Meeting minutes
<Wip> transcriber-bot, help
<Wip> transcriber-bot, connect
Will Abramson: Okay, I think we'll get started, and, you know, people can trickle in...
… Late? I know Otto is gonna be a few minutes late. Uh
Agenda Review
Will Abramson: Oh, no, great. It transcribes what it's doing. Okay, yeah, so I'll just do a brief agenda review
… I wasn't here on Friday, so I'm maybe not quite up to all the context, but I did have a chat with Otto. um
… Today, you know, we gotta discuss the CR exit criteria, according to the Charter, for did resolution. I raised an issue a couple weeks ago, and I just want to say what people think about it
… Uh, and then we'll review the issues, just to finalize and check that we think we're ready for TR. I think all of them have now been labeled, so thanks for that, Stephen
… Uh, and then, hopefully, with the rest of the time, we're just gonna focus on the updated diagram for DID resolution, although I don't yet see Joe on the call, so maybe I'll just ping him. Um
… And I see Manu on his handle
Manu Sporny: Uh, is the transcriber transcribing? I don't see it doing anything...
Will Abramson: Oh. It's transcribed… I can see it transcribing...
… Huh. That's alright
Manu Sporny: Sorry, I was scrolled up. My bad...
Will Abramson: Um...
… Okay, that was… that was it, right? Yeah, good. Cool. Yeah, actually, I'm just gonna ping Joe, and then we'll
CR Exit Criteria according charter
Will Abramson: Uh, okay. Alright, for that
<Wip> w3c/
Will Abramson: So, first topic is CR exit criteria, so this is… Issue 343, um
… There's really a question from
… me that I had when I was reviewing the charter about, like, what it means to be successful, like, how do we advance to proposed recommendations, and maybe I'll just read out what it says. It says, in order for a did resolution to advance to proposed recommendation, it is expected
… that each of the independent implementations mentioned above support at least two bid methods
… Which one have an open specification, freely available specification, fully implementable, and two are implemented interoperably by more than one of the independent implementations mentioned above. This means that there will exist at least two openly specified did methods that are each implemented by at least two independent implementations
… Now, like, I just flagged that because… and I'd like to have a conversation about it, because I
… kind of wasn't aware that that was what we were going for as our exit criteria, which, you know, maybe just changes. It's kind of saying that every resolver that is an implementation should support 2D methods to be considered valid. That's what I'm reading it as. Uh, and also, we need
<Wip> > In order for DID Resolution to advance to Proposed Recommendation, it is expected that each of the independent implementations mentioned above support at least two DID methods which 1) have an open specifications (freely available specification, freely implementable), and 2) are implemented interoperably by more than one of the independent
<Wip> implementations mentioned above. This means that there will exist at least two openly-specified DID methods that are each implemented by at least two independent implementations.
Will Abramson: two resolvers to export two of the same, these methods. But, uh, I'd love to hear what people think. You know, maybe we just drop this text in there
… Wow. Uh, and
Manu Sporny: Yeah, it, um… I mean, we can debate what it means, um, but I think what we could just do is make sure that at least...
Will Abramson: While doing that, I see Anu on the queue. Okay. Mm-hmm, mm-hmm...
Manu Sporny: there are two resolvers that support DIDKEY and DIDWEB, and call that a day, right? I think that… that would meet...
Will Abramson: Okay...
Manu Sporny: the requirements in the charter. Um, so that's my concrete proposal. um...
… As far as, like, yeah, I had forgotten that we had put that language in the charter as well. I think we were being very hopeful when we created the charter because we were like, surely, you know, we will have, you know
… enough time to create enough resolvers, and surely, you know, DID method, the standardization process will be well on its way at that point, and because of that, it'll be easy to achieve that. Um, you know, um
… things have not worked out in that way, uh, and so I think we're at a point where it's a little more difficult
… Um, to achieve that, but if we just do the did key, did web thing, and have at least two resolvers do that, Digital Bazaar can commit to having our resolver do that. I think Marcus's resolver already does that, so we can already meet that exit criteria. Um
… I think it would be very difficult for us to argue that thing that's in our charter away, so I don't think we should do that. I think we should find, you know, a way that we can be kind of true to
… the, you know, what we were going for there. Do we as a group believe that that is necessary for interop? I don't think that we do. I'd be surprised if we thought that that was the bar or we think that that has to be the bar. Um
… So, you know, I don't know if it really, you know, if we really need to kind of debate that or not. I think there's a
… Semi-easy out here, and we should take it and just proceed. That's it
<Wip> https://
Will Abramson: Yeah, uh, I've got myself on the queue to say, I mean, I think that is fine. I mean, it's still a bit like a box-ticking exercise. I guess, like, I'm wondering...
