[{"author": "Suresh Krishnan", "text": "<p><a href=\"https://notes.ietf.org/notes-ietf-122-aipref\">https://notes.ietf.org/notes-ietf-122-aipref</a></p>", "time": "2025-03-20T06:06:58Z"}, {"author": "Jonathan Hoyland", "text": "<p>Not great that we're ruling bot authn/z out of scope IMO. It informs everything else.</p>", "time": "2025-03-20T06:07:12Z"}, {"author": "Martin Thomson", "text": "<p>My position is that it is great that we haven't.  Authenticating bots is not responsive to the problem we are looking at addressing.</p>", "time": "2025-03-20T06:08:15Z"}, {"author": "Farzaneh", "text": "<p>+1 Martin</p>", "time": "2025-03-20T06:09:09Z"}, {"author": "David Schinazi", "text": "<p>We can just say \"if you know who this bot is...\"</p>", "time": "2025-03-20T06:10:09Z"}, {"author": "Martin Thomson", "text": "<p>We should start from the assumption that assertions about preferences are public statements.  Or that any access control is exercised orthogonally to this.</p>", "time": "2025-03-20T06:10:25Z"}, {"author": "Andrew Campling", "text": "<p>No doubt we can leave enforcement of acting on it to the protocol police :-)</p>", "time": "2025-03-20T06:10:45Z"}, {"author": "Jonathan Hoyland", "text": "<p>I wasn't actually planning to spark a whole discussion, I was just complaining in chat <span aria-label=\"sweat smile\" class=\"emoji emoji-1f605\" role=\"img\" title=\"sweat smile\">:sweat_smile:</span></p>", "time": "2025-03-20T06:11:12Z"}, {"author": "Farzaneh", "text": "<p>interpreting what we mean by \"who\" is tricky. and identification of bots is not always a good thing</p>", "time": "2025-03-20T06:11:26Z"}, {"author": "Max Gendler", "text": "<p>Bot Authentication would be nice, but as a publisher rep - we live in a world where even with extensive tooling, bot auth creates problems and is not fully  accurate. I'd love to fix that, but we need to work on this now - I don't think we can wait for later.</p>", "time": "2025-03-20T06:12:09Z"}, {"author": "Mike Bishop", "text": "<p>Enforcement is a policy question, not a protocol question. Content owner states their preference/licensing; if the consumer violates it the recourses are legal rather than technical.</p>", "time": "2025-03-20T06:12:39Z"}, {"author": "Chris Needham", "text": "<p>+1 Max, coming from large publisher perspective</p>", "time": "2025-03-20T06:12:59Z"}, {"author": "Alissa Cooper", "text": "<p>\"relatively\" is doing a lot of work on this slide</p>", "time": "2025-03-20T06:13:16Z"}, {"author": "Martin Thomson", "text": "<p>+1 Max also</p>", "time": "2025-03-20T06:13:23Z"}, {"author": "Jonathan Hoyland", "text": "<p><span class=\"user-mention\" data-user-id=\"5788\">@Max Gendler</span>  So I want to make fully accurate bot auth (or at least cryptographic bot auth). If you live in a world with perfect bot auth, then your design space is very different to one where you don't.</p>", "time": "2025-03-20T06:13:30Z"}, {"author": "Martin Thomson", "text": "<p>We should talk about \"relatively\"</p>", "time": "2025-03-20T06:13:36Z"}, {"author": "Max Gendler", "text": "<p><span class=\"user-mention\" data-user-id=\"453\">@Jonathan Hoyland</span> Good luck! I hope you succeed, but publishers and internet users can't wait around for that, we need better protections now. Frankly, we needed better protections many years ago, but I'll take the enthusiasm now</p>", "time": "2025-03-20T06:14:58Z"}, {"author": "Martin Thomson", "text": "<p>+1 to Alissa</p>", "time": "2025-03-20T06:15:29Z"}, {"author": "Jonathan Hoyland", "text": "<p><span class=\"user-mention\" data-user-id=\"5788\">@Max Gendler</span>  I mean, I have been beating the bot auth horse for a couple of years now, so I totally agree with you.