Post AwSdS4fKHZrnyPXn60 by artlog@agora.l0g.eu
(DIR) More posts by artlog@agora.l0g.eu
(DIR) Post #AwSLUMhIPh8nInPMci by bortzmeyer@mastodon.gougere.fr
2025-07-24T10:57:14Z
0 likes, 0 repeats
Good morning, Madrid! (Yes, I'm late, I was walking the Retiro park.) Fourth day of #IETF123 https://datatracker.ietf.org/meeting/123/Today, for me, post-quantum crypto talk, and formal methods
(DIR) Post #AwSMbuVhCDyc9V8zCK by bortzmeyer@mastodon.gougere.fr
2025-07-24T11:09:51Z
0 likes, 0 repeats
Talk "Migrating Telecom to Quantum-Resistant Cryptography on a Global Scale" by John Preuß Mattsson#IETF123 #PQcrypto
(DIR) Post #AwSOB6HXWNCGkcrlke by bortzmeyer@mastodon.gougere.fr
2025-07-24T11:27:25Z
0 likes, 0 repeats
Reminder that CRQC (really useful quantum computers) are far away in the future (the speaker does not expect them before 15-30 years). But government regulations ask for a migration before, even if useless.#IETF123
(DIR) Post #AwSP3PqN0Ufd3g3czw by bortzmeyer@mastodon.gougere.fr
2025-07-24T11:37:14Z
0 likes, 0 repeats
Reminder that QKD (Quantum Key DIstribution) is only good for making press releases (don't believe the marketing and the hype).Quantum random number generation is not mature.Quantum sensors are cool. (Very expensive so probably only for the military.)The rest of the talk will focus on quantum-resistant crypto, not n quantum devices.#IETF123
(DIR) Post #AwSPmHacXf5WUS5Kgi by bortzmeyer@mastodon.gougere.fr
2025-07-24T11:45:20Z
0 likes, 0 repeats
The focus of the talk is actual deployement of quantum-resistant crypto in actual mobile operators' networks. 5G depends a lot on IETF standards using crypto (such as TLS 1.3 but also lesser-known IETF standards). Lot of places to upgrade!There are also non-standard but very important uses of cryptography such as securing firmware updates for 5G nodes.6G will be (promise!) fully quantum-resistant from the start (2027? 2028?)#IETF123
(DIR) Post #AwSQq8FBKazCcKSMdM by bortzmeyer@mastodon.gougere.fr
2025-07-24T11:57:14Z
0 likes, 0 repeats
Saying "we must be quantum-resistant" is one thing, deciding what algorithms to use is something else. Specially in a highly regulated industry like the telecoms, and with the desire to be used world-wide with as few changes as possible.An example: hybridization or not? #IETF123
(DIR) Post #AwSUFkdxzyUCoWxPmK by bortzmeyer@mastodon.gougere.fr
2025-07-24T12:35:29Z
0 likes, 0 repeats
A bit of formality, now. "Usable Formal Methods Research Group" https://datatracker.ietf.org/meeting/123/materials/agenda-123-ufmrg-00#IETF123
(DIR) Post #AwSbnVYjHvBRYUYESm by bortzmeyer@mastodon.gougere.fr
2025-07-24T14:00:00Z
0 likes, 0 repeats
Presentation of Microsoft's EverParse system, which generates safe parsers from formal and validated specifications.#IETF123
(DIR) Post #AwSc4M9hVJ5FJTsV28 by bortzmeyer@mastodon.gougere.fr
2025-07-24T14:03:04Z
0 likes, 0 repeats
Many Internet protocols are already specified in EverParse (but not #DNS, alas).#IETF123
(DIR) Post #AwSciAop4aQ5ji6Bm4 by bortzmeyer@mastodon.gougere.fr
2025-07-24T14:10:15Z
0 likes, 0 repeats
Discussion in the group about writing all RFCs as just formal specification language and then use a LLM to generate a natural language description.#IETF123 #AI #RFC
(DIR) Post #AwSdS4fKHZrnyPXn60 by artlog@agora.l0g.eu
2025-07-24T14:14:57.262213Z
0 likes, 0 repeats
@bortzmeyer The case of rfc of for cookies come to my mind, where if i have well understood formal specification by example indicate SameSite while natural language description explains it can be mixed casing like sAmeSIte
(DIR) Post #AwSdS5qLu4HTcti6vA by bortzmeyer@mastodon.gougere.fr
2025-07-24T14:18:18Z
0 likes, 0 repeats
@artlog Another example is the bug I mentioned: https://www.rfc-editor.org/errata/eid8305