… Kind of how our test suite would be updated to, like
… represent that, because, like, for us, I think our definition of conformance is obviously conformance against the spec for any DID method, right? Like, you have… you have a DID that can
… exercise all of the normative statements in the spec, and successfully, you know, successfully pass the test that we've got for those normative statements. There's a separate thing here, which is, like, not just that, but you… your implementation has
… is conformant for two DID methods
… Two different did methods
… And there is more than one implementation that also is compliant for those two. I think your suggestion is great, like, I mean, I'm sure we can get that there. Like, I would love to, like, move to that direction, like, maybe we can get… I see we've got digital bazaars in, but I'm pretty sure it's the old one, and I know you guys have done
… some work, so, like, maybe we can just. Put some time into trying to get to this passing bar
… Uh, as soon as possible. I know also, like, the BTCR folks, like, uh, Dan and stuff, have been working on a number of different libraries in different language, you know, programming languages. I'm sure they could come up with, uh, two resolvers for
… their DID method, and then I guess it's, like, maybe you just add the profit key to pass the bar. But I guess that's the first thing I want to say, like, I mean, I think we're not saying that you are conformant if you support two DID methods, we're just saying there's this extra bar that we've put in our charter that means
… we can pass, uh, to become a proposed REC if we have two resolvers that are compliant to two of the same DIVM
… I think is what we said. Mane?
Manu Sporny: Yeah, I think that's right. Um, the other, uh… I think the way that we can do it is in… if I remember correctly, the… the way that we register DID resolvers is that you specify your endpoint, and then you specify the DIDs...
… the URLs that should be used to resolve against your resolver. And we just need to make sure that these two resolvers use the same did key value and the same did web value and result in equivalent
Will Abramson: Yes...
Manu Sporny: did documents for some definition of equivalent, right? And so, I think that's how we concretely test it. Um, we can add something in the test suite that's like...
Will Abramson: Mm-hmm...
Manu Sporny: you know, how many of the resolvers, when I ran the entire battery of tests, how many of the resolvers resolved this specific DID URL? And we just need to… and were the documents equivalent? And we just need to get that number up to 2. for two different DIT methods, I think...
Will Abramson: Great...
… Okay, yeah, that all makes sense. I guess, like, I would just like to make that a priority, and I mean, I guess I kind of took on the test suite, and I'm happy to do that, but I'd love to work with maybe you, Mano, or someone from your team, and I'll get in touch with Marcus, just to see if we can
… uh, get to a point where we're testing that bar and, like, seeing which kind of, uh, test cases are still failing. I think Marcus had a few based on, like, the updated errors
… to the URLs, and I also have a task to kind of review the test suite and just check that the normative statements I was testing are still in play
… Um, because I think they've shifted around a little bit, but I think… I think they're mostly still good, so
… I'll, I'll try and report that back on that soon
… Does anyone else have any comments about this? Like, is anyone concerned about this path? I mean… I know it's kind of
… kind of just came up, and it wasn't, like, on our radar, but I think it's good just to tick this box and have done with it
… Steven?
Stephen Curran: Yeah, my comment there. I'm just. I can't figure out why this is difficult. So I'm struggling with that. And I'm I can. Is it? Is it the test suite itself getting it running like the universal did resolver. Obviously. Resolves, like, 40 different types of DID, so it should be...
… One of them as a slam dunk, and
… And so I'm struggling with what is the bar and what needs to get done. I did look at the test suite and quite frankly, couldn't figure it out, too much. And so didn't quite know what to do
… to try to help move this forward. So if I can help, let me know. But I'm not quite sure, because I don't understand what the what the test suite is asking for. The the did resolver one seems to be straightforward. Here's an endpoint, and here's some dids
… Um, and if that's all we need. Doesn't seem hard. That's why I'm confused. Thank you
Will Abramson: Yeah, thanks, Stephen. I mean, I could quickly, uh, you know, take a moment and just kind of do a quick review of how the test suite works. Always my understanding of it. I haven't been in it for a little while, but maybe that would be helpful. Um, the other thing I'll say is, is, like, my hes… it's not hard, but my hesitation is,
… like, for me, this test suite tests conformance to the spec...
… And this thing that we're trying to prove is not about components to the spec, I don't think, it's about this extra bar of, is this spec worthy of being promoted based on the success criteria that we defined in the charter? uh
Stephen Curran: Yeah, no...
Will Abramson: So it just feels like a little bit of a different thing too...
… uh
Stephen Curran: Okay. I understand that. I'm just saying it doesn't seem hard. And anyway, yeah, enough. I'll I'll stop...
Will Abramson: Yeah. So I'll hand to Manu, and then, uh, I'm happy to take just two quick minutes, like, a few minutes just to review the test suite for folks, in case they don't understand it. So, Manu?...
… Yes
Manu Sporny: Um, yeah, let's, um… and I'm gonna ask to go off the record here in a second, Will, just FYI. This is on the record. Um, I mean, this came from, effectively, a very long-standing request from the Google review that basically asserted that, you know, if we could not prove that...