</p>", "time": "2025-03-20T06:15:36Z"}, {"author": "Martin Thomson", "text": "<p>Not all conceptions of simple are the same</p>", "time": "2025-03-20T06:15:38Z"}, {"author": "Martin Thomson", "text": "<p>That 'if uncertain, out\" is more useful guidance than what is on the slide, but what Alissa is saying is very relevant</p>", "time": "2025-03-20T06:16:48Z"}, {"author": "Jari Arkko", "text": "<p>+1 to Alissa. I think some of the proposals we have are too simple, but let's get to the bottom of it by discussing the features the vocabulary must support. E.g., will a simplest set enable research (e.g., classification of content) to happen?</p>", "time": "2025-03-20T06:16:59Z"}, {"author": "Farzaneh", "text": "<p>so simple means there is a lot of agreement on it and not controversial? or simple in technical implementation</p>", "time": "2025-03-20T06:17:04Z"}, {"author": "Andrew Campling", "text": "<p>Simple is pretty subjective, relatively simple more so....</p>", "time": "2025-03-20T06:17:09Z"}, {"author": "Chris Needham", "text": "<p>Strong +1 to Alissa</p>", "time": "2025-03-20T06:17:23Z"}, {"author": "Martin Thomson", "text": "<p>time to move on to the next question</p>", "time": "2025-03-20T06:18:07Z"}, {"author": "Max Gendler", "text": "<p>+1 to Alissa - we have 'simple signals' now. That has brought the problems.</p>", "time": "2025-03-20T06:18:13Z"}, {"author": "Martin Thomson", "text": "<p>we could talk about these in the abstract forever</p>", "time": "2025-03-20T06:18:15Z"}, {"author": "Martin Thomson", "text": "<p>robots.txt is kinda bad, but we can't ignore the fact that this is what bots are looking at</p>", "time": "2025-03-20T06:21:24Z"}, {"author": "Max Gendler", "text": "<p>Robots Exclusion Protocol is a useful vehicle for connection to URL, but content moves in other vehicles as well. REP does not follow the content off site</p>", "time": "2025-03-20T06:22:14Z"}, {"author": "Jonathan Hoyland", "text": "<p><span class=\"user-mention silent\" data-user-id=\"26\">Martin Thomson</span> <a href=\"#narrow/stream/412-aipref/topic/ietf-122/near/158205\">said</a>:</p>\n<blockquote>\n<p>robots.txt is kinda bad, but we can't ignore the fact that this is what bots are looking at</p>\n</blockquote>\n<p>I mean, some of them. <a href=\"https://blog.archive.org/2017/04/17/robots-txt-meant-for-search-engines-dont-work-well-for-web-archives/\">https://blog.archive.org/2017/04/17/robots-txt-meant-for-search-engines-dont-work-well-for-web-archives/</a></p>", "time": "2025-03-20T06:23:28Z"}, {"author": "Thomas McCarthy-Howe", "text": "<p>So, I'm not sure this is the time, but I'm part of the vCon WG and we're looking at using the definitions from this group to record consent grants for customer recordings. Not a URL, not a robots.txt.  Are we too far outside the intent?</p>", "time": "2025-03-20T06:23:36Z"}, {"author": "Farzaneh", "text": "<p>but I wonder Max if this is a topic we should address at the IETF or should we design a robust protocol that can consider that but it would be addressed elsewhere</p>", "time": "2025-03-20T06:23:54Z"}, {"author": "Richard Barnes", "text": "<p>It seems not very DEI that we are talking about only <em>excluding</em> the robots.</p>", "time": "2025-03-20T06:24:37Z"}, {"author": "Thomas McCarthy-Howe", "text": "<p>We do have the opposite problem in vCons.</p>", "time": "2025-03-20T06:25:06Z"}, {"author": "Alissa Cooper", "text": "<p>Thomas the vocabulary should be reusable</p>", "time": "2025-03-20T06:25:22Z"}, {"author": "Thomas McCarthy-Howe", "text": "<p>@alissa, I would think so.</p>", "time": "2025-03-20T06:25:37Z"}, {"author": "Max Gendler", "text": "<p><span class=\"user-mention\" data-user-id=\"4750\">@Farzaneh</span> I think it's why the vocabulary workstream is so critical - if the IETF can be a good place to standardize vocabulary and apply it to <em>some</em> protocols, it will be easier to get other protocols/mechanisms to use it in other areas that the IETF can't cover</p>", "time": "2025-03-20T06:25:40Z"}, {"author": "Farzaneh", "text": "<p>Thanks Max.