 Yeah, I can hear you Can you hear me? Awesome Yeah, can you listen? Yeah, sure. Let me Okay, yeah, thanks a nation. I Oh awesome, I hope I'm So everyone I'm Saik Silkar and it's good to be here. I'm sorry. I could not be there in person But yeah So today I'll be talking to you about leading and growing an open source community and So to start things off, I want to talk about What open source is and what does community leadership mean? So as contributors to a project become more familiar with how the project works They may wish to become more active in helping the project work Ultimately, this means that they will be taking on more responsibilities within the project So today we are going to discuss in more detail what this means what it means that is to lead an open source community or project and The fact that contributing to an open source project and leading that project up to different modes of engaging with that project Uh So As an contributor to an open source software project your duties to the project may more or less be straightforward You fix bugs answer questions in the project's communication channels and Don't know so when assuming the position of The leadership in the project your role becomes even more comprehensive. You may drive technical direction Begin speaking with authority for the project in public venues like Conferences initiate programs So why do we want to do that? Why do we want to take responsibility for a project and the community success? So there are many reasons to take on leadership position in your favorite open source project Ringing from the needs of your employer to your own personal joy And as you consider taking on a leadership position in the project You need to take some time to think through your motivations for doing that So having a good understanding of why you want to leave will eventually help you to choose the best places For you to apply your skills and interest in that project so It might be in terms of conflict resolution or better triaging issues and Or it might also be to make sure that the users are getting help more quickly Now it's no secret that leading an open source project has many tangible benefits when it comes to Seeking employment, but many individuals choose leadership position for A very simple reason that they just appreciate the social connections that they may in within the project community and After a while, they want to be of greater service to Other folks who might be in that community, right? So in this During the course of this session, we will explore leadership in open source projects The responsibilities leaders have the transition pathway from being a contributor to a leader And some ways in which excellent leaders Help their projects avoid common social changes so Our goals for today is more in terms of understanding common challenges which are associated with expanding maintaining and popularizing an open source project and The next Goal would be to expose steps new community leaders can take for building addressing those challenges and Finally identify steps that leaders can take to achieve project success So one additional note, uh, I can't guarantee that you will leave together like we will leave today with Like everyone equipped with everything you could possibly need to know about being a project leader or a community leader Every project is different and I don't know every bit of advice that can help. Uh, you lead each of these are different communities You will need to basically translate this advice into your own context So what we are going to do today is uh, we are going to examine the most common avenues for meeting the challenges new leaders face and general pathways for getting that For getting started, uh, and by using the The things that we learned which we find to be the most fruitful, right? Uh, so let's begin by stressing that anyone interested in assuming a leadership role in the community driven project Will need to begin asking three fundamental questions about that project and these questions are How you can expand the project? add contributors and increase participation in the the project maintain the project Ensure that your growth is sustainable and that the project stays on course In the pursuit of its mission and vision And popularize the project. How do you popularize the project increase its visibility? Make it an appealing destination for users and contributors and keep it relevant in ever-growing ecosystem, right? So the the remainder of our session today will focus on each of these three questions uh so Going forward, uh, let's begin with the challenge of expanding the community increasing community participation and contribution so What are the common challenges? uh, some of the common challenges Most common challenges that promote in communities are challenges from undocumented or insider information or folk system uh excessive An undocumented insider information usually makes new commerce uh feel Like the lack context and usually that makes them feel excluded um Then there's the challenge of unwelcome social dynamics. How welcome do your contributors feel? Perhaps the project is not as inclusive as it could be Then there is the challenge of undifferentiated contributor ways communities lacking a diverse contributor may not seem welcome into new commerce Uh, especially those from underrepresented groups. Uh, we mean diversity along all types of access over here So it might be the types of skill sets that they have a vendor diversity number of industries represented gender diversity racial diversity Diversity So when working to grow your community if it appears that the community is only comprised of one type of contributor people who are not existing who are not Like the people who already exist in the community the existing contributor base may not feel very comfortable joining that community um, then uh, finally the biggest Or the most common challenge that uh folks face is the challenge of language barriers Uh, so communities that communicate in only one language mostly English is the most common language that is found These