… you know, did methods were interoperable, and it was a bit vague about what they necessarily meant by that, um, that, uh, we did not have interoperability. And this, it was a request that they put into the charter that said, if you can't prove this this time around
<Wip> transcriber-bot, pause
Manu Sporny: Then you shouldn't be able to have a REC, right? So that is, that is the, the thing that, you know, was, uh, you know, expected. Um, it should be a, it, as you said, it should be a fairly easy bar. to hit… um
<Wip> transcriber-bot, resume
Will Abramson: Great, thanks. Did you have more to say? Um, no. Oh. Perfect...
Manu Sporny: No, other than this should be a fairly easy bar to hit. Did key, did web, same values. Let's put the, you know, show us kind of, Will, where in the test suite we can, uh, do that, and...
Will Abramson: Sorry, Miles, like, some of your content got on the record. I don't know, maybe, like, the queue got blurred at all? I think it's okay, right?...
Otto Mora: It gets substituted...
Will Abramson: Well, maybe we can remove it if not...
… Okay, um
Manu Sporny: Let's move on...
Will Abramson: So… I take what you said, thank you. Um...
Joe Andrieu: Ah...
Will Abramson: Yeah, great. So, I'm just gonna… Uh, what do people want? Is it worth me taking a quick...
… 5 min. Just dive into how this test suite works. Or, Stephen, do you want me to catch up with you separately and just go over it together. I'm happy to do either. I I would like to get to the dereferencing diagram today. So so maybe not. But if if everyone on the group thinks that would be useful, I'm happy to do it
… Yep
Manu Sporny: Can we skip it? Like, it's an implementation detail. I mean, Steven, if you're really interested in it, maybe you can chat with Will offline about it...
Stephen Curran: Yeah, that's fine...
Will Abramson: Yeah, reach out to me, Steven, we can… we can… we can get a call...
Stephen Curran: I I'm just trying to figure out how I can contribute and thought that might be a way. But so we'll let's chat...
Will Abramson: Great. Thanks, Rob. Okay...
Review of Issues for CR
<Wip> https://
Will Abramson: So, next up, we just want to have a quick review of all of the issues on the resolution
… Uh, to confirm that
… Oh
… We are ready for CR. And I think we just discussed 343, and for me, that is the only issue that's left that isn't labeled during CR or future or blocked
… And I… so I think that is great. I just wanted to confirm with folks that that's our agreement, and I see, Manu, you owe from this 349 issue, which is kind of a review of, like, what we need to do to get ready for CR
… I think we're going to go over that tomorrow
… Um… So maybe we don't spend much time on this at all, but… ah. I just want to check that the group's in agreement, like, we are
… Ready for ZR, and there are no issues left that we need to address, other than getting responses from the
… reviewers that say, yes, you know, we're happy this is complete. And we have reached… we have reached out yet again to all of the reviews that are outstanding, and asked them for comments
… So, you know, it's just chasing them up and then hopefully getting on VA to chase them up too. So
… You know, maybe we don't need to discuss that at all, but if anyone has a comment, please jump on the queue
… Okay, great, not seeing anyone
… Oh, there you go for one
Manu Sporny: Sorry, I'm struggling to find the window, Q+...
… Um, Joe, I know you did a pass, and do you feel like we're in good shape? The only thing that I saw missing there was usually in the status of the document section, you specify explicitly what the exit criteria from candidate recommendation is supposed to be. I don't know if the W3C process has changed so that that's not in
… That's not listed anymore. I know you took a look at it. Thoughts?
Joe Andrieu: Um… yes...
… Can you hear me? Oh, you can't hear me. I'm live. OK
Manu Sporny: Yes. Yep...
Joe Andrieu: I did not get that in. The other thing we needed was the URL for the test suite...
… So that is a… you caught a thing I had forgotten. I just grabbed the exit criteria from, uh… Um
… Well, actually, just changing it to CR added a bunch of boilerplate language, and I thought that that took care of it, but you are correct. I did not add the specific exit criteria
Manu Sporny: Yeah, the boilerplate doesn't specify what the exit criteria are, because that's up to the working group, so we have to explicitly state it...
Will Abramson: Great, yeah, and I just put myself on the queue just to mention about the test suite. I mean, that is on my plate, and I've reached out to PA, who needs to be the one who transfers that across, right, so hopefully soon...
… I mean, I've made him admin of that CCG repo, he will be transferring it across to the W3C repo, and
… changing the name to Bid Resolution Test Suite, right? We all agreed we wanted to get rid of the marker. So yeah, I hope to have that with you soon, um, and I guess… We don't need to have the test suite
… ticking all the boxes, right? Like, in… that's something that you can do in CR, right? Get the implementations, and uh… the test suite, is my understanding, is needed to be passing to be exiting CR. I think that's right
… So I think we're in good shape with the test su. Basically
… Okay. Uh, nothing else from that, then I will move on
DID Dereferencing Diagram
Will Abramson: Okay, next topic, last topic for today, we'll probably hand over to you, Joe, for the DID dereferencing diagram, uh, DID URL dereferencing diagram, I guess it should be
Joe Andrieu: Okay. Let me share my screen. The, uh, one thing I want to ask and confirm is, um...
