Session Date/Time: 02 Sep 2026 14:00
[00:00:06] Barry Leiba: Okay. We are at the top of the hour, and it looks like we have a good set of people on. I think we can get started. So the first thing on the agenda is discussing anything we need to discuss from the last call of the working group last call of the serialization document. So, Laurence, I'll send it over to you. Tell us what what the status is, what you need, and what you think needs being discussed, and then we can open it up to comments that others think we need to discuss. Go ahead.
[00:00:46] Laurence Lundblade: Okay. Can you can you hear me okay?
[00:00:49] Barry Leiba: Yes. You sound great.
[00:00:51] Laurence Lundblade: Okay. So there were comments good sets of comments, I think, from Joe and Martin and, Michaela. So I've been working through those. There's a couple of PRs for, to address some of those, but there's still a ways to go. So I'm actively working on that, and, you know, we'll produce another draft in in probably another week or so. Probably two things to highlight here today for some maybe some discussion, although Joe is not here. I mean, one of them is some of the clarifications that Joe was interested in that I I don't quite understand what he was after. And there's been some discussion on the list about this and what's checking decoder and what's not. So that's one. And then the other is whether this is an update of 8949 or not. I believe the only thing that makes it an update of 8949 is the change in rules about tag numbers. Then the last thing I will ask about or I'll mention is the the CDDL for deterministic and preferred plus. I don't. I'm not an implementer of of CDDL and where I am of all the other stuff. And so I as far as I know, nobody has implemented these. There's there's a fellow, I don't remember his name right now, that does that may be implementing some of this stuff. He's he's made some really useful comments, I think, Steve. So I'll probably try to spend some time with people that are that implement CDDL checkers to try to to validate that, you know, they've at least looked at the possibility of an implementation of these. So that's so so three things to to, you know, kind of that are open issues right now to maybe have some discussion or thinking on, and that's the checking decoder and what Joe was after. That's the first. Second is whether this is an update to 8949, and then the third thing is getting some CDDL implementation experience. So comments on those?
[00:03:58] Barry Leiba: The floor is open. So
[00:04:02] Carsten Bormann: can you hear me? Yes. Yes. Just to to comment on one of these three things. The main issue for implementers of the CDDL extension is that right now, there is no requirement for a CDDL validator to actually retain information about the serialization variant that was used in an incoming data item. So that is something that actually makes this a larger amount of work than meets the eye immediately. And that's why I haven't really started trying to do something here. But clearly, extending CDDL is very important, and I sure would like to make some contribution to that.
[00:05:05] Laurence Lundblade: Carsten, are you saying that it's maybe more difficult, but it's still a good thing to do and and we should do it? Or are you saying it's so difficult, we shouldn't do it?
[00:05:21] Carsten Bormann: Well, I'm sure I can do that. So it's not too difficult for me. Other implementers maybe indeed should speak up as you you said, about how that works in in their environment. I'm just I was just trying to say this is not going to happen within a week or so because it does make wide fundamental changes to how my specific processor works on this. And there are also other things that are not currently clear. For instance, what happens with nested serialization variant declarations and and all this stuff. So we we we need to develop an actual CDDL extension here that probably cannot be done just in a chapter of of this document.
[00:06:17] Laurence Lundblade: So I do believe that it is I mean, there is there is text in there that says what happens for nesting. There's maybe some cases that need to be where there's a conflict that maybe need to be addressed. But, I mean, it seems reasonable what's there. So you're saying again, are you saying we should take it out, put it somewhere else?
[00:06:53] Carsten Bormann: I think that that would be the the best way to handle not just this, but a few other places in in the document where there are drive by shootings on on other components of the CBOR ecosystem. And right now, what's in there doesn't even pass. So I think some work is is required there, but this is this is the easy part. The The interesting part is actually finding out what we mean by this and whether there even are any conflicts or there is a natural way in which the nesting of these things actually resolves and so on. And typically, you you start to fully understand the implications of that when you have implemented it. So that that's what what I plan to do.
[00:07:48] Laurence Lundblade: Okay.
[00:07:56] Barry Leiba: Other comments on any of the three topics that Laurence brought up? I I guess I'll make a comment on the update part, which is not something we need to resolve immediately. But the reason we might want to make it make it an update even if it doesn't actually make real updates is if we wanna make sure that everybody who reads the other document reads this one. So consider whether we want that or whether this really is an optional piece. And if it's an optional piece, then we don't need the updates for it.
[00:08:37] Laurence Lundblade: Yeah. That's a good reason to make it an update in my opinion.