communities can be intimidating for those who don't speak that language Now going forward, uh, what are the technical challenges in this regard, right? So lack of clear onboarding documentation on ways to get starting to get started can deter newcomers from aligning contributors with project needs uh Is one of the most common problems that leaders face So aligning contributors with different skills to the most appropriate domains inside the project can be really challenging and then finding additional contributors with specific domain knowledge of Very specialized use cases can be very difficult, especially when it comes to very niche court basis um, and speaking of court basis The project might be relatively niche, right? So locating additional contributors who share its typical use case is usually Uh, a bit more difficult So let's talk about contributor pathways, right? So every project and community will address Uh, we need to address these challenges in different ways. Uh, that are the most effective for them However, in general, we suggest beginning, uh to address these challenges by examining their own uh, community's contributor pathways So Okay, uh, I think we're there So I hope I didn't Uh, let's keep anything So yeah, let's continue So, uh, yeah, we were talking about contributor pathways, right? So opportunities for Volunteers to begin lending their unique talents to an open source project are Usually what's called the project's contributor pathways and the greater the number of contributors Pathways to your project features the more likely it is to recruit participants with various skills Required for the project success. So when you're thinking of ways to expand your project, uh, focusing on contributor pathways is a great place to begin and Yeah, so what are the typical contributor pathways, right? So let's take a look at these First here are some pathways to the community focus Documenting software and processes onboarding and mentoring new members Localizing content into various languages copywriting managing social media organizing events Uh, you might want to ask does your project currently offer new and existing contributors opportunities to contribute rewarding or even big ownership of in each of these pathways the next thing is Things with more technical Pathways So adding new features and documentation fixing existing bugs and rising issues refactoring existing work to improve it performing quality assurance improving user interface and With uiux in general these engineering creating and maintaining project roadmap for and user interface So these are some of the most common Contributor pathways with a more technical focus. So you might want to ask yourself again And does your project currently offer new and existing contributors opportunity to contribute? Rewardingly to or even take ownership of working each of these pathways If not one general way to begin expanding your project is by making a concerted effort to formalize refine and document and advertise this contributor pathways So I'm already good. I was not sure if Okay Okay, yeah, thank you So, okay, uh, let's uh continue with Exploring the challenges associated with maintaining a project sustaining its growth as our prior expansion efforts succeed and Look at the common challenges with maintenance, right? So additional infrastructure and overhead is One of the most Common is again So adding additional tooling and infrastructure creates additional work of coordinating participants time spent otherwise might Be used for enhancing project work for smaller leaders May have to take on responsibilities for deploying and maintaining these new systems in addition to coordinating volunteer work Other than this preserving community intimacy is The challenge so preserving the intimacy that makes the community so appealing requires much much more care As the community keeps expanding. So few of people may really know one another but What that results in is that the social wants into meeting As the community keeps growing The next challenge is keeping a growing number of participants informed So keeping all participants informed about project development requires more time and effort if that community is large so Managing competing visions for the project as a community goes competing visions for the project might create contributor tension Architecting and scaling a mission of the project's mission and vision So as projects begin as personal bodies They don't always have you know Very explicit clearly defined Community mission and vision statements and without these the projects are usually don't scale So I think one of the biggest problems here is Then when projects start when communities start in the initial stages Uh A lot of community leaders don't really have that Mindset or they don't probably imagine the common challenge that they might face in the future and this is what results in much of these projects coming up. So Hence me. I'm trying to Highlight this point over here Uh now moving forward. Let's talk about governance models. So every project and community will need to address These challenges in ways that are most effective for them However, in general, we suggest beginning to address these challenges by examining your project's governance models, right? so the specific combination of rules and customs that define who gets to do what and how They're supposed to do it is called a project governance model And the better you understand a project's governance model the greater are your chances of successfully helping that project So as you look at Ways to help sustain the projects both and success It's highly recommended that you begin by examining your projects governance models, right? So every project has a governance model even The ones that say that they don't So you might want to begin by filling down specific details of the way the project is running now here is a Way of just six types of open source governance models We don't really have much time to explore the individual models. However, it's suffice to say that our knowing your project is running and Who