… Uh, I believe this image I'm about to share
Will Abramson: Mmhm...
Joe Andrieu: is in the current PR, but you won't be able to see it with PR preview, so...
… Someone will have had to download the latest PR to double-check that, but I'll share my screen so we can all see it
… But if someone could confirm that it is in the PR, if folks have seen it
Otto Mora: Uh, I'll do it. I'll check it...
Joe Andrieu: Okay...
… So let me just walk through this head to toe. I incorporated a bunch of changes
… Uh although I don't have a crisp change list which I wish I had. Um so we'll just start at the beginning Uh first of all if there's a service query parameter. Sure
Will Abramson: Can you… can you zoom, Joe, as you go? It's a little hard to follow. Better thing...
Joe Andrieu: Sure...
… Um, okay, is there a service query parameter? If there is, we check to see if there are multiples
… Yes is an error because there should only be one service parameter. If there is a service parameter, I mean, if there's only one service parameter, then we use a service selection algorithm to figure out if it applies
… Um, which is, uh, simply checking the ID property, um
… I'm sorry, it is checking the value of the service property against the ID property of the service object in the DID document. And if that service type has a handler strategy that's been defined and published, then we use that
… If it hasn't, because nobody really realized that this was something we'd need to do, then we have a default service handler strategy. And the algorithm for this is completely malleable. I put something together in the spec, but with regard to the flow, I think we should have a default handler. What you do if you don't, um
… have one specifically defined for a specific type. And that ends with you have two handlers. You have set the handler strategy on both those routes
… If we did not have a service query parameter, then we check the pass part. So if there's a pass part, then we apply the path handler path matching algorithm
… Um, and if there's at least one path handler that matches the current DID URL, i.e. one that has basically indicated, hey, uh, this is the… the path range that I am, uh, to be invoked for
… um, then that's good, we continue on. Uh, if we don't, then we go to the next step, which we'll come back to later, but if we do have a path handler, then we check to see if there are multiple equivalent matches
… Um, if there are, unfortunately, that's an error. Um, if there aren't, then we get a path handler… Uh, we checked to see if the type
… That the path, how do I say this? We check to see if the path handler object also has a type that specifies a particular path handling strategy. So you might have a type is path handler and path service
… And path service is how you would specify the type of the path handling
… If that's true then you set the handler to that defined by that secondary type. Otherwise you raise an error. Um
… And, uh, if I didn't say earlier when we raise an error, let me actually see if I can close that. Yeah. If we raise an error, uh, then we use the default DID document strategy, which is you just return the DID document. In any case of any error in this file. Okay
… Uh, I am not looking at the queue if someone's managing that
Will Abramson: No, don't over here...
Joe Andrieu: No one, okay...
… All right, if we did not have a path bar here, then we check to see if there are document properties that defined a handling strategy. And this is not
… Um, oh, that's a very interesting gap here
<ottomorac> Diagram link for minutes: https://
Joe Andrieu: Maybe that's not a gap
… So the thing I was contemplating so you guys could think through it is, is this definition of document properties based on legacy? I we just understand we know through outside knowledge that there are path there are properties like linked resources. Or, is it because there's a path, um
… Uh, pan… path handler type in the object
… And I think it ends up not mattering. Um
… Because the the path handler type only applies if there's a path so that would have been an addressed up here
… Um, so this is for document properties that are not, uh, do not have a type of path handler. So these are legacy document properties. And if there are any that apply
… Then you you go and check to see if there are multiples that apply. And if they're not multiples then you set it to the strategy defined by that property So if there's one linked resource that's applied you set the handling strategy to whatever it is for linked resources
… If there are multiple salient properties, then that's an error
… And we get the DID document back
… But if there are no legacy properties, then we move on to check the did method specific handling strategy next
… And that is also one of the legacy fallbacks of if there are, then we check if the did method specific strategy applies to this did
… right? This particular did URL may not have the information in it. That would make that method apply. But if it does, then that you set the handler to that defined by the did method. If it doesn't, you go to the next legacy fallback
… Which is you check query parameters. So there may be query parameters that, change the path handling strategy. So if you know about those and you see them, then you can handle them. You set this handler to that defined by the query parameter specification, and that, ends. the flow
… If there are no query parameters, so at this point there are no document properties no did method specific and no query parameters. Finally we check the metadata properties So this is the metadata that that comes from the resolution. Um
… And if there are then we check if there are multiple. That's another error case we might have If there are multiples that just goes to an error If there's just one then this might be did link resources, then we use that And so we set the handler to that defined by the property
… And that ends the process. If we get through all of this, and we have no specific way to handle this particular DID URL for dereferencing, we don't have a handler strategy, then we go to the default. Uh, which is a DID document strategy
… So hopefully this incorporates all the feedback we got Um I know Steven you had quite a few things I think I I got that in those in here. But please let me know
Will Abramson: Yep, I see, uh, Manu on the queue. Manu?...
