "time": "2025-07-24T15:32:35Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"549\">Bas Westerbaan</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176602\">said</a>:</p>\n<blockquote>\n<p><span class=\"user-mention silent\" data-user-id=\"5873\">Mike Ounsworth</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176595\">said</a>:</p>\n<blockquote>\n<p>IMO the path-of-least-resistance to aligning the LAMPS and CFRG hybrid KEMs is to:</p>\n<p>1) Have the CFRG thing RECOMMEND a joint-seed-based keygen() but allow separable keygen()s.<br>\n2) Have the CFRG thing add back the HMAC-SHA2 combiner for QSF as an alternative to SHA3.</p>\n</blockquote>\n<p>I prefer 1. Phrasing is important. I'd make it SHOULD.</p>\n</blockquote>\n<p>To be clear, the LAMPS Composite needs both of those changes to the CFRG doc.</p>", "time": "2025-07-24T15:32:59Z"}, {"author": "Eric Rescorla", "text": "<p>I mean, it's fine that NTRU is flexible, but if CFRG pubishes it, it should publish very few parameters</p>", "time": "2025-07-24T15:33:02Z"}, {"author": "Thom Wiggers", "text": "<p>I think that we should rate the flexibility as such</p>", "time": "2025-07-24T15:33:20Z"}, {"author": "Dmitry Belyavskiy", "text": "<p>@Ekr ideally exactly one set of...</p>", "time": "2025-07-24T15:33:43Z"}, {"author": "Christopher Patton", "text": "<p>what's the difference between NTRU and NTRU+?</p>", "time": "2025-07-24T15:33:46Z"}, {"author": "Thom Wiggers", "text": "<p>we could e.g. do a 112-bit secure IoT profile if we really needed that for something. We couldn't do that with ML-KEM because of the IPR (it'd be outside of the standard)</p>", "time": "2025-07-24T15:33:51Z"}, {"author": "Simon Josefsson", "text": "<p>Chempat hash all components and offers no paramrization.  This seems to be what ekr and other engineers wants from us.  QSF seems to me MLKEM-specific and could live on its own</p>", "time": "2025-07-24T15:33:56Z"}, {"author": "Eric Rescorla", "text": "<p>@Dmitry: I could see multiple parameters for different security levels, like ML-KEM does</p>", "time": "2025-07-24T15:34:03Z"}, {"author": "Dmitry Belyavskiy", "text": "<p>Ah. Yes, also an option</p>", "time": "2025-07-24T15:34:22Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"5063\">@Simon Josefsson</span> there's a very good chance that HQC will be patched to also do C2PRI</p>", "time": "2025-07-24T15:34:24Z"}, {"author": "Alexey Melnikov", "text": "<p>No questions time after the NTRU presentation. We will have 15 minutes discussion slot for both NTRU and Classic McEliece after the Classic McEliece presentation</p>", "time": "2025-07-24T15:35:00Z"}, {"author": "Christopher Wood", "text": "<p>(Thanks, chairs)</p>", "time": "2025-07-24T15:35:11Z"}, {"author": "Mike Ounsworth", "text": "<p>Also, Richard has a reference implementation of the concrete CRFC hybrid kem. I have a reference implementation of LAMPS Composite. We're gonna get together next week and see if they are inter-compatible, I suspect they will be.</p>", "time": "2025-07-24T15:36:00Z"}, {"author": "Richard Barnes", "text": "<p>His citations demonstrate that they don't need an RFC to be referenceable</p>", "time": "2025-07-24T15:36:13Z"}, {"author": "Martin Eker\u00e5", "text": "<p>With respect to KEM combiners: From the perspective of the Swedish NCSA, we would like to see a conservative option such a FrodoKEM being available. Also a symmetric key input and some greater flexibility with respect to what elliptic curves are supported would be welcome.</p>", "time": "2025-07-24T15:36:31Z"}, {"author": "Richard Barnes", "text": "<p>@Mike - note that the reference implementation has diverged slightly from the spec</p>", "time": "2025-07-24T15:36:48Z"}, {"author": "Richard Barnes", "text": "<p>but the implementation is authoritative, we plan to adapt the spec to match</p>", "time": "2025-07-24T15:37:06Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"526\">Richard Barnes</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176641\">said</a>:</p>\n<blockquote>\n<p>@Mike - note that the reference implementation has diverged slightly from the spec</p>\n</blockquote>\n<p>Sure, but I think we'll learn something interesting by doing that exercise.</p>", "time": "2025-07-24T15:37:16Z"}, {"author": "Richard Barnes", "text": "<p>yeah absolutely</p>", "time": "2025-07-24T15:37:24Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"6380\">@Martin Eker\u00e5</span> I want to strongly push back against a symmetric key input. That would make it a protocol, not a primitive.</p>", "time": "2025-07-24T15:37:26Z"}, {"author": "Uri Blumenthal", "text": "<p>It was my understanding that ML-KEM had tighter/stronger proofs than NTRU, and that was one of the factors that drove NIST\u2019s choice of the winner. </p>\n<p>Also, they are similar enough that if MLKEM fails, the likelihood of NTRU failing too is rather high.</p>", "time": "2025-07-24T15:38:02Z"}, {"author": "Watson Ladd", "text": "<p>additional curves feel like it's make another entry in an IANA registry, not something the draft necessarily needs to do even with concrete settings</p>", "time": "2025-07-24T15:38:12Z"}, {"author": "Martin Eker\u00e5", "text": "<p>@Thom Wiggers I do not understand why it would make it a protocol if you have one more input.</p>", "time": "2025-07-24T15:38:41Z"}, {"author": "Uri Blumenthal", "text": "<p>Due to relationship between lattices and codes - how likely is it that if an attack against lattices is found, it will break code-based KEMs as well?</p>", "time": "2025-07-24T15:39:11Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"6380\">@Martin Eker\u00e5</span> it would no longer fit in any KEM definition</p>", "time": "2025-07-24T15:39:13Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"6380\">Martin Eker\u00e5</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176638\">said</a>:</p>\n<blockquote>\n<p>With respect to KEM combiners: From the perspective of the Swedish NCSA, we would like to see a conservative option such a FrodoKEM being available. Also a symmetric key input and some greater flexibility with respect to what elliptic curves are supported would be welcome.</p>\n</blockquote>\n<p>I agree that a symmetric key input is a protocol issue, not a primitive issue.<br>\nFor example, the LAMPS KEMRecipientInfo for CMS (RFC9629) exposes a UserKeyMaterial (UKM) input at the CMS protocol layer, which I think is the right place to put it.</p>", "time": "2025-07-24T15:39:16Z"}, {"author": "Bas Westerbaan", "text": "<p>Additional concrete hybrids are explicitly out of scope for this document, but can certainly be defined in a future document if there is a need for it.</p>", "time": "2025-07-24T15:39:20Z"}, {"author": "Brandon Luo", "text": "<p>Is it true that there are no attacks on Classic McEliece?</p>", "time": "2025-07-24T15:39:28Z"}, {"author": "Bas Westerbaan", "text": "<p>There is a recent key distinguisher, but that one is very theoretical and not concerning to me at all.</p>", "time": "2025-07-24T15:40:07Z"}, {"author": "Thom Wiggers", "text": "<p>I think one or two bits have been shaved for McEliece, but not practical</p>", "time": "2025-07-24T15:40:28Z"}, {"author": "Christopher Patton", "text": "<p><span class=\"user-mention silent\" data-user-id=\"6315\">Brandon Luo</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176665\">said</a>:</p>\n<blockquote>\n<p>Is it true that there are no attacks on Classic McEliece?</p>\n</blockquote>\n<p>I believe what was said by is that cryptanalysis has not improved much in the history of the scheme, whereas we've seen lattice attacks get stronger.</p>", "time": "2025-07-24T15:40:35Z"}, {"author": "Watson Ladd", "text": "<p>No; there are several. but they are very, very expensive. and if you change the kind of code used you can end up with lots of problems</p>", "time": "2025-07-24T15:40:43Z"}, {"author": "Martin Eker\u00e5", "text": "<p>@Watson Ladd I would like to avoid unnecessary restrictions. The current proposal seems to be very locked down. Support for Brainpool curves and  randomly selected curves would be welcome.</p>", "time": "2025-07-24T15:40:56Z"}, {"author": "Daniel Gillmor", "text": "<p>martin: \"avoiding unnecessary restrictions\" is synonymous with \"expecting consumers of the spec to make choices they might not understand\"</p>", "time": "2025-07-24T15:41:47Z"}, {"author": "Stavros Kousidis", "text": "<p>You can instantiate QSF with ECDH using a prime-oder EC that you assume is secure.</p>", "time": "2025-07-24T15:41:52Z"}, {"author": "Deirdre Connolly", "text": "<p>'[Classic McEliece]'s underlying public key encryption scheme\u00a0allows a ciphertext CT to decrypt to the plaintext message under\u00a0<em>any</em>\u00a0decryption key', not just\u00a0the\u00a0decryption key associated with the public encryption key used to generate the ciphertext.'</p>\n<p><a href=\"https://eprint.iacr.org/2021/708.pdf\">https://eprint.iacr.org/2021/708.pdf</a></p>", "time": "2025-07-24T15:42:48Z"}, {"author": "Dennis Jackson", "text": "<p><span class=\"user-mention silent\" data-user-id=\"6380\">Martin Eker\u00e5</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176681\">said</a>:</p>\n<blockquote>\n<p>@Watson Ladd I would like to avoid unnecessary restrictions. The current proposal seems to be very locked down. Support for Brainpool curves and  randomly selected curves would be welcome.</p>\n</blockquote>\n<p>Martin: I think there's a reasonably large fraction of the community against a smorgasbord of options.</p>", "time": "2025-07-24T15:42:59Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"3903\">Stavros Kousidis</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176689\">said</a>:</p>\n<blockquote>\n<p>You can instantiate QSF with ECDH using a prime-oder EC that you assume is secure.</p>\n</blockquote>\n<p>We have one of those in the -concrete document</p>", "time": "2025-07-24T15:43:03Z"}, {"author": "Stavros Kousidis", "text": "<p>The prime-order stuff is actually stated in Section 6.4.2 of the (generic) hybrid KEMs draft.</p>", "time": "2025-07-24T15:43:20Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"6380\">Martin Eker\u00e5</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176681\">said</a>:</p>\n<blockquote>\n<p>@Watson Ladd I would like to avoid unnecessary restrictions. The current proposal seems to be very locked down. Support for Brainpool curves and  randomly selected curves would be welcome.