Tuesday, November 9, 2021< ^ >
Jim Fenton (the other one) has set the subject to: DMARC @ IETF 108: 28 July 2020
Room Configuration
Room Occupants

[00:17:45] Glen joins the room
[00:17:49] Glen leaves the room
[15:34:28] Meetecho joins the room
[15:35:11] Meetecho has set the subject to: DMARC @ IETF 112
[15:36:10] Yoshiro Yoneya joins the room
[15:45:03] Robert Stepanek_web_285 joins the room
[15:45:03] Yoshiro Yoneya_web_894 joins the room
[15:46:38] Robert Stepanek_web_285 leaves the room
[15:46:53] Robert Stepanek_web_440 joins the room
[15:49:27] Tim Wicinski_web_194 joins the room
[15:49:36] alexamirante joins the room
[15:51:54] Barry Leiba_web_299 joins the room
[15:55:15] Gustavo Lozano_web_257 joins the room
[15:56:53] Seth Blank_web_157 joins the room
[15:56:55] Sean Croghan_web_474 joins the room
[15:58:05] Murray Kucherawy_web_631 joins the room
[15:58:44] Trent Adams_web_579 joins the room
[15:58:45] Taiji Kimura_web_953 joins the room
[15:59:06] Scott Rose_web_828 joins the room
[15:59:14] Peter Koch_web_952 joins the room
[16:00:04] Todd Herr_web_386 joins the room
[16:00:13] Jim Fenton_web_146 joins the room
[16:00:13] John Levine_web_192 joins the room
[16:00:28] Gustavo Lozano_web_257 leaves the room
[16:00:32] Gustavo Lozano_web_370 joins the room
[16:00:53] David Goldstein_web_759 joins the room
[16:00:54] joins the room
[16:00:54] Michael Breuer_web_949 joins the room
[16:01:20] Ulrich Wisser_web_762 joins the room
[16:01:25] Gustavo Lozano_web_370 leaves the room
[16:01:30] Alessandro Vesely_web_844 joins the room
[16:01:47] <John Levine_web_192> Ale is here
[16:01:49] Korry Luke_web_945 joins the room
[16:02:08] Gustavo Lozano_web_903 joins the room
[16:02:14] <Tim Wicinski_web_194> WE MUST GET TO ITEM #4
[16:03:14] Tim Hollebeek_web_119 joins the room
[16:03:21] Craig Pearce_web_522 joins the room
[16:03:31] Vasilis_web_695 joins the room
[16:03:35] Emiliano Spinella_web_673 joins the room
[16:03:37] Craig Pearce_web_522 leaves the room
[16:03:40] Tim Hollebeek_web_119 leaves the room
[16:03:41] Craig Pearce_web_778 joins the room
[16:03:42] Alexey Melnikov_web_595 joins the room
[16:03:44] Tim Hollebeek_web_986 joins the room
[16:03:44] Gustavo Lozano_web_903 leaves the room
[16:04:04] Ned Freed_web_746 joins the room
[16:04:13] Ulrich Wisser_web_762 leaves the room
[16:04:18] Autumn Tyr-Salvia_web_750 joins the room
[16:05:01] Emiliano Spinella_web_673 leaves the room
[16:05:05] Emiliano Spinella_web_141 joins the room
[16:06:16] Jörg Backschues_web_387 joins the room
[16:06:30] <John Levine_web_192> Drat
[16:06:45] Vasilis_web_695 leaves the room
[16:06:52] <Tim Wicinski_web_194> I am sad
[16:07:01] Antoin Verschuren_web_943 joins the room
[16:07:02] <Trent Adams_web_579> But I came for Topic 4!