Manu Sporny: Yes, uh, this is great. Uh, awesome. Thank you, Joe. Um, definitely addressed the concerns I had, uh, and it is… I feel like...
… really easy to follow. Uh, I think this is the first time I've been able to really follow, you know, um, the entire algorithm. Um… I don't know
… So it's great. I mean, I think we should merge this, unless there are, you know, details that… Steven, I'm deferring to you on finding any, you know, details here that are missing, but even as a first pass, this is really good. Like, this is way better than a first pass, right? So getting this in, and then… You know, updating the algorithms
… to follow
… Uh, feels like a
Will Abramson: Hmm. Okay. Thanks, man. I'll do...
Manu Sporny: Good concrete next step...
Joe Andrieu: Right...
Otto Mora: Yeah, just to confirm, Joe, on my side, I did check the PR, the diagram is there, so… so, yeah, this diagram, as you're showing it right now, it is there. Obviously, you just adjusted that line, so that's obviously not going to be there yet, but uh...
… Yeah, there, uh, on that last, uh, diamond, uh, that multiple affecting meta, meta, maybe you just say multiple affecting metadata properties, just to… Make it more readable
Joe Andrieu: Yeah. That's I'm happy to make those adjustments. Yeah. It's easy...
… Thanks
Will Abramson: Great...
… Uh, Steven?
Otto Mora: But other than that, yeah, it looks good. I echo Madeline's words. Great job...
Stephen Curran: Yeah, the last one still bothered me just slightly in that it feels like if you just put a bare did in...
… And you had a DID linked resource
… You would
… you would say, oh, there's a deadly resource here, so I'm gonna use it, versus if it's just a
… did, if there did, you ought to be returning the did doc. So there still seems to be a slight
… issue there. There's gotta be some trigger… to me, there's gotta be some trigger. in the URL
… to do something other than just return the did doc. So that's the only in there. Most of my comments were about the text. So
… obviously wouldn't come up here. Um, so there's still things in the text, I think, that
… But… and… and I didn't know if you'd address those… the comments in the text
<Zakim> JoeAndrieu, you wanted to talk about text
Will Abramson: Thanks, Steven. Uh, Joe's next, I think. Joe?...
Joe Andrieu: Yeah That's a great add Steven I did get salience in for most of the the other threads but not for here So let me I don't know if I wanna put it before or after this multiple probably before because I think that's the pattern I I do up here...
… Um, so that's a good add. Thank you for that. Um, the text has not been updated. Um, I didn't want to go through and do all the edits for a particular graph if we were going to have changes here. So my commitment, um, as many of you know, I will be out the next two weeks, um, but my commitment is to get this document updated and get the text
updated
Joe Andrieu: this weekend before I turn into a pumpkin. So we will have an update of the draft. I welcome feedback and comments on that. But I'm going to try and address everyone's comments, or I will flag and respond to them and say, I'm not sure how to resolve this, or this doesn't quite seem right
… Um, but I will do that thoroughly now that we've had this conversation, and blessed a version of this graph that just has one edit. I think that's a really easy edit to get in there
Will Abramson: Great. Uh, thanks. I… yeah, thanks, Joe, this is… this is awesome. I… I just had one last… one comment about, like, what… I mean, what do we do about this, sort of, like, overhanging text? Is that, like, you know, like, the green triangle's kind of got a bit outside? And the yellow triangle, also, like, it's a bit hard to
… read. I mean, does. And it's kind of very minor. Right...
… Right around. Okay, well, I think it's fine. Okay. Thank you
Joe Andrieu: So there's almost nothing I can do about it without spending design resources and turning this into an Illustrator file and handling over just custom currentness...
… Um, or we could recreate this flowchart in another tool. I'm… I am super annoyed by this tool. I love it, but, uh, in this case, I don't know why it renders that text in those diagrams so badly
Will Abramson: Okay, I mean, I think overall, I think it really doesn't matter, right? Like, this is so you have some hooks to attach the text to, right? Probably, yeah...
… Yep
Joe Andrieu: That's right. This will all be described immediately following in order. So, you know, I think it'll be right there as a more detailed explanation...
Will Abramson: Great...
… Uh, I don't know, Otto, if you want to talk to that, feel free
Otto Mora: Yeah, I mean, I have Figma. I mean, once we do have a more buttoned down version, I'm happy to...
… uh, converted into Figma, so probably a little more, uh
… Uh, like, it
Joe Andrieu: Cool. And, and...