</p>\n</blockquote>\n<p>Remember that CFRG does frameworks, not protocols. You need go to an IETF Protocol WG to get something that you can actually use in a protocol.<br>\nFor example, the LAMPS thing has all the Brainpools:<br>\n<a href=\"https://datatracker.ietf.org/doc/html/draft-ietf-lamps-pq-composite-kem-07#name-algorithm-identifiers\">https://datatracker.ietf.org/doc/html/draft-ietf-lamps-pq-composite-kem-07#name-algorithm-identifiers</a></p>", "time": "2025-07-24T15:43:28Z"}, {"author": "Eric Rescorla", "text": "<p>Maybe we should distinguish between mechanisms which in principle allow for replacement (i.e., they are flexible) and ones where we allow replacement/options by policies.</p>", "time": "2025-07-24T15:43:40Z"}, {"author": "Stavros Kousidis", "text": "<p>@Deirdre: I know, I was too slow to write that :-)</p>", "time": "2025-07-24T15:43:59Z"}, {"author": "Dennis Jackson", "text": "<p><span class=\"user-mention silent\" data-user-id=\"6380\">Martin Eker\u00e5</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176638\">said</a>:</p>\n<blockquote>\n<p>With respect to KEM combiners: From the perspective of the Swedish NCSA, we would like to see a conservative option such a FrodoKEM being available. Also a symmetric key input and some greater flexibility with respect to what elliptic curves are supported would be welcome.</p>\n</blockquote>\n<p>Can you say more about FrodoKEM vs other options not based on LWE? For a 'conservative' approach, FrodoKEM seems like a very small step from ML-KEM.</p>", "time": "2025-07-24T15:44:07Z"}, {"author": "Eric Rescorla", "text": "<p>I think it's good to have mechanisms which are cleanly factored so they are safe for replacement.</p>", "time": "2025-07-24T15:44:17Z"}, {"author": "Christopher Wood", "text": "<p>+1 Richard -- I have not seen a problem statement</p>", "time": "2025-07-24T15:44:21Z"}, {"author": "Deirdre Connolly", "text": "<p><a href=\"https://www.ietf.org/archive/id/draft-irtf-cfrg-concrete-hybrid-kems-00.html#name-qsf-mlkem768-p256-sha3256-s\">https://www.ietf.org/archive/id/draft-irtf-cfrg-concrete-hybrid-kems-00.html#name-qsf-mlkem768-p256-sha3256-s</a><br>\n<a href=\"https://www.ietf.org/archive/id/draft-irtf-cfrg-concrete-hybrid-kems-00.html#name-qsf-mlkem1024-p384-sha3256-\">https://www.ietf.org/archive/id/draft-irtf-cfrg-concrete-hybrid-kems-00.html#name-qsf-mlkem1024-p384-sha3256-</a></p>", "time": "2025-07-24T15:44:25Z"}, {"author": "Eric Rescorla", "text": "<p>I also am unclear about the problem statement</p>", "time": "2025-07-24T15:44:30Z"}, {"author": "Martin Eker\u00e5", "text": "<p>@Thom Wiggers Maybe that is a valid point, but it is also a valid point that is very good to be able to input a pre-placed shared secret in use cases where this is feasible. If there is no support in the KEM combiner for this, then where would it be supported? In each individual protocol? I fear that there is then a risk that it will be forgotten.</p>", "time": "2025-07-24T15:45:45Z"}, {"author": "Stavros Kousidis", "text": "<p>@Martin Ekera: I guess you could instatiate your favorite hybrid KEM from the generic doc and use 'label' for any private input you'd like to have.</p>", "time": "2025-07-24T15:45:54Z"}, {"author": "Watson Ladd", "text": "<p>label is not for applications: libraries would very much not like that</p>", "time": "2025-07-24T15:46:23Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"3903\">@Stavros Kousidis</span> I would not encourage this because that will result in proof weirdness for your upper layer protocol</p>", "time": "2025-07-24T15:46:24Z"}, {"author": "Christopher Wood", "text": "<p>The new CFRG process would suggest that proponents of these things produce a problem statement -- is someone willing to do that work? We could revisit these questions after that exists.</p>", "time": "2025-07-24T15:47:23Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"6380\">@Martin Eker\u00e5</span> yeah, I think you will need to do this in every individual protocol. But each protocol would need to define how to handle that PSK input anyway. I don't think adding K = H(PSK | KEM_HYBRID_K)  is so different</p>", "time": "2025-07-24T15:47:29Z"}, {"author": "Richard Barnes", "text": "<p>Even if people need diversity of computation assumptions -- there are existing, citeable, analyzed things out there with multiple approaches</p>", "time": "2025-07-24T15:48:13Z"}, {"author": "Richard Barnes", "text": "<p>What's missing?</p>", "time": "2025-07-24T15:48:22Z"}, {"author": "Deirdre Connolly", "text": "<p>Oh I should show up and register some sick new KPQC KEMs :D</p>", "time": "2025-07-24T15:48:35Z"}, {"author": "Watson Ladd", "text": "<p>well, target HPKE instead which does let you mix in keys iiuc</p>", "time": "2025-07-24T15:48:37Z"}, {"author": "Deirdre Connolly", "text": "<p>FUD indeed! crypto scawy</p>", "time": "2025-07-24T15:48:50Z"}, {"author": "John Gray", "text": "<p>Side Note to main discussion point:  We have an expired draft in LAMPS which externalizes the public key in an X509Certificate.  Use of McEliece would make it very practical in X.509 land as it is a mechanism that allows the public keys to be externalized (and thus cached), saving the 1MB download bandwidth on subsequent use.   We let the draft expire because we didn't have anyone telling us they needed it.  If McEliece is standardize then maybe it is worth resurrecting this draft?     <a href=\"https://www.ietf.org/archive/id/draft-ounsworth-lamps-pq-external-pubkeys-05.html\">https://www.ietf.org/archive/id/draft-ounsworth-lamps-pq-external-pubkeys-05.html</a></p>", "time": "2025-07-24T15:48:55Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"6380\">@Martin Eker\u00e5</span> I think Watson may have a great point, that HPKE might already be the kind of primitive-with-psk-support that you are looking for</p>", "time": "2025-07-24T15:49:29Z"}, {"author": "Thom Wiggers", "text": "<p>well-analyzed as well</p>", "time": "2025-07-24T15:49:35Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"526\">Richard Barnes</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176761\">said</a>:</p>\n<blockquote>\n<p>Even if people need diversity of computation assumptions -- there are existing, citeable, analyzed things out there with multiple approaches</p>\n</blockquote>\n<p>I think, they are either hidden/not open (ISO) or they are 'merely' written specs, not blessed by any standards body</p>", "time": "2025-07-24T15:49:39Z"}, {"author": "Christopher Patton", "text": "<p>To refine my point: It feels like different WGs are making different decisions about PQ agility. One thing CFRG could do is try to guide everyone towards the same considerations.</p>", "time": "2025-07-24T15:49:59Z"}, {"author": "Simon Josefsson", "text": "<p>I hear criticism in some IETF WGs that the non-IETF/NIST external references are not stable enough to be valid for use.  IIRC, there was a question about this in SSHM the other day.</p>", "time": "2025-07-24T15:50:01Z"}, {"author": "Bas Westerbaan", "text": "<p>When we still thought that this eprint ( <a href=\"https://eprint.iacr.org/2024/555\">https://eprint.iacr.org/2024/555</a> ) was correct, we were scrambling to see what would be a good replacement for ML-KEM in TLS for the Web. It looked like it would be a more conservative variant of ML-KEM instead of FrodoKEM or HQC. Not saying it will always be like that.</p>", "time": "2025-07-24T15:50:15Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"5063\">Simon Josefsson</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176787\">said</a>:</p>\n<blockquote>\n<p>I hear criticism in some IETF WGs that the non-IETF/NIST external references are not stable enough to be valid for use.  IIRC, there was a question about this in SSHM the other day.</p>\n</blockquote>\n<p>this</p>", "time": "2025-07-24T15:50:15Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"603\">@Deirdre Connolly</span> or even just mathematical descriptions with some code of questionable clearness</p>", "time": "2025-07-24T15:50:30Z"}, {"author": "Stavros Kousidis", "text": "<p>@Thom Wiggers: There are pre-shared key options in TLS, so Martin Ekeras question is a viable question.</p>", "time": "2025-07-24T15:50:30Z"}, {"author": "Rebecca Guthrie", "text": "<p><span class=\"user-mention silent\" data-user-id=\"86\">Thom Wiggers</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176779\">said</a>:</p>\n<blockquote>\n<p><span class=\"user-mention silent\" data-user-id=\"6380\">Martin Eker\u00e5</span> I think Watson may have a great point, that HPKE might already be the kind of primitive-with-psk-support that you are looking for</p>\n</blockquote>\n<p>IKEv2 and TLS 1.3 also have a way to incorporate a PSK at the protocol level</p>", "time": "2025-07-24T15:50:40Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"549\">Bas Westerbaan</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176790\">said</a>:</p>\n<blockquote>\n<p>When we still thought that this eprint ( <a href=\"https://eprint.iacr.org/2024/555\">https://eprint.iacr.org/2024/555</a> ) was correct, we were scrambling to see what would be a good replacement for ML-KEM in TLS. It looked like it would be a more conservative variant of ML-KEM instead of FrodoKEM or HQC. Not saying it will always be like that.</p>\n</blockquote>\n<p><span class=\"user-mention\" data-user-id=\"549\">@Bas Westerbaan</span> did that look like different params?</p>", "time": "2025-07-24T15:50:43Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"3903\">@Stavros Kousidis</span> and TLS specifies how to handle PSK, the key exchange is separate</p>", "time": "2025-07-24T15:50:49Z"}, {"author": "Deirdre Connolly", "text": "<p>Exactly what Watson is saying</p>", "time": "2025-07-24T15:51:07Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"603\">Deirdre Connolly</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176765\">said</a>:</p>\n<blockquote>\n<p>Oh I should show up and register some sick new KPQC KEMs :D</p>\n</blockquote>\n<p>SM2! SM2!