</p>", "time": "2025-03-20T06:26:25Z"}, {"author": "Thomas McCarthy-Howe", "text": "<p>@max +1</p>", "time": "2025-03-20T06:26:31Z"}, {"author": "Max Gendler", "text": "<p>REP is already a 'both' system, but there is an assumption that it's opt-out only from many players in the space</p>", "time": "2025-03-20T06:26:41Z"}, {"author": "Mirja K\u00fchlewind", "text": "<p>Remote people: feel free to join the queue and also participate in the in-room discussion</p>", "time": "2025-03-20T06:27:04Z"}, {"author": "Farzaneh", "text": "<p>so Richard I think your \"inclusion\" point can be discussed here</p>", "time": "2025-03-20T06:27:05Z"}, {"author": "Farzaneh", "text": "<p>I always thought it is not up to us to decide opt in or opt out</p>", "time": "2025-03-20T06:27:57Z"}, {"author": "Andrew Campling", "text": "<p>No signal = abstain?</p>", "time": "2025-03-20T06:28:01Z"}, {"author": "Max Gendler", "text": "<p><span class=\"user-mention\" data-user-id=\"811\">@Mirja K\u00fchlewind</span> thank you for the welcoming prompt, but it's 230 in the morning here, so I don't think you'll all enjoy my tired, croaky voice. :)</p>", "time": "2025-03-20T06:28:03Z"}, {"author": "Mirja K\u00fchlewind", "text": "<p>No need to turn on the video. <span aria-label=\"wink\" class=\"emoji emoji-1f609\" role=\"img\" title=\"wink\">:wink:</span></p>", "time": "2025-03-20T06:28:29Z"}, {"author": "Farzaneh", "text": "<p>no Andrew. the interpretation of this depends on the jurisdiction.</p>", "time": "2025-03-20T06:28:46Z"}, {"author": "Andrew Campling", "text": "<p>@Max there may be some tired, croaky people here too</p>", "time": "2025-03-20T06:28:49Z"}, {"author": "Mirja K\u00fchlewind", "text": "<p>But thanks for joining in the middle of the night</p>", "time": "2025-03-20T06:28:50Z"}, {"author": "Jonathan Hoyland", "text": "<p>There's a difference between no preference expressed and no preference.</p>", "time": "2025-03-20T06:29:01Z"}, {"author": "Martin Thomson", "text": "<p>jonathan: yes, but it's not necessarily something we need to encode specifically</p>", "time": "2025-03-20T06:29:36Z"}, {"author": "Farzaneh", "text": "<p>yes but the interpretation of that does not depend on this group, does it?</p>", "time": "2025-03-20T06:29:37Z"}, {"author": "Martin Thomson", "text": "<p>You can know about the preference, then say nothing</p>", "time": "2025-03-20T06:29:48Z"}, {"author": "Martin Thomson", "text": "<p>You can even express the empty set of preferences</p>", "time": "2025-03-20T06:30:00Z"}, {"author": "Martin Thomson", "text": "<p>That's possible in the draft I put out</p>", "time": "2025-03-20T06:30:08Z"}, {"author": "Jonathan Hoyland", "text": "<p>We just need to be able to distinguish all three levels opt-in, no preference expressed, opt-out.</p>", "time": "2025-03-20T06:30:26Z"}, {"author": "Richard Barnes", "text": "<p>always worth writing down what Richard says</p>", "time": "2025-03-20T06:31:17Z"}, {"author": "Farzaneh", "text": "<p>can you write it down here richard?</p>", "time": "2025-03-20T06:32:26Z"}, {"author": "Jari Arkko", "text": "<p>We've already \"specified\" the \"don't know\" model... it's when there's no preference in any shape or form. And that doesn't need a spec. However, is the aipreference related to the existence of a robots.txt preference? If you said \"no bots at all\", then ai bots seem to be excluded along everything else as well.</p>", "time": "2025-03-20T06:32:37Z"}, {"author": "Martin Thomson", "text": "<p>the barnes' unertainty principle</p>", "time": "2025-03-20T06:32:46Z"}, {"author": "Christopher Wood", "text": "<p>Isn\u2019t it Martha\u2019s uncertainty principle?