[16:07:04] Gustavo Lozano_web_880 joins the room
[16:07:15] <Tim Wicinski_web_194> We are all here for Topic #4
[16:07:17] <Tim Hollebeek_web_986> IV. I like roman numerals
[16:07:37] <Trent Adams_web_579> Agreed!  Especially VII! ;)
[16:09:51] <Jim Fenton_web_146> Anyone that wants to help with minutes, feel free to join me on HedgeDoc
[16:10:15] Gustavo Lozano_web_880 leaves the room
[16:10:19] Gustavo Lozano_web_828 joins the room
[16:10:28] <> thanks to the note takers
[16:11:13] Yang Zhao_web_703 joins the room
[16:12:34] <Murray Kucherawy_web_631> +1 to John, that's what I wanted to say at the mic
[16:12:36] <Barry Leiba_web_299> +1 to John
[16:12:47] <Seth Blank_web_157> +1 as well
[16:12:50] Yuji Koyama_web_109 joins the room
[16:13:03] Yang Zhao_web_703 leaves the room
[16:13:07] Yang Zhao_web_623 joins the room
[16:13:08] <Tim Wicinski_web_194> +1 to Mr Levine.
[16:13:38] Kenneth Murchison_web_502 joins the room
[16:13:53] <Tim Wicinski_web_194> I am not the DNS community, but do feel that tree walks are not considered horrible anymore
[16:13:54] Taiji Kimura_web_953 leaves the room
[16:14:26] <Tim Hollebeek_web_986> Yes, I'm in favor of anything that helps the public suffix list go away.  Walks are more intuitive.
[16:14:41] <John Levine_web_192> The DNS crowd used to have religious objections to tree walk, but since the CAA record they seem to have lost faith
[16:14:47] Yang Zhao_web_623 leaves the room
[16:14:56] Sergey Myasoedov_web_494 joins the room
[16:15:40] Kenneth Murchison_web_502 leaves the room
[16:15:44] Kenneth Murchison_web_597 joins the room
[16:15:58] <Tim Wicinski_web_194> I also think there have been enough operational improvements more standardized that assuages feelings
[16:16:17] Kenneth Murchison_web_597 leaves the room
[16:16:21] Kenneth Murchison_web_557 joins the room
[16:16:53] <John Levine_web_192> dead horse metaphor
[16:17:33] <Seth Blank_web_157> Agreed on that as the rough consensus.
[16:18:09] Kenneth Murchison_web_557 leaves the room
[16:18:48] <Murray Kucherawy_web_631> @John: Good point.
[16:18:50] <Seth Blank_web_157> Tim-- what was the bug with the original CAA implementation detail?
[16:19:08] <Peter Koch_web_952> load is only one issue with the tree walk (and maybe it isn't in practice), the implicit concept of inheritance, which isn't part of the DNS, is the protocol design issue; the PSL is a kludge, of course - so I'm waiting to the the "next slide" for the limitation of the walk ;-)
[16:19:11] <Murray Kucherawy_web_631> Also a good thing to record in some appendix to explain why a walk is fine now but wasn't before.
[16:19:15] <Tim Hollebeek_web_986> The original CAA tree walk tried to follow CNAMEs, instead of just going up the prefered tree
[16:19:39] <Tim Hollebeek_web_986> The results were ... hilarious and wrong.
[16:19:48] <Seth Blank_web_157> Ah, so you could get referential loops, weeee
[16:19:53] <Tim Hollebeek_web_986> yep
[16:20:30] <Murray Kucherawy_web_631> A very minor point: The layout of the bullets needs work; seems like 4-6 should be indented under 3.
[16:20:31] Gustavo Lozano_web_828 leaves the room
[16:20:35] Gustavo Lozano_web_592 joins the room
[16:21:55] <John Levine_web_192> 5 came from the deepest name in the existing PSL
[16:22:21] <Seth Blank_web_157> Data was shared to the list as to why 5 was chosen: the longest suffixes in the PSL have 4 parts, so 5 covers all existing suffixes
[16:22:38] <Murray Kucherawy_web_631> Seems legit.
[16:24:09] <Jim Fenton_web_146> Thanks for clarifying the choice
[16:24:30] <John Levine_web_192> paramater makes things not interoperate
[16:25:01] <Jim Fenton_web_146> +1
[16:28:21] <Seth Blank_web_157> I think the max depth of the walk is a security consideration, and has some rough consensus on the list
[16:28:39] <Alessandro Vesely_web_844> if 5 comes from the PSL, we're still using the PSL.
[16:29:06] <Autumn Tyr-Salvia_web_750> It makes sense for there to be a minimum and a maximum for security and for people to understand what they can depend on definitely working, but I suppose there could be flexibility between min and max....