</p>", "time": "2025-07-24T15:51:46Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"5873\">@Mike Ounsworth</span> <span aria-label=\"flag china\" class=\"emoji emoji-1f1e8-1f1f3\" role=\"img\" title=\"flag china\">:flag_china:</span></p>", "time": "2025-07-24T15:51:58Z"}, {"author": "Bas Westerbaan", "text": "<p><span class=\"user-mention silent\" data-user-id=\"603\">Deirdre Connolly</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176798\">said</a>:</p>\n<div class=\"codehilite\"><pre><span></span><code>@**Bas Westerbaan** did that look like different params?\n</code></pre></div>\n<p>No, it'd been much more than bumping the dimension.</p>", "time": "2025-07-24T15:52:21Z"}, {"author": "Rich Salz", "text": "<p>Dual-EC would like to join the room.</p>", "time": "2025-07-24T15:52:21Z"}, {"author": "Deirdre Connolly", "text": "<p>But what does 'established algorithms' mean? These schemes are not 'standardized' in other bodies, they are mostly the left over submissions from the NIST competition or closed bodies like ISO</p>", "time": "2025-07-24T15:52:26Z"}, {"author": "Aron Wussler", "text": "<p>Don't we already describe ECC in 7748? What would be different here?</p>", "time": "2025-07-24T15:52:42Z"}, {"author": "Uri Blumenthal", "text": "<p>Luckily for all of us, that eprint turned out to be full of turd. <span aria-label=\"smiley\" class=\"emoji emoji-1f603\" role=\"img\" title=\"smiley\">:smiley:</span></p>", "time": "2025-07-24T15:52:53Z"}, {"author": "Eric Rescorla", "text": "<p>+1 to what Watson is saying. What is the CFRG's involvement bringing to the party here? Is it making the algorithms better? Is it making the documents better? Is it just a way to get it into an RFC?</p>", "time": "2025-07-24T15:52:54Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"1247\">Aron Wussler</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176822\">said</a>:</p>\n<blockquote>\n<p>Don't we already describe ECC in 7748? What would be different here?</p>\n</blockquote>\n<p>This is part of why we're in this predicament</p>", "time": "2025-07-24T15:53:21Z"}, {"author": "Uri Blumenthal", "text": "<p>@Deirdre +1</p>", "time": "2025-07-24T15:53:26Z"}, {"author": "Dennis Jackson", "text": "<p>It seems like there's nothing for CFRG to do here</p>", "time": "2025-07-24T15:53:29Z"}, {"author": "Rich Salz", "text": "<p>@Ekr, historically the latter is a big part.</p>", "time": "2025-07-24T15:53:30Z"}, {"author": "Watson Ladd", "text": "<p>right, but what do we do when we do that?</p>", "time": "2025-07-24T15:53:51Z"}, {"author": "Eric Rescorla", "text": "<p>@Rich: If we think the documents are in good shape, then this seems like what the ISE is for.</p>", "time": "2025-07-24T15:54:00Z"}, {"author": "Eric Rescorla", "text": "<p>The other thing the CFRG might be doing is <em>endorsing</em> ana lgorthm.</p>", "time": "2025-07-24T15:54:28Z"}, {"author": "Rich Salz", "text": "<p>++ ekr</p>", "time": "2025-07-24T15:54:37Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"810\">Eric Rescorla</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176845\">said</a>:</p>\n<blockquote>\n<p>@Rich: If we think the documents are in good shape, then this seems like what the ISE is for.</p>\n</blockquote>\n<p>Eliot may yelp again at this</p>", "time": "2025-07-24T15:54:40Z"}, {"author": "Eric Rescorla", "text": "<p>I.e., by saying it's good and people should use it. That's what 7748 did</p>", "time": "2025-07-24T15:54:42Z"}, {"author": "Dmitry Belyavskiy", "text": "<p>@Ekr++</p>", "time": "2025-07-24T15:54:55Z"}, {"author": "Simon Josefsson", "text": "<p>I see significant value in having CFRG publish RFC on a couple of alternatives for which there are no stable enough reference out there.   The CFRG is a good body to provide a good security considerations and other engineering-related advice.</p>", "time": "2025-07-24T15:55:08Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"810\">Eric Rescorla</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176860\">said</a>:</p>\n<blockquote>\n<p>I.e., by saying it's good and people should use it. That's what 7748 did</p>\n</blockquote>\n<p>wait, what is distinct between 7748 and adopting either of these ?</p>", "time": "2025-07-24T15:55:12Z"}, {"author": "Watson Ladd", "text": "<p>7748 came after wide deployment, and was a disasterous mess because people thought they could change what it said</p>", "time": "2025-07-24T15:55:19Z"}, {"author": "Eric Rescorla", "text": "<p>@Deirdre: I think adopting one of these <em>would</em> be like 7748. And so I think the question is \"do we think that it's important for the CFRG to endorse a new PQ KEM\"</p>", "time": "2025-07-24T15:55:54Z"}, {"author": "Watson Ladd", "text": "<p>to be clear, I think e.g. describing classic McElice or NTRU prime because people are using them and are out there is good</p>", "time": "2025-07-24T15:55:59Z"}, {"author": "Aron Wussler", "text": "<p>I also believe there is no need to re-define an algorithm defined at ISO but at CFRG we could still make sure there is a guidance, and avoid an excessive proliferation of algorithms?</p>", "time": "2025-07-24T15:56:41Z"}, {"author": "Watson Ladd", "text": "<p>I think there are valid reasons to rewrite things at ISO that are widely used. About 235 of them, some cases more :P</p>", "time": "2025-07-24T15:57:15Z"}, {"author": "Christopher Patton", "text": "<p>I'd rather rely on other standards body, but we'd take it on a case by case basis: if the Review Panel thinks the scheme is sound as specified, then we can endorse it.</p>", "time": "2025-07-24T15:57:19Z"}, {"author": "Dmitry Belyavskiy", "text": "<p>I think CFRG should provide the requirements/limits/other framing documents for PQ KEMs</p>", "time": "2025-07-24T15:57:28Z"}, {"author": "Richard Barnes", "text": "<p><span class=\"user-mention\" data-user-id=\"1247\">@Aron Wussler</span> that could be part of a proble mstatement</p>", "time": "2025-07-24T15:57:28Z"}, {"author": "Daniel Gillmor", "text": "<p>I'd like to hear from someone who said the problem statement is <em>not</em> well defined, and also the CFRG is the right place to do the work.</p>", "time": "2025-07-24T15:57:32Z"}, {"author": "Eric Rescorla", "text": "<p>So as a hypothetical, suppose that some other SDO published a completely public spec for Classic McEliece. Would people still think it was valuable to publish it here?</p>", "time": "2025-07-24T15:57:35Z"}, {"author": "Christopher Patton", "text": "<p>Like we can specify something if we decide we need to fix it</p>", "time": "2025-07-24T15:57:39Z"}, {"author": "Aron Wussler", "text": "<p>Daniel Gillmor</p>\n<p>I'd like to hear from someone who said the problem statement is not well defined, and also the CFRG is the right place to do the work.</p>\n<p>+1</p>", "time": "2025-07-24T15:58:10Z"}, {"author": "Rich Salz", "text": "<p>@CPatton: does fix mean change language or change technical content?</p>", "time": "2025-07-24T15:58:33Z"}, {"author": "Martin Eker\u00e5", "text": "<p>@Thom Wiggers So I am mainly concerned with products and RFCs/specifications/other similiar documents having more conservative options available in practice. Otherwise, we may end up with a very restricted set of options.</p>", "time": "2025-07-24T15:58:36Z"}, {"author": "Uri Blumenthal", "text": "<p>We should not describe NTRU here, as it offers no advantages against ML-KEM.</p>", "time": "2025-07-24T15:58:55Z"}, {"author": "Simon Josefsson", "text": "<p>Paywalled ISO specifications are sometimes claimed to not be open enough for use as a stable reference in IETF work</p>", "time": "2025-07-24T15:59:02Z"}, {"author": "Christopher Patton", "text": "<p><span class=\"user-mention silent\" data-user-id=\"11\">Rich Salz</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176907\">said</a>:</p>\n<blockquote>\n<p>@CPatton: does fix mean change language or change technical content?</p>\n</blockquote>\n<p>could be both. On the one hand we might find that some claimed security property doesn't hold, but we don't actually need to fix anything to fullfill requirements</p>", "time": "2025-07-24T15:59:09Z"}, {"author": "Watson Ladd", "text": "<p>I am sympathetic to that, but I do think they are enought to get a codepoint from which is what you need</p>", "time": "2025-07-24T15:59:24Z"}, {"author": "Richard Barnes", "text": "<p>Specifically, we need to know what is insufficient about the universe of PQ KEM specifications that exists right now, such that CFRG could address that insufficiency.</p>", "time": "2025-07-24T15:59:31Z"}, {"author": "Uri Blumenthal", "text": "<p>McEliece has tighter proofs, and use cases that ML-KEM may not be optimal for - this would be a reasonable justification for describing it here.</p>", "time": "2025-07-24T16:00:12Z"}, {"author": "Simon Josefsson", "text": "<p>I think the problem is that the universe of CFRG-approved PQ KEM specifications is empty</p>", "time": "2025-07-24T16:00:15Z"}, {"author": "Eric Rescorla", "text": "<p>Why is it a problem that the universe of CFRG-approved PQ KEMs is empty? The IETF is using ML-KEM.</p>", "time": "2025-07-24T16:00:44Z"}, {"author": "Dennis Jackson", "text": "<p>I don't think the CFRG should be here to stamp such things. Usage in deployed protocols is what counts.</p>", "time": "2025-07-24T16:00:46Z"}, {"author": "Eric Rescorla", "text": "<p>Without any CFRG stamp</p>", "time": "2025-07-24T16:00:52Z"}, {"author": "Mike Ounsworth", "text": "<p>I voted \"No\" to \"Is CFRG the right venue\" on the basis that A) full cryptanalysis of a novel primitive really needs a national competition level of activity; CFRG is not that. B) if an algorithm has sufficient attention through such a competition, then surely a specification of it exists that IETF WGs could register codepoints against? Like, IETF has absolutely registered codepoints for algs that don't have CFRG specs.</p>", "time": "2025-07-24T16:01:06Z"}, {"author": "David Waite", "text": "<p>I don\u2019t know if CFRG is really necessary to publish an algorithm defined elsewhere as an informational RFC, if the challenge is document access</p>", "time": "2025-07-24T16:01:11Z"}, {"author": "Uri Blumenthal", "text": "<p>Paywalled specs should not be considered here, period, full stop.</p>", "time": "2025-07-24T16:01:17Z"}, {"author": "Richard Barnes", "text": "<p><span class=\"user-mention\" data-user-id=\"849\">@Uri Blumenthal</span>  out of curiosity, what do those use cases look like?  (McEliece &gt; ML-KEM)</p>", "time": "2025-07-24T16:01:27Z"}, {"author": "Eric Rescorla", "text": "<p>I actually don't agree with Dennis here. I think we need to have a set of primitives that the IETF can depend on. It's very convenient when other SDOs take care of that for us, but if they don't, then CFRG should step into the gap.</p>", "time": "2025-07-24T16:01:36Z"}, {"author": "Simon Josefsson", "text": "<p>The CFRG has stamped algorithms before, to provide a community service with stable specifications and an errata process around them.  I think this would be useful here too</p>", "time": "2025-07-24T16:01:55Z"}, {"author": "Daniel Gillmor", "text": "<p>ekr: and if there's a documentation gap, we should be clear about what specific gap we're filling.</p>", "time": "2025-07-24T16:02:12Z"}, {"author": "Dennis Jackson", "text": "<p>Sure, we need the option of a research group when we have a problem we don't know how to solve. But KEMs is not that.</p>", "time": "2025-07-24T16:02:16Z"}, {"author": "Eric Rescorla", "text": "<p>@DKG: yes, absolutely</p>", "time": "2025-07-24T16:02:20Z"}, {"author": "Richard Barnes", "text": "<p><span class=\"user-mention\" data-user-id=\"5063\">@Simon Josefsson</span> that process is equally well served by the ISE</p>", "time": "2025-07-24T16:02:23Z"}, {"author": "Richard Barnes", "text": "<p>(those goals)</p>", "time": "2025-07-24T16:02:39Z"}, {"author": "Richard Barnes", "text": "<p>(and with much less work!)</p>", "time": "2025-07-24T16:03:11Z"}, {"author": "Eric Rescorla", "text": "<p>To clarify my position, suppose there was consensus that we needed a PQ KEM that filled a point in the design space that ML-KEM did not and there was no other SDO standardizing something. I think it would make sense for CFRG to run a process to select and document one. The question here is whether that is the case</p>", "time": "2025-07-24T16:03:23Z"}, {"author": "Daniel Gillmor", "text": "<p>@Barnes: unless you're making a claim in the document about cryptographic properties of the algorithm in the document; the ISE is not well equipped to verify such a claim.</p>", "time": "2025-07-24T16:03:26Z"}, {"author": "Uri Blumenthal", "text": "<p>@Richard, constrained communication links <em>and</em> ability to pre-load public keys. Which makes McEliece 100% unacceptable for ephemeral cases.</p>", "time": "2025-07-24T16:03:26Z"}, {"author": "Simon Josefsson", "text": "<p><span class=\"user-mention silent\" data-user-id=\"526\">Richard Barnes</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176947\">said</a>:</p>\n<blockquote>\n<p><span class=\"user-mention silent\" data-user-id=\"5063\">Simon Josefsson</span> that process is equally well served by the ISE</p>\n</blockquote>\n<p>I disagree - had RSA, ECC, EdDSA, Argon2 etc been published through ISE it would have carried less weight than if these were done in harmony with CFRG processes</p>", "time": "2025-07-24T16:03:39Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"849\">Uri Blumenthal</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176936\">said</a>:</p>\n<blockquote>\n<p>Paywalled specs should not be considered here, period, full stop.</p>\n</blockquote>\n<p>i think this is a major part of the problem</p>", "time": "2025-07-24T16:03:45Z"}, {"author": "Thom Wiggers", "text": "<p><a href=\"https://datatracker.ietf.org/doc/draft-ietf-pquip-pqc-engineers/\">https://datatracker.ietf.org/doc/draft-ietf-pquip-pqc-engineers/</a> also exists, in the context of \"what are KEMs with different security properties\"....</p>", "time": "2025-07-24T16:04:02Z"}, {"author": "Richard Barnes", "text": "<p>@dkg - Simon was just asking for a stable reference and errata.  ISE does that fine.</p>", "time": "2025-07-24T16:04:23Z"}, {"author": "Richard Barnes", "text": "<p><span class=\"user-mention\" data-user-id=\"5063\">@Simon Josefsson</span> \"Weight\" is a different requirement from stability + errata</p>", "time": "2025-07-24T16:04:42Z"}, {"author": "Dennis Jackson", "text": "<p>I've always been a bit confused about pqwip being a WG rather than an RG</p>", "time": "2025-07-24T16:04:45Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"810\">Eric Rescorla</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176953\">said</a>:</p>\n<blockquote>\n<p>To clarify my position, suppose there was consensus that we needed a PQ KEM that filled a point in the design space that ML-KEM did not and there was no other SDO standardizing something. I think it would make sense for CFRG to run a process to select and document one. The question here is whether that is the case</p>\n</blockquote>\n<p>I 80% agree, with a footnote questioning whether CFRG is really qualified to be the judge of such a competition. I watched the NIST competition. They had a full staff of PhD cryptographers. CFRG doesn't have nearly the same level of resources.</p>", "time": "2025-07-24T16:04:59Z"}, {"author": "Richard Barnes", "text": "<p><span class=\"user-mention\" data-user-id=\"2103\">@Dennis Jackson</span> i am also confused as to why PQUIP exists</p>", "time": "2025-07-24T16:05:08Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"526\">Richard Barnes</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176967\">said</a>:</p>\n<blockquote>\n<p><span class=\"user-mention silent\" data-user-id=\"2103\">Dennis Jackson</span> i am also confused as to why PQUIP exists</p>\n</blockquote>\n<p>because I wanted a soapbox, obviously!</p>", "time": "2025-07-24T16:05:33Z"}, {"author": "Simon Josefsson", "text": "<p><span class=\"user-mention silent\" data-user-id=\"526\">Richard Barnes</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176960\">said</a>:</p>\n<blockquote>\n<p>@dkg - Simon was just asking for a stable reference and errata.  ISE does that fine.</p>\n</blockquote>\n<p>I think CFRG contributed some value to the publication of previous crypto primitives, and offer a place for discussion of erratas, which would not be served by going the ISE route.</p>", "time": "2025-07-24T16:05:40Z"}, {"author": "Rich Salz", "text": "<blockquote>\n<p>had RSA, ECC, EdDSA, Argon2 etc been published through ISE it would have carried less weight</p>\n</blockquote>\n<p>I think the world has changed a lot since then. (And I notice AES isn't in that list.)</p>", "time": "2025-07-24T16:05:48Z"}, {"author": "Dennis Jackson", "text": "<p><span class=\"user-mention silent\" data-user-id=\"5873\">Mike Ounsworth</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176968\">said</a>:</p>\n<blockquote>\n<p><span class=\"user-mention silent\" data-user-id=\"526\">Richard Barnes</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176967\">said</a>:</p>\n<blockquote>\n<p><span class=\"user-mention silent\" data-user-id=\"2103\">Dennis Jackson</span> i am also confused as to why PQUIP exists</p>\n</blockquote>\n<p>because I wanted a soapbox, obviously!</p>\n</blockquote>\n<p>LAMPS wasn't enough!? :-P</p>", "time": "2025-07-24T16:05:56Z"}, {"author": "Christopher Wood", "text": "<p>LOL</p>", "time": "2025-07-24T16:06:04Z"}, {"author": "Thom Wiggers", "text": "<p>tbh Mike's soapbox comment is not wrong. We are doing some stuff in PQUIP that I am not sure fits anywhere else and I think/hope it's useful.</p>", "time": "2025-07-24T16:06:08Z"}, {"author": "Daniel Gillmor", "text": "<p>oh the S in LAMPS stands for Soapbox?</p>", "time": "2025-07-24T16:06:25Z"}, {"author": "Richard Barnes", "text": "<p>And <span class=\"user-mention\" data-user-id=\"5873\">@Mike Ounsworth</span> just colonized ACME too!</p>", "time": "2025-07-24T16:06:28Z"}, {"author": "Rich Salz", "text": "<p>PQUIP is where mike takes off his instigator mask and turns into a reporter</p>", "time": "2025-07-24T16:06:49Z"}, {"author": "Richard Barnes", "text": "<p>I grant that some of the terminology that PQUIP has done has been helpful.  i just had my Always Be Closing hat on.</p>", "time": "2025-07-24T16:07:32Z"}, {"author": "Rebecca Guthrie", "text": "<p><span class=\"user-mention silent\" data-user-id=\"86\">Thom Wiggers</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176976\">said</a>:</p>\n<blockquote>\n<p>tbh Mike's soapbox comment is not wrong. We are doing some stuff in PQUIP that I am not sure fits anywhere else and I think/hope it's useful.</p>\n</blockquote>\n<p>if I remember correctly, pquip was intended to serve as a place to work on PQ migration for protocols that did not have an active WG associated with them</p>", "time": "2025-07-24T16:07:47Z"}, {"author": "Rebecca Guthrie", "text": "<p>was compared to curdle a few times</p>", "time": "2025-07-24T16:07:56Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"606\">@Rebecca Guthrie</span> while that was discussed I don't think that is in charter</p>", "time": "2025-07-24T16:08:06Z"}, {"author": "Christopher Patton", "text": "<p>BBS talk: what do the coefficients of this polynomial represent?  (what is <code>n</code>?)</p>", "time": "2025-07-24T16:08:31Z"}, {"author": "Daniel Gillmor", "text": "<p>n is number of pseudonyms?</p>", "time": "2025-07-24T16:08:49Z"}, {"author": "Flo D", "text": "<p><span class=\"user-mention silent\" data-user-id=\"86\">Thom Wiggers</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177002\">said</a>:</p>\n<blockquote>\n<p><span class=\"user-mention silent\" data-user-id=\"606\">Rebecca Guthrie</span> while that was discussed I don't think that is in charter</p>\n</blockquote>\n<p>It's in there are a \"last resort\"</p>", "time": "2025-07-24T16:08:53Z"}, {"author": "Rebecca Guthrie", "text": "<p><span class=\"user-mention silent\" data-user-id=\"86\">Thom Wiggers</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177002\">said</a>:</p>\n<blockquote>\n<p><span class=\"user-mention silent\" data-user-id=\"606\">Rebecca Guthrie</span> while that was discussed I don't think that is in charter</p>\n</blockquote>\n<p>from the charter!<br>\n\"The WG will also provide a venue of last resort to discuss PQC-related issues in IETF protocols that have no associated maintenance WGs.