</p>", "time": "2025-03-20T06:32:50Z"}, {"author": "Martin Thomson", "text": "<p>*uncertainty</p>", "time": "2025-03-20T06:32:52Z"}, {"author": "Jonathan Hoyland", "text": "<p><span class=\"user-mention silent\" data-user-id=\"128\">Christopher Wood</span> <a href=\"#narrow/stream/412-aipref/topic/ietf-122/near/158298\">said</a>:</p>\n<blockquote>\n<p>Isn\u2019t it Martha\u2019s uncertainty principle?</p>\n</blockquote>\n<p>Well, maybe?</p>", "time": "2025-03-20T06:33:05Z"}, {"author": "Richard Barnes", "text": "<p><span aria-label=\"person in lotus position\" class=\"emoji emoji-1f9d8\" role=\"img\" title=\"person in lotus position\">:person_in_lotus_position:</span></p>", "time": "2025-03-20T06:33:18Z"}, {"author": "Martin Thomson", "text": "<p>(Did Martha take Richard's name?  I was there, but I forget.)</p>", "time": "2025-03-20T06:33:24Z"}, {"author": "Farzaneh", "text": "<p>Internet is not free for all anymore? WE NEED TO KEEP THE INTERNET FREE</p>", "time": "2025-03-20T06:33:28Z"}, {"author": "Richard Barnes", "text": "<p>(No)</p>", "time": "2025-03-20T06:33:29Z"}, {"author": "Martin Thomson", "text": "<p><span aria-label=\"+1\" class=\"emoji emoji-1f44d\" role=\"img\" title=\"+1\">:+1:</span></p>", "time": "2025-03-20T06:33:40Z"}, {"author": "Jonathan Hoyland", "text": "<p><a href=\"https://xkcd.com/552/\">https://xkcd.com/552/</a></p>", "time": "2025-03-20T06:33:44Z"}, {"author": "Richard Barnes", "text": "<p>The Internet is not free as in unconstrained by things like copyright</p>", "time": "2025-03-20T06:33:53Z"}, {"author": "Chris Needham", "text": "<p>The default / baseline is set in regulations, which may vary around the world. There's a case to be made for regulatory alignment, but it's in their scope, not for this group</p>", "time": "2025-03-20T06:34:01Z"}, {"author": "Richard Barnes", "text": "<p>The Internet is in fact constrained in some quite complicated and ambiguous ways</p>", "time": "2025-03-20T06:34:13Z"}, {"author": "Farzaneh", "text": "<p>yeah the Internet has been constrained by copyright for so long...</p>", "time": "2025-03-20T06:34:26Z"}, {"author": "Martin Thomson", "text": "<p>speed of light is pretty unambiguous</p>", "time": "2025-03-20T06:34:30Z"}, {"author": "Farzaneh", "text": "<p>and many other things..</p>", "time": "2025-03-20T06:34:37Z"}, {"author": "Martin Thomson", "text": "<p>copyright is less ambiguous than some of the constraints we operate under</p>", "time": "2025-03-20T06:34:53Z"}, {"author": "Farzaneh", "text": "<p>I think we are making this too complicated. we just have to make it possible for people to signal their preferences. I dont think we should focus on opt-in opt-out</p>", "time": "2025-03-20T06:35:35Z"}, {"author": "Thomas McCarthy-Howe", "text": "<p>Yes yes yes.</p>", "time": "2025-03-20T06:35:37Z"}, {"author": "Lucas Pardue", "text": "<p>thx tommy, good phrasing I agree with</p>", "time": "2025-03-20T06:35:49Z"}, {"author": "Mike Bishop", "text": "<p>How is this different from multiple rules in robots.txt?</p>\n<p>deny: /<br>\nallow: /ai-go-here</p>\n<p>Anyone can indicate their default with a global rule to begin with.</p>", "time": "2025-03-20T06:36:16Z"}, {"author": "Richard Barnes", "text": "<p>The AI operator / consumer of this information is going to have some decision calculus that takes in these preferences in addition to laws, contracts, whatever else</p>", "time": "2025-03-20T06:36:21Z"}, {"author": "Martin Thomson", "text": "<p>Farzaneh, this is something we do a lot in standards.  We get a little bit more abstract so that we can agree on principles, so that when it gets to details, the course is soother</p>", "time": "2025-03-20T06:36:32Z"}, {"author": "Max Gendler", "text": "<p><span class=\"user-mention\" data-user-id=\"147\">@Mike Bishop</span> I think that's part of the usefullness of looking at REP. It's pretty simply extensible from my read.</p>", "time": "2025-03-20T06:36:56Z"}, {"author": "Alissa Cooper", "text": "<p>Not clear why we have to set this list in advance -- if people want to write drafts that propose other embedded mechanisms for other protocols/formats along the way, they should bring them to the WG.