 Good morning. Let's give it until five after and perfects to arrive. Hello. Hey, are you saying something because I'm going to hear you. No, you can hear. I did start. Can you hear me? Yeah, now we can hear you. All right. Yeah. Good morning to you. It'd be good afternoon for Tom. And we will get, we can get started five after I guess, Tom, if you're good with that. Yeah. Yeah. Right. Let's make a stop afternoon. Good morning. Maybe wait a couple more minutes if anyone joins. Still first for us at the moment. Right. Well, let's make a start. So. Um, I noticed last week, Victor, you said you were attending a KCD in Columbia. Sounded like that was quite good. About that event. So, um, I mean, it was, it was weekly. It was the first one, which was made in LaTan in person. So. So, um, actually open source communities in. In LaTan are not material. The Europe. Um, the good thing about it is, um, Greece attended that conference. Most of the audience was students. So. Um, and there were like a few, um, I mean, one is, uh, sharing, um, experiences and. Another things. Not too much. With you. With your methods. But I guess what they could begin in. So. So we're planning to do a. A KCD maybe in Mexico or like, uh, Guatemala. But yeah, I mean, this is a. This is basically for LaTan. Which is behind. Compared with other countries around the world. Yeah. Yeah. Um, that's pretty much all the feedback about that. The particular events. So. Not too much right with the, the telcos. Not too much. Well, they were. She, she says she's about, uh, Kubernetes, but they were just touching the basics. Um, about. Mostly how to deploy Kubernetes. How to use containers. Like, uh, As I mentioned before, like. That time is not. At least in that particular is not. Let's see. Okay. Yeah. And Tony, you said you spoke with quite a few. That's, um, CSPs. And then there's a big 5G event. There was a lot more there than I expected. Um, including some good. Presentations and also panels where you had. Um, Folks from different Latin America and other parts of. Canada. There is a good showing from Canada as well. I'm going to try to get. Reach out and talk with some of the folks. Spoken with and see if. We can get someone. To come and talk with us here. Okay. Um, so moving on to the. Gender. Um, So the Kubecon North America CFP closes this Sunday. I don't know whether we've got any. Anyone who comes to this meeting is. Doing a telco specific talk or. Proposal for that. So. There is the. One thing to keep in mind for us, anyone we're speaking with. There, there's going to be a new track at this starting at this Kubecon. What's that? A new track for network and edge. So. Potentially just the main Kubecon, we could end up having a lot more. Related to what we're talking about. In addition to. What we would do at plantative telco day or anything like that. Um, So I'd say we should encourage folks to. Submit for this one. It'd be nice to have that track filled. Yeah, that would be good. Do we. Obviously not today, but generally do we have money? Um, North American. Telcos. Vendors. Involved in the working group. Um, I don't know about other European ones. I'm. I'm struggling with travel. Um, still. It's more likely to be able to get there. Yeah. So Taylor, do you know if that. New track is going to be at the same time as a telco day or like. I mean, it's not going to overlap. I, it goes, this is some primary main Kubecon. So this isn't a co located again. This would be. Um, Concurrent. So there's the concurrent talks that happened during Kubecon. And there are a lot when they're selecting talks and. Um, for Kubecon. They do it based on filling in. Different topics. So that track. So different areas. So they've added. This whole set, uh, topic area. For sets of talks to fill in. And that's the networking edge. You know, I'm not sure what they are, but maybe there'll be like an environmental sustainability track. And then you would have potentially if, if that was a track, you could have. A networking edge talk at the same time as an environmental. Those will be happening during the main Kubecon. So. This isn't. This isn't the co located event. We still need sponsorship and push for the sponsorship and get. Uh, things going for a. Helco day. And I think we should fill that up separately from anything that happens, of course. If. You know, if someone. Submits a CFP for, you know, some type of networking. Or telecom. Focus talk. At main Kubecon. And they also do something at cloud native. So today. Or their company or whatever, then it's probably more likely that company is going to pay for. That group to, to go. So I think that's fine. Got it. Thanks. Okay. So. I noticed the KCD Texas was just added as a thing. Is that a newly scheduled event? Oh, I didn't even say that one. Well, this is an August and October. Yeah, I just thought in the drop down. I was looking for the KCD. Mexico or Guatemala. And I noticed. There's a Texas one. Just opened. Last week. And