\"</p>", "time": "2025-07-24T16:08:54Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"606\">Rebecca Guthrie</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/176997\">said</a>:</p>\n<blockquote>\n<p>if I remember correctly, pquip was intended to serve as a place to work on PQ migration for protocols that did not have an active WG associated with them</p>\n</blockquote>\n<p>I'm fairly sure that PQUIP is explicitly NOT chartered to do protocol work. It's a place to coordinate PQC-related protocol stuff between WGs that otherwise don't have a good contact-point.</p>", "time": "2025-07-24T16:09:10Z"}, {"author": "Rich Salz", "text": "<p>PQUIP discusses issues and transitions but DOES NOT specify algs or how to use them in protocols.</p>", "time": "2025-07-24T16:09:14Z"}, {"author": "Thom Wiggers", "text": "<p>ah right</p>", "time": "2025-07-24T16:09:17Z"}, {"author": "Rich Salz", "text": "<p>@mike jinx!</p>", "time": "2025-07-24T16:09:26Z"}, {"author": "Thom Wiggers", "text": "<p>I think SSH was the main example though and there is now SSHM</p>", "time": "2025-07-24T16:10:01Z"}, {"author": "Daniel Gillmor", "text": "<p>LAMPS: Let's Appreciate Mike's Post-quantum Soapbox</p>", "time": "2025-07-24T16:10:12Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"637\">@Daniel Gillmor</span> more believable than \"limited\"</p>", "time": "2025-07-24T16:10:25Z"}, {"author": "Flo D", "text": "<p>Aside from anything else I think PQUIP's a good venue to get protocol engineers and cryptographers together on PQ issues that aren't specific to a protocol.  I think MediaOps and IOTOps were mentioned as similar constructs, but having never been to either of those I can't comment on how good that parallel is.</p>", "time": "2025-07-24T16:10:32Z"}, {"author": "Uri Blumenthal", "text": "<p><span aria-label=\"joy\" class=\"emoji emoji-1f602\" role=\"img\" title=\"joy\">:joy:</span></p>", "time": "2025-07-24T16:10:32Z"}, {"author": "Sofia Celi", "text": "<blockquote>\n<p>I 80% agree, with a footnote questioning whether CFRG is really qualified to be the judge of such a competition. I watched the NIST competition. They had a full staff of PhD cryptographers. CFRG doesn't have nearly the same level of resources.</p>\n</blockquote>\n<p>I 100% agree on this. CFRG does not have many cryptographers as in the NIST process. And I see even less on the cryptanalysis part. Perhaps a competition could encourage that but it will require community building</p>", "time": "2025-07-24T16:10:48Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"637\">Daniel Gillmor</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177028\">said</a>:</p>\n<blockquote>\n<p>LAMPS: Let's Appreciate Mike's Post-quantum Soapbox</p>\n</blockquote>\n<p>Bash. I prefer \"Amplify\".</p>", "time": "2025-07-24T16:11:01Z"}, {"author": "Uri Blumenthal", "text": "<p><span aria-label=\"+1\" class=\"emoji emoji-1f44d\" role=\"img\" title=\"+1\">:+1:</span></p>", "time": "2025-07-24T16:11:12Z"}, {"author": "Watson Ladd", "text": "<p>the last time we had a contest we did a lot of damage to cryptographers feeling participating was worthwhile</p>", "time": "2025-07-24T16:11:17Z"}, {"author": "Sofia Celi", "text": "<p>100% watson</p>", "time": "2025-07-24T16:11:27Z"}, {"author": "Rich Salz", "text": "<p>++Watson</p>", "time": "2025-07-24T16:11:28Z"}, {"author": "Christopher Wood", "text": "<p>@Watson are you referring to the curve wars?</p>", "time": "2025-07-24T16:11:31Z"}, {"author": "Watson Ladd", "text": "<p>yes</p>", "time": "2025-07-24T16:11:35Z"}, {"author": "Christopher Wood", "text": "<p>(We also did the PAKE competition, which was much more pleasant)</p>", "time": "2025-07-24T16:11:47Z"}, {"author": "Uri Blumenthal", "text": "<p>CFRG cannot perform full analysis comparable to that of NIST, BSI, etc.</p>", "time": "2025-07-24T16:11:55Z"}, {"author": "Sofia Celi", "text": "<p>but also PAKE</p>", "time": "2025-07-24T16:11:56Z"}, {"author": "Daniel Gillmor", "text": "<p>Mike: as long as it's not \"Avoid\"</p>", "time": "2025-07-24T16:11:57Z"}, {"author": "Deirdre Connolly", "text": "<p>be still my heart, 'we need PQ ZK schemes going forward'</p>", "time": "2025-07-24T16:12:28Z"}, {"author": "Thom Wiggers", "text": "<p>does ISO do much analysis, though? I think they just require that there's a paper that's been out a few years.</p>", "time": "2025-07-24T16:12:41Z"}, {"author": "Uri Blumenthal", "text": "<p>however, it can evaluate and pick from the \u201capproved\u201d algorithms those that erotic site specific IETF use cases best/better.</p>", "time": "2025-07-24T16:12:56Z"}, {"author": "Watson Ladd", "text": "<p>WGs are much better suited to understanding what works and what doesn't for them</p>", "time": "2025-07-24T16:13:28Z"}, {"author": "Christopher Patton", "text": "<p>big fan of IOPs</p>", "time": "2025-07-24T16:14:04Z"}, {"author": "Dan Harkins", "text": "<p>@Uri, \"erotic site\"?</p>", "time": "2025-07-24T16:14:06Z"}, {"author": "Deirdre Connolly", "text": "<p>Ligero is nice</p>", "time": "2025-07-24T16:14:07Z"}, {"author": "Dan Harkins", "text": "<p>to tell!</p>", "time": "2025-07-24T16:14:12Z"}, {"author": "Thom Wiggers", "text": "<p>(I may be wrong about ISO, not like I can join their mailing list)</p>", "time": "2025-07-24T16:14:12Z"}, {"author": "Uri Blumenthal", "text": "<p>@Watson, yes, <em>but</em> they often lack cryptographic expertise and awareness of subtle issues.</p>", "time": "2025-07-24T16:14:15Z"}, {"author": "Christopher Patton", "text": "<p>FWIW draft-irtf-cfrg-vdaf uses a \"fully linear\" IOP (for proving things about secret shared data)</p>", "time": "2025-07-24T16:14:38Z"}, {"author": "Deirdre Connolly", "text": "<p>&lt;1s on a phone is great</p>", "time": "2025-07-24T16:14:40Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"86\">Thom Wiggers</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177062\">said</a>:</p>\n<blockquote>\n<p>does ISO do much analysis, though? I think they just require that there's a paper that's been out a few years.</p>\n</blockquote>\n<p>Yeah ... but ... you know what you're getting with an ISO spec. RFCs have a different reputation.</p>", "time": "2025-07-24T16:15:06Z"}, {"author": "Uri Blumenthal", "text": "<p>@Dan - darn the spell checkers! \u201cInventor of spellchecker died - may he roast in piss!\u201d</p>", "time": "2025-07-24T16:15:29Z"}, {"author": "Thom Wiggers", "text": "<p>Mike: a 500 CHF invoice, right?</p>", "time": "2025-07-24T16:15:33Z"}, {"author": "Richard Barnes", "text": "<p>oof, C++</p>", "time": "2025-07-24T16:15:45Z"}, {"author": "Dan Harkins", "text": "<p><span aria-label=\"rolling on the floor laughing\" class=\"emoji emoji-1f923\" role=\"img\" title=\"rolling on the floor laughing\">:rolling_on_the_floor_laughing:</span></p>", "time": "2025-07-24T16:15:47Z"}, {"author": "Deirdre Connolly", "text": "<p>Man this is really making a fan out of me</p>", "time": "2025-07-24T16:15:51Z"}, {"author": "Simon Josefsson", "text": "<p><span class=\"user-mention silent\" data-user-id=\"205\">Sofia Celi</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177035\">said</a>:</p>\n<blockquote>\n<blockquote>\n<p>I 80% agree, with a footnote questioning whether CFRG is really qualified to be the judge of such a competition. I watched the NIST competition. They had a full staff of PhD cryptographers. CFRG doesn't have nearly the same level of resources.</p>\n</blockquote>\n<p>I 100% agree on this. CFRG does not have many cryptographers as in the NIST process. And I see even less on the cryptanalysis part. Perhaps a competition could encourage that but it will require community building</p>\n</blockquote>\n<p>+1 -- what CFRG could do well is to publish specifications as RFCs.  And let other WG's make context-specific decisions what kind of primitives they want to use, choosing from the available options of stable references.</p>", "time": "2025-07-24T16:15:58Z"}, {"author": "Deirdre Connolly", "text": "<p>I should go impl LIgero</p>", "time": "2025-07-24T16:16:01Z"}, {"author": "Sofia Celi", "text": "<blockquote>\n<p>FWIW draft-irtf-cfrg-vdaf uses a \"fully linear\" IOP (for proving things about secret shared data)</p>\n</blockquote>\n<p>I guess the full difference is speciallized over secret-sharing schemes, rather than more general ones over full commitments as this presentation</p>", "time": "2025-07-24T16:16:10Z"}, {"author": "Uri Blumenthal", "text": "<p>^\u201d those that fit specific IETF use cases better \u201c</p>", "time": "2025-07-24T16:16:37Z"}, {"author": "Deirdre Connolly", "text": "<p>good question Watson</p>", "time": "2025-07-24T16:17:03Z"}, {"author": "Christopher Patton", "text": "<p><span class=\"user-mention silent\" data-user-id=\"205\">Sofia Celi</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177106\">said</a>:</p>\n<blockquote>\n<blockquote>\n<p>FWIW draft-irtf-cfrg-vdaf uses a \"fully linear\" IOP (for proving things about secret shared data)</p>\n</blockquote>\n<p>I guess the full difference is speciallized over secret-sharing schemes, rather than more general ones over full commitments as this presentation</p>\n</blockquote>\n<p>you can use our FL IOPs for encrypted or Pederson-committed data as well!</p>", "time": "2025-07-24T16:17:05Z"}, {"author": "Deirdre Connolly", "text": "<p>a generic proof system</p>", "time": "2025-07-24T16:17:13Z"}, {"author": "Deirdre Connolly", "text": "<p>not specific circuit</p>", "time": "2025-07-24T16:17:17Z"}, {"author": "Christopher Patton", "text": "<p>I would love to see if there is a lattice friendly version of this</p>", "time": "2025-07-24T16:17:21Z"}, {"author": "Uri Blumenthal", "text": "<p><span aria-label=\"+1\" class=\"emoji emoji-1f44d\" role=\"img\" title=\"+1\">:+1:</span></p>", "time": "2025-07-24T16:17:29Z"}, {"author": "Sofia Celi", "text": "<blockquote>\n<p>you can use our FL IOPs for encrypted or Pederson-committed data as well!</p>\n</blockquote>\n<p>yes, but in the draft is specified over secret-shared data, no?</p>", "time": "2025-07-24T16:17:33Z"}, {"author": "Sofia Celi", "text": "<p>(as far as I remember)</p>", "time": "2025-07-24T16:17:42Z"}, {"author": "Christopher Patton", "text": "<p>yes, the draft is specific.