is making decisions is a perfect way to think about To actually start thinking about making the project more sustainable Now if the project relies on too few people is that decisions aren't being made in the most effective ways Then you might want to begin by addressing these issues first So So the next thing that I wanted to talk about is dimensions of governance Fundamentally, however, you will need to begin thinking about the various roles people play in the community And the various policies and procedures that govern and direct people in those roles Now roles are specific functions that contributors perform in and for the project Now roles have rights and responsibilities and expectations associated with them So you might want to make sure that these are explicitly documented And the next thing is policies and procedures now policies and procedures are the specific rules and processes that direct people in particular roles and define the limit of You know acceptable behavior for these projects. It's it's best to Identify these its best practices and so on So you need to make sure that these are also explicitly documented so simply documenting roles policies and Procedures can go a long way towards helping your your project become way more sustainable in the long run So next and let's move on to popularizing The project how we can do that So increasing a project's visibility while making it an appealing destination for users and contributors and keeping it relevant growing Ecosystem is key to popularizing a project. So what are the key challenge and challenge challenges, right? so some of the challenges for popularizing a project into Increased computation for contributors time attention and energy So competing for contributors time attention and energy Is usually more difficult than ever as the number of open source projects keeps on increasing globally Then community materials are available in limited number of languages So projects become popular when they connect with diverse groups of people But often project materials are only available in a limited number of languages Now if your community is not communicating in English, many people may not even be aware that it exists, right? So You might want to take a look at this and The next thing is the ever-growing threat of maintainer burnout. Now, this is a very very big challenge So the project becomes more and more popular maintainers might experience burnout While you're trying to keep pace with it um The next point is around mismatched expectations between enterprise users and hobbyists So many popular open source projects are relied upon by large enterprises But they're maintained by volunteers leading to a mismatch of expectations between Users who may be expecting enterprise support and developer hobbyists who are working in this way So when people are using the project for enterprise work work, uh, but Uh, the software is maintained by hobbies. There can be an expectations mismatch that leads to friction in the community uh, the last point is around the preparation of platforms for user engagement and finding the best way, you know to Preach your potential users and contributors can be time consuming given that many users many So given that many there are many ways by which people choose to consume information And how the users expect to hear about things on online probably on places like reddit and on facebook and twitter uh, so figuring out where to meet people when when where they are And also trying to figure out Are you there or not is Kind of difficult and it might go for project leaders So as with other challenges we address so far we could discuss many many other ways You know to we can address in popularization challenges But in general we have found That focusing on a project's reward models is a great place to speak. So let's talk a bit about reward models So strategies for acknowledging uh, and honoring various contributions to an open source project are called the project's reward models The more reward models your project constructs the better it can engage with participants in different motivations for contributing right For example, uh, here the form of most common models spotlighting community contributors swag big chunk of community contributors Love swag. So meet the community features community awards and You know each of these models seems to connect to the certain set of motivations That participants may have for contributing to a project Now everyone contributes to a project for a reason and the most popular projects are those that make contributors feel rewarded right valued their And the contributors feel respected in those countries in those countries, right? So generating those feelings in contributors is a great way to build your project's reputation as a rewarding and welcoming So let's unpack this a bit more Uh spotlighting community contributions are using uh point systems badges and leaderboards to chart and public identify contributions Usually rewards rewarding participants by visualizing The magnitude of the contributions, right? So this is a great option for connecting with people who tend to value opportunities to show the technical price of it um then things like You can be a blog post or video or series to spotlight contributors reward participants by You know helping them to grow the network with their personal brand and also to grow them Also by helping them to grow their professional networks, right? So this is a great option for connecting with people who see community participation as an opportunity to gain visibility maybe even find future or employment opportunities, right? swag This is the biggest one Giving contributors various items Signifying the affirmation of the community things like shirts hats key chains, etc These allow participants to feel a sense of connection with the project This is a great option for connecting with people who see community In participation as central to their identities and as a fundamental part of their identity building Right. So it basically allows them to show off their connections with the community