they're, they remain open until August 30th. So it's a better time of year for something here. Yeah. In which part of Texas is going to be. Have no idea. It's new Dallas. Irving Texas. Yeah. All right. That'll allow. Easier. And flights more flights going into. The Dallas Fort Worth. The FW Metroplex. Victor, are you going to get an open compute? San Jose in October? Maybe. Yeah, I haven't. But yeah. As long as the restriction is free, I can. I meant to say actually one of my colleagues. Spoke out. Cloud native meetup in London. I'm going to find a link for it. Navigate meetup.com. So there's a challenge in the chat. Anyone's interested. I don't know if the presentation is available. Sure. Any other updates on the other events in terms of. Things we should be pushing people to. Propose. Sponsorship for cloud native software day. Yeah. So it, you know, if, if we don't get sponsorship. I keep hearing some people kind of say they're interested in it. It's not moving forward. It's possible that it doesn't get supported and added. And that seems. Not good since it does seem like we have a good number of people. It's frustrating that we don't have virtual because we've had. On the first one, there was a larger showing with people there. And people are met. Sponsorship for that. Sponsorship is quite a virtue. Yeah. And a sponsorship at the level where we could. Get virtual. A bunch of small ones would be as. You know, work out. We need that. And then. If it weren't to happen, or even if it does happen. Do we want to do something like the. Telco community gathering that we did in. Amsterdam. Where we had an additional time. So. The reason why we did that was. We had a lot of talk a day was only half a day and it seemed like there's enough impulse to keep talking beyond. Yeah. Time. And that was a very successful. Event. Or whatever you call that. And the community gather. It was good. On conference. Yeah. And conference. So do we want to. Have another one of us. I'm saying we like. We called it the working group. Heading forward something. Do we want to do that? Plan for it. Yeah. I mean, I think we should. It was. I thought the one in Amsterdam was brilliant. It was good having some input from other people. I'm just hearing people's views about the things that are being discussed. I think it's a little bit more relaxed atmosphere. Right. It does. Go ahead. No, the other thing that I was thinking is about the audience, I guess, like. It was more. You know, it was a little more informal. I guess that the people feel more free to share ideas. Like, for example, when they were saying about. They were like, we didn't have any. Project in the. But. We have a. At least a good explanation about the project and goals and things like that. So. Yeah, plus one with. Gathering events. So it's, it seems to me that. If. If there's a good chunk of. I don't know how much. Publicities happen for the KubeCon networking edge track. But if we have more folks that are. Coming in and. And putting their presentations in under that track. And maybe they're more. Networking telecom related as a result. Then it's likely we'll have more people. That might be interested in. The community gathering. That would be one thing I'd want to take advantage of. If we put it forward. It sounds like Tom, you probably wouldn't be coming. To the. To that. You know, if it's happening at KubeCon North America. North American. So that. Maybe you could help. Help us with organizing and stuff. If we're going to do it. I want to try to, it's, it's frustrating that. The CFPs coming up. Soon for the KubeCon, because it feels like. We need to push to say. Get your, you know, your telecom. CFP and so you're there and get a large group of. Folks ready to talk with us, but. Maybe all this. We should do that. Post. We can. Post on. Various. Places and. Reach out to colleagues. Yeah. Anyone that can get. A CFP and for KubeCon is more likely to. Get. Support. From their organization for going to the. Conference. Having people just show up for the. Community gathering is very unlikely. Yeah, great. Specifically in that location. Think of it as like Santa's a. Somewhere in the bay area we may. Be able to pull people in even if. They weren't speaking. Let's try to do that this week. You know. Put some stuff out. Say. Like to see it. KubeCon. Maybe we need to. Even put a. A Google form together. Yeah. Simple as. Are you interested? Yeah. To the CFP. For them for the main. Yeah, the main. The main one. Yep. Yep. Yeah. Yeah. Yeah. Yep. Yeah. Yeah. Yeah. Yeah. I. To review. Regarding the PR. Well, one thing that I noted is he hasn't. Signed. He's a comment. Something that we have to tell him. to tell him. And given that it is his first contribution into this project, we have to manually approve the execution of the CI. So just approve it. Why not? He's running the CI. The last thing that I checked, he didn't update some references, so hopefully it's fine. Yeah, it looks like there's some spoke