 Hello, hey Clemens. How's it going? It's going well. Oh, that's good Um Already registering my regrets for the next two weeks. Oh vacation. I assume. Yes Going a place exciting or just hanging out. I'm going to A big lake in northern Berlin For a week and then No firm plans for the for the for the second week, but I might be driving around a little bit in the convertible Just nice to get away Yes, yeah, I've been I've been doing this and I've been posting pictures on the weekends from my Loops that I did through Germany seeing old towns, etc I've also put a bunch of stuff on Flickr. Cool. That was good during the COVID during the COVID weeks Yeah and unfortunately It's not looking like it's getting better for you guys Well, it's interesting I keep seeing different things right on the one hand I keep hearing Reports about the number of cases going up But I also at the same time see charts that show the number of people dying is going down Yeah, that's a trailer indicator though because people are in the ICU for like two or three weeks and then they die Yeah, that's that's what I'm waiting for is to see what it what it looks like in two weeks. Yes so Dove the better number to look at right now is actually the positivity rate, right? Which is the number of positive cases versus how many tests are being done? Not is going up along with hospitalization. So those are those are the more leading indicators interesting It's yeah, it's all terrible Hello Christoph Hello, I hear you having very nice conversations. Yeah, we're talking about death Before we get to the serious things We actually don't have a lot on the agenda today. That's that's good Well, so yeah, and I have not I just I just realized That I have not yet Tweeted about the fact that we did merge the skewer registries back. There you go tweet away. I will All right, Tommy. Are you there? And Ray, are you there? Oh wait, no microphone yet What a ginger you have a microphone yet? No ginger It's fascinating sometimes watching how long it takes some people's computers to like wake up and get in gear Hey Colin Good morning morning and ginger you have a microphone yet Yes, I'm here Doug. Excellent and Ray microphone yet. Yes. I excellent perfect Hello Nick morning Lance Hello, hey class Remy, are you there? Yes, hello, Brian. Are you there? Hello, good morning. Good morning. Hey, and how about Eric? Good morning. Good morning The name you there Yes, good morning. Let's see if I can spell your name right. Yeah, sorry It seems forever Sergey's Mike's just will not turn on how funny Oh, hey, Mark, how's it going? Awesome. Excellent. All right Boy, it feels like these three minutes are taking forever. I think we're into like month four now, so You're talking about the virus, aren't you? We're very We already had that discussion. That was a morbid discussion. You missed all that I was just saying that it takes forever for my clock to switch over to 1203 so I just I just pasted the tweet Cool. So you can spend the time retweeting it. Oh, here we go. Let's see what the tweet is while we're waiting Hopefully it's safe for work. Yeah I just randomly click on it while recording. There we go. Excellent. All right I'm missing anybody. I hope it's safe for work. Yes. I've had It happens Sergey you there Yep, I'm here. All right All right three after let's go and get started Let's see 17 All right, so anything from the community if you want to bring up I think we all old timbers on the calls and nobody knew. All right STK we do have an STK call scheduled after this one I believe they're just a couple of things on the agenda if we have enough people to make a quorum So please stick around for that if you're interested I don't see anybody from the STK working group. I'm just gonna check. Nope. I don't see anybody. So no update there Okay, before we jump into PRs anything from the agenda that I forgot to add that we should talk about first All right in that case Klaus you get to go first Come on. There we go. Yeah, so after the last week's discussion. I just did the changes I was asked to do how we discussed. So I added Nometer statement that it must be a level one template now I put it under type and Adjusted the hierarchies of those nested Headlines I also further don't I also added it to the examples Yep cool Any questions comments? Okay. Any objection to approving? Perfect. Thank you Klaus All right. So the next three are technically Too soon for us to approve. So I just wanted to draw your guys attention to it The I believe the first two or maybe they're all three of them. Yeah, the first two are all about SDKs in particular governance stock and Common procedures and stuff. We don't necessarily need to go through it here unless somebody hasn't they'd specifically like to bring up about these Like I said, I just want to draw people's attention to it. So you can review it comment on it We are trying at least in some cases to have a Consistence process and governance across the SDKs just for some consistency. So we will be looking at putting documents in the main repo under community So anyway, when you when you get a chance, please review those comment Okay, and we'll be looking to approve those next week, I guess Okay, any questions or comments on any of those people want to talk about? Well, sorry was someone trying to see something in there Okay, in that case moving on this one very minor Since it just came in this morning. Otherwise, I would have just only approved it Someone wanted to add into the dark demos doc a little demo. They had I'm assuming no one has any objections to this Normally, I would have just approved this but since it just came in I thought I put it up there. Okay No objection. I'll go ahead and get that one in there Thank you everybody. Oops All right slinky, are you on the call? No, he is not. Oh, yes. Yes, there is there you are of these three I think these are all yours. Are any of these three you would like to discuss today? No Okay, at some point we probably need to decide at what point do you want to discuss those or to close them out one of the two Okay, we don't have to decide right now. Just going forward. Okay Um So that was technically