</p>", "time": "2025-03-20T06:37:31Z"}, {"author": "Max Gendler", "text": "<p>+1 Martin</p>", "time": "2025-03-20T06:38:04Z"}, {"author": "Jari Arkko", "text": "<p>+1 Alissa. And we can do this one by one anyway.</p>", "time": "2025-03-20T06:38:07Z"}, {"author": "Farzaneh", "text": "<p>so I think this is separate from the vocab work right? so we dont have to finish attachment mechanisms by august.</p>", "time": "2025-03-20T06:38:28Z"}, {"author": "Lucas Pardue", "text": "<p>I propose a Cuckoo filter</p>", "time": "2025-03-20T06:38:53Z"}, {"author": "Jari Arkko", "text": "<p>We don't have to finish all of them, but one at least (REP++ perhaps)</p>", "time": "2025-03-20T06:38:54Z"}, {"author": "Chris Needham", "text": "<p>We need attachment to image and video formats, so an example IETF format is Matroska</p>", "time": "2025-03-20T06:39:09Z"}, {"author": "Richard Barnes", "text": "<p>@Farzanah - Seems separable, yeah</p>", "time": "2025-03-20T06:39:09Z"}, {"author": "Suresh Krishnan", "text": "<p>@Alissa, good point. The idea is to focuss on a few items for the interim</p>", "time": "2025-03-20T06:39:17Z"}, {"author": "Suresh Krishnan", "text": "<p>Not really to restrict future work</p>", "time": "2025-03-20T06:39:23Z"}, {"author": "Jonathan Hoyland", "text": "<p>Hmm, maybe something SCITT could think about? <span aria-label=\"thinking\" class=\"emoji emoji-1f914\" role=\"img\" title=\"thinking\">:thinking:</span></p>", "time": "2025-03-20T06:39:47Z"}, {"author": "Farzaneh", "text": "<p>maybe this group can also have a section on future work that can be done. and we drop stuff in that bucket</p>", "time": "2025-03-20T06:40:05Z"}, {"author": "Jari Arkko", "text": "<p>+1 for REP + HTTP header</p>", "time": "2025-03-20T06:41:14Z"}, {"author": "Jari Arkko", "text": "<p>(to begin with)</p>", "time": "2025-03-20T06:41:45Z"}, {"author": "Jonathan Hoyland", "text": "<p>Will the HTTP headers be structured fields?</p>", "time": "2025-03-20T06:41:57Z"}, {"author": "Martin Thomson", "text": "<p>jonathan: my proposal is</p>", "time": "2025-03-20T06:43:30Z"}, {"author": "Farzaneh", "text": "<p>well we probably need a liaison with all the open source AI developers</p>", "time": "2025-03-20T06:43:39Z"}, {"author": "Farzaneh", "text": "<p>and nonprofit ones!</p>", "time": "2025-03-20T06:43:48Z"}, {"author": "Martin Thomson", "text": "<p>The HTML thing probably is  best solved with &lt;meta&gt; and http-equiv if we have a header field</p>", "time": "2025-03-20T06:43:55Z"}, {"author": "Farzaneh", "text": "<p>but I also think we can separate this from the vocab work</p>", "time": "2025-03-20T06:44:18Z"}, {"author": "Lucas Pardue", "text": "<p>like Link MT?</p>", "time": "2025-03-20T06:44:19Z"}, {"author": "Martin Thomson", "text": "<p>The W3C is aware of this work and we have active discussion back and forth already.</p>", "time": "2025-03-20T06:44:20Z"}, {"author": "Andrew Campling", "text": "<p>Liaison with other SDOs I think</p>", "time": "2025-03-20T06:44:20Z"}, {"author": "Martin Thomson", "text": "<p>Lucas: not Link specifically</p>", "time": "2025-03-20T06:44:39Z"}, {"author": "Lucas Pardue", "text": "<p>sorry, I meant how Link allows different wire format for the smae $logicalthing</p>", "time": "2025-03-20T06:45:16Z"}, {"author": "Jonathan Hoyland", "text": "<p>Abstract grammar, with lots of potential encodings.