when interacting with others and also makes membership in that community part of their identities Uh, and then again, we have community awards like allowing the community members to recognize and Reward others among them helps communities to develop shared identities and social bonds So, uh, we are almost out of time, but let me try to take a moment to review what uh might be coming next for you Gathering community feedback and making your success a really critical component and since I don't have much time Let me try to go through it really really fast. Uh, so, you know leading an open project means taking Making yourself more available to address community ideas and concerns and more So your feedback is critical to your success and you're positioning yourself as the party that is most likely to act on community feedback so you'll need continually collective and uh, here are some of the four Most common methods you can have open feedback forms. You can do fireside charts community calls and community q&a again I need to explain you this much. Uh, so let's move on. Uh And uh, you will need to then be able to measure the success of the of the work with the community and depending on your situation This could be important for two reasons. Uh, community communicating success with the community. Uh, how will you narrate your community success in achieving its mission and Objectives and you will need to be able to help your community measure its progress chart exact accomplishments and stay focused on our achievements Right, how will you do that? And the next one is communicating success with organizational stakeholders How will you explain the value of working with your community to others in your organization? If you're leading a community because of our project is uh, Integrated to your organization, then you will need to be able to explain to the value of your work internally to right To the organization's stakeholders um, and here are just a few more potential metrics you might want to take a look at pertaining to technical contributions and So and also here are a few more potential metrics you might want to Pertaining non-technical contributions, right? So once you determine how you are going to begin addressing leadership challenges These will be your next steps so in summary I think uh, I can just uh go through this really quickly People choose to lead open source projects for various reasons including organizational need a sense of responsibility to the community and our personal or professional fulfillment, right? By choosing to lead an open source project or community you assume increased responsibility for that project's success or failure Uh, new community leaders face challenges related to expanding maintaining and popularizing their projects Uh, they can begin to address those these challenges by examining their project's contributor pathways governance models and reward models and they will also need to Determine the best ways to have the feedback on their communities and uh to Communicate the value and success of their work to various stakeholders, right? So, yeah, that was it for me. Uh, thank you and I'm sorry if I went too fast at the end. I just uh, we were perfect on time So, uh, thank you for listening to me and if anyone has any questions, I Can take that. Yeah So I have not heard about the book um, I don't have book recommendations as much as I would probably uh Suggest going through some of the Literature that's already available online. So I think I think there are a few people From the initial days of the internet, uh, that you might want to look up You might want to look up at the stories of some of the most popular open source communities and how they work for example, there is the Federal communities that the mozilla community is there and You know, if you look at the the genesis of any of these projects, you will find that there are a lot of leaders And you might want to I don't want to specify anyone's name right away, but I would Uh, strongly suggest like going through how those leaders actually started Uh, those projects and how they went about it. It's very openly documented So a quick search on the internet can really help you with it I I'm not a big fan of suggesting books for leadership for listening leadership in open source I am a strong believer in uh, experience based uh understanding so If you are not already participating, I I think you already are but if you Like try getting into open source communities read read about the The genesis of each of the the communities and you will get to learn a lot more Uh, does that answer your question? Excellent question actually um you know this uh a I just have a counter question to that before I answer that Is are you asking about you as an individual who has a certain disability? How you can do it or are you talking about uh, how can we increase participation in general? Yeah Yeah, I've already on this one. Thank you everyone. It's great. Uh being here Can you hear me? Mine's on. I don't know about yours. It's on a little bit. Well, did he say it was on? Yeah This is better. Well, it's pretty loud now All right, uh, welcome to our talk. Uh, today's talk is uh, what's new and what's next at the door chorus? Uh, Michael Neon and this is Dustin Abe Uh, so today's agenda We're gonna talk about what is for door chorus What happened last year? What's new since last year? And what's coming soon? So, uh, we'll start with uh, we'll take questions at the end. So Please hold them until we're finished with the presentation so What is for door for us? Uh, it's an emerging for door edition. It came from the merging of like two communities uh chorus inks container linux and project atomics atomic hosts And what we did was we basically picked what we liked from each, uh, project From container linux. We took the philosophy the provisioning stack and the cloud native with expertise And from atomic and this mostly revolves around a tool called admission. We'll talk about it a little bit later Uh, from atomic host We took uh, the for door foundation, uh, the update stack using rpm