[{"author": "Joel Halpern", "text": "@Darren as I recall, in the late 90s we crossed over 2000 people.
", "time": "2022-03-24T12:00:30Z"}, {"author": "Nicolai Leymann", "text": "Yes, Orlando was a huge IETF.
", "time": "2022-03-24T12:00:49Z"}, {"author": "Ron Bonica", "text": "IETF 49 (San Diego) had 2800 in person attendees
", "time": "2022-03-24T12:01:13Z"}, {"author": "Tony Li", "text": "Since someone mentioned IETF 30, I just pulled it up.
", "time": "2022-03-24T12:01:48Z"}, {"author": "Jeffrey Haas", "text": "Quoting some sarcasm from other colleagues: Work on the Internet is done. Go home!
", "time": "2022-03-24T12:02:08Z"}, {"author": "Tony Li", "text": "Toronto 1994, ~~800 syyrmfrrd
", "time": "2022-03-24T12:02:28Z"}, {"author": "Tony Li", "text": "~~800 attendees
", "time": "2022-03-24T12:02:40Z"}, {"author": "Andrew Alston", "text": ":) I prefer \"Work is done on the internet - I don't need to be in the office\" ;p
", "time": "2022-03-24T12:03:03Z"}, {"author": "Darren Dukes", "text": "@andrew indeed amazing to see how the world is now working on what we've specified and built - even the IETF ;)
", "time": "2022-03-24T12:08:35Z"}, {"author": "Tony Li", "text": "More fun: calculate how much business is done on the Internet, daily.
", "time": "2022-03-24T12:09:56Z"}, {"author": "Tony Li", "text": "It works out to be ~~$3B.
", "time": "2022-03-24T12:20:04Z"}, {"author": "Gyan Mishra", "text": "@Kireeti I can bring this up on 6MAN thread related to deprecation of router alert for mpls ping
", "time": "2022-03-24T12:36:22Z"}, {"author": "Gyan Mishra", "text": "thank you
", "time": "2022-03-24T12:36:24Z"}, {"author": "Gyan Mishra", "text": "I will reference your draft and this deck for the thread, thanks
", "time": "2022-03-24T12:36:56Z"}, {"author": "Zafar Ali", "text": "@Tony claim 2 is as correct as 2+2 = 4 math. Jag's presentation in PALS WG explains the point on how reusing EL reduces the label stack size. In fact it is the other way round. If you add a new SPL, you add two additional labels. I.e., a simple 2+2 = 4 Math
", "time": "2022-03-24T12:38:29Z"}, {"author": "Ahmed Abdelsalam", "text": "@Tony Claim 6 : Adding two copies of EL in MPLS header is complex to implement in hardware, prone to inconsistent behavior and increased label stack (wasting a label).
", "time": "2022-03-24T12:41:29Z"}, {"author": "Rakesh Gandhi", "text": "@Tony: Claim 1 and 4 have wrong conclusions. We are still developing the protocol extensions for ELI after 10 years. It will take years to develop new protocol extensions for new SPL. Adding a bit in ELI signaling is definitely less work and faster deployment.
", "time": "2022-03-24T12:46:54Z"}, {"author": "Francois Clad", "text": "@Tony, I don\u2019t follow your point on the NPL. If it is not defined, then it should be defined. To me, that could be any user-configured label.
", "time": "2022-03-24T12:49:49Z"}, {"author": "Greg Mirsky", "text": "I'd note that the draft Bruno presented at the joined meeting is limited to 11 behaviors. I believe that's the proposed solution is too restrictive. If the authors make that solution extensible, it seems that we have Jags' draft. Am I missing something?
", "time": "2022-03-24T12:54:18Z"}, {"author": "Darren Dukes", "text": "Tarek, that is false
", "time": "2022-03-24T12:54:31Z"}, {"author": "Wim Henderickx", "text": "@greg, my point is if we boil the ocean it will take years and most likely we will need new HW to support it
", "time": "2022-03-24T12:55:20Z"}, {"author": "Darren Dukes", "text": "Please describe how you see any change in hashing.
", "time": "2022-03-24T12:55:37Z"}, {"author": "Ketan Talaulikar", "text": "@wim +1
", "time": "2022-03-24T12:55:52Z"}, {"author": "Jaganbabu Rajamanickam", "text": "@Darren +1
", "time": "2022-03-24T12:56:06Z"}, {"author": "Ketan Talaulikar", "text": "@kireeti we can make progress in steps instead of waiting for a leap?
", "time": "2022-03-24T12:56:11Z"}, {"author": "Stewart Bryant", "text": "We wrote this a long time ago
", "time": "2022-03-24T12:56:28Z"}, {"author": "Tony Li", "text": "@Francois MIAD should not have a dependency on SR. We should support other signalling infra too. This is already in our requirements.