</p>", "time": "2025-07-24T16:17:52Z"}, {"author": "Sofia Celi", "text": "<p>and the security model is def different</p>", "time": "2025-07-24T16:17:52Z"}, {"author": "Simone Onofri", "text": "<p>I support longfellow, as it solves a lot of issues in the Identity space (and not only). And it is great that it is post-quantum. I should just adjust the framing of post-quantum results from non post-quantum credentials</p>", "time": "2025-07-24T16:18:02Z"}, {"author": "Stavros Kousidis", "text": "<p>Concerning PQ KEMs. I think the basic idea of e.g. draft-longa-cfrg-frodokem-00 is to pull that in front of the paywall. Same idea for Classic McEliece I would guess.</p>", "time": "2025-07-24T16:18:20Z"}, {"author": "Deirdre Connolly", "text": "<p>Nice and boring :D</p>", "time": "2025-07-24T16:18:49Z"}, {"author": "Daniel Gillmor", "text": "<p>boring is exciting \u2665</p>", "time": "2025-07-24T16:19:00Z"}, {"author": "Simone Onofri", "text": "<p>yes, boring but it solves a huge problem</p>", "time": "2025-07-24T16:19:10Z"}, {"author": "Simone Onofri", "text": "<p>:)</p>", "time": "2025-07-24T16:19:14Z"}, {"author": "Michele Orr\u00f9", "text": "<p>For the record, Ligero was published in 2022.</p>", "time": "2025-07-24T16:19:25Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"3903\">Stavros Kousidis</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177129\">said</a>:</p>\n<blockquote>\n<p>Concerning PQ KEMs. I think the basic idea of e.g. draft-longa-cfrg-frodokem-00 is to pull that in front of the paywall. Same idea for Classic McEliece I would guess.</p>\n</blockquote>\n<p>So, I think the general sentiment is that \"put in front of the paywall\" does not answer the question of \"Why CFRG? Why not Independent Stream?\".<br>\nIn my opinion, people expect a certain level of analysis when something comes through CFRG. That is not required to rubber-stamp an ISO spec.</p>", "time": "2025-07-24T16:19:33Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"5673\">Michele Orr\u00f9</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177146\">said</a>:</p>\n<blockquote>\n<p>For the record, Ligero was published in 2022.</p>\n</blockquote>\n<p>yep, the Google application is the recent thing</p>", "time": "2025-07-24T16:19:39Z"}, {"author": "Deirdre Connolly", "text": "<p>CAT</p>", "time": "2025-07-24T16:19:51Z"}, {"author": "Daniel Gillmor", "text": "<p>2022 is pretty recent though compared to the other elements</p>", "time": "2025-07-24T16:20:07Z"}, {"author": "Deirdre Connolly", "text": "<p>(or am I imagining cat?)</p>", "time": "2025-07-24T16:20:12Z"}, {"author": "Thom Wiggers", "text": "<p>:shown-cat:</p>", "time": "2025-07-24T16:20:18Z"}, {"author": "Deirdre Connolly", "text": "<p>no i am not! I think that is Bunn the Cat</p>", "time": "2025-07-24T16:20:30Z"}, {"author": "Deirdre Connolly", "text": "<p>a very good cat</p>", "time": "2025-07-24T16:20:32Z"}, {"author": "Richard Barnes", "text": "<p>She needs to introduce her cat</p>", "time": "2025-07-24T16:20:40Z"}, {"author": "Richard Barnes", "text": "<p>I don't make the rules</p>", "time": "2025-07-24T16:20:44Z"}, {"author": "Brent Zundel", "text": "<p>Strong +1</p>", "time": "2025-07-24T16:21:09Z"}, {"author": "Richard Barnes", "text": "<p>If there's time during Q&amp;A, i might ask</p>", "time": "2025-07-24T16:21:22Z"}, {"author": "Christopher Wood", "text": "<p>It is indeed Bunn</p>", "time": "2025-07-24T16:21:32Z"}, {"author": "Deirdre Connolly", "text": "<p>Yeah sigma protocols are old school including slapping a Fiat-Shamir on them</p>", "time": "2025-07-24T16:22:17Z"}, {"author": "Sofia Celi", "text": "<p>zkps are very interesting nowadays and very timely: they are being proposed quite heavily for digital credentials and age verification, and we need that they are specified correctly and in a way that really preserves the privacy</p>", "time": "2025-07-24T16:22:17Z"}, {"author": "Uri Blumenthal", "text": "<p>No, for stuff that\u2019s behind a paywall - I strongly object against CFRG doing <em>anything</em> with it.</p>", "time": "2025-07-24T16:22:21Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"849\">Uri Blumenthal</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177184\">said</a>:</p>\n<blockquote>\n<p>No, for stuff that\u2019s behind a paywall - I strongly object against CFRG doing <em>anything</em> with it.</p>\n</blockquote>\n<p>including doing their own version?</p>", "time": "2025-07-24T16:22:45Z"}, {"author": "Uri Blumenthal", "text": "<p>Correct. Unless there\u2019s an authoritative source elsewhere that IETF could adopt - we have no business touching it with a 10-foot pole.</p>", "time": "2025-07-24T16:23:36Z"}, {"author": "Sofia Celi", "text": "<p>@deidre sigma protocols are the best!! they come from complexity and logic foundations: they are soo amazing</p>", "time": "2025-07-24T16:23:37Z"}, {"author": "Watson Ladd", "text": "<p>NIST Round 4 should be approving stuff soonish right? we might not have to worry about everything on lattices</p>", "time": "2025-07-24T16:23:41Z"}, {"author": "Sofia Celi", "text": "<p>also a variant of signatures are indeed sigma ;)</p>", "time": "2025-07-24T16:23:54Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"2983\">Watson Ladd</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177199\">said</a>:</p>\n<blockquote>\n<p>NIST Round 4 should be approving stuff soonish right? we might not have to worry about everything on lattices</p>\n</blockquote>\n<p>Round Four finished, we're getting HQC</p>", "time": "2025-07-24T16:24:12Z"}, {"author": "Deirdre Connolly", "text": "<p>the signatures on ramp is continuing</p>", "time": "2025-07-24T16:24:17Z"}, {"author": "Uri Blumenthal", "text": "<p>@Watson, maybe. Let\u2019s wait and see. I for one hope for HAWK or such, but let\u2019s not hold our breath.</p>", "time": "2025-07-24T16:24:22Z"}, {"author": "Sofia Celi", "text": "<p>woo!! go, Cathie!! amazing!</p>", "time": "2025-07-24T16:24:28Z"}, {"author": "Deirdre Connolly", "text": "<p>They've selected ~14 signature candidates</p>", "time": "2025-07-24T16:24:36Z"}, {"author": "Sofia Celi", "text": "<p>the work of Lindsey on Tor is super amazing: def worth checking</p>", "time": "2025-07-24T16:25:03Z"}, {"author": "Quynh Dang", "text": "<p>many of those 14 have amazing stuff about zkp.</p>", "time": "2025-07-24T16:25:26Z"}, {"author": "Deirdre Connolly", "text": "<p>these are good updates</p>", "time": "2025-07-24T16:26:01Z"}, {"author": "Sofia Celi", "text": "<p>ah, yes, the fishlin technique can also be used instead of Fiat Shamir</p>", "time": "2025-07-24T16:26:06Z"}, {"author": "David Waite", "text": "<p>I thought I was the only one confused</p>", "time": "2025-07-24T16:26:15Z"}, {"author": "Deirdre Connolly", "text": "<p>p256 is also widely available, p384 less so</p>", "time": "2025-07-24T16:26:57Z"}, {"author": "Sofia Celi", "text": "<p>+1 to not including the compressed part, maybe as discussion</p>", "time": "2025-07-24T16:27:58Z"}, {"author": "Rich Salz", "text": "<p>Very impressive set of responses to the initial call feedback</p>", "time": "2025-07-24T16:28:14Z"}, {"author": "Deirdre Connolly", "text": "<p>I'll be keeping <a href=\"https://eprint.iacr.org/2025/118.pdf\">https://eprint.iacr.org/2025/118.pdf</a> in my pocket as I read these docs</p>", "time": "2025-07-24T16:28:46Z"}, {"author": "Daniel Gillmor", "text": "<p>this is the clearest summary+response i've ever seen to a call for adoption.</p>", "time": "2025-07-24T16:29:29Z"}, {"author": "Robert Moskowitz", "text": "<p>agree.</p>", "time": "2025-07-24T16:29:33Z"}, {"author": "Sofia Celi", "text": "<p>@deirdre: don't we all love negative results? HAHAH</p>", "time": "2025-07-24T16:29:35Z"}, {"author": "Deirdre Connolly", "text": "<p><span aria-label=\"nod\" class=\"emoji emoji-1f642-200d-2195\" role=\"img\" title=\"nod\">:nod:</span> <span aria-label=\"nod\" class=\"emoji emoji-1f642-200d-2195\" role=\"img\" title=\"nod\">:nod:</span> <span aria-label=\"nod\" class=\"emoji emoji-1f642-200d-2195\" role=\"img\" title=\"nod\">:nod:</span></p>", "time": "2025-07-24T16:29:37Z"}, {"author": "Christopher Patton", "text": "<p><span class=\"user-mention silent\" data-user-id=\"637\">Daniel Gillmor</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177251\">said</a>:</p>\n<blockquote>\n<p>this is the clearest summary+response i've ever seen to a call for adoption.</p>\n</blockquote>\n<p>it's kind of beautiful :)</p>", "time": "2025-07-24T16:29:47Z"}, {"author": "Uri Blumenthal", "text": "<p>Composability - certainly yes, if possible.</p>", "time": "2025-07-24T16:30:36Z"}, {"author": "Christopher Wood", "text": "<p>@Uri why?</p>", "time": "2025-07-24T16:30:45Z"}, {"author": "Richard Barnes", "text": "<p>Strong preference for the simplest set of mechanisms to address the current use cases.  You can always add complexity later.</p>", "time": "2025-07-24T16:31:25Z"}, {"author": "Daniel Gillmor", "text": "<p>Barnes++</p>", "time": "2025-07-24T16:31:36Z"}, {"author": "Uri Blumenthal", "text": "<p>@Chris, tends to be useful, eventually if not immediately</p>", "time": "2025-07-24T16:32:01Z"}, {"author": "Stavros Kousidis", "text": "<p>@Mike Ounsworth: I can imagine \"Why CFRG?\" stems from the believe \"CFRG is the authority for that stuff.\".</p>", "time": "2025-07-24T16:32:17Z"}, {"author": "Richard Barnes", "text": "<p><span class=\"user-mention\" data-user-id=\"849\">@Uri Blumenthal</span> seems pretty hand-wavy</p>", "time": "2025-07-24T16:32:20Z"}, {"author": "Christopher Patton", "text": "<p>but can't the composition of the proofs be expressed within the same proof?</p>", "time": "2025-07-24T16:32:24Z"}, {"author": "Daniel Gillmor", "text": "<p>@Uri can we add it later then, when it becomes necessary?</p>", "time": "2025-07-24T16:32:31Z"}, {"author": "Christopher Patton", "text": "<p>(naive question)</p>", "time": "2025-07-24T16:32:37Z"}, {"author": "Sofia Celi", "text": "<p>composition is hard: the majority of attacks have been on it</p>", "time": "2025-07-24T16:32:41Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"3903\">Stavros Kousidis</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177276\">said</a>:</p>\n<blockquote>\n<p>@Mike Ounsworth: I can imagine \"Why CFRG?\" stems from the believe \"CFRG is the authority for that stuff.\".</p>\n</blockquote>\n<p>\"They have 'Crypto' in their name\"</p>", "time": "2025-07-24T16:32:44Z"}, {"author": "Sofia Celi", "text": "<p>both implementation and security</p>", "time": "2025-07-24T16:32:49Z"}, {"author": "Christopher Patton", "text": "<p>thanks, sofi!