 Okay. Hello, everyone. Yeah. Okay, so welcome to DevCon US 2022. And this is the first session in this afternoon. And I'm Jingtao Wang from Red Hat. I'm a moderator today. So let me introduce the speaker. The speaker is Renu Chauhan. She's a senior project manager at Red Hat. And that's welcome. Good afternoon, everyone. As my colleague introduced myself, I am one of the senior project manager in the back. I take care of the Red Hat business in the back region. I'll quickly get you through this talk. It mentions about if you can't measure it, you can't improve it. This is mainly focusing about the health status of our community. How can you contribute much more to our open source community? What are the ways using which you can include everyone out there rather than being the race alone? So this is how it is going to be. So how many of you recognize him? I am so happy to see all the hands up. So he's Mr. Bean who is trying to make a coffee. So right here what he's trying is he's trying to balance out the ingredients, you know, one little coffee which you want to make. But what is the appropriate proportion which you want to have in that? If you mess up with that, definitely you cannot make it a healthy coffee, a delicious coffee for yourself. This is the same rule which goes on with our community. So how can you contribute in a community and to take along with you? So this particular animation, it reminds that the role of a community is not to win the race alone. It is to make the whole community successful. So a very live example I would like to give you from my life. It was one of the times when we were a group of 35 folks who went out for tracking up on a hill and it was having a group from different regions of the country and we went out there and one of my closest friends, she says that, you know, all of a sudden in the middle she said, hey, I have reached the top of the hill, of course. And we was, okay, we just go ahead. We don't mind respecting her, you know, freedom to think on that basis. And then what we realized that she reached the top of the hill the soon, she was very happy, she was super excited. But when she was looking at the memories which we had in our phones, which we captured through the videos, she realized that she has missed a lot of the memories. She was not right in there in the picture. Though she could reach the top of the hill the soonest, she was the winner, what she felt like. But our community never does that. The leader of our community doesn't want to win the race. They always take its people all together. So now the question is how to be a good corporate citizen when participating in an open source of such kind of projects. So for this, I'll say that while I started back, I didn't know even if in case something like open source communities exist in this world, I was working in a service-based organization. I got to know about the open source committee, what I did the first, the earliest was I joined the community. At that particular point of time, I was not taking any task at my hand. I was not participating into any kind of a commitment where I can complete one task and I can deliver it to you now. Simply just join the community, take the belongingness for you or not. Even to give you a hint, it will take you approximately five, five months or sometimes it can give you a click within a few seconds to, it happens. So I joined multiple open source communities, but then I have realized after completing several projects to that now this is not a piece of work or an admiration for myself. I should move on to some other community. So wherever you feel that you are having an inclination towards it, just join the community. First thing which you have to do is start where you do not overcome it, do not try to exceed the expectations of the community. So in Red Hat, we do have a committee of practices in which we are all the project managers or who are aspiring to be other business analysts. Such folks have joined in, that is very much confined to this profession area. And then I have seen that people over commit, they get exhaust and they leave the community very soon. So always start small. The next thing is understand the governance. This is a very important piece while you are working with any kind of community. Different countries, different regions, they have different types of policies, they have different types of governance. When it comes to the governance, it is not always the typical policies which comes into the picture of the businesses, but it also the type of the governance which comes into the picture while handling the people, while handling the communications. So always try to understand them and then try to reach out to the bigger piece of communication then. This really helps. And the next thing which I really respect and I have also observed on one of the booths right here for the open source, open space technologies which has been connected by RJ, he's sitting right here, that diversity and inclusion. This is a very important piece right here. So how to build and sustain diverse communities. It is a very, very fundamental thing which you should take care while you are a part of any kind of open source communities and you are leading it. So one of the most fascinating thing which I could understand from RJ's booth is wherever you are or whatever could happen that that could happen, the only the case which could happen you know the thing which could which was the best which could ever happen to somebody. This is a very fundamental thing which you have to understand. So how does it mean is there have been certain examples certain circumstances in the committee task where people committed but they could not show up or they had very diverse ideas which were sometimes off the track but which were many times it was helping the community to grow but they were not being heard maybe because many of the times that they were very much new to the community and the leaders who were practicing it they should never ever understand that since the body is new they do not know the background of the committee they this particular idea should not be that no it is like having everyone being included being inclusive of your ideas. Always always provide organizational support to establish the identity groups what do I mean by that is for example in one particular committee there must be somebody who is doing who is very much excited to do the blogging thing or maybe somebody else is there who is very much excited to do the presentation thing but someone is not interested he can prepare the presentation he is very much reluctant to go out to the audiences and then to present it. So in such kind of scenarios always provide them the organization means that hey if you need to present if you need an assistance I can be a co-speaker which will give a confidence to that particular person next time maybe he can lead a mega event you never know this this really happens many other times right develop inclusive committee leadership models this is very much welcomed in Red Hat. This is the open framework open leadership framework so but actually this this is completely dependent on a person whether because no matter how many policies you have no matter how many leadership models you have but it should be somewhere down in your heart and in your mind when you are leading any open source project that it should have an open decision framework it should have an open leadership framework what do I mean by that is the very smallest example I'll give you is once I was I traveled to one of the events back in India in Bangalore and we were just about to go to hang out in a coffee shop the Starbucks we were just taking we were in a queue we were just taking our coffees outside and then I realized that one of the member he said that hey buddy can I go first and he was within our community then so in this particular thing we said yes of course we always say yes of course and then one of the leader who was available over there he took care of the fact that he was standing at the rear end of the queue and he was making sure that each and everybody who is available on his team they are getting the coffees first by this small act of kindness or giving a priority to your team it clearly showed that not only your ideas in the business models they work but in the personal spaces too where a good leader is very much open to include each and every of his team member the very first this really helps to build a very strong community the open source community the thing is Nick the strongest point which I find out is integrate dmi standards and best practices into the product life cycles this is very much imperative when we are releasing a new product life cycle and it happens when we are having very crucial business standards on the table and then we want to discuss so one of the was there I remember back when I was into some other team within that and I released there was an idea which was coming from an affiliate level and it was having a principal engineer the ideas which were which were brought up there made it convince coming from an associate level it was the right one and it got implemented no matter where the ideas come in because the inclusivity is the principle of the organization of the community it makes the decisions very good and the product life cycle is easy to get a release on okay so so this is this is very much interesting since the open source community they are very much geographically distributed they are across the different countries now the position many times it happens that though English is a common language they all speak English but are we sending out the right intent what you mean through your words what does it mean is many times somebody is in a problem within a committee when they want to complete one particular task they say it in simple words but some people are not very much comfortable relaying their emotions through their words you know so in such kind of activities we should have there should be a common language the standard language we should always try to break the language barriers out there because it helps to put people together for example many of the people say that i'm available on google chat but i'm not available on Slack but other region they work only on Slack but for the open source communities there should be a defined mode of communication that for this community if in case you have to come you communicate something this is the channel this is the right and the most adopted channel so this removes the confusion that shall shall go or i was speaking here the very first one