", "time": "2022-03-24T12:56:47Z"}, {"author": "Rakesh Gandhi", "text": "@Tarek: Bruno draft adding fixed SLID in EL does not alter ECMP hashing.
", "time": "2022-03-24T12:57:01Z"}, {"author": "Wim Henderickx", "text": "this is why i mentioned to go step by step iso big bang.
", "time": "2022-03-24T12:57:10Z"}, {"author": "Ketan Talaulikar", "text": "To clarify, I am not against progressing MIAD work. Let us not progress draft-bruno in parallel to get that initial set of functionality rolled out
", "time": "2022-03-24T12:58:44Z"}, {"author": "Greg Mirsky", "text": "@Wim If you consider Kireeti's proposal \"boiling the ocean\", then I can characterize Bruno's as \"kicking can down the road\".
", "time": "2022-03-24T12:58:45Z"}, {"author": "Ketan Talaulikar", "text": "i meant .. *Let us progress draft-bruno ;-)
", "time": "2022-03-24T12:59:23Z"}, {"author": "Wim Henderickx", "text": "i looked at all of them when i made my statement, because with kireeti's proposal we need a lot of changes before this is usable in a network. And some HW cannot support it today
", "time": "2022-03-24T13:00:11Z"}, {"author": "Darren Dukes", "text": "@tarek and chairs. one more note. The title of this presentation \"Redefining ELI considered
harmful;
NPL considered harmful\" was not supported by the content of the presentation. As Loa said, lets make sure we have draft text and not presentations like this.
", "time": "2022-03-24T13:00:26Z"}, {"author": "Wim Henderickx", "text": "@greg, this is why i said lets do something current HW can do and define something future HW can do
", "time": "2022-03-24T13:01:22Z"}, {"author": "Darren Dukes", "text": "Correction: Lets not have presentations like this, as loa said \"lets get a draft to discuss\" Sorry for the misquote Loa !
", "time": "2022-03-24T13:01:25Z"}, {"author": "Ketan Talaulikar", "text": "@ Tony perhaps \"not helpful\" would have been more appropriate than \"harmful\" ?
", "time": "2022-03-24T13:01:35Z"}, {"author": "Tarek Saad", "text": "meetecho allows us to stay bit longer - so hopefully we'll allow Haoyu to go over his deck
", "time": "2022-03-24T13:01:50Z"}, {"author": "Tony Li", "text": "@Ketan that would be breaking computer science tradition. :satisfied:
", "time": "2022-03-24T13:02:01Z"}, {"author": "Rakesh Gandhi", "text": "@Loa: New draft should be analysis draft focusing on approaches for Indicators and not using a biased view.
", "time": "2022-03-24T13:02:43Z"}, {"author": "Francois Clad", "text": "@Tony my point is that NPL does not have to be SR-specific just because some external documentation suggest to use it in that context. NPL needs to be properly defined.
", "time": "2022-03-24T13:02:46Z"}, {"author": "Greg Mirsky", "text": "@Wim, true, some HW cannot support MIAD. And the use of MIAD is optional. Besides, using ISD with Bruno's proposal breaks backward compatibility.
", "time": "2022-03-24T13:02:46Z"}, {"author": "Tony Li", "text": "@Francois What I could find claimed that NPL came from the SRGB. Kinda hard to do that with LDP.
", "time": "2022-03-24T13:03:27Z"}, {"author": "Zafar Ali", "text": "@Gred Bruno's proposal does not breaks backward compatibility. Bruno slides to design team explains it very well
", "time": "2022-03-24T13:03:43Z"}, {"author": "Zafar Ali", "text": "sorry @Greg
", "time": "2022-03-24T13:03:53Z"}, {"author": "Bruno Decraene", "text": "@Greg \"would require signaling\" seems like a better caracterization than \"breaks backward compatibility\"
", "time": "2022-03-24T13:03:59Z"}, {"author": "Rakesh Gandhi", "text": "Agree Bruno
", "time": "2022-03-24T13:04:25Z"}, {"author": "Jaganbabu Rajamanickam", "text": "@Burno +1
", "time": "2022-03-24T13:04:27Z"}, {"author": "Wim Henderickx", "text": "@greg, why do you claim bruno's proposal breaks backward compatibility? On what basis do you make this claim? I can tell on Nokia systems it doesn't. So before making such claim let verify it as it does not help any of this .
", "time": "2022-03-24T13:04:28Z"}, {"author": "Tony Li", "text": "Bruno's proposal is limited but doesn't have the problem of adding LSEs to ELI.
", "time": "2022-03-24T13:05:24Z"}, {"author": "Andrew Alston", "text": "Thanks guys :)
", "time": "2022-03-24T13:07:13Z"}]