check problems as well. Yeah, but probably that's something that we haven't fixed in the main branch. So my major concern was like with the broken references. Yeah, probably the only feedback. I'm going to put it in the PR. Yeah, it's regarding the DCO failure. It means like, I don't know if he has submit. Yeah, he has three commit messages, but only one was fine. Right, okay. And we can ask him to update them. The other thing that we can do is maybe we can squash them. I mean, modify all these things during the merge, like we can assure that things can. Yeah, because then it's DCO, we're on the score. Okay, let's see if it succeeds this check. Oh, yeah. Well, it's going to fail, because there are two things from the gathering event, markdown page, which is not following standard markdown. But yeah, again, those things are from the main branch, not with that particular PR. So sorry, I didn't quite follow what you said. Is there something we need to ask him to update in his PR? Is there something to update or do we squash and merge and then do it ourselves? I think we can squash and merge, just like a simple change of files from one location to another one. And eventually, we can fix the other things like the entity issues plus the start checker. I think that we can do that in subsequent PRs. And do we want him to fix these markdown formatting things before it merges or not? No, well, I don't know. Go ahead, Taylor. That was Tom. No, what I was saying is like, yeah, those like, yeah, we need to fix it later, like, I mean, it's not related with, especially with a PR, like, so I guess it's fine. Formatting things. Yep, I would agree. Do you refuse, does it need ad content and create PR for, so is there a PR created for this one? We're saying we need a PR. There's no PR yet. There's the draft document. Is there any other new comments? No, there aren't. All right, well. Okay, yeah, I suggest we get up. Play it as a PR, then, and see what. Trying to write up, add more content into here. Okay. That summary, that first one you're looking at, it's just a copy from there. The next part is content that was related to test suite, test suite, and other places. Just an example. But that first portion from Docker, it kind of just communicates what we're doing. So it's hard. I don't know. We can write something else. Of course, it talks about Docker and stuff, but this intent, I think, is what this practice is going for. One process, category, or type, whatever it's going to be called. Yeah, I agree. There's no point in trying to wordsmith something new. Right. If Docker's wording is fine other than the word Docker, then. Right, yeah. I mean, I think we can probably reference this, just add it in the reference section, but write something very similar. This bottom part is maybe a little bit redundant. It kind of explains, it's kind of expanding on it, but between these two, we could write something up. Maybe talking about a CNF running on Kubernetes or something like that, but take those ideas and put it like that. Anyways, I think the worst thing with this whole best practice is the words process type or process category is only slightly better. It's probably the worst part, where people get stuck on that versus focusing on the concern. Yeah. The thing is we're trying to help people with guided towards implementation a little bit more than being super high level and say, write good code. That's not going to work. I worry a little bit that if we change this to one concern per container, which would be the highest level, then someone would go, oh, our concern is this whole network function, which happens to require, I'm not going to say processes. I'm going to use another word, multiple applications to handle the concern of that network function service, which happens to be NGINX, Apache, whatever else, DNS server, and a ton of other things. And they go, our concern is this service, and it requires these applications to all run, which are all different processes, whatever. But that would kill the point of this. Then we go, well, your concern is too large. I'm trying to figure this out. The intention of saying one process category or one process type is it's more likely that when you get into a single process like the HTTP server or NGINX or DNS, or I can't use like a the session manager, even though that 5G core, it is breaking it down, but the session manager actually has multiple components when you look at like open 5GS or free 5GS, either one of those projects. So there's something there, it's easier to look at other applications like NGINX or something like that core DNS, where they were designed to handle one concern, truly one concern, like what are they actually doing? Core DNS may have a HTTP way of talking with it, because I know they have a lot of different ways of doing integration, but they're not trying to be a generic HTTP service, providing whatever you want to say on that. But that gray area where things start blending