risk which i have realized is the lack of the motivation over the period of time though somebody joins the open source community and they won't be motivated to do so and the second is sometimes the policies bring up into the people and this causes a risk so how can you overcome certain risks and enforce certain policies and processes this is very much being by the community there is an awesome leader you know who is taking care of the policies of the processes wherever there is some assistance whether it is about delivering a talk out there or applying to a different kind of a position within red hat the leaders the community leaders are always approachable the other ones who can even through resume is they just help you to know that which skills do you want to build so such kind of assistance is available but policies second is powerful in the other open software and also to map open source through vulnerability so this comes to the product life cycles whenever there is something into the upstream there is any kind of a code which you are pushing into the upstream and that is being utilized into the downstream all of these policies come over there and identify the licensing risk so licensing is a very very huge topic which certainly is out of the purview of this particular presentation because it is something like which can go up to two and three hours of discussion because licensing is a very key thing in the open source software development which which is a proprietary of a particular organization though it is open source but something which belongs to one organization and licensing is the key to contribution as well as to utilization of particular software so people ask me why the open source why not why not do go with any of the features of which are available in the market the open source communities the very first thing is they help you to retain talent there are an awesome pool of talent available out in the world they are very much open they do not want to stick to one organized but they want to make the the codes very much easy available to each and everybody so there is such kind of talent can be so much helpful to the open source party as well as in the community the maintenance cost of the open source committee is very much less to the parties out there as well as the it is open to all the motive is not to bring it to a success by delivering something but it is the motive is to influence a vision once you can influence somebody towards a vision you can bring in the success very easily but it takes a lot of courage as well as a lot of time to influence somebody and for this we constant contributions into open source communities so now how can I bring or make an impact in an open source the very first thing is build relationships at ease right now the example of that whenever you are having time try to build up connections with different people even if in case you have a Google chat or Google meet always make sure that you are building a connection feel that you know you need a reach somebody asking you any query try to give them an answer before they come back to you for a follow this really helps you to build a very strong connection the community is very much related to the software engineers who pushed out the code and they work regularly with the internal stakeholders so I can say that it is very much easy for me to catch up with someone to build a relationship and just to be in touch for a month or two but it is very much important that you keep a regular sync up with them to keep a connect then only you can make your relationship alive with them in the community practices you should be a part of the sync up team sync up or the team meetings very often at least once in a month this really helps you to build your relationships the next thing is how to create your open source contribution strategy this is very much important why are these contributions important what open source projects do we use within the organizations definitely you should always strategize or you should always focus on the strategies that do we already have relevant expertise or we need to upskill for it whenever you are making a contribution strategy you should always think of all these questions what should we promote our open source efforts how will we determine whether the plan is successful or not these are the things measurable things which you should measure before you go ahead and make any open source contribution strategy once you have all the answers for these after the self-realization then only you can go ahead and measure of this you can have a plan for improving it because if some of the questions are there which you do not have a answer for then you can certainly would like to go back and think about how can I make an impact or how can I make a contribution strategy you to this particular importance of information so now the most important is whenever I go to any event most of the people they walk to me and they ask me that hey apart from just the code how can I contribute to the open source because it has a very much tech oriented organization or most of the open source committees are for example four-mancible but what can I do so honestly speaking when I started into the open source contribution it was not like that I was only into the technical part though I was into technical but it was not only technical then I started making the blogs it really helped the people learn about the different products and the technologies the contributions the different events which were having what is the impact of the events all of this information was proven to be very much helpful to them whenever you you find a portion if you are into QE you can become a tester for the particular community for testing out a few codes if not then you can write down the documentation this is very much useful when it comes to the technical writing or the non-technical writing there are certain opportunities within the community where you can write down the product documentation if not then you have the you have the freedom for writing some of the feature specifications about product there that is the moment when you can use your writing skills then is definitely become a translator if you can it has helped me in handling some of the projects where I was working in a pack and was specifically to Japan where some of the community members they helped me as translator and it was a very very big success for us in the event whenever you need or you can help of the posters which they could make these are just contributions which they could help us out with also give a talk at a user group meeting wherever you feel there is an opportunity and it is very much relevant for your audiences go ahead and deliver a talk it really helps community members answer about the questions and the thing which I helped out is suggest a feature this is very much important whenever you are in a team meeting where there is something being pushed to the upstream now there is something to hesitate to contribute such kind of contributions need to very good product features in some of the ways so before I go ahead so are you ready to measure how you can contribute to the open source communities if yes you are good to go if not then please go ahead and think that how and you can make a contribution strategy by looking at some of the questions and then what are the improvement areas which you could figure out that's all from my side do you have any question for me so if I could understand your questions how can I get the metrics about the number of the yeah yeah sure so and how can you measure so they are certain parameters which are there which are very much specific and they vary from community the communities which I have led they are very much important when it comes to delivery of certain tasks to the upstream as well as the non-technical kind of thing which I have handled is people aspirations then what do they aspire for are they getting what they are aspiring for are they upskilling for that so if not then means the health of a particular community to go for if they are not hitting that particular thing in one year times when for example they are the people who tell me that the amount of time in which I am aiming that I should be there if not it means there is something which we need to tweak the way which in the fashion in which we are working within the community that needs to be traced then yeah so you do that by microphone now by surveys okay and do you do a regular survey of the community thank you everyone for attending the talk thank you I really appreciate your time thanks hello everyone welcome to devconf u.s. 2022 and jingtao on the moderator and this session the speaker is Emily Johnson quality engineer at redbat let's welcome her was from play the python but that was too many characters but I'm Emily Johnson I'm a quality engineer at red hat currently but as you may have surprised from the title and summary of the talk I actually got a degree in psychology initially and was a school teacher and made the rather abrupt pivot to tech so we're going to kind of get into how that happened uh so our basic contents are that story uh and then we're going to dive a little bit into massive open online courses also known as MOOCs uh and that is one of the things that I used that was completely free to get me into learning how to code uh and then we're going to go into some of the research about both MOOCs and uh open coding resources and how that looks demographically and uh what background people are coming from to use those so what can you do with a b.a. in psychology turns out especially north where I live you can teach preschool and that was great it was rewarding but after about a year I was getting burned out at 22 which is not a good time to be burned out so I was looking around for what else I could do you know if I wanted to stay in psych I could have gone to grad school gotten a phd sidey something like that I was kind of staring down many years of extra school in a field I wasn't entirely sure about anymore uh and my boyfriend at the time suggested trying to learn code and I was like yeah I'm not really mathy don't really do that and like turns out as we all might know the computer does the math for you you just have to know logic and be a little creative so I started with uh codecademy.org I learned Python uh it went pretty well the initial things were like building you know a text-based game based on the zelda games so I would like build links backpack it was all very fun um and yeah I realized I loved it I was decently okay at it learning at least so I moved on to