[00:08:41] Barry Leiba: That was my thought. Yeah. Any comments on that from others?
[00:08:49] Ira McDonald: This is Ira. I agree for that very reason. That's a good strong reason. I had market updates just to make sure that people read it for refinement or explanation of things that exist in a recording.
[00:09:12] Barry Leiba: K. So I we should probably why don't you, based on this discussion, add the add the update flag and put text in the in the introduction and abstract that say that this updates because it's an important addition to, what's already there. So it's clear why we're putting them in.
[00:09:40] Laurence Lundblade: Okay. Okay. Sounds good.
[00:09:46] Barry Leiba: Okay. Does anyone have anything they wanna discuss in in this other than the things that Laurence brought up?
[00:10:02] Вадим Гончаров: Carsten?
[00:10:03] Carsten Bormann: Yeah. Maybe one thing that we occasionally should bring up just for reference is what what is our timing that we see for this? And I think Laurence has already said a new draft is needed based on the comments. I will make some more comments in the course of this week. So that will contribute to a new version probably being needed. And what what is our time scale for that?
[00:10:38] Laurence Lundblade: I I think it's probably most optimistic would be sometime next week, but it's possibly more like two weeks for a for a new update.
[00:10:55] Carsten Bormann: Yeah. I I take that as an estimate and and not as a promise. So so thank you. Yeah.
[00:11:06] Barry Leiba: Okay. I I think we're set on this topic then. I'll just do one more call. Anything else we need to talk about about the serialization document? Okay. We're moving on then to the EDN. Carsten, anything you wanna say about EDN?
[00:11:34] Carsten Bormann: Sure. So right now, I'm mostly waiting for decisions. In in the one place where I actually could move forward, I have made a PR. And and I I see that Rohan has looked at that, but I haven't seen his comments yet. So that that will be one thread of discussion on on the mailing list eventually. If if there is any any natural stuff in there. And apart from that, I think we we need to decide which of the 11 proposals are going in there because it's usually a little bit of work to do that. So I would like to have a decision on those before I invest at work. I think we had consensus to remove stream stream, and we had consensus to for Christian's ellipsis removal and also removing CBOR 888 and 999.
[00:12:34] Barry Leiba: Yeah. The the Paul and Christian are the chairs who will be running through that. Unfortunately, neither of those could be on this call. So I'm kind of on the side on on the EDN part. The but they will they will have to go through the posts that have been made so far and and do their consensus decisions, and and then they'll get back to the working group on the list. Anything any other discussion anybody wants to have perhaps that arose from the list postings or something new to raise? Well, I think this is going to be a short call then. Go ahead. From
[00:13:42] Вадим Гончаров: what's the decision about new proposals made? It seems that end of summer is vacation season. So there were there were a few applies. But as I remember from June, the things that I pro proposed recently were supported much more that time.
[00:14:21] Barry Leiba: Well, if if one second. Have something in my throat. Okay. That's better. As I said, the Paul and Christian will have to go through the the posts that were made, the suggestions that were made, and decide which ones have consensus to remove. And when that list is posted, we can revisit the ones that did not get consensus briefly and see if there's more support for them. Yes. It's possible that we just lacked participation in August, so let's see what happens the first couple of weeks in September.
[00:15:16] Вадим Гончаров: If needed, I can provide a message with links to messages from beginning of summer when when these proposals has the ideas in my recent proposals have had more support.
[00:15:42] Barry Leiba: Okay. Let's wait for Paul and Christian to go through what they see. And if you think they missed something, you're encouraged to let them know what you think they missed and where they are. Does that work?
[00:16:01] Вадим Гончаров: So this is this means, first, to wait to what Christian and Paul will write write into list?
[00:16:13] Barry Leiba: Yes.
[00:16:13] Вадим Гончаров: All the okay. Correct.
[00:16:20] Barry Leiba: Okay. Any other discussion needed now? It looks like the EDN topic is done. Is there anything else that was not on the agenda that anyone wants to discuss now? Alright. Then I think we do indeed have a short meeting, and thank you all for coming. I hope that by the next interim, we can settle the EDN bit and move ahead with that document. And by then, we will also have another rev of serialization. Yeah. And and so thanks thanks again for coming, and we'll see you in two weeks and on the list.
[00:17:20] Laurence Lundblade: Thank you. Bye.
[00:17:24] Carsten Bormann: Thanks, everybody. Okay.
[00:17:32] Вадим Гончаров: Thanks. Bye.