mostry and mostry And enhanced security with su linux, which container linux did not have So, uh, the philosophy behind for door chorus, um One of the biggest features is, uh, automated provisioning All nodes basically start from the same point. We use the same essentially the same disadvantage for whatever, uh platform you're using. Uh, we only bake in like, uh specific platform ids into the image, but essentially it's the same disadvantage and the way we, uh The way we change that image and mutate that image is basically using, uh, this tool called ignition And that basically takes a configuration file Called an initial config and on first boot will Laydown files start system to units or and whatever you need To make it what you want Uh, the second feature is immutable infrastructure um We want you to automate the deployment and the system configuration and instead of like logging in and ssh in and uh making configuration making changes to the file system and stuff yourself Uh, we actually want you to do that through configs like the ignition config and The idea here is that if you have changes to a system, you can just Reprovision it with the the changes from your ignition config and you could have that stored in like a get or something like that Uh, if it's cheap enough to do so um another Philosophy behind is we we actually want the software that you run on for door for us to run be running containers We need this clean separation between what you're running and what the host os is in order to make host updates more reliable um And one of the biggest features we have is automatic updates. So automatic updates are actually enabled by default Um, there's no interaction required from the administrator um, so if there's new uh bug fixes or cvs you get them by default Uh, and obviously we don't want you to turn off automatic updates So the biggest things that we don't want you to do is actually to disable it. So If we don't want you to have to just uh to disable it, uh, we have to make sure that the updates are rock solid and None of that happens without a lot of ci tests passing Uh, everything has to pass that our ci's test every build that we do has to pass all these tests So, uh, we make sure that the updates are rock solid uh another way to help with the Stability of the systems is we actually have three testing streams In for door for us We have the next stream that can include the next major version of for door or uh, some new early access to new features And testing the table are pretty, uh tied together So testing is basically a snapshot of what will be in stable in about two weeks Uh, we try to release a stable release every two weeks So how you would have this setup in your your setup would be you would you would have like a whole bunch of stable systems A whole bunch of systems running for door for us on the stable stream And maybe like one or two that are running the testing and if there are any issues Uh, if anything goes wrong in the testing one, you know that the change will come down in two weeks So if you find any issues there like you can file an issue with for door for us And then we'll just like take a look at it and see if we could fix it before it actually makes its way down to stable and affect other users Uh for door for us is actually supported on a whole bunch of different platforms Uh, we have like over 10 listed here And the for door for us is directly launchable on aws and gcp Uh, we also have options for bare metal. We have a live iso installer You can run it in ram or you can just boot in and then you can start your installation from that live iso Uh, we also have a pixie boot And also have raw and 4k native images We also have support for multiple architectures x86 arge and s3 90x and we have ppc 64 le coming down the pipeline So, uh What happened last year in for door for us? We switched to c groups v2 by default um We did this a little bit later than the The than regular fedora. Um, that was because some of the container software wasn't ready for it most notably Uh kubernetes So, uh, that's which was made in uh I forget where was it was 34? Yeah. Yeah Yeah earlier last year Uh, we have another feature is a more reliable live changes to the system because Uh, rpm os 3 and os 3 for door for us is uh Is mostly immutable We now have new options to change the content live in a safe and atomic way um We basically this could man used to be experimental, but now it's fully supported. It's rpm os 3 install Live apply. So if you've ever used any open rpm os 3 based system before Whenever you may change as you have to reboot basically and With this you can install A package and just have it automatically switch into the deployment without a reboot, which is really nice Uh, another feature that was added in admission was support for adding removing and replacing kernel arguments This is applied on first boot and will trigger a reboot. Uh, so essentially like if any of you use AWS for a while, they didn't have a serial console. So there was basically no way to Change the kernel arguments on first boot You would have to like log into the running system then change the kernel arguments and then reboot With this with ignition we run early enough in their ram at us That uh, we can detect that if when there are kernel arguments And then reboot before and reboot and apply those kernel arguments before we actually get into the real root We made slash boot read only uh, manually modifying Slash boot is disparage It's really easy to make mistakes and find yourself in a really precarious situation And it's really just to it could it also has the advantage of like protecting people for making malicious Do malicious things on your system um The way we encourage users to Modify boot is actually with rpm os tree. There's a rpm os tree kr command And there's a if you want to choose change the boot order you can use rpm os tree rollback deploy or update We ask also added support for advanced root repartition reprovisioning Here's a couple examples of two things we added. This is just lux encryption for your your root partition and This