[16:29:26] <Seth Blank_web_157> The PSL is a registry of what people are using, so it's a proxy for what's needed in the real world. If 4 is the max used, that's a very good data point collected over the past decade.
[16:29:34] <Ned Freed_web_746> The argument that the PSL only has four levels seems persuasive to me.
[16:30:06] Craig Pearce_web_778 leaves the room
[16:30:39] Robert Stepanek_web_440 leaves the room
[16:30:44] <Ned Freed_web_746> There has to be a minimum to insure interoperability, and best interop is achieved if it's also the maximum. There has to be a maximum for security reasons.
[16:30:48] Gustavo Lozano_web_592 leaves the room
[16:30:52] Gustavo Lozano_web_305 joins the room
[16:31:18] Pete Resnick_web_156 joins the room
[16:31:44] <Ned Freed_web_746> To me this means that if we're going to switch to tree walk, we have to specify a single value, the only question is what that value should be.
[16:33:03] Bron Gondwana_web_194 joins the room
[16:34:40] <Murray Kucherawy_web_631> +1 Seth
[16:34:41] <John Levine_web_192> unless we're planning to rerun DBOUND I hope not
[16:35:25] <John Levine_web_192> I run a K12.NY.US, it's useful but it's only 3 deep
[16:35:36] <Peter Koch_web_952> RFC 1480 is long since obsolete, isn't it - but how up away would one expect a DMARC entry to appear?
[16:37:29] <John Levine_web_192> @peter I am the registry for a bunch of <place>.NY.US domains. The existing ones aren't going away
[16:37:32] <John Levine_web_192> but they're only 3 deep
[16:38:03] <Ned Freed_web_746> Another way to look at it: Any security-conscious implementation is going to impose a maximum, no matter what the standard says.
[16:38:33] <Jim Fenton_web_146> But the fact that there are currently 3 level domains makes me wonder if we should tie the spec to a 4-level limit in the PSL.
[16:39:21] <Peter Koch_web_952> @john nobody expects ... but then, in this particular example, NY.US and US would not be proper query targets anyway
[16:39:26] <Jim Fenton_web_146> s/there are/these are/
[16:40:14] Gustavo Lozano_web_305 leaves the room
[16:40:18] Gustavo Lozano_web_956 joins the room
[16:40:46] <Ned Freed_web_746> I understand the concerns here. But the constraints are what they are.
[16:41:15] <Autumn Tyr-Salvia_web_750> I worked with a few organizations that were using subdomain1.subdomain2.orgdomain.tld for sending, and a couple others that wanted to be able to do subdomain1.subdomain2.subdomain3.orgdomain.tld. Haven't heard of anyone wanting to send email beyond that at all.
[16:41:56] <John Levine_web_192> you can still send from those domains, you just can't have an org domain below level 4
[16:42:09] <Murray Kucherawy_web_631> I need to drop off.  Please let me know if I have any AD homework.
[16:42:47] <Ned Freed_web_746> To reemphasize John's point, I've seen plenty of domains with many levels, but none that belong on the RHS of the @ in a From: field.
[16:43:12] <Barry Leiba_web_299> And I think what Ned just said is the key point.
[16:43:45] Murray Kucherawy_web_631 leaves the room
[16:48:02] <Alessandro Vesely_web_844>
[16:49:10] Phillip Hallam-Baker_web_472 joins the room
[16:49:45] <Ned Freed_web_746> Sorry, can't seem to get the mic working.
[16:50:01] <Barry Leiba_web_299> OK, type here if you like, and I'll channel.
[16:50:26] <Ned Freed_web_746> OK, let someone else go while I'm typing something.
[16:50:31] Gustavo Lozano_web_956 leaves the room
[16:50:35] Gustavo Lozano_web_275 joins the room
[16:51:50] <Seth Blank_web_157> +1 to John-- seen a lot of drafts like this that haven't been able to make any headway
[16:52:12] Ned Freed_web_746 leaves the room
[16:52:16] Ned Freed_web_649 joins the room
[16:52:30] <John Levine_web_192> "Sometimes it works" doesn't seem like a strong path to interoperability
[16:52:59] <Pete Resnick_web_156> Yeah, at most an Informational "mitigations" document.