it for the the meat of the agenda Last night as I was thinking about Obviously the agenda. I was thinking that I mentioned the last week's call that we were talked there I mentioned we should probably start looking at doing some sort of Testing of the discovery spec itself whether that's a full-blown interop or just encourage you to implement it to see What kind of feedback they get or something I because I think I think the specs in a fairly good state for people to access their Playing with it. Um, and so I thought okay, maybe we could do some little interop type thing I was thinking something basic, right people choose to put up a service provider. Maybe use github as a sample trying to follow the The adapter that we set up for github with all this cloud of adapters mappings and stuff And then people can write clients that talk to the various services that are out there or servers that are out there I'm sorry discovery endpoints that are out there testing the various things that we define the spec Verifying responses or example You know if you if you do you get the same list of services from this query versus this query that kind of stuff You know does all the data in a data match? Does the searching seem intuitive that kind of stuff? Um, I don't know I was just sort of initial thought I had and I just wanted to sort of open the floor up for discussion See what people thought in terms of other ideas for moving forward on doing some sort of base level testing of this thing Remy your hands up if I may I was discussing with my team yesterday But like all cloud events and basically we are using so often github that they were like I'd like to see another type of example So it might be good to try to not use github for once and play it with something else I think it was a valid question from my staff Okay, any suggestions Yes Well anybody else have any suggestions or ideas in that space I say I guess to me let me ask a higher order question first Do people think it's too soon to start pushing people to do implementations and I it's my assumption incorrect It's never too early for writing code That's what good way to phrase it. I like that No, I mean like you came up with that Registry PR. Do you think it will be applicable to Asia as an example? This that the schema registry. Yeah, I like to give a full sample with like Asia and not It's you know, it's just it's good. It's just a document store in the end So we can you can use that for anything and for everything So my understanding my understanding is that Over do we have red hat people on the call? I lance there. Yeah. Yeah, so I think I think there's there's there's activity over in red-hand lands Eric had been had been making comments, I think I think there's some aspirations to go do some implementation and we have some aspiration to do some implementations obviously as well But yeah for the schema registry, there should be a You know a simple app Maybe a node or an ace p.net depending which which simply implements that interface Ideally just generated from the open API and then plums that on the simplest possible document store in the back Maybe just a blob store Yeah, that'd be great and I would do that if I found the time Right now, but I just don't So, I mean we're on where we are from a product perspective, we're looking and implementing that interface But it will be great to have not only one but actually multiple examples I keep people can go and build on that are not entangled in all the product stuff that we have But are just effectively clean start That's that's something that I would do but I'm also about to go on vacation so maybe maybe after but if there is someone who's Adventurers enough to go and venture off and validate that open API Document that would be great All right, anything else? Let me assume your hand is old, right? It was right. Yep. Anybody else have questions comments. I mean ideas on some sort of coding effort Is anybody willing to raise their hand and say yes, though Try to work on an implementation. I'm going to put mine up. I think it'd be kind of interesting Sorry descriptions all of here Your question is related to the schema registry implementation, correct? No Because I got misled with the previous sentence. Okay, so okay. Thanks. Yeah, no, I mean obviously, you know We could we can talk about schema registry to I just was focusing on the discovery one, but yeah Of course, sorry, sorry for interrupting that I think of the this discovery the discovery the subscription API and the schema registry are somewhat related and Having so so actually having a project that that tries to Build on top of whatever a simple database is tries to prototype them out all at once would not be terrible and It doesn't need to I don't think it needs to be needs to be, you know running in Kubernetes and being 400 lines of YAML deployment stuff Sorry if I'm insulting anybody They should be using Knative come on. Yeah. Yeah, I know so so I think if someone were to throw together Scott If if someone if someone were to pull together like a note app or or go up, I don't want to insult anybody Just keep on keep on Keep taking back by a very trivial little sequel database Or even a file store that would be just that would be just fine But I think we we want to go and see how that all looks and ultimately how that all looks in code whether whether someone who tries to implement it Ideally really someone who's not the author of the spec Can come up with something that that took basically to validate it. So I'm turns out I'm for two of the two of the three pieces. I'm probably not the person to go and do the implementation because I may have too much context in my head okay I think having a project that does that does all those things improves them out in the simplest possible way would be great Yes, I agree. I was just wondering whether you wanted to insult another project while you're at it, but that's okay We left provisions to be able to bind this data into other protocols, right? Yes, so we have the schema registries currently has it has a rest API because that's the lowing fruit But it should all of those things should work eventually