is where we get a lot of arguments and then we get a lot of blocking on something and not moving forward. So anyways, this is my point. I don't know how to name this, but I think we can work on the content for this best practice and not worry as much about the naming, and then we can come back whenever we have all the content fleshed out. So the summary, I think we can probably get something we like. Motivation, we need to replace what's here. This is something that we had before. So do you want to do the working changes in this Google Docs or in the PR? Yeah, let's go ahead and do it in the Google Doc. Okay, now we can just start making changes. They can be like this or people mark stuff out or do suggestions or just write stuff in. I don't think it really matters since we have history. And of course, we can have comments. This is linked out of the main document. I don't think I put a link to it in the ticket. I don't know if it's, let me see what the sharing is. Can you edit or suggest edits? No, it looks like it's anyone with link can comment. So theoretically we could put it into the ticket and worst cases we get spam suggest edits. That's really the only bad thing when you drop something like this. But as long as we don't make it edit directly, I'm going to get it access. Tom, your Gmail account? Do you do that or your Vodafone? My Vodafone account, please. All right. I got another, I'm going to grab another summary item. I was trying to find content to put in the motivation. I haven't found that yet. Let's see. It's best practice. Summary. I don't like this one. This is, scroll up into the summary section. What do y'all think of that one? And we can, maybe it could be like a CNS microservice or something. Yeah. Yeah, again, I think that the general gist is clear, isn't it? It's just making sure that the kind of definition of those terms is clear and understood. So there was a comment, I think in the ticket about not spawning external processes. I think Ian put it down about what about spawning like the IP commands or other, I'll say simple commands as part of, to facilitate the functionality of the main process. I'll say concern. And in that, it seems like it's, this is, I'm just going to put it back to us. Here we go. See my comment over there? Yeah. Yeah. So that seems okay. You know, these words in this are saying, don't do it. And again, this is like, what is your intent? So if we talk about what is, you know, we're trying to guide people on a best practice. And then what is the intent of that best practice? And what is the intentions when someone spawns the executable? So if the concern was something like a, it's, it's a IPsec microservice, like a, a, it's the endpoint where it's, you're connecting IPsec clients are connecting and this is a terminating IPsec endpoint. And you're trying to, you need to know what are the different addresses or whatever that maybe, maybe this pod has lots of different network service connections or something. I'm trying to come up with this story. So you have lots of different connections, however that is, multis, maybe it's the new fancy network object that has a pull request. So it's, it's actually native to Kubernetes, but you have a lot of endpoints. And you have an inbound connection to your service, you know, maybe that comes in through the normal flat network. Doesn't really matter. You receive this connection. You need to look at, and they're saying, Hey, I'm trying to connect to, you know, some, some other place and they're talking to you, right? Trying to connect in somewhere. Maybe you spawn the IP command and look at all of the available networks or something as part of it. Of course you could do this in some other fashion. There's, there's kernel calls. There's all sorts of stuff. But one way to implement it would be you're calling the IP command. I'm using this IP command because I think Ian pointed that one out. But yeah, that one to me would be part of the concern. And it's not something where you're stepping outside. It's you're using the IP command as if it was an internal function or whatever if you wrote it. That to me sounds okay. But if you said I'm going to spawn up a Postgres database, and then I have a web server running that's connected to Postgres database and that web server runs, you know, something to collect IP info and have that available to the IPsec service that then goes and talks to this network IP database. It seems like that IP database should be moved over to some other pod and you reach out and talk to it. Even if you said I'm going to feed it for my own pod. So here's all the connections I have, but something like that. Does that resonate with you? It doesn't make sense. Yeah. All right. So I agree with expanding to be okay with running executables that are within the concern. And I understand why it's written this way because I can see people just saying, well, I'm everything that I run is the concern and it becomes too expansive. So