is Read support. So you can do all this now through ignition on the first boot I'll hand it off to dusty Thanks, mike okay, so uh Mike kind of went over what happened last year We decided to include that just because we thought some of those things were pretty cool But let's talk about what's uh, what's been going on this year in fedora cora west what we've been up to Um, so first off. We're always adding new platforms and new architectures or at least trying to So we added a art 64 support Including aws images. So if you want to grab some of those grab a ton instances and run fedora cora west you can do it We also added s390x support and uh, if you Are like me you'd be surprised to know that you can actually launch an s390x instance in the cloud so Yeah, that actually made it accessible for the first time to me anyway We also added support for natanix. So that's a different platform that people were asking us for that we added support for And then one thing that's interesting is we also became the base for a podman machine so podman machine is kind of a way to Get up and running with a little container development environment on a platform that isn't linux I think you can do it on linux too, but if you're on linux, you might as well run podman directly We'll take questions at the end So fedora cora west was a good small like system That the podman folks decided to use because it includes podman To run on top of to create that vm To use for the development environment So like if you're on an m1 mac and you run podman machine, you're running ar64 fedora cora west underneath, which is kind of cool So we added dnf count me support So if you've ever been to a talk by matthew miller and he goes through all of those different slides Yes, he's in the back It goes through all those different slides and talks about numbers and how many people are on fedora 33 and 32 and all that stuff We were never included in any of those numbers primarily because os tree based systems don't hit young repositories We delivered via an os tree Repository typically so we wouldn't necessarily be counted in that information. So we decided to rectify that so we have a small system unit that runs periodically and checks in in the same way that systems that do dnf updates check in but like you know What have we gotten out of those? Statistics so here is about the last five months of information that we have for Fedora cora s nodes kind of checking in and if you see we've gone from somewhere around 17,000 nodes in april of this year to about 33,000 nodes and I broke this graph down based on systems that have been around for longer than a week and systems that have been around for less than a week so maybe ci systems and or cloud systems that spin up do a job and then spin down Which is nice in the cloud if you don't need it anymore. Don't pay for it So yeah, so it's about a half and half distribution between nodes that are just transient less than one week And nodes that are around for a long time Some other information that we've got out of those count me stats We can kind of see that two-thirds of the nodes that exist today Or at least that have checked in in the past week our own fedora 36 That's a good indicator that a lot of people are leaving automatic updates enabled Which is what we want them to do. We have one person with a node on fedora 32 if you uh If you are that person, please update your node So we have a few that are still on fedora 34 and then uh looks like about you know Maybe 30% or so still on fedora 35 that need to also update Uh as far as architecture breakdown We just recently added s390x and I obviously we probably won't have a huge base there either, but uh between ar64 which looks like it's probably about I don't know 20% maybe between 15 and 20 or something uh and x86 64 that pretty much takes the bulk of it Um for ar64, obviously we have people using pod mem machine on like the the m1 max We also have raspberry pi support And then there's aws and other clouds that have ar64 instances now. So we probably should see that grow um another thing since the last year so One thing we used to not do very well in fedora core west is follow Some of like the fedora change proposal process We were kind of just getting off the ground for a long time. So One thing that we missed actually was when fedora, uh, I think in 32 switched over to nftab enough tables as the back end for ip tables um We actually missed that because the way it was implemented was using alternatives and the way alternatives works doesn't really work very well with rpmos tree systems. We actually have a A fact entry for that at the bottom there But we just happened to miss it and somebody reported an issue one day and we're like Hey, why are we different than the rest of fedora on this? We have a test for that now Um, but yeah, so we had to like kind of figure out. Okay. We're in this situation How do we how do we manage to migrate our nodes that exist to? What the rest of fedora is so we don't have a delta there, right? We want to be as much like the rest of fedora as we can In most ways So that we're not a snowflake Another thing that collin and team have been working on in the background is Basically taking os tree commits and putting them in container images Why is this useful? So if you happen to be You know in an offline or disconnected environment and you're trying to mirror all this infrastructure into that environment So you can keep your nodes up to date You probably already have a container registry because you're running containers Otherwise, you wouldn't be running fedora core os most likely And to deliver os updates you would also have to have an os tree repository Which is just served over htdp. So it's not a huge barrier to entry, but it is something else that you have to do So we started looking at like how do we actually take an os tree commit and put it in a container image So you can put it in a registry and kind of mirror it around that way And i'll talk a little bit more about why this is useful here in a minute We also started building and testing against the raw hide