with with other protocols for MQP we have a path Via the MQP HTTP over MQP spec, which is something that we have in oasis over in oasis in the in the As a draft and I think that's those two things might compose really well for that kind of an API Yeah, but we have to go and think about how we can go and map that on to the other protocols I it's just something that I haven't done yet So the way you see it is like to create another repository where basically we put that simple implementation. Yeah If by any chance I have time And I start I should start just on my own repo like on github and then we merge later. Yeah Yeah, okay. I'm not promising anything because like No, and and and if you want to turn that into the thing that you're gonna sell to the world fine It's just I think I think that what we need is the proof point that the stuff that we're doing here in all theory is Actually holds together in practice. Yeah, I do agree. Yeah, so more so more than one implementation would be great But if we can start with one and get get an implementer implementers opinion that will certainly help Yeah, well, that was my intent here was try to get some more multiple pages of this thing going and I Agree the Clemens that we graded. We actually did all three specs I was just trying to do one step at a time and I thought there were discovery speck has had more Reviews than the other ones at this point. So that's why I was thinking that might be good place to start What's the debate you're thinking on that I Don't know Monday. No, I don't know I Mean they gives you a weekend. Well, come on Scott. Um, I don't know. Do you want to I mean we could push for I don't know a couple weeks. Maybe I can't imagine that the the discovery speck is actually that difficult To be honest. I think the hardest piece is that piece is actually just figuring out what all the metadata looks like to be honest Yeah, I'm interested in my favorite language Hmm, okay, so I'm hearing Scott as a possibility I'll stick my hand up there Anybody else? Just so I know who to reach out to to poke on Just yeah, which is your favorite thing just guts Sorry Yeah, well, no go is his favorite Thanks I mean, I'm gonna put your name there since you said you may do something in this space Not that it means anything just putting your name there just for fun Okay, I have to ask a few colleagues. Perhaps you can also Too late. You're in. Okay. I Think I could I could do something Excellent. Okay Probably check on type script Okay, well, this is a good start. Okay So, let's let's see what we can do in a couple weeks I won't nag I may be remind people next week, but I won't nag too much next week But then the following week, maybe we'll ask for status and see how people are doing And see how it goes, okay Yes, sir, do you foresee these these implementations one day inside the cloud events as the case or or do you think this should be just external That's a good question. I think that goes back to what Clemens was saying earlier because I think what Clemens was suggesting He's my own words was maybe we should have a reference in reference implementation of this available Yeah, that is something we have not talked about. I know the SDKs kind of come close to that But we haven't talked about that for these specs yet. I think it's definitely a valid option If the group wants to do that, I don't think there's anything that stops us from doing it. Okay well because they are not related to cloud events in some ways, but Well, they're not related to cloud events other than we're gonna be sending cloud events They are related to the other specs that we're working on exactly exactly. Yeah, that's what I wanted to say. Yeah Yeah, I mean if that's what the group wants we can definitely do it, okay Anything else on the Interop or implementation discussions. Does anybody want to volunteer to look at the other two specs yet or should we hold off on? nagging people about those Okay, I'm hearing anybody jump up and down yet. Let's focus on that. I guess discovery first see how that goes Okay, any other topics for the agenda because we are technically at the end of the agenda It's my daughter's birthday. So she will be happy that I'm gonna be go upstairs again Well, don't get too happy to show the SDK call after this. So that's gonna be quick to hopefully yes Okay, in that case, let's just do quick final agenda or attendee list Kent are you still there? Kent or Oleg. I'm here. Oh, okay. I got a leg about Kent Whoops Okay Anybody else that I missed for the attendee list All right in that case I'll get grants is he's on the process of joining in that case anybody who's not interested in the SDK You are free to leave and enjoy the rest of your day We'll start the SDK call in about a minute or so. Thank you everybody Grant you finally got a microphone Welcome Kent. I don't suppose you came back yet. Did you we're able to come off mute? All right, does anybody else have any other topics for the SDK call? These are the three that I had thought of I had a question I guess it was the leftover from last week that I met or last time was As part of the template of docs and licenses and all of that did you guys discuss? Whether or not to add a code of conduct and if so, which one I don't believe that officially came up as part of the discussion But we probably should add one and my recommendation would be to just point to or copy in the one from the CNCF Anybody else have any other ideas? I'm totally sure I have not read the CNCF So I have no opinion I actually was funny is I happened to be looking at just yesterday because The was the workflow spec is you know trying to go forward within as an s as a sandbox project and They got held up because they didn't have a code of conduct even though technically it doesn't require them to have a code of conduct So they quickly added one and they weren't sure which one to add so I pointed them to that one So it seemed reasonable to me The way they was little Conduct is not a place where we should be innovating Probably not But what's what's funny about this is the CNCF code of conduct in the little section that says, you know If you're concerned