edx.org where they host actual classes from universities I took harvard cs 50 or most of it as we'll get to uh and that is actually a move a massive open open online course um and other things I used for project Euler uh they have like little code challenges it's a pretty old school website but like it works really well for learning and some of the programming challenges on reddit actually were very helpful uh so then just continued my story I actually ended up after using all those free resources realizing I really wanted to go for it and went back to school for a second bachelor's degree I went to meredith college in rural north carolina and I attended part-time for two years got another degree uh and then in the first semester of that I was actually hired by IBM as a co-op I actually worked full-time at wild hunting school and the psychology degree A was good to get me in the door because they were like great she already has a degree B uh I was working as a support engineer so in order to deal with customers and write things that they will be able to understand it did help to understand how people think and it did help to be able to uh deal with difficult people both in you know my preschool background the parents and the kids and that really did translate to angry customers sometimes after that I hired full-time at IBM I then moved I got a job at a tiny place called systems prevention incorporated and then about a year ago I was hired at red hat there's my first day at red hat during the pandemic with my messy home office behind me zooming in a little bit on mooks so I took almost one mook um it was hard to get 50 which is pictures there so they are massive open online courses which means they are almost always free they're put together by universities occasionally companies um so there are some paid offerings they for example if you want to get a certificate that says yes I did complete this that you can show an employer or something I never went in for that but I think you can also put together like a degree program and that's a little bit more official but you have to pay for that but if you're just there to learn you don't have to pay anything um they gained popularity in the early 2010 and they're hosted on these various platforms edx in particular is the one I used it was put together by MIT and Harvard but a lot of other universities have the courses on there now um and they can be self-paced you don't necessarily have to start at a specific time it's not like oh this course runs you know April through whatever it's just start when you find it on the internet and many of them are auto graded which means you know you submit an assignment especially if you're writing code like does it do the thing it's supposed to do doesn't have the inputs outputs pretty easy and then for more subjective things there can be a pure feedback situation where like you'll grade five peoples and someone will grade yours um that has been shown to be a little bit less reliable but you know that's kind of how these things go so the early completion rates when people looked at MOOCs actually pretty bad um around 10 if you looked at people who intended to complete the courses it was close to the 30 percent which is still not great but as I mentioned I didn't actually complete the one MOOC I was enrolled in I did most of the work and kind of just dropped off as I got into other things and I still learned a lot like it put me on a great course um going back to school and I really you know had a solid foundation by the time I actually got to use the stuff completion is good but necessarily necessary um computer science courses have been found to have like the largest um they've been the largest magnet basically to these things and uh if people are moving on to other areas they often move from computer science so it brings people in and it gets them interested in other topics which is cool uh so study of the University of Pennsylvania's MOOCs on Coursera found that 60 percent of those enrolled were over age 30 79.4 had a bachelor's degree and 50 percent were employed full-time so the initial promise of MOOCs kind of you know bring education to the masses higher education make it free make it open everyone can do it these numbers show that it's a bit self-selecting it's probably people who have already had the opportunity to get education but it can also bring people in from different backgrounds and areas to learn something new like if you're bringing people in with computer science who already have a degree in something else so overall impacts and effectiveness are still you know all studies are being done by people who are at traditional universities learning traditionally so there's a little bit of like hand-ringing and hemming and hauling like oh no is this going to be disruptive or is it going to be as high quality but nobody's really sure and everybody kind of agrees that having things you know open and accessible is very convenient and has still some promise so who is using these resources um this is survey data from free codecamp.org so not a MOOC but a place where you can come and learn various languages do exercises for free uh 28 percent of their new coders were women non-binary transgender or gender fluid these aren't going to be apples to apples numbers comparisons because the other two I think we're just divided into women and men but 19 percent of comm side-degree recipients at the same time were women and 24 percent effect jobs by way I think that's in the US so it's maybe a little better not like a silver bullet or anything but you know you can still see the potential to get people in the door um of those in the US 45.4 percent were non-white and 38 percent of American tech jobs are currently held by non-white employees so again a little better um of those who were not already developers 90 percent were interested in software development jobs which means people from different fields and different backgrounds trying to get into the field all good and 43 percent already had a bachelor's degree so MOOC numbers to be like well you know it's people who are already uh educated on a higher level but that can bring in different perspectives tech field so let's talk about benefits to the learners from these courses you can get skills and knowledge without spending any money no costs of traditional degrees or even boot camps which can be like 10 to 20 thousand dollars um self-paced courses easier to fit in busy schedules if you already got a job obviously uh there's no intimidation factor you don't show up and you know you're around a bunch of other people when one kid's already raising their hand and they know everything like there's nothing like that you just get to sit on your computer and learn and then you can sample a field without commitment you know you can try of course in something that's completely new to you that you've never heard before and be like is this for me before you have to go enroll in something or pay somebody money for companies uh this has the potential as we saw you know it's not revolutionizing yet but the potential to result in uh employees with more diverse backgrounds and that has shown to result in better team performance is actually a study that shows that software projects specifically do that benefit from first team and uh company financial performance overall can be fair uh and then beyond that this can create a larger pool of candidates overall we all might know that it is really hard to hire people who know how to write code right now would be better if there were more uh so companies can help by providing their training materials for free you know necessarily everything they've got but if they have a field where they're like wow we really wish we had more people who knew x put the materials for x out there uh and not requiring formal degrees for technical roles which you know everybody has that little part on the job description that's like or equivalent experience they could spell that out they could be okay if you've taken the initiative to go like learn a software thing that nobody has or even that we need um we can you know take that into account rather than just being like okay do i have a ba do i not like spell it out for people uh so those are all my sources i've got these slides i think posted to the sketch listing if anybody would like to go dig into those but i have spoken very quickly and finished early does anyone have any idea for anecdotes about preschool and door dealing with customers request for anecdotes that's very fair um yeah i think i've had preschool parents and tech customers try to like bait me in the same way um preschool parents can you know come on this academic thing and let's just be teaching before you say that in school and then was like you know what are their vocab words um so then i was actually for a little while um working at the preschool and going to school part-time for tech and then i immediately switched over like one friday i stopped at preschool and the next monday i started ibm uh and when i dealt with customers sometimes they would come to me like you know this product project is just so critical to our business that like by you not immediately helping me you are killing my small business okay you're installing a server i think it's life or death it's not even up yet i really don't understand how i could be hindering your development in such a way so you can see the skills really translated obviously that's a customer service that sometimes should have come up with more anecdotes but oh yes go ahead of course i keep going back on the line and in the job descriptions expanding the language around that were there i mean i really like that example that's a that's a that's a classic one that one would sort of like gloss over privilege and go oh sure i can expand in my mind that would mean but can everybody do that expansion not silly and i know that there's some efforts underway like at red hat for example the company do about like having a review of job descriptions around inclusive language and so were there some other aspects were there other that you found i mean that was that was clearly one that jumped out for you are there the ones that had jumped out for you are the things that are within that inclusive language and ways of making people especially with the non-traditional roots on how how often we just did there can be these biases in the systems that are written in there nobody thinks to change it kind of thing so if you want to expand on more thoughts on that yeah with the company's language and everybody being aware like if you just have a