… In fact, I could give you the export in XML here, and uh… you could probably get AI to translate it into a variety of formats. I didn't think about that option. I'm still integrating AI into my life
Otto Mora: Yeah, no, no, if you give me the XML, as long as I have the text, it should...
Joe Andrieu: Um...
Otto Mora: Straightforward for me to just spend 30 minutes or so...
Joe Andrieu: Cool. Yeah, I'll update this, and then I'll send you the resources I can pull out of it...
Otto Mora: Go on. Sounds good...
Will Abramson: Great, thanks. I see Manu on the queue...
Otto Mora: Oh… Yes...
Will Abramson: Oh...
Manu Sporny: Yeah, just real quick on the Figma thing. Every time we use a proprietary something to generate these things, the next editor pays the price because they don't have Figma, right?...
… You know, draw, having a tool that is available to everyone and will continue to be available to everyone over the next decade is important. Alternatives include things like draw.io, mermaid diagrams
… Um, and as long as you are able to represent it in those things, then you can hand it over to AI to make it look prettier and whatever, and generate the PNG. Um, also remember that for this diagram, the accessibility people inevitably will basically be like, where's the accessible description of this diagram?
… And we could have two responses. One of them is there's five pages of alt text that describes this beside the image. Or we could say, well, no, this image really is described by the algorithms
… I don't think they're going to like that answer because any image they want to be backed up with, you know, something that someone that is visually, you know, has sight needs can read through. So for all those reasons, you know, minus one to Figma, we need to keep it. kind of in
Otto Mora: Yes. SharePoint, thank you...
Will Abramson: Okay, great...
Manu Sporny: tooling that is going to be available to future editors. That's it...
Will Abramson: Thanks, Mike. Yeah, good point. Uh, Joe?...
Joe Andrieu: Yeah, I actually disagree with you, Manu. I think Figma is just as proprietary as Mermaid, um, and the other one you mentioned. Um, it is widely available, people use it globally, there's no reason to expect they're going out of business next year. To your advocacy, if you had pointed to something like C4, I'd have a different
… response. I think I would love to have an open source resource...
… Um, that we can use
… But you did not mention one. So I think going from Figma to Mermaid doesn't buy us more openness for water source
<ottomorac> excalidraw?
Manu Sporny: Well, how is Mermaid not open source? There are open source libraries that render it. Like, I don't understand how Mermaid's not open source...
<ottomorac> excalidraw/
Joe Andrieu: Um… Wow. Let me double check my fax...
Manu Sporny: That's why I suggested Mermaid is because there are tons of the open source renderers and at least I'm fairly positive that I'm fairly certain there were. I'll check as well...
<swcurran> +1 to mermaid
Joe Andrieu: So, right, so that's maybe a good call out to my strong pushback. The Mermaid Foal Editor...
… Uh, it's not open source, and it is not compatible with their rendering in GitHub. So, if you use their… their… if you turn off auto-update and start doing layout in their editor, it does not export to GitHub
… So, they're playing some interesting, weird games with their mermaid.live service, but there is an open source layer, so I… I… Maybe it was a little more harsh than, uh, necessary
Will Abramson: Okay, thanks. So, I mean, yeah, I think this is great. I guess I have a question, which is, like, what, you know, are there things that we can usefully… I mean, that was the agenda. Like, we got through it, I expected this to be a little bit more bumpy, but it seems to have gone very smoothly. So, I mean, we've got 20 minutes, we
… could give 20 minutes back to everybody. Um… I don't know. Are there other things that the group thinks we could talk about? I mean...
Otto Mora: Yes...
Will Abramson: The other things on our plate are obviously the threat model, and...
Otto Mora: Transition request, maybe...
Joe Andrieu: Oh, yeah, I have one thing, Will. It's just… If we have the URL...
… For the new test suite, which should just be a matter of, um, us deciding what it should be, and
… Either getting that blessed by staff, or just believing that staff will follow that lead
Will Abramson: And...
Joe Andrieu: then I can put that URL in the CR, because that's the last holdup on that particular PR. Um...
… And then once I'm gone, we can… if there are any further edits we need to that PR, you guys could just go through that and… and accept it. Um
<Zakim> manu, you wanted to URL for test suite and note transition request details.
Will Abramson: Mm-hmm. So… Yeah, I see Connor. Connor?...
<Wip> https://
<Wip> Is it that?
Manu Sporny: Yeah, plus one to that, let's just decide what the URL should be so Joe can finish his work, um, and then we'll just, you know, make sure that, uh, Pierre-Antoine, uh, updates. the repo to do that. Um, was the repo still in CCG land?...
Will Abramson: Yes, it is...
Manu Sporny: Okay, so we need to resolve to move it into...