about something that happened, you know notify us well when it in the notify us section it points to Kubernetes not to the CNCF, which I thought was kind of amusing So I did ask about that and they said yeah, that's a work in progress to fix that So that's the only thing that's a little bit awkward about their their code of conduct I have a lot of things to complain about it Kubernetes, but I don't think you want that in our in our spec Are we still picking on Kubernetes is that part of the same thread here that will never stop. Yes, okay All right. So I yes, I agree with you Doug that we should just use the CNCF on okay so let's do this Use CNCFs and So the quick question I Would prefer to just have one Code of conduct and have all the SDK repos point of for that consistency factor You folks okay with that and that way we can only do one. We only have to worry about one PR instead of like seven or eight. Yeah Yeah, okay in that case, I'll take the action and Just so that I can nuance that are you saying they each read me Should have a section that would point to a common code of conduct. Yes, okay Yes, I want to make sure that there is a pointer from each repo. Yes Yeah, whether the repo whether it's in the read me or their governance doc or something we can figure out that later But yes, each repo should point back to the main CE repo for the governance or the code of conduct. That's unfair. Yep So technically yes, I still have a gazillion different PRs to do so yeah Okay, cool All right Anything else before we jump into the other stuff Okay So these two actually are two PRs I put here not because I thought there was anything we had to discuss But just I wanted to bring them up to see if there's anything you folks wanted to talk about Are there any I act to be honest I have not followed any discussions going on on those two So I don't know if there's anything controversial or worthy of discussion If not, we can skip it, but I just want to give you folks the opportunity to Bring something up if you want to talk about it Lance should the the other one the process docs also Be brought up. Oh, did I did I skip one? Wait a minute? You got the protocol buffer representation. Oh Sorry I'm sorry, I grabbed the wrong one. Okay, there you go. Thank you. So are there any Anything on these two PRs and if he wants to talk about One thing about mine is that after I finished to write it I I recognize that I implicitly assume that Who owns the rights the right rights to for a repo as also the rights to To publish the releases and that's not true because we don't have automation in I think in most of the SDKs We don't have automation to perform the releases. So I think I think that's something that was already brought up in past and I don't know if you could in my opinion every SDK maintainer should take like an action item to To to to set up a release automation. What do you think about that? Anybody want to comment on that? Yeah, we're John is Set up the So there's some automation on these C sharp SDK ready But we are what we're missing to be complete in the net world is strong names strong name signatures And so John has taken the action item to go in and do that and I think once we have strong names Then we can basically just go in and Automate builds to and push them straight to to Nougat So first we're going to do that Okay, cool. Well, I think I'll take the action item for the rust SDK and if Scott if you can help me We can do it together the one On the goal and SDK Yeah, I was taking a look at that right now and I'll do also the one for John Okay Anything else about these two PRs people want to talk about? Quick question for everybody in the call. I was going to Have the full cloud events group approve these Is there any reason that we should not do that and limited to just the SDK maintainers? We might use the new voting process, which is explained in the PR That is an option Anybody have any opinion on whether it's requires everybody in in cloud events or just the SDK people Lance I don't have strong feelings about it, but it really only affects the SDK books right And remember correctly. What did you say in here for the process? Was it a week offline or something like that? I Was talking to me Doug. I don't know whoever wrote this PR up Who wrote it? No, no, so I said that depends on the situation Like the end over All the numbers are definitely provisional. I just throw them. I I don't have a strong opinions on the numbers. So Yeah, I was right. Yeah, I'm expecting from the people there some feedback on that But end over is one week and archive a project is two weeks Okay Well, so tell you what are both these PR is actually ready to go or they're still outstanding comments Think mine is ready to go and yours was this one, right? I did raise a question in there which Francesco suggested maybe just become an issue and that's that to do should we consider changing the default branch name? That's probably not worth talking about in this PR, but Something to think about maybe for the future. You're not master, but something else. Yeah Yeah, we can talk about that at some point Okay The reason I was asking whether they're ready to go is because if we want to do you know that one week time period Then we could say okay today's the day where no significant changes are planned Maybe typo type stuff, but between now and next Thursday is that one week period and we can just Nag the appropriate SDK maintainers to vote on on the two PRs Is that sound fair? So Doug? I I Shouldn't bring this up, but I will okay. Is there a limit to the number of SDK maintainers Is there a limit not that I'm aware of so one one language could have 50 maintainers building That is technically true the way I correct me wrong Francesco, that's the way you have it set up, right? We don't We don't try to say one one One SDK can overrule everybody else because they're so large you don't you don't deal with that situation, right? I don't deal with that. I know I don't deal with the situation or with the situation of adding maintainers I mean I I just left this for another discussion or time because it's a bigger discussion and I think Your