talk with your team about like hey where'd everybody come from because not everybody necessarily you know graduated high school went to college got the bachelor's degree and exactly what they were going to do for the rest of their lives and started doing that job so you can kind of just gain awareness around you of you know where different people might be coming from and yeah i guess if you wanted to like you know more formalize it really make it an effort like if we're trying to you know expand our teams into more diversity like put that in you know some of your DEI research and goals to be like hey you know look at where people are coming from and make it an actual point to look different you know engineering centers or you can just you know go to a liberal arts college and see what's up um but yeah excellent question anybody else right well i think thanks very much welcome everyone to the next talk in terse lounge my name is karmish i'll be the moderator for this session and let's give a warm applause to eric for his talk thank you so when i when i booked the flight to come out here um red hats travel software informed me that the process of flying myself from phoenix to uh boston to be in front of you today i'm injected about a half a metric ton of carbon into our atmosphere um i'm curious how many people have seen this kind of a carbon report on their travel bookings yeah okay um i expect this to become more and more common as these things progress um so my name is eric ronson software engineer at red hat and today i'm going to talk to you about scaling the open source climate community um and so i'll be describing sort of like briefly like what oh its climate is and what it's trying to do and then i'm going to talk about of course uh what we're doing to scale um and of course we talk a lot about scaling with things like uh compute and data and i'll be talking some about that as well um but i'm also going to talk about scaling um the ability of the community to manage its own deployments and um scaling the actual community itself and actually scaling governance um uh in a sense overall the rest of these um elements um if you find find these topics interesting um i was at another conference recently um and saw a great talk um by these two gentlemen on the perils of building a democratic data platform and um it was a they had their own um perspective on a lot of the same issues i'll be dealing with and so if you find these things fascinating um i highly recommend checking out this talk there's a link at the bottom um when the deck circulates so what is the goal of uh the os climate community um the estimated total sum of financial activity uh in investment um over the planet earth um circa around 2020 was some number north of like 200 trillion euros um so os climate basically wants to align as much of that human economic activity as possible um with the goal of controlling global temperature rise um so in a sense we're trying to scale the climate impact of human investment um so what does that what does that look like um there's a lot of ways you can slice this problem on one of the ones that were actually got working um relatively early it was called asset alignment um so what does that mean well suppose you have um an investment portfolio and you know you can actually with this tool which we're deploying on the os climate clusters um you know select a benchmark like down and in the uh lower left here um you know the standard like 1.5 degrees celsius temperature rise people are trying to keep um keep our carbon emissions so it stays under this limit i mean you can see over here um they actually measure if you look at all the emissions pertaining to the companies in this portfolio um they're not meeting the goal so um this is you know providing this kind of tool to the people who create investment products and also to the people who purchase them um is one of the main functions that we're hoping to support um so asset alignment consists of a couple different things um there's physical risk alignment and transition risk alignment um physical risk is probably the one people are be intuitively familiar with um you're basically trying to federate a lot of different kinds of data data about potential hazards like when wind speed temperature rise floods um combined with like the actual location the physical locations of corporate assets um and link the again to like the financial investment data that i just showed um so the idea is of course that if you're a company um that maybe owns another company and it sits on a river and that river starts flooding more often because of climate change that's a physical risk impact to the value of your investments the transition risk is slightly more um esoteric it has to do with the risks we impose on ourselves um by the policy choices we make as governments um or societies um and so here again you're trying to you know federate data about um energy power energy source choices um and you know the policies like say we pass legislation to you know increase wind power and stuff like that or the failure to do those things um and what the actual impacts might be again to all of our you know physical assets that embody all of our economic activity so as you can see oh this climate works in the esg arena which is the increasingly popular acronym for uh environmental social and government's concerns so there are obviously tons of challenges to addressing this kind of problem um you know regulatory roadmaps are constantly changing um again i mentioned like you know there's physical risk and transition risks um that are you know challenging to federate the data for and keep track of um and there's a lot of just data challenges in terms of like you know actually banks are unable to um often get access to this data or don't have the resources unless they're very large to properly process it in the way that they can actually make use of in the ways i've been describing um and it's even worse than that of course because not all um not all stakeholders are necessarily honest um you know they misreport their values or report them in misleading ways um so you're having to deal with the possibly you know dishonest data or incorrect data just by accident um that problem has become so acute that uh just this may the sec submitted a proposal um do a number um broad things quote unquote with respect to a registered esg fund so you must now provide um investors with information in these investment prospectuses like what esg factors are actually considering um and uh then disclose you know like what the data they're using to compute the values they're reporting and what metrics they're actually using which includes things like you know what you're necessarily porting in um so you can see that uh you know computing these for something like greenhouse gases um or other kinds of pollutants um require a lot of complex relations um there's different scopes to this scope one is sort of like direct emissions like if you're a company that owns a factory and it has a smokestack you know that's scope one that's pushing carbon out scope two is things where you know you're basically purchasing energy that may have emissions so again like you know my flying out here could be construed as a scope two impact for red hat um or when we purchase you know computing power out on the cloud and from some ways like aws and aws is purchasing um you know electricity from somebody that's kind of like a scope three kind of a level so like there's a different levels of indirection and so in order to compute these things um you actually need to be able to track you know what companies own who um and what you know what companies are in the supply chains um so one one piece of data we've actually stood up relatively early on our platform is the global legal entity identifier foundation data set glyph um which contains basically nothing but these relationships it's like what companies exist what are their legal identifier um numbers and you know who who they own who owns them etc um so you can see uh being able to walk through this data and trace it is crucial to being able to sum up scope two and scope three um climate impacts the oaf climate it's an open community um and it's open in a sense of open source but also open data and open ops and open governance and I'll be trying to talk about all these factors as I go so by the numbers um this number slightly out of date there you have just shy of 200 individual members now I think it's 195 um there's around 10 community projects that have been started and around 50 repositories and probably more than 20 data sets at this point um and the general story is that over the last six months or so all of these have just about doubled in size so it's a fairly rapidly growing community um and um if you want to explore further like what oaf climate is doing what we're all up to I encourage you to uh check out these links um that's the oaf climate website and our actual github org I'll not be talking too much more about like the specifics of the tasks I'm going to be talking about how we actually scale some of the uh community activities so at the lowest level is one we're often familiar with um you know scaling compute and so our model um is that we run on a kubernetes cluster and it's running on top of a aws and so of course kubernetes allows you to scale your workloads using um you know pods deployments that can be expanded or contracted and aws provides the actual scalable compute hardware um this is a fairly common um organization of compute activity these days show of hands like how many people use this kind of um environment to scale their own stuff okay maybe a third of you so again you know at a you can talk tell aws to increase the size of your uh number of servers and I can ask kubernetes to you know scale out the number of pods and my trino deployments so of course that was scaling out there's also scaling up and so again I can ask aws to give me you know servers with more cpus or more ram and just literally scale the hardware up and similarly I can do the same thing at the kubernetes control plane and say look I just need this pod to have more memory to do my data science or more cpus um so um in the case of aws climate the particular flavor of kubernetes is an open shift and aws um actually donated the sort of like compute credits to run the cluster but as we all know um running these things uh cost money um and so again it's a here's a here's a useful example of cultivating relationships for your community to obtain resources that allow you to scale because otherwise we'd be having to