[16:53:04] Peter Koch_web_952 leaves the room
[16:53:14] Peter Koch_web_581 joins the room
[16:53:40] <Ned Freed_web_649> Sorry, had to restart things and couldn't hear the discussion. Anyway, in case it hasn't already been said, while I wish it were otherwise, I don't think this covers enough of the mutations to provide sufficient benefit given the costs.
[16:54:38] <Ned Freed_web_649> I *really* wish this weren't true.
[16:54:47] <Pete Resnick_web_156> +1 to both points.
[16:54:59] <Bron Gondwana_web_194> we now need to have an indirect discussion about it on the mailing list
[16:55:13] <Pete Resnick_web_156> :white_frowning_face:
[16:55:18] <Seth Blank_web_157> This is me singing, for anyone wondering.
[16:55:24] <Trent Adams_web_579> Woot!!  Dancing time!
[16:55:25] <Autumn Tyr-Salvia_web_750> This seems pretty cool, but as others have said, perhaps not comprehensive enough to make it worth the work to try to overcome resistance.
[16:55:28] <Bron Gondwana_web_194> beautiful alto voice
[16:55:30] <Todd Herr_web_386> excellent falsetto, seth
[16:55:34] <Jim Fenton_web_146> yup
[16:55:35] <Pete Resnick_web_156> Oh heavens, let there be no dancing.
[16:55:46] <Alessandro Vesely_web_844> was that point 4?
[16:55:53] <Pete Resnick_web_156> Indeed it was.
[16:55:54] Sergey Myasoedov_web_494 leaves the room
[16:55:59] <Bron Gondwana_web_194> @Seth you should have heard my soprano descant to Ding Dong Merrily On High at choir tonight
[16:56:13] <Bron Gondwana_web_194> it was... well, it was mostly at pitch.  I'll give it that
[16:56:30] <Seth Blank_web_157> thank you all!
[16:56:38] <Tim Wicinski_web_194> thanks boss
[16:56:40] <Bron Gondwana_web_194> thanks all!
[16:56:42] <Autumn Tyr-Salvia_web_750> Thanks!
[16:56:43] <Ned Freed_web_649> thanks everyone
[16:56:45] Scott Rose_web_828 leaves the room
[16:56:46] Barry Leiba_web_299 leaves the room
[16:56:47] Pete Resnick_web_156 leaves the room
[16:56:47] Yoshiro Yoneya_web_894 leaves the room
[16:56:48] Todd Herr_web_386 leaves the room
[16:56:50] Seth Blank_web_157 leaves the room
[16:56:51] Alexey Melnikov_web_595 leaves the room
[16:56:56] Tim Wicinski_web_194 leaves the room
[16:57:07] Yuji Koyama_web_109 leaves the room
[16:57:11] Alessandro Vesely_web_844 leaves the room
[16:57:21] John Levine_web_192 leaves the room
[16:57:21] Korry Luke_web_945 leaves the room
[16:57:24] Trent Adams_web_579 leaves the room
[16:57:25] Tim Hollebeek_web_986 leaves the room
[16:57:26] David Goldstein_web_759 leaves the room
[16:57:31] Michael Breuer_web_949 leaves the room
[16:57:32] Ned Freed_web_649 leaves the room
[16:57:34] Autumn Tyr-Salvia_web_750 leaves the room
[16:57:35] Gustavo Lozano_web_275 leaves the room
[16:57:37] Bron Gondwana_web_194 leaves the room
[16:58:19] leaves the room
[16:58:48] Yoshiro Yoneya leaves the room
[17:00:18] Antoin Verschuren_web_943 leaves the room
[17:11:07] Jim Fenton_web_146 leaves the room
[17:14:14] Emiliano Spinella_web_141 leaves the room
[17:25:08] Jörg Backschues_web_387 leaves the room
[17:30:40] Peter Koch_web_581 leaves the room
[17:32:06] Meetecho leaves the room
[17:42:37] Sean Croghan_web_474 leaves the room
[18:05:11] Paolo Saviano_web_551 joins the room
[18:05:43] Phillip Hallam-Baker_web_472 leaves the room
[18:05:43] Paolo Saviano_web_551 leaves the room
[18:21:08] alexamirante leaves the room