point is part of the discussion. I think I'm not sure I don't know Mark is that something you think we need to resolve before we merges PR. Is that something we should look at later? I Think we you know just in terms of governance We need to understand wait wait. We're saying that we're going to improve this so for approving it What does that really mean? I'm not sure I understand the question I mean the rules are what they are in the sense that if we follow this it's right fine the question rules Then the SDK SDK maintainers will Be voting to to Make any changes, right? True. Yes Do you really need to vote on it or? Use it sort of a different inverse approach where okay people commented at some point of time. There is no conflicts there is no contradictions and That effectively makes it because I don't assume that each document is Once it's merged. It's not changeable. I mean things will change things will evolve and we'll enjoy those changes will be a Separately so for example looking at this document Sure, let's say merge it right now. It doesn't mean that it's not going to change But if there's no immediate conflicts, why do we need to wait and and let's say for the case if we do have 50 maintainers like get everybody to vote people, you know Just throwing it out there see if it sticks not 100% sure what you're suggesting We don't need to even vote on this or you're saying we could just vote on it and worry about these issues later No, I'm suggesting not to vote in it rather say, okay Well, it's been out there for let's say significant amount of time in five six days people the activity in terms of comments has stopped There is no conflicts. There's no contradictions It's it's mergeable. Okay. Yeah, okay, so thinking I'll get to you in a sec the reason I think this is Different than say a PR in one of the SDKs itself is because first of all, this is going to the cloud events repo itself and technically the process we try to follow is Unless it's so blindly obvious that As a co-chair I can do it myself like syntax type changes, right anything of any significance gets approved by the group All right, and that usually means Thursday phone call people get to review it at least two days in advance and we basically take a vote like we do on everything else so technically This should follow that it's actually in process. The only reason we're doing slightly different here is because This is really only about the SDKs, right? And so I feel like we still should have some sort of vote around this thing It's just the scope of who gets to vote. I'm just saying that like I kind of almost got it from your example Where you sometimes you say does anybody disagree and there's a silence and that means everybody agree. So Okay, I see what you say. Okay. Yeah. Okay. Okay. I apologize The reason this is the reason this is different is because people understand that when they on the Thursday phone calls We're gonna do votes, right? And if they choose not to vote if they choose not to join the call They know they can speak up on the on the PR to raise an objection These I don't think we gave fair enough warning if I if we had known in advance that we were gonna prove these today Then yes, we could have done that if I had mentioned it two days ago So people could have raised an objection. That's the only reason that this is a little bit funky So that might be an interesting thing which is Should we put into this governance doc that if there are major issues with the SDK governance that it can be escalated to the main cloud events group For discussion in voting at that level as well So in other words, as long as there's no problem with how the SDKs are maintaining and agreeing then the main group doesn't have to get involved, but Would the main group be an escalation process? Yeah, it's interesting what appeal that because I know for Francesco on the last week's call you You mentioned the possibility of having to introduce some kind of TOC kind of thing I think mark what you're suggesting there is something a little bit similar, right? The authority above the SDK it's in the each individual SDK. So it's like a your hands up Yeah, so two things first Note that at the end of the document it is clearly states that to change these rules We need a vote following the voting process that I that I stayed here. So I Mean if we if we pass these rules, then let's make sure we like it and then I See what mark you're talking about and I wish to address those issues too, but I did this draft just to Talk about security patches Quality of service which are which were long-standing issues In this community I Really wish to talk about these other things, but I think that we may need other PRs for that I'm here. Is that okay for you? I mean in general for the community if you want to take action items separately to To fill this document with the order bits we need like the one that you said about should when we have a problem should we escalate or The other one that you said about The what happens if one SDK maintain us overall the other so Can you comment to that? Sorry, I think what you're saying is that I should just propose additional language that I think see this PR Yeah, no what I'm saying is that just Or we add things to this PR now or we or we merge this PR and then in the next Iterations of this document we add these things. I mean as you prefer But I think that these things should be addressed as well in this document. It's just that I didn't All right. Well, let me look over a little bit more because I had some other feedback in different areas so Let me propose something Okay, the the the other thing with with this document That I was gonna comment on but since we're on the call. I'll comment here. It talks a lot about security patches and Doug do we have a secure way of Being being able to handle security patches Not where one but we have a do we have a mailing list that someone could use if they find a security issue and Not have to go through the PR issue path Yeah, correct me wrong, but I don't think we have anything like that set up yet, right anybody So, you know, I think I think that's likely something that we will need