find a way to pay you know pay that money ourselves um the next we'll talk about scaling data um the main data architecture currently operating is um trino which is a scalable database um and it's using s3 buckets um as its backing store and of course we are also just literally exposing some s3 buckets directly for certain purposes um so we're using the we're using the s3 object stores in three different scalable ways the first is sort of like a physical landing bucket we provide one um community members can also bring their own bucket um so that's an example of basically you know community data community data scaling um and again as I said we're using it as the backing store for um all of our trino sequel databases um and also um when we do uh kufla pipelines and other kinds of pipelines that's generally provided in the backing store for intermediate results um so trino is um designed to expose a sequel interface um to a user but under the hood um it scales out the computation of all the queries that you give it um using you know pods on some kind of cluster um and so in this respect um anybody who's ever worked with something like uh spark structured streaming might find this kind of similar to you know exposing running sequel queries against spark spark can do basically the same thing however trino um unlike spark was designed natively to run in a cluster and do this and so it turns out that it's somewhat easier to get this kind of deployment working to do scalable you know database like operations um another way it scales is federation it comes with over three dozen connectors at this point and they're writing more all the time um and so what it allows you to do again somewhat like spark is um declare a connection to somebody else's data and as long as it's in one of the formats that it knows how to talk to um I think the some of the ones that are particular interest to us are um like Google BigQuery and iceberg um and Kafka may be coming to and then also of course it knows how to ingest uh csv and parquet files and so again suppose you want to um take something like a csv file and get it into a database table um on our trino deployment which is running on top of an iceberg connector so it's basically leveraging icebergs data versioning at a very fundamental level um I wanted to provide the community with some standard tooling to do this and so I wrote a little um python class called trino batch insert and you can use it as a method argument to um the standard sort of like panda's two sequel method and I just taught it how to talk to trino um and so this is an example again of community scaling um via providing people standard tools so I wanted to say look I'll figure out how to do this and I'll give you guys a function to do it so nobody has to like reinvent the wheel um so how did this work well the UX was good you can see it's actually pretty simple you just give it a simple call like this lots of people know how to do pandas two sequel um unfortunately we found out relatively fast that the scalability is terrible um it was fine for like the testing I did but when people tried to start ingesting real data like into the you know hundreds of megabytes or gigabyte class um it immediately started taking multiple days to cycle through um so that was unattainable um so I went back to the drawing board and tried a different method um another another connector that trino can use is hive tables and it knows trino knows how to do a clever trick with hive tables um if you push a bunch of stuff onto s3 you can simply tell trino hey here's some parquet files in a structure I would like you to treat this as a database table and that's effectively instantaneous um and then you can run a sequel insert command from the hive into the iceberg and because that's basically letting trino do the thing it's designed to do very well it's much faster now ux obviously slightly more complicated but it is literally 70 times faster to ingest data so it was a massive win um and again I exposed this I took all this and wrapped it up in a function so that people don't necessarily have to manually do all this and call function it does all this under the hood um so let's talk about scaling deployments um we have some goals um you know we want deployments to actually be configurable by all the osc community members um and discussion reviews and approval um of all these kinds of deployment changes should be all full in the open um and also durable um so the way we did this was leveraging the operate first community their purpose is to basically take open source principles and apply them to actual um deploying in operations in the open so operate first um so how does that work um suppose you're open source climate community member um you want to change the deployment on our clusters you do this by submitting a pull request via git um and the operate first github org and it usually basically means modifying things like customize overlays and yaml the you know the kinds of things that kubernetes clusters know how to talk um so here's an example of one I did I just wanted to upgrade the uh database on the trino database to a new version and so I submitted a pull request you can see it right here it's a very simple one um but the main point the main point is it's all sitting up on kit you guys could all go look at this and see the history see who talked about it who approved it um and so again you basically get the entire open source process of discussion and review um in a durable way up on github um and once the discussion finishes and you merge something we have an argo deployment running through the operate first project that detects these changes and says okay I got a new thing to sync up and resyncs the deployment out on the actual physical cluster um and you have a change and it's all been done using open source principles um so again this is a great way to you know scale out the community's ability to make changes to things um themselves but have review and full durability it's like a five minutes I better hurry up um so the last thing I actually wanted to talk about was scaling community um the primary way we've done this uh is with self-service environment so as an OS climate community member um I can basically authenticate to our cluster via my github ID so it's easy for everybody anybody the github and that also gives me um authenticated access to Jupiter environment superset um which both are connected to Trino which is also using the same authentication methods um and all this is being uh installed here uh using the open data hub project so basically you have unified authentication which is a great way for people to you know get access to a lot of tooling without a lot of overhead um so one github ID authenticates to all the uh tooling we're running we also am scaling with github templates so somebody wants to on board previously you know you could either talk to some osc admin with a bunch of channels um none of it was organized and it was worse than that because the admin was usually me um so what I did was I created some github issue templates where people can basically say I want on board I'm going to use this template and so it would get them to give me the required information um so things like you know the github team names their username it gives a little task list for people to make sure everything happens it gives some standardized labels uh and assignees so again basically this allows you know things to be organized far better in the future we want to take this and do expand it to actually doing some automation on github um so scaling governance um at the community level we're doing this fundamentally like a cloud native compute foundation where you know you kind of have a rough one-to-one relationship between participants a data product they want to provide a github project or uh you know an actual trino schema schema that they can use for their purposes um and we're running out of time I'm going to skip over that um so future work we want to um you know again leverage leverage some of our uh domain specific languages to improve automation of the trino credentialing we need to do metadata management um integrate code with like you know data versioning either via things like paki term or a combination of dvc and dbt um so that was a lot of work to do and we had trouble so the last thing I want to talk about is actually scaling human participation sometimes even if you're using a very clever technique if the problem is big enough you just got to throw a lot of people at it um so how did we do this well we fostered a relationship and became a member of the linux foundation um and this allowed us to access some of the linux foundations um dev ops people to assist the operate first community and doing some of our deployments um one of the ones that freed us up to do recently was open metadata so now our data tables are actually browsable um whereas before they weren't so again this has been a massive community scaling uh win um by acquiring new dev resources so to summarize how to scale your community you should build on platforms that are designed to scale federate your systems like trino and your data and also your governance principles try to make your environment self-service so you're not trying to like do things for people that can do things for themselves and access tools um and introduce complexity only when necessary otherwise try the simplest thing first um and publish libraries to standardized patterns of operations um build build on solutions that actually facilitate automation um and cultivate strategic relationships that give you access to resources um so anyway i hope that inspired you to go forth and uh scale your own communities um and thanks a lot for tony thanks uh well kind of shouting to the mic it was a great talk um one of the i i wanted to draw back to a point that you had made a moment there where you made a um a form to have people fill out to do that part for them and what i wanted to catch on that is there um that same kind of process of uh adding yourself to the github group like somebody can go in and make a poll request for themselves to do this piece and this was an example of where you identify it seems like you identified that that was a barrier to people to entry to people getting that initial setup and then you chose to do a human position here as opposed to a self-service action because it accomplished a goal and and i think i mean i don't want to tell your story because i'm curious what what led you to hear and why you did this particular piece benefits just in terms of organizationally but what i