stream So this really helps us Find and fix issues before they ever even get close to our production next testing and stable streams We found a lot of issues since we implemented this it also means that every time we put out a new version of Or a new build happens on the raw hide stream all of those tests pass If something fails we find like the culprit say it's a new rpm that came in And we'll actually pin on the older version of that rpm while it gets fixed So we can then re-spin the build with the pinned rpm and see if anything else is broken Because one thing about raw hide is a lot of times it it might take like if the if the pungy raw hide compose is broken More things that are broken keep building up and you don't actually know it because there was no composed to test Right, so like this is a way that we can Actually say okay, we know this is broken pin on that And then try it again, right and we can find other things that are broken at the same time And we've also been doing a lot of boring stuff So we've made a lot of investments in our automated testing And our automated builds so our release engineering processes All of this stuff allows us to be more efficient and to do more things like adding new platforms and architectures Because without that we're not able to do the other stuff So there's a lot that's been going on behind the scenes that you know users don't necessarily see Um, okay, so what's coming soon? Uh, so I mentioned some things about fedora in there. Basically, we're trying to become a top level fedora edition I'm hoping that's going to happen for fedora 37 Um, we've done a lot of hard work to get to this point So we've done a lot of work To get closer to like when a fedora ga release happens switching fedora core os nodes over to that latest version of fedora ga A long time ago or when we first started out We would wait a lot longer because we do automatically update nodes and it's kind of scary It's like, you know, okay fedora 37 just came out. Is everybody going to go now that can be scary But at this point we've done a lot of investing in our automated of testing our ci We also run test days. We also have multiple streams that people test out Um that might mention earlier. So we're starting to get to a point where we have a lot more confidence so now When fedora 37 goes beta our next stream will get switched to fedora 37 When it goes ga our testing stream will get switched to fedora 37 and then two weeks after ga Hopefully if you know, everybody smiles on us Our stable stream will switch over to fedora 37 too So we've done a lot of work to get a closer proximity to fedora releases So we are seen as more like related to fedora and not just a one off We've also been working a lot closer with fedora qa to kind of like Set some release criteria and kind of be involved in the go no go meetings for for fedora major releases We've been participating more in the fedora changes process So reviewing those changes as they come in and also proposing new ones I think we had three that were proposed for this last cycle that actually came from Our group or a collaboration between our group and other groups And then also just continuing to do what we do best which is deliver fedora core west and try to be as stable as possible We're also working on more enablement more platforms. So azure and gcp have armaments This is now that we want to kind of put out images for Azure community galleries is a thing that exists which allows us to create an image in azure and share it with other people Previously we weren't really able to do that unless you have like a commercial relationship with microsoft To be in their marketplace We're adding power pc support that should be in the next round of releases that happens We're adding support for cube vert which is kind of like launching of the um in kubernetes or open shift And then secure execution support for s390x if you like mainframes Okay, the big one core os layering Uh, something we've been called calling core os layering uh and or os tree native containers Um, so as I mentioned earlier fedora core os or os trees in general have typically been served from an os tree repository um But now We have the ability to actually ship that as a container So now it's in a container registry and it's shipped up like a container like can I treat it as a container? Well, we've been actually been working To get to a point where you can at least treat it You know and derive from it as a container I don't know if you necessarily want to run it as your base for your application But at least as far as treating it like an image that you can derive from Uh, yes, and I'll show you a little bit more about that here So, okay, so we have a container in a container registry. This is a docker file and or a container file depending on what you want to call it um, so we derive from an image in a container registry this example basically uh Creates a yum repository file installs a package enables a service and then commits the container so in this case, it's tail scale Um, so it basically grabs the tail scale repo Does an rpmo history install a tail scale and then system ctl enables tail scale and then commits that container So if we take this docker file and we run a build Uh of that container image and we push that containment container image to a registry You can then take your systems and rebase to it Um, so you can actually reference the container in the container registry Rebase to it and then any updates to that container image in that container registry Uh, we'll you know, if you rpmo oestry upgrade in the future, it'll essentially look at that So this is really useful. Uh, we put it in the coming soon section Because there's a lot of little details that still need to be worked out one of the benefits of fedora coro s is We have an update graph, which means that You know, if we identify that an update it was actually bad after the fact we can actually remove that from consideration um from our update server that tells nodes what to update to uh, or we can You know, basically send people through barriers. So for