</p>", "time": "2025-07-24T16:34:20Z"}, {"author": "Sofia Celi", "text": "<p>yes!!! brave zkps aahah</p>", "time": "2025-07-24T16:34:48Z"}, {"author": "Stavros Kousidis", "text": "<p>@deirdre: and 'Forum'</p>", "time": "2025-07-24T16:35:19Z"}, {"author": "Sofia Celi", "text": "<p>&lt;3, cris ;)</p>", "time": "2025-07-24T16:35:20Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention\" data-user-id=\"603\">@Deirdre Connolly</span>  ASK ABOUT CAT</p>", "time": "2025-07-24T16:35:49Z"}, {"author": "Sofia Celi", "text": "<p>ahhh, datatracker sometimes malfunctions on it</p>", "time": "2025-07-24T16:36:10Z"}, {"author": "Christopher Wood", "text": "<p>Leaving the queue. I wanted to say that this is great, but please do not over generalize this work unnecessarily. Let's please focus on the use cases in front of us.</p>", "time": "2025-07-24T16:36:14Z"}, {"author": "Cathie Yun", "text": "<p><a href=\"https://datatracker.ietf.org/doc/draft-orru-zkproof-sigma-protocols/\">https://datatracker.ietf.org/doc/draft-orru-zkproof-sigma-protocols/</a></p>", "time": "2025-07-24T16:36:28Z"}, {"author": "Brent Zundel", "text": "<p>We would like to see the cat as well</p>", "time": "2025-07-24T16:36:44Z"}, {"author": "Christopher Patton", "text": "<p>scope creep is something to worry about here</p>", "time": "2025-07-24T16:36:48Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"5873\">Mike Ounsworth</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177308\">said</a>:</p>\n<blockquote>\n<p><span class=\"user-mention silent\" data-user-id=\"603\">Deirdre Connolly</span>  ASK ABOUT CAT</p>\n</blockquote>\n<p>BAH I'M SORRY</p>", "time": "2025-07-24T16:37:01Z"}, {"author": "Christopher Patton", "text": "<p><span class=\"user-mention\" data-user-id=\"2103\">@Dennis Jackson</span> cat</p>", "time": "2025-07-24T16:37:04Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"6359\">Cathie Yun</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177314\">said</a>:</p>\n<blockquote>\n<p><a href=\"https://datatracker.ietf.org/doc/draft-orru-zkproof-sigma-protocols/\">https://datatracker.ietf.org/doc/draft-orru-zkproof-sigma-protocols/</a></p>\n</blockquote>\n<p>Awesome thanks, yeah it's not showing up because it doesn't have -cfrg- in the slug</p>", "time": "2025-07-24T16:37:30Z"}, {"author": "Thom Wiggers", "text": "<p>and/or the metadata seems to say \"networking working group\"</p>", "time": "2025-07-24T16:37:54Z"}, {"author": "Rich Salz", "text": "<p>no, the RG chairs can add it to the list</p>", "time": "2025-07-24T16:37:57Z"}, {"author": "Richard Barnes", "text": "<p>I missed it, what is Rocca-S ?  a hash function?</p>", "time": "2025-07-24T16:38:03Z"}, {"author": "Richard Barnes", "text": "<p>An AEAD?</p>", "time": "2025-07-24T16:38:13Z"}, {"author": "Daniel Gillmor", "text": "<p>it's fine that it doesn't have -cfrg- in the slug, fwiw.  there is no expectation that it needs to.</p>", "time": "2025-07-24T16:38:14Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention\" data-user-id=\"6359\">@Cathie Yun</span> pics of Bunn? <span aria-label=\"pleading face\" class=\"emoji emoji-1f97a\" role=\"img\" title=\"pleading face\">:pleading_face:</span></p>", "time": "2025-07-24T16:38:20Z"}, {"author": "Thom Wiggers", "text": "<p><a href=\"https://datatracker.ietf.org/doc/draft-nakano-rocca-s/\">https://datatracker.ietf.org/doc/draft-nakano-rocca-s/</a></p>", "time": "2025-07-24T16:38:32Z"}, {"author": "Thom Wiggers", "text": "<p>I guess</p>", "time": "2025-07-24T16:38:36Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"86\">Thom Wiggers</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177333\">said</a>:</p>\n<blockquote>\n<p>and/or the metadata seems to say \"networking working group\"</p>\n</blockquote>\n<p>That's the default in the MT template, if you don't change it, right?</p>", "time": "2025-07-24T16:38:40Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"637\">Daniel Gillmor</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177339\">said</a>:</p>\n<blockquote>\n<p>it's fine that it doesn't have -cfrg- in the slug, fwiw.  there is no expectation that it needs to.</p>\n</blockquote>\n<p>It is fine, I think the auto-detection in datatracker does match on it</p>", "time": "2025-07-24T16:38:40Z"}, {"author": "Richard Barnes", "text": "<p>For the record, the presence of the cat  definitely helped me pay good attention to the preso</p>", "time": "2025-07-24T16:38:48Z"}, {"author": "Valery Smyslov", "text": "<p>This is unfair to other presenters :-)</p>", "time": "2025-07-24T16:39:32Z"}, {"author": "Richard Barnes", "text": "<p>Definitely noticing the lack of feline content in this presentation</p>", "time": "2025-07-24T16:39:37Z"}, {"author": "Frank Denis", "text": "<p>These performance numbers are for encryption. Unfortunately, ROCCA-S decryption is very slow <a href=\"https://github.com/jedisct1/zig-rocca-s?tab=readme-ov-file#benchmark-results\">https://github.com/jedisct1/zig-rocca-s?tab=readme-ov-file#benchmark-results</a></p>", "time": "2025-07-24T16:39:37Z"}, {"author": "Daniel Gillmor", "text": "<p>Bunn didn't even register for the IETF</p>", "time": "2025-07-24T16:39:44Z"}, {"author": "Mike Ounsworth", "text": "<p>Following on the \"Why should CFRG do NTRU / Classic McEliece / FrodoKEM\" discussion.<br>\nSame question for ROCCA-S and HiAE.</p>", "time": "2025-07-24T16:39:45Z"}, {"author": "Thom Wiggers", "text": "<p>I know of a few protocols PhDs walking around but symmetric cryptography PhDs ....</p>", "time": "2025-07-24T16:39:55Z"}, {"author": "David Waite", "text": "<p>I forgot the WG slug my first time submitting too</p>", "time": "2025-07-24T16:40:11Z"}, {"author": "Uri Blumenthal", "text": "<p>Please remind who did the evaluation of ROCCA-S?</p>", "time": "2025-07-24T16:40:20Z"}, {"author": "Thom Wiggers", "text": "<p><a href=\"https://link.springer.com/chapter/10.1007/978-3-031-50594-2_12\">https://link.springer.com/chapter/10.1007/978-3-031-50594-2_12</a></p>", "time": "2025-07-24T16:41:03Z"}, {"author": "Rich Salz", "text": "<p>Yes, you <em>can</em> set up search rules (Regexp) pattern to automatically sieve out documents.  But the Chairs can manually add things.</p>", "time": "2025-07-24T16:41:21Z"}, {"author": "Thom Wiggers", "text": "<p>but I don't think ESORICS, as a security venue, is necessarily the best qualified in the review process...</p>", "time": "2025-07-24T16:41:39Z"}, {"author": "Deirdre Connolly", "text": "<p>if we're going to tweak this existing AEAD what about context commiting?</p>", "time": "2025-07-24T16:43:23Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"603\">Deirdre Connolly</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177392\">said</a>:</p>\n<blockquote>\n<p>if we're going to tweak this existing AEAD what about context commiting?</p>\n</blockquote>\n<p>I'm not well-versed on symmetric, but is \"context committing\" the same as \"with Additional Data\"?</p>", "time": "2025-07-24T16:44:21Z"}, {"author": "Frank Denis", "text": "<p>There's also a chacha20-poly1305-siv proposal that is context committing <a href=\"https://github.com/C2SP/C2SP/blob/main/chacha20-poly1305-siv.md\">https://github.com/C2SP/C2SP/blob/main/chacha20-poly1305-siv.md</a></p>", "time": "2025-07-24T16:44:59Z"}, {"author": "Uri Blumenthal", "text": "<p>@Thom, thank you. And yes, it\u2019s nice being published - but I\u2019d like to know what cryptanalysts (persons or organizations) evaluated it.</p>", "time": "2025-07-24T16:45:22Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"5873\">Mike Ounsworth</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177400\">said</a>:</p>\n<blockquote>\n<p><span class=\"user-mention silent\" data-user-id=\"603\">Deirdre Connolly</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177392\">said</a>:</p>\n<blockquote>\n<p>if we're going to tweak this existing AEAD what about context commiting?</p>\n</blockquote>\n<p>I'm not well-versed on symmetric, but as context commiting the same as Additional Data?</p>\n</blockquote>\n<p>no the design of a context-commiting AEAD 'binds' to the AD, and the IV, and the key</p>", "time": "2025-07-24T16:45:41Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"3094\">Frank Denis</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177405\">said</a>:</p>\n<blockquote>\n<p>There's also a chacha20-poly1305-siv proposal that is context committing <a href=\"https://github.com/C2SP/C2SP/blob/main/chacha20-poly1305-siv.md\">https://github.com/C2SP/C2SP/blob/main/chacha20-poly1305-siv.md</a></p>\n</blockquote>\n<p>yeah I think this is valuable, unfortunately that's not this proposal</p>", "time": "2025-07-24T16:46:45Z"}, {"author": "Deirdre Connolly", "text": "<p>can hear you over here</p>", "time": "2025-07-24T16:47:41Z"}, {"author": "Deirdre Connolly", "text": "<p><span aria-label=\"sob\" class=\"emoji emoji-1f62d\" role=\"img\" title=\"sob\">:sob:</span></p>", "time": "2025-07-24T16:48:12Z"}, {"author": "Deirdre Connolly", "text": "<p>pet Gus</p>", "time": "2025-07-24T16:48:18Z"}, {"author": "Christopher Patton", "text": "<p>GUS :(</p>", "time": "2025-07-24T16:48:18Z"}, {"author": "Richard Barnes", "text": "<p>Gus is Chris's french bulldog, for those unfamiliar</p>", "time": "2025-07-24T16:48:38Z"}, {"author": "Rich Salz", "text": "<p>Gus &gt;&gt; Bunn.  Fight me.  :)</p>", "time": "2025-07-24T16:49:22Z"}, {"author": "Cathie Yun", "text": "<p>Can Gus be the PAKE mascot</p>", "time": "2025-07-24T16:49:31Z"}, {"author": "Yuto Nakano", "text": "<p><span class=\"user-mention silent\" data-user-id=\"849\">Uri Blumenthal</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177373\">said</a>:</p>\n<blockquote>\n<p>Please remind who did the evaluation of ROCCA-S?</p>\n</blockquote>\n<p>Done by Indian Institute of Technology\u00a0Madras and University of Rennes 1</p>", "time": "2025-07-24T16:49:34Z"}, {"author": "Daniel Gillmor", "text": "<p>@Rich: by mass or by cuteness?