… Well, you're a chair there as well, Will. This is us officially asking you to, you know, move it. I don't know what we need to do. I don't think we need to publish it as a report, or finalize it, or
<Wip> This is the repo - w3c-ccg/
Manu Sporny: because it doesn't have IPR implications, I don't think, um, and so we could just, you know, announce to the CCG that we're moving it
… and see if
Will Abramson: Yes...
Manu Sporny: I don't see if there are any objections. I don't know if they can object and then put it into W3C space and then my suggestion will is to remove the mocha in the name like the implementation doesn't matter. We just need a...
… You know, it's just gonna be did-resolution-test-suite. Um
… I don't remember if one already exists at W3C with that name. If so, we need to
… Rename that, archive it, and put the new one in place
… Um, and then the way we do that is we put Pierre-Antoine as an admin on the CCG repo, and he moves it into the right place in the… W3C organization
… And that should be the move over that we need there
<JoeAndrieu> w3c/
Manu Sporny: And then I'll re-queue to talk about the transition request
<manu> ^ +1 to the notion that the name is available and Joe should probably just use that name
Will Abramson: Yeah, okay, great. I mean, I tried to explain this earlier, but I didn't do the best job. I mean, I haven't announced the CCG, so I should have done that, but I have already asked PA to, you know, I made him admin of this CCG repo, and asked him to. The form T...
… uh, transition to W3C space. So, after this call, I will shoot out an email to the CCG
… and, I guess, notify them that it's gonna happen, and ask if anyone has concerns, but I expect most people aren't gonna
… be offered, right? Like, I mean, this is a repo that we've kind of created in this working group, I guess. Maybe it existed before, but… um
… I don't know if it did. Anyway, um, I can't say there's gonna be an issue with that, but I will send that email out, courtesy at least
… But yeah, I'm hoping this is gonna happen soon. So, yeah, the repo… the URL you can use, Joe, is the one that you said, I think
… I would expect that to work
<Zakim> JoeAndrieu, you wanted to confirm the TR url is the right one https://
Will Abramson: Oh, right, yeah, all these TR, yeah. So, uh… can I go to Joe first, and then you, man? So I just want to mention about that
Joe Andrieu: Yeah. I just wanted to confirm. Is is that is that right? We're gonna use TR, not the, like, GitHub?...
Will Abramson: It is, yeah...
Joe Andrieu: I.O...
… Like, we are… then… because there are two… that's two different processes. So, one is getting PA to
… Create a repo, which we could access with the github.com URL
… Um, the second thing is to take what's in that repo and release it under TR. And those feel like two different procedural steps. So I'm not sure we are
Manu Sporny: To simplify things, Joe, let's just use the GitHub page, not TR space. I think that's totally fine...
Joe Andrieu: Cool. Cool, that's easier...
Manu Sporny: From a process perspective...
… No
Will Abramson: Uh, great, and just last on that, like, is going to TR with this something that we should be doing eventually, or is it not? Okay, great...
Manu Sporny: No...
<JoeAndrieu> I'll use w3c/
Will Abramson: um...
<Zakim> manu, you wanted to discuss transition request
Will Abramson: Okay, so we use that WCC, the GitHub repo URL, yeah, the one that you just put in, Joe, and
… I'll let you know when Pierre's got round to them. Okay, Manu, over to you
Manu Sporny: I'm, uh, looking at what we did for other… Um...
… Hold on one second. I'm looking at what we did
… Yeah, we… that's what we did for other… I'm looking at, like, um, the data integrity test suite, and… and we… we published… Ah
… not as a note, just as, like, a un… a bare document. Respec has a… I forget what we… it has, like, a base property, so no W3C anything goes on it. And we use the GitHub pages. Uh, link. Oh
… So, that should be, you know… uh, anyway, that's what we… that's what we did. For other things, it should work here as well. Um, I queued to mention the transition request. I think we should… well, in the… in the remaining time, I think we should go through
… all of the fields that we need to fill out in the transition request, so will you and Otto know what information PA's gonna need to fill out, because he is almost certain to ask you
… Uh, or Joe, you, uh, about what he should put in each field
Will Abramson: uh… Yep...
<Zakim> ottomorac, you wanted to discuss what we discussed with PA about transition request
Manu Sporny: I don't know if y'all have already gone through that with him, but, you know, let's make sure that...
… There are no extra questions, you know
Will Abramson: That's great. I see Arthur, maybe he has some thoughts on that. Arthur?...
Otto Mora: Yeah, uh...
… I mean, yeah, I mean, I understand that there's, uh, quite a few things to fill out over there. What we discussed with, uh, with Pierre was that, uh
… we as long as like, obviously, this needs to be. The horizontal stuff needs to be in in purple like I follow up on each of these individual ones to like. Ask folks if we can close the issues, and we're waiting to hear back from them
… But on the on the transition request that I know you provided this sample that was done by Ivan for the the Sid spec. What he said is that he's fine to do it, to to go ahead and do it as long as we have what he called the pre snapshot, which I guess is. It's the CR, um
… like, a snapshot that Joe will eventually generate, and then as long as he has that, he should be able to do it. But I think, yeah, your question is that all of these fields here, right, make sure that we have all of that
… So that, like, basically appeared so I could just almost copy and paste and just go ahead and put it in, right?