example If we do some sort of migration where we're switching you from I don't know ip tables nft or you know, the old legacy ip tables to ip tables nft We can essentially Make you take a particular update path. So we can say okay update to this version first Which we know has a script that'll do run some migration for you and then update to that version later, right? Um with this stuff all of that kind of goes away because now it's just like following a container in a container registry Right. So we kind of need to figure out some of that stuff before we really tell a lot of people to use this Um, but it's there today. You can already do these things But you might lose some of like those extra features as a result um Yeah, so why would you want to do something like this though like One example is like what what that example was add a new rpm You could take that container and push it to like a staging Container Registry location and run ci tests against that like in your own infrastructure And then say okay, that was good then push it to prod right right now You have two options if you want to do something similar you could build your own fedora core west Which is a little more heavy weight. It's not too bad or You can just client side package layer all that stuff, right? um, which You know you're not necessarily testing that Before you ship it out to all your nodes, right? They're just individually all doing that You can also package up like configuration Right if you happen to want your node to have different crypto policies or something like that You can do all of that here instead of doing it on the node itself, which has some benefits Okay, let's take some questions you So right now fedora itself doesn't support risk five and I feel like uh, our part of the reason for that is there's not really like a server Like equivalent risk five that you can put in a rack and has some sort of out of band management, right? um, because fedora's infrastructure is in a data center without people there to like Push a button to turn it on and stuff like that. So basically I think what's blocking risk five in fedora is just having proper Server like hardware that can be used and then once we get that we'll probably get some donations Of that hardware and we can actually add it But yeah, there is there is like a shadow version Of uh people working on risk five support, but it's not official and we can't do anything unless it's official Dan Dan's heckling me Yeah, so uh in that A layering example that I just mentioned Where you add a package you can also remove a package the same way um, so like in the future if podman wanted to create their like Their own podman machine version of fedora or west Like to get rid of the bits He don't know if the the resulting I mean it would just drop out of our package list Right, if it's no longer in fedora anymore, then It would become something that just got removed in a particular release of We could always have a discussion about it, right? Dan wants to get rid of a mobi engine in fedora coral s Any other questions? I had one back there. Uh, can you repeat the question? Sorry anybody hear that better? So is there any crash dump telemetry? I don't think so Uh, I think fedora has like a server setup where like crashes will send Like data back to a server so we kind of can analyze If there's like a common pattern But I don't think our systems do any of that. Um, yeah, I don't think any of the the infrastructure for that is is installed by default I think I had you right here So The analog to fedora coral s within red hat is something called rel coral s Which a lot of people just abbreviate as rh coral s There's no productized version of specifically that Because it is part of open shift. So the only way to consume that is to basically consume open shift There is some effort within the okd the upstream open shift community community to Have like a syntos stream based coral s But that's kind of like just getting off the ground. So right now Yeah, we keep fedora in in front of fedora coral s because that's just the most explicit way and obviously it Allows for there to be other options in the future too Yeah The uh, when you showed adding via a container file Um, can I put a jar or any any type of file and an os tree will handle it? What is that? Possibly so you'd also have to install java, right right But anytime I could put any blob any type of file. Yes, we obviously prefer you not to necessarily go crazy with it Um Because it makes it less supportable, right? So like if you do a bunch of stuff Uh, some of that could rely on things in the host one good example for that Workflow is actually something like uh kernel modules, right? Like if you happen to have a kernel module that's not In the you know in the kernel itself. Obviously, it's an external module You can now derive from fedora coral s check what version of the kernel is in this fedora coral s Build your module install it to the right place and commit the container, right? That's an example where you can't really do that in a container very easily Like there's some examples where people do it, but it's kind of janky. I've done it myself Um, but like as far as like installing a java application. Yeah, that was just yeah Yeah, but we'll like you can do that. But like it's also one of those things where like you You you if you break it you get both pieces, right? You can't really come to us and be like hey I installed java Yeah, anybody else here I use oracle cloud does coral s have instances have Does the coral s have an ability to make a cloud virtual clients on oracle cloud? In other words Do we support oracle cloud? Uh, like we create an image for oracle cloud. We don't have one right now I think there is an open issue for requesting that And I think there is some benefit to doing that because I think they have a free tier That actually has arm instances and stuff like that which could be interesting to add Yeah, it's very effective for that purpose. Yeah, so we're not opposed and um, you know, we we encourage contributions in that area too Okay, thank you everybody