to address at some Do you mean like a mail like a security at? Cloud events. Yeah, yeah, probably we need it. I Assume I assume that we discover CVs from GitHub the panda boat. I know that the panda boat does Those PRs private PRs that only mountainous can see Right, I'll reach out to the CNCF infrastructure folks to see if they can Set up another email address just for security type of questions And we'll make that global for all of cloud events not just for a specific SDK and make it routed to the appropriate people Yeah, I you know, it's It may be unlikely that we see this in the in the main cloud events Code but as we're talking about subscription and discovery Those things might be more open to the internet. Mm-hmm. Yep Okay So I'm hearing that there may be some changes in particular to you. Well either one of these PRs based upon People looking them over like mark. You said you may want to suggest some burning changes. Um, I I'm thinking about the the one week thing that we've been talking about here our people comfortable with Waiting and to start that one week timer, which means we probably won't make next Thursday Or do we want to say, you know, wait one week But let's turn that on after emerges PR and for this PR We'll just go back to the normal rules that cloud events has which is as long as that the changes are done by Tuesday evening I don't I don't it's up to you guys depending on how it's how quickly do you want to force these things through I Think I think we should go a little bit slow on that. So yeah, that's okay Okay, so we won't start this seven day timer until we're all sure that everybody's in the reviews and has other comments And so we'll hold off and I assume that's true for both PRs. Is that true everybody? Okay Okay I'm sorry. I'm trying to go ahead. Let's go said that this PR primarily was Obviously is going to expand but primarily it's for security patches. Maybe it would help because for example I'm looking I'm reading it for the first time and I could provide a few comments or or areas questions and That could prolong the process Maybe we should slim down these PRs to say listen, this is a security patch part of it And then eventually we agree on that we agree on something else and so on and then at some point of time We agree on this chunks and then we can assemble the entire SDK governance document because We have things that are related to security patches here. I'm assuming and then things that are not related and that's that good. I don't know Coping issue that's just Anybody want to comment on that I have a comment, but a lot of people go first So so the way I kind of view that is I Don't I don't personally have a problem with splitting this out into several PRs, but I'd rather Wait until someone does review and says, you know what this particular issue is such a Scary beast that let's pull this out to have a separate discussion because I can't even live with this as a starting point Right, and I can't wait for a follow-up PR to fix it I'd rather wait until we come up to that situation rather than just assume we have to split it out right now That's just the way I kind of look at it Anybody else want to okay. Anybody else want to comment? I Mean, I'm not a unit of maintainer, so I don't get a vote. So it's up to you guys That was my immediate reaction was Was exactly what you just said is there that right now? Okay Okay Anything else related to those two PRs then Okay Whoops, hold on. There we go. All right, so I mentioned In one of the I think it was in the job SDK and one of the PRs there that we talked about this on this call Francisco, I don't have the issue in front of me, but you opened up an issue where I'm sorry you opened up a PR Where you created a milestone for the job SDK and then you merged that Relatively quickly and I my interpretation of what happened there because some people raised some concerns about that and my interpretation was that a Milestone is a relatively light thing. It's it's it's a little more real than just so just telling people to go grab You know the head of the of the branch But it's not Like a release candidate where he says hey, this is what we think is the final thing It's just sort of a well, I kept using the term snapshot in my head But I wanted to bring up that discussion because I know there were some concerns there. So let me let me get off and let other you will talk This is Oleg good just coming from the world where You correct the word snapshot has a very distinct meaning the word Milestone has a very distinct meaning and the world of this kind has a very distinct meaning So it's not just oh grab it from the head. No, we have snapshots for that for example I mean a lot of projects published snapshots daily nightly Every after each commitment and so on so forth there is different rules to do that at least in a job all right So milestone means you have some significant again Means you have something that can feature that you want a community to start Demonstrating so you announce it you explain what it is you say this is why are we doing this milestone? it could change but There it is out there people can start using it It's actually a step above the snapshot because we have something important to say so it's not we can't just release milestones just because enough time went by and then obviously as you said release candidates and so on they So and with that there are certain patterns Which people are familiar with especially within a particular community that they're going to be looking for and Eventually when we and so Sergei's comment was kind of related to that say listen There's maybe not not an issue for losing milestone because we don't some significant work that maybe you should change it to M1 Sorry about that something that can Well Understand and we'll accept and so on so forth. That's all so and it didn't go the right directions Yeah, so it seems to me that that there was just a For lack of better phrase just sort of miscommunication in terms of the process or the terminology and stuff And so the first thing that ran through my mind was okay The best way to solve this going forward is to just document what