Will Abramson: Hmm. Okay...
… Yeah, I think so. So
Otto Mora: Mm-hmm...
Will Abramson: I mean, I guess, also, me and you can have that chat with him, hopefully on Tuesday, like, I'm hoping he… I mean, he's been away this week. Again, I'm not sure if he's gonna be here on Thursday. If he is, we can have a little talk about it then, but certainly on Tuesday, we should go through all of this, because hopefully by. You
… know, we are ready to open this issue, I would hope...
Otto Mora: Yeah, makes sense...
… Not sure if Manu wanted to say something about the data here in these fields
Manu Sporny: No, it just… do you… do you guys have any questions on what you would fill out for each one of these? I just want to make sure that...
Will Abramson: Mm-hmm...
Manu Sporny: I scanned through it, it's fairly… I think, I mean, I think we have answers for all of those things. Do you guys have any questions about what to fill out?...
… Because we don't want, like, another week of delay to, you know… Put this in
… My only concern is the wide review, our liaison. Do y'all have a story for that? Did
Will Abramson: Oh, jeez...
Manu Sporny: Charter. So have we...
… What is, uh, who are our liaison?
… Coordinate — oh, this is great. No. Is this our charter? End date September 2021. No. Where are we? It's the newest version
… Okay, end date October 2026
… And our liaisons
… are the IETF Security Area Directorate, the CFRG
… No, is this right? Why are these people listed? They have nothing to do with… in the DIF Identifiers and Discovery Working Group. So we have people
… Yeah, we… Grace has shown up at meetings and knows what we're doing here, and we publish
… our findings. The ITF Security Area Directorate and the Crypto Forum Research Group are the only ones
Will Abramson: Mmhm...
Manu Sporny: That we don't… we have not been in communication with. Um...
Will Abramson: So you should reach out to them...
Joe Andrieu: Could… could you scroll up, Manny?...
… There was… there's something that triggered for me, but then he scrolled it off. Um
… It's about this request to transition. So the first request we did, I think, wasn't correct enough. We didn't have your guidance
… So… I think we do still need a proposal. Right, that a group is going to vote on. It's alright
Manu Sporny: That's correct. Yep, good call out, because we didn't specify what our exit criteria are...
… In the last proposal. And we didn't, you know, you usually specifically point to… The static snapshot
Will Abramson: Mm-hmm...
Manu Sporny: So we, in our resolution, we really should point to the static snapshot URL. We should say we want to publish it as a candidate, uh, first candidate recommendation...
Will Abramson: Mm-hmm...
… Okay
Manu Sporny: We should specify what the short name should be, and we should specify what the exit criteria are. Yeah, that thing on screen, that is exactly what we would want to see...
Will Abramson: And so, is that something we should try to run tomorrow? Like, will we have this snapshot URL? Alright, hold on...
… The URL is there
Manu Sporny: We already have it. I think it's not ... updated, Joe, but we can pass the resolution with what's there right now...
Will Abramson: Okay...
Joe Andrieu: Um...
… We did… we… well, my challenge is we did merge something in, um, but it was… it was wrong. Um
Will Abramson: Mm-hmm. Mm-hmm...
Joe Andrieu: So there is… because we merged in the PR from the other week, so there is a file that is at that URL that is not the actual snapshot we want approved. Um...
… Uh, but I can… I can get something today
… I can because what we need in the in that PR is the exit criteria and the URL for the test suite. So, I can have that ready for tomorrow
Will Abramson: Okay...
Manu Sporny: Excellent. That's all...
Will Abramson: Great...
Joe Andrieu: Cool...
Will Abramson: So tomorrow, we will run a proposal that looks like that, but specialized for bit resolution...
Otto Mora: Mm-hmm...
Will Abramson: Okay...
Joe Andrieu: Yep, that sounds good...
Will Abramson: Okay, great...
… Um, well, I think with that, let's finish 5 minutes early. I mean, we're supposed to finish 5.52 anyway, we don't often, but this time we shall. Uh, thanks everyone, I appreciate all the effort. We're nearly there. Hopefully we'll pass this resolution tomorrow, and then we can raise this issue. Ah, to transition
… Have a great day. See you tomorrow
Joe Andrieu: Thanks, Will...
Will Abramson: Cheers...
Otto Mora: Thanks...
<ottomorac> transcriber-bot, please excuse us
<TallTed> s/not updated, Joe
<TallTed> s/… this weekend before I turn/Joe Andrieu: this weekend before I turn
<TallTed> s/... this weekend before I turn into a pumpkin/Joe Andrieu: this weekend before I turn into a pumpkin/