</p>", "time": "2025-03-20T06:45:36Z"}, {"author": "Shigeya Suzuki", "text": "<p>+1 to Colin</p>", "time": "2025-03-20T06:45:45Z"}, {"author": "Martin Thomson", "text": "<p><a href=\"https://html.spec.whatwg.org/#pragma-directives\">https://html.spec.whatwg.org/#pragma-directives</a> is what I was thinking of for HTML</p>", "time": "2025-03-20T06:45:49Z"}, {"author": "Lucas Pardue", "text": "<p>/nods</p>", "time": "2025-03-20T06:46:22Z"}, {"author": "Mirja K\u00fchlewind", "text": "<p><span class=\"user-mention silent\" data-user-id=\"4750\">Farzaneh</span> <a href=\"#narrow/stream/412-aipref/topic/ietf-122/near/158365\">said</a>:</p>\n<blockquote>\n<p>well we probably need a liaison with all the open source AI developers</p>\n</blockquote>\n<p>Hopefully we can reach out to same and talk to them without being overly formal</p>", "time": "2025-03-20T06:47:36Z"}, {"author": "Lucas Pardue", "text": "<p>@chris needham, are you thinking of preference definition at the media manifest level?</p>", "time": "2025-03-20T06:48:03Z"}, {"author": "Farzaneh", "text": "<p>I think they have a group actually. Common Crawl and Open Future know more</p>", "time": "2025-03-20T06:48:06Z"}, {"author": "Martin Thomson", "text": "<p>conveniently for liaison setup: our chairs are also on the IAB</p>", "time": "2025-03-20T06:48:08Z"}, {"author": "Suresh Krishnan", "text": "<p>Not by design :D</p>", "time": "2025-03-20T06:49:40Z"}, {"author": "Martin Thomson", "text": "<p>PLUS</p>", "time": "2025-03-20T06:50:18Z"}, {"author": "Chris Needham", "text": "<p>@Lucas Yes, definitely need solutions for streaming A/V media. Haven't thought through yet about signaling via manifest or embedding in the media container (or both)</p>", "time": "2025-03-20T06:51:37Z"}, {"author": "Farzaneh", "text": "<p>so when do we wanna finish that one Suresh?</p>", "time": "2025-03-20T06:52:08Z"}, {"author": "Chris Needham", "text": "<p>... or if HTTPs headers enough</p>", "time": "2025-03-20T06:52:14Z"}, {"author": "Max Gendler", "text": "<p>I am heavily skeptical on this deliverable, but would like to hear other people's thoughts</p>", "time": "2025-03-20T06:52:52Z"}, {"author": "Farzaneh", "text": "<p>I thought we were going to finish the vocab by august</p>", "time": "2025-03-20T06:52:55Z"}, {"author": "Martin Thomson", "text": "<p>Max: I have a solution for this deliverable in my draft</p>", "time": "2025-03-20T06:53:05Z"}, {"author": "Martin Thomson", "text": "<p>it's pretty straightforward: if you see a NO anywhere, it's NO, otherwise, if you see a YES anywhere, it's YES, otherwise, UNKNOWN.</p>", "time": "2025-03-20T06:53:38Z"}, {"author": "Max Gendler", "text": "<p><span class=\"user-mention\" data-user-id=\"26\">@Martin Thomson</span> I'm embarrassed to say I haven't read yours yet. But that summary is not ideal from my point of view. I'll hold off on a more in-depth thought on that until I've read it and also when it's not 3 in the morning</p>", "time": "2025-03-20T06:54:20Z"}, {"author": "Jari Arkko", "text": "<p>Martin does. I guess I am a bit puzzled why we think the combination process is seen as the most complex. I think the base vocab work is going to be more challenging to agree, combination work ... just  logic. Or am I missing something?</p>", "time": "2025-03-20T06:54:57Z"}, {"author": "Martin Thomson", "text": "<p>Happy to discuss more, but it's not clear to me that there are many options</p>", "time": "2025-03-20T06:55:03Z"}, {"author": "Martin Thomson", "text": "<p>Jari: I agree that the vocab is the hard part</p>", "time": "2025-03-20T06:55:27Z"}, {"author": "Max Gendler", "text": "<p>I agree with Jari as well....and big +1 to Sarah on exactly why I'm heavily skeptical on this deliverable!</p>", "time": "2025-03-20T06:56:19Z"}, {"author": "Farzaneh", "text": "<p>relatively simple vocab Martin :)</p>", "time": "2025-03-20T06:56:22Z"}, {"author": "Martin Thomson", "text": "<p>simple, not simplistic :)</p>", "time": "2025-03-20T06:56:35Z"}, {"author": "Jonathan Hoyland", "text": "<p><span class=\"user-mention\" data-user-id=\"764\">@Jari Arkko</span> this would be the ultimate bikeshed. Which thing is the most complex is actually not that important.