really hoped to do which i haven't gotten to yet is to follow this up with you know again automating so like you could write a github action that would detect the creation of this and then it could do some of these things automatically that we are doing manually um this part has not happened yet but this was basically a strategic positioning to get some benefit immediately but then set us up for actual automation we're gonna create a problem huh so rather than leaving the individual because a person could go make that same kind of poll request to themselves against the um i'm sorry is it is it pre-blows right that's doing the it's doing the github organizational changes when so there's a the the the the onboarding request form that you did there some of that included pieces that somebody could theoretically go and set up for themselves if they knew how to do all the pathways right like going and making that's an actual github issue form you know i can like tell github to give standard forms when you get so like a large part of that they can't do for themselves they have to add you have to be like an administrator of the github org to make certain changes to put them as memberships of special groups that govern the open shift access so they pieces that they couldn't do that themselves uh yeah some if you can do it with the poll request they can do it themselves this was active so yeah it's like it's like github console stuff so it's a little different but you know there's like a lot of different get here can you go a little bit into the different kinds of consumers of this data it seems like you're focused on the like etf issuers but it also seems like there's a whole swath of different audiences maybe really a lot of audiences i mean like on an example um i'm trying to pursue right now is of course people are trying to bring esg reporting like into the kubernetes control plane there was a talk yesterday here about the pro kepler project which i recommend people go check out um but of course if you want to increase the fine-grained reporting um of what your compute races are actually using let's say you wanted to know hour by hour like how much solar versus how much coal my compute was using you actually have to have access to like where the physical location of the uh you know compute resources were what power plant they were power grid they were drawing from so like all of that becomes like a combination another federation of things like you know physical locations and reported data about grid power you know sources um so i would like to see that hosted on the us climate community i think it's highly aligned but it would actually be being consumed by kepler that's one example but the other many i like i much more frankly than i'm probably even aware of how are students in educational institutions um working to like enhance the os climate community and what you were mentioning student activity um however it is um again a fully open community and so we if there are like student organizations or even individual students that want to contribute in this space um it's all fully it's a fully open community and it's fully open source and so like i would well welcome that kind of involvement also um microphone problem um also i am wondering um uh do you know how um the community grew in the last year do you have any statistics on how the os climate community grew the size here it's actually slightly larger now these are numbers that are about eight weeks old but um again i think basically those doubled all roughly doubled in the last half year awesome but thanks thanks everybody this one i think one of these everybody hear me all right okay hello hi how are you my name is ifa maloney if you want to flick to the the directors so that one is me uh you pronounce the name ifa backwards drop a couple of the letters and i am the um product owner on the community platform engineering team i'm going to talk to you today about how we've changed our planning styles we're still going to we still plan our work but we just do it in what we find is a little bit more fun and a little bit more human uh i'm here today as well with uh one of the cp managers i'll leave you to introduce yourselves thanks ifa yep my name is stefan matthew so um the accent and the name they don't they don't uh match but anyway um so yeah look i've over 20 20 years of experience in it uh i joined the cp team a couple of years ago uh i'm when i came in i had some new ideas maybe to bring to the team so this is what we're here for today okay move on so yeah so look just the schedule so we start off a little bit of history just around you know where the team was what we were doing and we then move on to auditions so some of the resources we used the trial and erroring that we kind of that we did uh the dailies then just the detail our new approach uh editing what did the whole thing look like when we put it all together because there was a few different parts of it um some awards maybe um so just take all her feedback and then maybe some questions and answers in the interview section so that's it on we go okay so you've probably been listening to a lot of technical talks for the last two days maybe three days if you're here for the dojo as well this is not going to be technical at all but it's important if you plan with technical teams so and you need to have people talking because sometimes we're not very good at communicating with each other especially not when we're on calls and remote and the best way to make good decisions good planning is actually talk through and be comfortable with people that you're with talking about things so this is kind of a rough overview of the timeline that we were looking at in the cpe team the whole planning process is based on agile concepts and a lot of the team are allergic to agile when it came in so we've had to kind of water down an awful lot not water for a little lot but um so we uh we were introducing this concept to the team didn't like us I came on they liked me and then we were looking at changing our processes and they were awful because they were way too agile for us so then we figured out our own way and now we're here today just to share it with everybody because we really enjoy our planning now and I would like to think the team do as well they might admit it but I think they do okay okay so this is why we were bored out of our minds with the original approach because because it's it's a it's a spreadsheet like it's a spreadsheet about technical stuff and yep very effective you can see the names of the projects we were doing it was linked to more documentation um how long it was going to be who was on the team what they were going to need to know and then a very loose kind of Moscow voting system and we all turned up four times a year for two and a half hours and went yeah okay open a new tab mic off camera off boring yep boring it was it was terrible it was really formal so I mean if you're looking at a purely from a business perspective you were going to say yes you're ticking the boxes and we were all going boxes are shit sorry about the swearing and we were talking what we found was we were talking to the people and nobody was offering anything back to us just you know a plus one in the chat which are mic and camera off you know that where how are you supposed to plan properly if people are not talking back to you and giving you their honest opinion because I was hearing their honest opinions three and four weeks later when we were in the project and somebody would say I don't know I don't think we should really have done this yeah yeah where were you four weeks ago oh you were on a different tab probably on Netflix anyway next slide this is going to go fast because we have a lot of pictures yeah so I think there's someone yes so look what were we going to do then how are we going to change that form and scripted kind of nature of the planning process you know it was all about engagement so we needed engagement from the stakeholders engagement from the team and you know we needed to get those real insights from people it was we weren't getting anything when you look at a spreadsheet like if you said people tune out after five minutes and that was it you got nothing so you know we wanted to kind of bring this give this a bit of personality is what it was so it was all about that engagement bring the personality in and make sure that we kind of try and engage people in the way that they'd actually want to talk to us so we iterated slowly because nobody likes change and we just played around with a couple of ideas of like adding a little bit of color to the spreadsheet because it was boring as to tears we were looking at maybe how could we get this this whole process to be a bit more visual as well because a lot of people like there's three kind of learning styles there's the our visual and kinesthetic most people learned with their eyes you're going to look at something and say whoa we're not going to do that or well that looks really interesting so we were trying to apply that to presenting project information to stakeholders and teams to see it to visualize it and to want to talk back to us about it so we did this and they didn't hate it so we kind of went all right we could probably do a little bit more now that we haven't gotten a bad reaction from it so she just want to talk about that a little bit there the stick man stick person stick person sorry it's just it's a term um just how we added that and that was kind of like the first little thing we got a bit of a reaction from it and things that yeah we could do like I mean yeah we tried to show the people behind the work so we used stick people and that was good so people were like okay that takes three of the people rather than just the number three which yeah it's not going to weigh too much into here but I think as well what happened there was we actually had a lot of engagement with that just people started speaking about oh the stick people are there now that's new something new so yeah so we doubled down yep so I suppose like I said when I joined the team I had some previous experience in another part of the organization where we kind of maybe used more of the open practice library and I don't know if people know that but that's just a basically a collection of really you know interesting agile techniques and processes and so we kind of went back to that and we had to look and we said look what is there something here that can help us with this and what we came across was an interesting idea called it was originally called cover story so it was to kind of