something with we need to capture the intent. But I think between these three, that's a good start. Do y'all have any thoughts on, I mean, if you want, we can keep going here. Motivation, goals, all. I'm going to struggle to carry on past the top of the arrow, but something I can put some time aside this week to look at and make some suggestions. I'm going to highlight the sections that are have maybe Al needs new content made place on content. Yes. It's basically like Loram Ipsum. I don't know how to say it. I'm going to say that. Replace the Loram Ipsum bill there. This is content from a different goal. So I'm going to do this for each one of these. We know which ones to work on. Yeah, like this. The proposal, this one should be more straightforward other than the same issues that we're having where people can say, wait a second, can I do this? Isn't that part of the concern? And I think we'll just have to keep expanding on it and add this sort of thing would be okay. We may need to actually write something about intention. Yeah. Intention is important. I put a comment up there by Ian's my paraphrasing of Ian's words. Workload context. Can you scroll down there? All right. This one should be something that we could knock out right now. Yeah. So that's, I mean, I would just say all pod types. Yeah. Is there any reason why it should be system pod types should do this? Victor, you had comments in the last best practice where we actually thought we needed to exclude system ones, whatever, wherever that came from. But what about this one? No, I guess like, I mean, there is no reason to not follow him. I mean, at least in the system pods, I think they also have to follow the same in practice. All right. That's what I'm thinking. They should follow it. So the other thing that I was thinking is like, is there is a way to validate that? Like, I mean, like, like using EVPF to track which are the processes running by the OS, like, you know, if there is a tool to, I don't know if there should be another way to describe these as a, for example, like the number of these goals to execute new processes. Mostly, I'm concerned about, like, people confused in terms like threads and processes. Yeah, okay. Notes, caveats. Yeah, something here to do, add information about threads. So is this like threads and processes? Add information about child threads or processes? Y'all can't see what I'm doing. Okay. I just put it down there under notes, caveats, sorry. Oh, yeah. So the, that we may need to expand on that. I did try to put something up under the summary section where, under the summary, if you scroll up there, Tom, I'm trying to write something about that. So the docker content, it actually talks about it. It's okay to have multiple processes, but avoid, it's talking about the responsibilities and then forking multiple worker processes. So that's child processes and threads. Well, the way that it's, I think they actually become non-child the way, I think you can enable that in Apache. So they actually are no longer a child process, but running by themselves and then they can have child processes. I would still consider them part of that same concern. That's an implementation of those, the worker process is probably a good idea. But the threads, whether it's a thread, whether it's a child process or separate process, but all from that same concerning concern for the application, I think all of that's legitimate. And we can cover that both in whether it's in the summary or the notes. There is going to be some, there's going to be some point where we have to need some expectation that they understand what we're talking about. If it gets down and they're like, I don't know what you mean by a thread or a child process, it becomes an education and programming beyond probably the scope. We go, go look up this and we can give them links and stuff. We can put some definitions and then references. And then we have to have, there needs to be some assumption that the reader is going to understand those things. So they need to go look them up. In my opinion, is that cover it, Victor? No, yeah, yeah, that's perfect. Because it's a perfect way to, I mean, it's a perfect place to put all these extra information. All right. This makes me think that we're going to need to come back through all of our content and try to look for shared definitions and terms and pull things out or copy things or whatever into our known shared area, but also consider which items would be good to donate upstream to the CNCF glossary. Yeah. Of course, there may be new stuff that we should go look and reference. That'd be the other side. All right. Well, workload context. I'm going to delete all the user stories for supply chain attack and then we can add those back in. Okay. Well, I'll look to provide some comments, content, suggestions, whatever, this week. But I need to go now for another meeting. Yeah, sounds good. Okay. Thanks. Thanks, Phil. Cheers. See you next week. Bye.