</p>", "time": "2025-03-20T06:56:37Z"}, {"author": "Richard Barnes", "text": "<p>+1 @Jari -- if we structure the rules language right, combination should follow naturally</p>", "time": "2025-03-20T06:57:16Z"}, {"author": "Mirja K\u00fchlewind", "text": "<p>Do we already have a list of ISO groups that is relevant here (expect MPEG is which is mostly separate now)?</p>", "time": "2025-03-20T06:57:42Z"}, {"author": "Farzaneh", "text": "<p>so Thom from common crawl has an ID on that if I am not mistaken</p>", "time": "2025-03-20T06:57:51Z"}, {"author": "Alissa Cooper", "text": "<p>Leonard sent a list: <a href=\"https://mailarchive.ietf.org/arch/msg/ai-control/hEJBp912Q5Xbx596jlF73SqW-GE/\">https://mailarchive.ietf.org/arch/msg/ai-control/hEJBp912Q5Xbx596jlF73SqW-GE/</a></p>", "time": "2025-03-20T06:58:10Z"}, {"author": "Farzaneh", "text": "<p>is this a 90 min meeting? I have to get on a flight soon.</p>", "time": "2025-03-20T06:58:18Z"}, {"author": "Mirja K\u00fchlewind", "text": "<p>thanks!</p>", "time": "2025-03-20T06:58:25Z"}, {"author": "Farzaneh", "text": "<p>agreed Suresh</p>", "time": "2025-03-20T06:58:33Z"}, {"author": "Jonathan Hoyland", "text": "<p>It finishes in 30 minutes.</p>", "time": "2025-03-20T06:58:34Z"}, {"author": "Martin Thomson", "text": "<p>it's 90 min</p>", "time": "2025-03-20T06:58:39Z"}, {"author": "Martin Thomson", "text": "<p>Safe flight!</p>", "time": "2025-03-20T06:58:49Z"}, {"author": "Farzaneh", "text": "<p>I think thats very important Suresh</p>", "time": "2025-03-20T06:59:10Z"}, {"author": "Mirja K\u00fchlewind", "text": "<p>(for SC29 and SC2 we have formal liaison relationships already, so it's easy to reach out)</p>", "time": "2025-03-20T07:00:00Z"}, {"author": "Max Gendler", "text": "<p><span class=\"user-mention\" data-user-id=\"958\">@Glenn Deen</span> Tuxedo for you</p>", "time": "2025-03-20T07:00:57Z"}, {"author": "Farzaneh", "text": "<p>I also think policymakers should know that we cannot accommodate everything (which is  a part of managing the \"expectations\")</p>", "time": "2025-03-20T07:01:52Z"}, {"author": "Farzaneh", "text": "<p>another thing about policymakers is that we shouldnt just focus on certain part of the world but I understand why we are starting from that certain part of the world</p>", "time": "2025-03-20T07:03:12Z"}, {"author": "Jonathan Hoyland", "text": "<p>From the RFC</p>\n<blockquote>\n<p>This document specifies and extends the \"Robots Exclusion Protocol\" method originally defined by Martijn Koster in 1994</p>\n</blockquote>", "time": "2025-03-20T07:06:04Z"}, {"author": "Suresh Krishnan", "text": "<p>@Farzaneh, agree. It is not a statement of relative importance but one of relative urgency</p>", "time": "2025-03-20T07:06:50Z"}, {"author": "Suresh Krishnan", "text": "<p>Will clarify this</p>", "time": "2025-03-20T07:06:56Z"}, {"author": "Farzaneh", "text": "<p>+1 Lucas. Important point. Alissa did provide a good response for that on the mailing list</p>", "time": "2025-03-20T07:10:13Z"}, {"author": "Lucas Pardue", "text": "<p>I'll take a task to read things before YOLOing into a session</p>", "time": "2025-03-20T07:10:39Z"}, {"author": "Andrew Campling", "text": "<p>Not reading the docs is an IETF meeting tradition surely</p>", "time": "2025-03-20T07:11:25Z"}, {"author": "Farzaneh", "text": "<p>hehe didnt mean that at all. we need to discuss important points over and over :)</p>", "time": "2025-03-20T07:11:26Z"}, {"author": "Suresh Krishnan", "text": "<p>@Lucas: Check this out <a href=\"https://mailarchive.ietf.org/arch/msg/ai-control/YtzrhC-1h7MeQ01NdeFJIke2HFY/\">https://mailarchive.ietf.org/arch/msg/ai-control/YtzrhC-1h7MeQ01NdeFJIke2HFY/</a></p>", "time": "2025-03-20T07:11:36Z"}, {"author": "Jonathan Hoyland", "text": "<p>What's the use case of expiring constraints?</p>", "time": "2025-03-20T07:11:43Z"}, {"author": "Farzaneh", "text": "<p>ok I have to go. Happy spring  and nowruz people.