Tom Hill - 2025-10-20 15:39:56
No-one mention YANG Push
Antonio Prado - 2025-10-20 16:03:41
Hello, for questions use the Q&A button :)
Shane Kerr - 2025-10-20 16:05:53
"draft standard" in the IETF means that it is an actual RFC and can be implemented... it's not quite as "drafty" as in other contexts...
Shane Kerr - 2025-10-20 16:06:43
Oh, maybe on that graph "draft" meant Internet draft? Which isn't an RFC. Hm.
Shane Kerr - 2025-10-20 16:15:01
"Est. Verification CPU Time (s)"... 3729.5 for SHAKE-128f!!!
Shane Kerr - 2025-10-20 16:15:48
838MB expands to 1.4 GB. Oef.
Shane Kerr - 2025-10-20 16:15:57
For FALCON, that is.
Shane Kerr - 2025-10-20 16:18:26
Quantum computing means you need to double the size of hashes right? So maybe SHA512 instead of SHA256?
Antonio Prado - 2025-10-20 16:18:56
😂
Shane Kerr - 2025-10-20 16:19:09
I'm not sure if that matters, since the difference is only 32 bytes per record...
Marco d'Itri - 2025-10-20 16:20:11
I get it, he had to pick some thesis work, but this is totally pointless for short lived signatures
Shane Kerr - 2025-10-20 16:20:51
How long do RPKI signatures live?
Shane Kerr - 2025-10-20 16:21:36
NIST will basically disallow algorithms that are no post-quantum safe in 10 years.
Marco d'Itri - 2025-10-20 16:21:37
weeks? the upper level certificates are longer, but it's not like we will have actually useful quantum computers next year
Shane Kerr - 2025-10-20 16:21:57
And because it takes so long to change software and so on, probably it is important to start now.
Marco d'Itri - 2025-10-20 16:22:06
NIST has to deal with store now and decrypt later scenarios, which are not relevant for either DNSSEC or RPKI
Marco d'Itri - 2025-10-20 16:22:27
it takes long for DNS, not for RPKI (RPKI validators are a tiny number compared to DNS resolvers)
Marco d'Itri - 2025-10-20 16:22:53
which I think is the major flaw of this work: validators will be replaced much faster when needed
Shane Kerr - 2025-10-20 16:23:15
Fair. So RPKI should let the PQC technology mature before worrying about it?
Marco d'Itri - 2025-10-20 16:23:23
that's my opinion :-0
Marco d'Itri - 2025-10-20 16:23:27
:-)
Antonio Prado - 2025-10-20 16:24:18
any comments, questions for Dirk?
Marco d'Itri - 2025-10-20 16:25:20
me, I will go to the mike
Antonio Prado - 2025-10-20 16:26:02
great
Shane Kerr - 2025-10-20 16:33:51
Finally a DNS presentation!
Shane Kerr - 2025-10-20 16:34:00
😉
Franziska Lichtblau - 2025-10-20 16:34:27
That's what we've ALL been waiting for!!!!
Antonio Prado - 2025-10-20 16:35:04
it's a bgp presentation!
Shane Kerr - 2025-10-20 16:35:50
"It's always ___" A: BGP B: DNS
Antonio Prado - 2025-10-20 16:36:06
hehe, you're right :)
Franziska Lichtblau - 2025-10-20 16:38:50
C: RPKI
Sascha Lenz - 2025-10-20 16:49:26
wrong, it's always MTU 🥸
Antonio Prado - 2025-10-20 16:49:31
any questions for Lefteris?
Franziska Lichtblau - 2025-10-20 16:58:14
We are very on time, so come ask questions to our LT speakers :)
Marcus Gerdon - 2025-10-20 17:02:26
Always been wondering how people differentiate between congested and faulty from the outside.
Shane Kerr - 2025-10-20 17:08:17
If you're measuring user experience, congested vs. faulty does not matter.
Shane Kerr - 2025-10-20 17:08:30
If you're measuring YOUR infrastructure from the user point of view, it matters a lot.
Antonio Prado - 2025-10-20 17:15:11
any questions for Christoff?
Marcus Gerdon - 2025-10-20 17:17:58
Agree. I was more triggered by again and again seeing matter-of-fact "congestion" statements being made by research publications based purely on remote dataplane samples. Makes me go "yeah, always blame someone but dont raise questions on own methodology".
Shane Kerr - 2025-10-20 17:19:51
I'm confused. Did the Secret WG move to Monday? 😉
Franziska Lichtblau - 2025-10-20 17:20:03
I told him to not use his funny voice :P
Antonio Prado - 2025-10-20 17:20:25
that's DNA
Simone Basso - 2025-10-20 17:20:52
Marcus, Shane: part of the answer is understanding and differentiating can be found in the interconnection study methodology (https://www.measurementlab.net/publications/M-Lab_Interconnection_Study_US.pdf) and also there is a link in the slides I presented that lands to Matt’s presentation which also discusses some of these topics.
Marcus Gerdon - 2025-10-20 17:22:32
Thanks for th elink. I'll have a look for sure. :)
Michael Richardson - 2025-10-20 17:24:36
"single gateway" implies the existence of "married gateways", and/or dating sites for gateways :-)
Antonio Prado - 2025-10-20 17:26:47
any questions for Remco?
Antonio Prado - 2025-10-20 17:29:12
single gateway? don’t worry, it’s just waiting for VRRP to put a ring on it
Michael Richardson - 2025-10-20 17:29:18
I for on, welcome our new 192.0.0.11 overloads.
Michael Richardson - 2025-10-20 17:29:50
This is not yet posted as an I-D. "draft-ipv6", please us draft-book-ipv6...
Michael Richardson - 2025-10-20 17:29:59
s/book/mook/