the document what the process is right document what we think It's okay document what a milestone is document release candidates are and as long as everyone understands and agrees on what they are How big it created when they get created with the process by which they get created That at least we don't have that miss that miss communication So I think it's more of a document documentation of our process issue more than anything else. Is that fair? Yeah Okay, so I have a I sort of a high order question here Is this the kind of thing that we should look for a commonality across all the SDKs or should as each SDK be allowed to do its own thing Oh, I definitely each does because patterns these patterns and expectations from the community Could be different in different language environments and so on so we can just put it under the single blanket and I Would be personally strongly against it. Okay anybody disagree with each SDK having their own process for this stuff My answer coming up muted you want to chime in? I Like commonality. I get the point that every language, you know has has its own idioms But but we're talking about things like milestones and stuff like that right and that to me seems like it should be a cross cutting sort of You know sort of thing So let me ask this it's my hands up Oleg When you talk about each repo each SDK should have their own that have the freedom to do what they need to based upon their particular language Style Do you see that being more As a release kind of a thing The reason I'm asking is because I'm wondering whether it makes sense to have commonality just in terms of terminology and and not necessarily when things happen So for example, let's say we defined what a milestone is and the rules for who can create a milestone when we define What a release candidate is how that process happens and then we decide What a release is and who gets to vote on whether release passes or not But we don't necessarily say that every single SDK has to have all three levels, right? Some may choose to have all three some may only have releases But at least then there's commonality in terms of how you approve each one of those or what they actually mean does that make any sense Yes, it does make sense and I agree with that. I guess just to clarify To lens my point It's not about go has snapshot Java has not shots, but go doesn't or we should not have it or whatever It's more about for example. Yes snapshot has Meaning we can define what that is and it doesn't necessarily have to follow a particular thing But once we agree what the snapshot is How do we identify the snapshot artifact itself within Java world versus within other languages? So and that's that's where I come in because for example with the Java world It's not just what we want. It's there are tools there are products out there who will do parsing Who will do just so give me the latest one and they will know how to pick the latest one So we need to be very careful whether we do it lowercase uppercase and one or milestone and so on So there is much more to discuss there. We who've been doing, you know Within spring community for past 20 years some years We actually had a meeting right now a couple weeks ago and had to introduce certain minor changes to our naming schema Because of certain things that I don't want to hijack the meeting for that as well We neither but I'm just saying that we can't just throw some names out there and and say that's that's how we do There is much more to these names then Okay, so so you're okay. It sounds like with defining the common terms and even possibly the common sort of Process by which those release are and those artifacts get approved and stuff like that It's just you want each SDK to have the freedom to do its own naming for lack of better phrase kind of stuff. I Mean, yeah that because yeah, that's there's just Moving parts that outside of our control that We may pay the price if if we don't follow certain because if everybody expect milestone to be an M1 and You don't then your thanks. You effectively remove yourself from The features and capabilities of that particular system of that particular process and if you try to build community I mean, why would you want to do that in the first place? Right. So is it so what is what I wrote here fair actually? Before I ask that question, does anybody else want to chime in? Because what I think what I wrote down is what I'm hearing from at least all I get a searcher. Jay, go ahead Okay, so as a off-tariff Commend the request I just wanted to point out that my main concern is not that was like the selection of the wording For example beat milestone or release candidate or whatever, but rather and we're seeing it not the first time but lack of any process in the Java SDK of Reviewing and accepting big changes like it was just dropped as a pull request and merged instantly without any Approval of any other maintainer and if I remember correctly, we do have more than one maintainer in the cloud events Java SDK which kind of suggests that is some formal process of reviewing should be done and my comment was just an attempt to kickstart the discussion But and it's okay if it would be ignored But I was surprised that other maintainers did not approve the pull request before it got merged and it went to the Mavin central by the way after second attempt not on the first one, so I believe Could review could catch that issue as well and Yeah, it's the same concern Maybe new file, but the same concern like how should we proceed with the Java SDK and How is it maintained and It's maintained it actually. Yeah, and I think that's what we're trying to address here is by defining that that process that way People are not surprised, right Mm-hmm. Okay So does this do these two bullets summarize we want to come and process Come in Artifacts right milestones versus release candidates versus releases But each SDK is gonna have the freedom to figure out, you know How they package those up what the names are whether they actually use all three or how many we have those kind of things Is that sound fair and accurate to what we talked about? as long as we are