lay out kind of you know information in a kind of a magazine type format where you would have you know a cover image information about and things like that so we said okay that's that's interesting let's see if we can adapt that a little bit to what we might need so we kind of drew it up similar to this at the time and you know we had a few basic requirements we wanted you know we wanted to use something from the open practice library like we said we needed to use it online because all our team is pretty much remote we were in the middle of the pandemic as well so we couldn't come together anyway so it had to be some sort of online white boarding solution or whatever it was and we needed the information to be all in one space you see the spreadsheets that were there earlier you were brought off the other documents you had to see many many different things to be able to um understand what the basic idea was for the initiative we were looking at and we just wanted a more dynamic and fresher approach to it to try and get people engaged and make sure that they would be interested in actually working with us so what we came up with was initiative stories we call it so we tweaked it slightly from what was there in the cover stories of practice library and maybe I just run through each section so we had a cover image obviously and this became quite important for us and this is where the movie kind of concept came in and we had like you know coming to a screen near you which was like an estimate by quotes from people which were quite good as well brainstorm and ideas we had the big headlines which is basically what are people saying about this initiative what do they want to know about it and the critic reviews then were from people in the team where they would basically add their commentary about what they knew about this initiative and we had things like the cast how many people would take we did credits then which brought you to those other documents if you wanted to see them but it was generally all contained on this screen and then probably the most important one there was we had a bottom line we just call it so one mistake or a perspective did they want this to proceed or not were they interested in actually doing it or not so that was kind of the basic makeup of the of the information we wanted to gather and it looked like that so a lot of color and the teams stakeholders were presented with this top template first and purely because I think I think it was around Oscar season when we started this but Stefan is into movies a lot of the team were into movies I like Oomah Thurman and we thought it might be really fun to see what it would look like if we did the movie poster element and we presented this to the team it's a project it's one of the projects that was in our quarterly planning but we have it as kill bill it's actually that number data grep or it's it's two services the database in the grep service so like it has nothing to do with kill bill but it was kind of funny to see it represented as such and we just had our teams then fill out using stickies and we had a lot of calls about it we used to call we call them insight sessions so they're calls that we set facilitated calls that are optional but people do turn up because they wanted to see what the projects what the movie posters for the projects were and it was a great springboard into oh I've never seen that movie before and somebody else would pipe up and say you've never seen that movie before and it would just spiral off like that and it's very easy then once the conversation starts to kind of guide it back into well put it on your watch list but like have you any experience in data grep or can you tell me a bit more about that oh yeah yeah yeah use that for and we already were getting these organic conversations very easily just by giving something to talk about so yeah it was a lot more colorful so I mentioned about the insights and you saw umith herman but she only represented one project and when you're doing a prioritization call a quarterly planning you're usually talking about a lot of different work packets that you can choose from so we did movie posters for all of the projects that we were going to choose from that quarter we had Teenage Mutant Ninja Turtles we had Jurassic Park for a PDC application I'm not sure does anybody know what that is but when we were when we were talking about what ones and the guys like it's a jungle app I was like well it's going to have to be Jurassic Park then for that one so it was just refreshing to have these colors these posters instead of spreadsheets and the spreadsheets are all linked down here in the credits because they are still important there are you do need to have a source of your information we're trying to contain as much as we can on the boards but if you do need further reading we wanted to make that make sure that was accessible for you as well so yeah like we said you know we got it's visual and eye-catching we were getting the conversation started people wanted to offer their opinions this time rather than checking out for an hour to catch up on other work or emails and we got a lot of in-depth analysis from the applications because there was people who either knew about it and willing to share and people who didn't know about the project would be then working on it and was able to learn from first-hand people who may not have been so we were kind of we were already winning before we even got into the deciding call yeah so I suppose we had all this information how can we make some decisions based on it so we have limited resources in the team and you know we have five or six projects that come up every quarter we might be able to do two or three of them maybe and so what do we do how do we make these decisions so previously we have kind of tried different voting kind of mechanisms Moscow voting things like this but we had our stakeholders which basically and you know we're letting us know what they were interested in but I suppose for that we we found that we didn't get a great response a lot of response was I don't really care and so how do we prioritize that so that was difficult so we said and we look back at the the Open Practice Library again and another technique that I'd used on a few other engagements previously and one of the main things as well was we we did not want to be the drivers of this discussion we just wanted to facilitate it but we didn't want we found that we were talking the whole time we were kind of making decisions almost we wanted those decisions to come from the stakeholders instead so that we were clear on what we need what we needed to work on so yeah back to the Open Practice Library again this technique very simple technique priority sliders it's called and basically you know each project or initiative that we had can only have one kind of score on this not the five scale that we've got here so you know every stakeholder had to kind of discuss we went through and we discussed each initiative okay so where do you think that this should should belong is it a two is it a three is it four is it five where's it go and the stakeholders then started actually discussing amongst themselves I think this is a one for this reason I think this is a two whatever was so we had some really good engagement some good discussion actually happened and we just faded into the background more than anything and just said okay guys we just used terms like when when people were saying oh I think that's a one which said well can you explain that explain that to the rest of the group so we can understand what you mean and so yeah so you know I've used this technique before on a few engagements like I said and it gets to you now and unanimous agreement at the end of the plan and also it shows where people have to make compromises as well based upon resources that are available so I think it's it's a really good technique to use if you have any type of contentious decisions or whatever amongst stakeholder groups where you can you can really easy and visually put it out like this and you know get the conversation going so that's that's kind of what we did for that in terms of um yeah getting the results that we needed a non-confrontational way so we had done our insight sessions we'd gathered all the information about the projects these really heavy technical projects that I as somebody who sits in this weird gray area of not an engineer not a manager but still responsible for kind of both sometimes it was great to be able to get the information all from the team and the stakeholders and then then be able to circulate it back out before we went into this decision and to be honest because I'm in those conversations from from the get go from about two weeks before we decide on the work plan I already nearly know what's going to happen because people have been so forthcoming in those calls because they're in a call where they feel safe and they're comfortable it's not recorded but like it is facilitated so the commentary is is described but because it's not recorded people are a little bit freer with their discussion and they feel safer to do so and they can challenge opinions respectfully so most of the time it only happened once that it didn't but they were able to do that and it still didn't derail conversations and we still had all this wonderful insights about the projects that we didn't have in those documents beforehand so we put it all together we send it out we send a summary email capturing the perspectives coming up to the call then we have that call that Steph has mentioned that we use the priority sliders where we discuss and rank the projects and see this is what it looks like this is your one to five wish list we can do one to three what does that look like for you and everybody was everybody felt like they had had their fair say and then we landed on the result and we were able to say okay done we have our next quarter planned and then it gets sent out again so i hate emails i absolutely hate them i'd rather get on a call with a hundred people and tell us something in five minutes but they are unnecessary evil evil because time zones so the information is important to get back out and we were able to do that quite easily and certainly like i said from somebody who doesn't come from a technical background i was able to give the right level of information to the stakeholders before and after the prioritization call and i think just looking at this screen as well you can see all the information is contained there when we had those calls we could easily flick back to any of the any of the