</p>", "time": "2025-03-20T07:12:11Z"}, {"author": "Lucas Pardue", "text": "<p>thank you</p>", "time": "2025-03-20T07:12:22Z"}, {"author": "Suresh Krishnan", "text": "<p>Bye Farzaneh. Have a safe flight</p>", "time": "2025-03-20T07:12:27Z"}, {"author": "Lucas Pardue", "text": "<p>part of my question is motivated by things like WebRTC or MOQ, where we have content contribution towards an architecture that might then publish it all in real time</p>", "time": "2025-03-20T07:13:23Z"}, {"author": "Lucas Pardue", "text": "<p>\"I don't want my voice forwarded to a transcription AI\" - is a preference</p>", "time": "2025-03-20T07:14:22Z"}, {"author": "Martin Thomson", "text": "<p>I really don't think that \"this might be nice\" is a good reason to build machinery</p>", "time": "2025-03-20T07:14:54Z"}, {"author": "Martin Thomson", "text": "<p>Gary suggested the use of Cache-Control. but my understanding is that that is not a good fit</p>", "time": "2025-03-20T07:15:51Z"}, {"author": "Chris Needham", "text": "<p>@Lucas Both things to add to the \"IETF defined protocols\" list from earlier</p>", "time": "2025-03-20T07:15:59Z"}, {"author": "Max Gendler", "text": "<p>I'm very skeptical of the embargo use case, and +1 martin on 'might be nice' is not a good idea for building things, especially given pressure on publishers that tends to happen when things don't get thought through</p>", "time": "2025-03-20T07:16:42Z"}, {"author": "Martin Thomson", "text": "<p>I can't imagine a way in which a <em>time</em> helps common crawl better</p>", "time": "2025-03-20T07:17:18Z"}, {"author": "Martin Thomson", "text": "<p>max-hops, perhaps</p>", "time": "2025-03-20T07:17:37Z"}, {"author": "Martin Thomson", "text": "<p>that is, you can use this. but you can't pass that license to another</p>", "time": "2025-03-20T07:17:49Z"}, {"author": "Sarah Jennings", "text": "<p>Is there a use case for RAG for news and returning contemporary info? Just spitballing</p>", "time": "2025-03-20T07:18:25Z"}, {"author": "Martin Thomson", "text": "<p>I would prefer to split RAG out as a different type of usage than try to use time or something oblique to manage that usage</p>", "time": "2025-03-20T07:19:37Z"}, {"author": "Max Gendler", "text": "<p>+1 Martin</p>", "time": "2025-03-20T07:20:40Z"}, {"author": "Max Gendler", "text": "<p>RAG is it's own separate purpose to my mind.</p>", "time": "2025-03-20T07:20:53Z"}, {"author": "Chris Needham", "text": "<p>The whole question of news and gen AI answers is ... complicated. Will share some research we've done on the list</p>", "time": "2025-03-20T07:21:09Z"}, {"author": "Thibault Meunier", "text": "<p>@martin do you think rag should be specified as a preference in the registry defined in your draft, or let as future work?</p>", "time": "2025-03-20T07:21:13Z"}, {"author": "Martin Thomson", "text": "<p>Thibault: that will depend on discussion.  I am open to it, but it's presently under genai</p>", "time": "2025-03-20T07:22:01Z"}, {"author": "Chris Needham", "text": "<p>+1 Martin on RAG. It's an important usage to address</p>", "time": "2025-03-20T07:22:05Z"}, {"author": "Martin Thomson", "text": "<p>The draft is extensible.</p>", "time": "2025-03-20T07:22:09Z"}, {"author": "Max Gendler", "text": "<p>Thanks to all for the conversation, looking forward to more - sorry to be missing Belgium with W3C meetings.</p>", "time": "2025-03-20T07:23:37Z"}, {"author": "Chris Needham", "text": "<p>Same here, the scheduling for the interim is unfortunate</p>", "time": "2025-03-20T07:23:59Z"}, {"author": "Suresh Krishnan", "text": "<p>Apologies Max and Chris. Finding a universally working time was pretty difficult.</p>", "time": "2025-03-20T07:24:57Z"}, {"author": "Suresh Krishnan", "text": "<p>We would love to catch you up after, and work on a further interim if needed</p>", "time": "2025-03-20T07:25:44Z"}]