talking about to process like release process and Like the maintains process or a projects process. Yeah, that's that's Well thought we are talking about yes Yeah, well, it's both it's process, but then defining the actual artifact types themselves. Yes Because I think it'd be kind of weird if one SDK has milestones, but another one has Fuse, I don't know some are the word that implies milestone right snapshot, right? If one does snap shot one does milestone, are they the same thing or not? That's that's kind of weird to me Not the same things and just for your to complete your example There's also we don't use it, but there's some people use it with working progress, right? Right or something that is better than snapshot That's just you know one of those things, but I think the surrogate point is He raised a greater point The the the schema question whether it's milestone or M1 is just a particular incident But Sergey's point. I think you're correct me if I'm wrong that There was really no discussion. No debate nothing. It was just BAM BAM done Right, no, I understand and that's why we're trying to write down the process that way. There's common understanding Code or yeah, yeah, sorry. Yep Okay, um Okay, would somebody like to take an actual item to take a first stab at defining these two bullets? I Mean maybe part of the the governance stock thing they were talked about anyway slinky your hands up if you are gonna if you're going to automate the process then I Think we don't even need to document it. I mean That's that's what I've wrote on on on an issue after I did the release Well, even if so my my it seems to me that even if you have an automated process It still seems you need to it still seems like you need to write down The fact that it's automated or when it happens or when the automation gets kicked off, right? Just so people understand Oh, yeah, of course. Yeah, I'll to use the automation. Yeah, of course And when it happens who has the authority to kick it off that kind of stuff. I I think the bigger concern here isn't No It's just people you see you understand how the mechanics of the group is are going to be right How when things happen by by who when and stuff like that So that somebody who doesn't understand where isn't who's in part of the group they can read this document say oh, okay I understand why this milestone was automatically created, right? It's because the process says it gets created every day, you know every every Friday at 3 p.m. Right, whatever as long as it's documented. I think that's what people are more concerned about and then We need to publish somewhere. What are we actually calling that like like two series actually original comment Yeah, and I think I think that goes back to what you were saying earlier where each SDK was going to probably decide The the name of each of our the the syntax of the artifacts for their particular SDK Right Okay, does somebody want to take an action item or volunteer to to write up the proposal for these we can try it Okay, because I'm not sure I'd be I'm not the right person to do it Okay, thank you guys And to be clear this will be the initial pass will be at the cloud events level so it's cross SDK right Once we have that in place then each SDK can decide on their own what they're going to do with this first bullet, right? sure and You know, it's going to be pretty much what you already have there It's just going to be in a more formal document and like I said I'm not using other languages so other people can contribute in terms of what other typical artifacts are Being released or expected by the community within that language environment Yeah, okay Lance did you want to say something because I know she came up mute there for a second Yeah, I mean you kind of addressed it I was just wondering where this proposal was going to show up I'm assuming it would show up on this back repository, but then that sort of makes me start to wonder if there shouldn't be a Governance repository because there's a lot of governance stuff going into this spec repo Do you think we need a whole repository as opposed to just a governance directory Yeah, I don't know that's I'm not sure Really, I just wanted to know where this was where this proposal was going to end up and and I'm assuming it's in the spec Right. Yeah, I think for now. That's where it has to go if it's cross-cutting. Yeah Okay, anything else on this topic then that we need to bring up. Okay, any other topics for the SDK call All right. Um, I just want to say Looks like there's been some great progress on the typescript repo So great job Lance and team If there's I know there's a lot of rapid changes I Haven't been too involved but If yeah, I guess I'm curious as to like if there's any timing that you're thinking of Like it's publishing the module Yeah, I was in general, I mean You know actually some of this stuff that we're talking about right now in terms of You know artifacts and processes That that's why I'm keenly interested because I don't want to just publish it, you know, right now. It's a major change and And so I'm a little concerned about doing that But I also know that You know getting it out sooner rather than later would be would be great because it is Completely different than the existing API Okay Yeah, I guess The current as the last publish was 17 days ago So I guess does that include Some of the more recent changes with changing the API or would we need a major version bump The changes that went into place with the typescript rewrite were breaking API changes, so it would bump the Module to 3.0 Okay, I think we probably might want to consider Continuing to maintain 2.0, but I don't know. Okay cool Well, okay. Thanks Grant. Yeah, I'm glad to hear what you said about the typescript stuff I'm not following it too close. I just sort of see the emails go flying by so it's glad to hear the things are going well That's good Yeah, thanks for doing well. Yeah, cool great. All right any other topics then on the call. Otherwise, we will adjourn All right, we are done then. Thank you everybody and have a good rest of the day and weekend since it's almost Friday No, bye ready. Bye