</p>", "time": "2025-07-24T16:49:37Z"}, {"author": "Richard Barnes", "text": "<p><span class=\"user-mention\" data-user-id=\"11\">@Rich Salz</span> no need to have a competition here</p>", "time": "2025-07-24T16:49:42Z"}, {"author": "Christopher Patton", "text": "<p><span class=\"user-mention silent\" data-user-id=\"6359\">Cathie Yun</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177441\">said</a>:</p>\n<blockquote>\n<p>Can Gus be the PAKE mascot</p>\n</blockquote>\n<p>I think the chairs should call for a hum</p>", "time": "2025-07-24T16:49:57Z"}, {"author": "Richard Barnes", "text": "<p>all animals are welcome, even you <span class=\"user-mention\" data-user-id=\"11\">@Rich Salz</span></p>", "time": "2025-07-24T16:49:57Z"}, {"author": "Cathie Yun", "text": "<p>Gus is definitely denser and intenser, they may meet eventually... will keep the group posted</p>", "time": "2025-07-24T16:50:01Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"11\">Rich Salz</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177440\">said</a>:</p>\n<blockquote>\n<p>Gus &gt;&gt; Bunn.  Fight me.  :)</p>\n</blockquote>\n<p>Please submit benchmarks.</p>", "time": "2025-07-24T16:50:04Z"}, {"author": "Uri Blumenthal", "text": "<blockquote>\n<p>Done by Indian Institute of Technology Madras and University of Rennes</p>\n</blockquote>", "time": "2025-07-24T16:50:26Z"}, {"author": "Rich Salz", "text": "<p>We need to change the RFC template to add an optional  GIF of the animal co-author</p>", "time": "2025-07-24T16:50:39Z"}, {"author": "Deirdre Connolly", "text": "<p><a href=\"https://eprint.iacr.org/2025/1343\">https://eprint.iacr.org/2025/1343</a></p>", "time": "2025-07-24T16:50:39Z"}, {"author": "Deirdre Connolly", "text": "<p><span class=\"user-mention silent\" data-user-id=\"11\">Rich Salz</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177454\">said</a>:</p>\n<blockquote>\n<p>We need to change the RFC template to add an optional  GIF of the animal co-author</p>\n</blockquote>\n<p>I'll have to pick a corgi</p>", "time": "2025-07-24T16:50:57Z"}, {"author": "Uri Blumenthal", "text": "<p>Are they known for cryptanalytic skills?</p>", "time": "2025-07-24T16:50:58Z"}, {"author": "Richard Barnes", "text": "<p><span class=\"user-mention\" data-user-id=\"603\">@Deirdre Connolly</span> i think we can design it to accommodate multiple corgis.  they're small</p>", "time": "2025-07-24T16:51:22Z"}, {"author": "Rich Salz", "text": "<p><a href=\"https://eprint.iacr.org/2025/1237.pdf\">https://eprint.iacr.org/2025/1237.pdf</a> where Peter's dog did PQ factoring</p>", "time": "2025-07-24T16:51:24Z"}, {"author": "Deirdre Connolly", "text": "<p>Oh, Quake!</p>", "time": "2025-07-24T16:51:50Z"}, {"author": "Deirdre Connolly", "text": "<p><span aria-label=\"shaking face\" class=\"emoji emoji-1fae8\" role=\"img\" title=\"shaking face\">:shaking_face:</span></p>", "time": "2025-07-24T16:51:58Z"}, {"author": "Yuto Nakano", "text": "<p><span class=\"user-mention silent\" data-user-id=\"3094\">Frank Denis</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177360\">said</a>:</p>\n<blockquote>\n<p>These performance numbers are for encryption. Unfortunately, ROCCA-S decryption is very slow <a href=\"https://github.com/jedisct1/zig-rocca-s?tab=readme-ov-file#benchmark-results\">https://github.com/jedisct1/zig-rocca-s?tab=readme-ov-file#benchmark-results</a></p>\n</blockquote>\n<p>Thank you for this information. We can optimise decryption.</p>", "time": "2025-07-24T16:52:01Z"}, {"author": "Dan Harkins", "text": "<p>it's PQ OEKE</p>", "time": "2025-07-24T16:52:29Z"}, {"author": "Yuto Nakano", "text": "<p><span class=\"user-mention silent\" data-user-id=\"526\">Richard Barnes</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177335\">said</a>:</p>\n<blockquote>\n<p>I missed it, what is Rocca-S ?  a hash function?</p>\n</blockquote>\n<p>It is an AEAD</p>", "time": "2025-07-24T16:52:55Z"}, {"author": "Deirdre Connolly", "text": "<p>ooo Kemeleon</p>", "time": "2025-07-24T16:53:05Z"}, {"author": "Frank Denis", "text": "<p>By the way ROCCA-SX2 performs pretty well on ARM.</p>", "time": "2025-07-24T16:53:08Z"}, {"author": "Deirdre Connolly", "text": "<p><a href=\"/user_uploads/2/1d/rntxUpdt-VGGCjbtN9GMSizL/image.png\">image.png</a></p>\n<div class=\"message_inline_image\"><a href=\"/user_uploads/2/1d/rntxUpdt-VGGCjbtN9GMSizL/image.png\" title=\"image.png\"><img data-original-content-type=\"image/png\" data-original-dimensions=\"1010x912\" src=\"/user_uploads/thumbnail/2/1d/rntxUpdt-VGGCjbtN9GMSizL/image.png/840x560.webp\"></a></div>", "time": "2025-07-24T16:53:30Z"}, {"author": "Deirdre Connolly", "text": "<p>that's an interesting newish property</p>", "time": "2025-07-24T16:53:57Z"}, {"author": "Dennis Jackson", "text": "<p>NoIC should be pronounced 'Noice'?</p>", "time": "2025-07-24T16:54:20Z"}, {"author": "Uri Blumenthal", "text": "<p>In ROM, but not QROM - probably not critical?</p>", "time": "2025-07-24T16:54:24Z"}, {"author": "Renzo Navas", "text": "<p>@Uri, Rennes  / INRIA has many cryptographers <a href=\"https://team.inria.fr/capsule/\">https://team.inria.fr/capsule/</a> INRIA creators of CryptoVerif (and others I think), co authors of FALCON</p>", "time": "2025-07-24T16:54:31Z"}, {"author": "Dan Harkins", "text": "<p>ccccccvcevkeultfnullguvlbkdevbkfutndikincdkv</p>", "time": "2025-07-24T16:55:03Z"}, {"author": "Christopher Patton", "text": "<p><span class=\"user-mention silent\" data-user-id=\"849\">Uri Blumenthal</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177495\">said</a>:</p>\n<blockquote>\n<p>In ROM, but not QROM - probably not critical?</p>\n</blockquote>\n<p>My understanding is hat ROM \\approx QROM modulo some factor that matters for concrete security</p>", "time": "2025-07-24T16:55:03Z"}, {"author": "Dan Harkins", "text": "<p>oops, damned yubikey</p>", "time": "2025-07-24T16:55:14Z"}, {"author": "Richard Barnes", "text": "<p><span class=\"user-mention\" data-user-id=\"1661\">@Dan Harkins</span> you really should upgrade to WebAuthn</p>", "time": "2025-07-24T16:55:31Z"}, {"author": "Christopher Patton", "text": "<p>\"CPaceOQUAKE\" is pronounced \"space quake\"? This is getting ridiculous gang</p>", "time": "2025-07-24T16:55:42Z"}, {"author": "Thom Wiggers", "text": "<p><span class=\"user-mention\" data-user-id=\"3787\">@Christopher Patton</span> ehh there's stuff that we don't know how to prove in the QROM</p>", "time": "2025-07-24T16:55:44Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"1661\">Dan Harkins</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177501\">said</a>:</p>\n<blockquote>\n<p>oops, damned yubikey</p>\n</blockquote>\n<p>What were you trying to log into? #replay ?</p>", "time": "2025-07-24T16:55:46Z"}, {"author": "Bas Westerbaan", "text": "<p><span class=\"user-mention silent\" data-user-id=\"1661\">Dan Harkins</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177499\">said</a>:</p>\n<blockquote>\n<p>ccccccvcevkeultfnullguvlbkdevbkfutndikincdkv</p>\n</blockquote>\n<p>This messages was brought to you by Yubico.</p>", "time": "2025-07-24T16:55:48Z"}, {"author": "Rich Salz", "text": "<p><a href=\"https://www.reddit.com/r/vintageads/comments/18d3xaw/quisp_vs_quake_breakfast_cereal_tv_commercial/\">https://www.reddit.com/r/vintageads/comments/18d3xaw/quisp_vs_quake_breakfast_cereal_tv_commercial/</a></p>", "time": "2025-07-24T16:55:54Z"}, {"author": "Dan Harkins", "text": "<p><span aria-label=\"+1\" class=\"emoji emoji-1f44d\" role=\"img\" title=\"+1\">:+1:</span></p>", "time": "2025-07-24T16:55:55Z"}, {"author": "David Waite", "text": "<p>There\u2019s a lot of options in the QUAKE Arena</p>", "time": "2025-07-24T16:56:51Z"}, {"author": "Cathie Yun", "text": "<p>We're making Space Quake happen, gang</p>", "time": "2025-07-24T16:56:52Z"}, {"author": "Mike Ounsworth", "text": "<p><span class=\"user-mention silent\" data-user-id=\"3787\">Christopher Patton</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177508\">said</a>:</p>\n<blockquote>\n<p>\"CPaceOQUAKE\" is pronounced \"space quake\"? This is getting ridiculous gang</p>\n</blockquote>\n<p>Still not worse than the ICAO \"PACE\" pronounced \"Pach\u00e9\".</p>", "time": "2025-07-24T16:56:56Z"}, {"author": "Daniel Gillmor", "text": "<p>that's just Latin, Mike</p>", "time": "2025-07-24T16:57:12Z"}, {"author": "Uri Blumenthal", "text": "<p>@Renzo, impressive, thanks. Any publication from them on actually breaking something? Or are they Defense- related, so not about to brag about successes?</p>", "time": "2025-07-24T16:57:59Z"}, {"author": "Robert Moskowitz", "text": "<p>@Mike, be aware the time it takes ICAO to agree on something.   I am living this for the past 5 years...</p>", "time": "2025-07-24T16:59:37Z"}, {"author": "Dan Harkins", "text": "<p>non hybrid version, yes</p>", "time": "2025-07-24T17:00:02Z"}, {"author": "Deirdre Connolly", "text": "<p>Yes a pure PQ PAKE</p>", "time": "2025-07-24T17:00:20Z"}, {"author": "Uri Blumenthal", "text": "<p>@Dan &amp; @Deirdre, +1. And worth doing here (CFRG).</p>", "time": "2025-07-24T17:01:12Z"}, {"author": "Dan Harkins", "text": "<p>I've implemented it, I can provide test vectors!</p>", "time": "2025-07-24T17:01:29Z"}, {"author": "Yuto Nakano", "text": "<p><span class=\"user-mention silent\" data-user-id=\"849\">Uri Blumenthal</span> <a href=\"#narrow/channel/267-cfrg/topic/ietf-123/near/177458\">said</a>:</p>\n<blockquote>\n<p>Are they known for cryptanalytic skills?</p>\n</blockquote>\n<p>I assume this is to me. Yes, they are well-known in symmetric crypto.</p>", "time": "2025-07-24T17:02:19Z"}, {"author": "Dennis Jackson", "text": "<p>This looks good. I think pure would be ideal. Hybrid also seems like a riskier bet in terms of side channels (and with a PAKE we have a long term, low entropy secret to worry about).</p>", "time": "2025-07-24T17:02:35Z"}, {"author": "Renzo Navas", "text": "<p>@Uri, sorry I do not know (I think they are mostly designers so mostly defensive), I am not cryptographer... but I can ask them, as I am also from Rennes ans have them across the street :). Anyway if the ESORICS is the only things that has only peer reviewed by a handful of people not a proper analysy</p>", "time": "2025-07-24T17:02:52Z"}, {"author": "Deirdre Connolly", "text": "<p><span aria-label=\"wave\" class=\"emoji emoji-1f44b\" role=\"img\" title=\"wave\">:wave:</span></p>", "time": "2025-07-24T17:02:58Z"}, {"author": "Renzo Navas", "text": "<p>bye!</p>", "time": "2025-07-24T17:03:00Z"}, {"author": "Nick Sullivan", "text": "<p>Thanks, everyone!</p>", "time": "2025-07-24T17:03:16Z"}, {"author": "Nick Sullivan", "text": "<p>Especially Scott for being the scribe!</p>", "time": "2025-07-24T17:03:35Z"}]