initiatives talk about you know particular issues with them or particular things that we wanted to dive into a bit more and just keep flicking back and forth to make those decisions on the prioritization call so the vast vast majority of the information was contained on the one board we didn't really need to go anywhere else to have those discussions or make those decisions so that was a big part of it so they loved us they really loved us we had a couple of them a couple of nice commentary back from our main stakeholders Matthew was one of them um but yeah it everybody felt like it was it was a really nice way to approach quarterly planning because those those of you who are in teams that do it you probably have felt the board and kick in after 20 minutes of looking at a spreadsheet in a document we were able to do our quarterly planning calls in 45 minutes was our best record because we had these really comfortable engaging useful conversations coming up to it um what I did like on I think it was Bex he had a movie list watch out of it so he has now a list of movies that he got he now has to watch which he felt was great um from stakeholders of our team who are not part of our team but close enough for into our team that we see every day they felt like they were becoming part of it that they had gained friends from it so I mean if you if you work with people who you consider friends are you really working and it was it was what we were trying to get we were trying to get this really comfortable useful conversation coming into our quarterly planning sessions so we could then decide on the best pieces of work that we should devote time engineering time people time too and it would generate the most value for our stakeholders so but we did it in a kind of a nicer way did match Miller Miller actually say that there you go he did okay very good um yeah look and and that was the basic principle behind it we've just gone through it super quick there because we only had 20 minutes this talk we've given for an hour before but um yes look I think that's it that's the end we've keep we've right on time I think so that's good so uh roll the credits yeah step on to take it I was like no this is the venue for Monty Python and we have a few minutes for questions hey great talk lads um yeah that's really interesting and and I'm involved a lot in continuous planning as part of a different engineering section so really interesting stuff in this two quick questions uh for the priority sliders if you have different stakeholders with different requirements you have product security coming in with something compliant to SML you have QE coming in you have PM coming in engineering coming in who's decide if they do not come to unanimous decision on the priority for a specific piece of work who it makes that ultimate decision not where that lands so look they should make the decision that that's who that's that's we were trying to teach our stakeholders as well that you know they have to make the decisions themselves we're a team that can support the decisions that they make whatever but they need to make the decisions themselves they're there compete there's there's always going to be competing you know priorities for everyone but what we found previously was we were making the decisions like I said we would just say okay so look it looks like you want to do that so we just do it but what we found over time a little bit was and this didn't just happen overnight either this whole process was probably about six months and you know from when we started it to when we wrote it out and so what I'd say is the time is up and so what I say is it's more about teaching the stakeholders that they will have to compromise they will have to make decisions themselves and it will be all in their best interest to do that I agree so you have you you've never had a scenario where there's been just pure disagreement on the priority of the guide no no not really I'm like to double down on what what Steph means is that the idea of the priority slider is called is that all stakeholders are present and they have to listen to every each perspective yeah so the idea would be that if there's vastly competing priority or priorities or requirements by the time ProdSec has had their say as to why this is more important if it truly is actually more important to the entire project the other stakeholders should should understand where they're coming from and maybe change their decision it's about influencing other people's decisions and you kind of keep going until everyone feels unanimous on us and I suppose what it was as well was you then see people like so someone might say okay this is the most important priority now and they might get that and then you see them compromising a little bit on the second and the third and all that so you kind of do get a fairly good feel for what's going on yeah hello I do not have time I'll I'll give it to I want to ask my second question but yeah great talk I think being able to bring people together in that way using planning is pretty awesome so thanks go easy now Karsten okay um the yeah actually that was really interesting that what you described was a consensus decision-making model where you drive towards solving people's problems instead of getting a 51 49% vote thing right yeah that's um the question I wanted to ask was about the the evolution of the the the display did you reach a point where it got did you do anything that was too much visually people on the team that said was there anybody who had you know colorblind this or other elements that made it or had a hard time parsing that with visual information and it got to be a little overwhelming were there any that come at all and you have to make any adjustments that's actually really good um a really good comment to be honest nobody has said they were colorblind to us but I never even thought of it no but yeah no I did get I did get a bit of feedback on the site um not in not an any confrontational way of just like jeez even there's too much color in that thing I can't look at it I'm like well don't don't look at it just come on the call just like share your your thoughts you don't have to interact with Miro there was one guy who said every time I log on to the thing my laptop is going to explode I was like okay just join the call just be part of the conversation I don't need you to put your thoughts on a sticky that's why it's a facilitator call I'll do that I'm listening to all of the conversation I'm I'm catching those commentary as and when they come up so if you're not comfortable logging into Miro because your laptop's gonna blow up if you're not comfortable looking at the screen just just stop by and listen and be a participant so that's um yes that did happen but it didn't actually harm what we wanted out of the finish and look it was all about engagement getting people engaged to join the call to speak to do it instead of just sitting there so that's pretty much what it was if you were a project manager or product owner that hasn't had a lot of experience with software engineering and you're tasked with trying to plan software technical work I highly recommend that you find a really neutral way to like engage with your team because trying to meet them down at the code level it's not going to work for you and they are not going to be able to explain things to you on a one-to-one basis very well you're better off with a semi big comfortable space where everybody can just join and converse when you were talking about movies like Kill Bill and Jurassic Park and things like that you said that they were good conversation starters did you actually use that movie metaphor for you know the movie you chose in other ways like drawing parallels with the project beyond the yeah I think you might yeah you made a little joke about Jurassic Park but yeah I think we had a little bit of fun coming up so we looked at the project briefs and we seen you know what what what they were in tail and some of the projects we didn't even know what they were and sometimes it was just a name I think it was one called rpm autospec and we used like the Fast and the Furious and things like that so it was just we just kind of made connections like that yeah so we just make connections like that and then sometimes we did sometimes there was you know maybe people added a few you know a few wise cracks or a few jokes about the movie representing the the the actual project in the end you know or or or appearing like it did so we didn't put loads into it but you could do a whole thing with that as well you could do a whole thing with ask the team you know what movie this would be represented evolve and things like that but yeah it gets a bit tricky trying to match software projects with movies just try and google a few a few movies movies about replacing fedora messaging systems it's not a lot comes back off that google search yeah I did I did go and toy with that one how can we get these pretty graphics and fun movie things into the cpe quarterly report or something like that that gets published for the community because it's long and not exciting even though there's exciting stuff going on we are any question yeah that's that's a good that's a good point I'm pleased to say we are working on infographics rather than the weekly text report to go out I think michelle kineshi is actually spearheading that so you'll definitely see more infographics from our team but yeah we it's the answer to you it's just a miss on my part to actually put it through we use a tool called mirror which is an enterprise tool so it's not open source but we have put one or two of the pictures of the discussion up on I think it's on my fedora fedora people page so it's just a matter of like tidying up on our end to make sure that the information is not just within our team and stakeholders but gets back up to the upstream as well I've made a mental note to remind myself question do you have a link to the mirror template that other teams could potentially use all of this great stuff yep we do so there's there's a blank one up there so we can we can share that around later unfortunately it's probably contained within the red hat space only yeah so we could potentially publish it I'll just have to figure out how to publish it as a template in mirror I think they have to accept it and things like that so 100 percent and some supporting document work obviously but some supporting documentation would be really useful for this is a big company and I think a lot of other people could make good use from this we did also have a write-up on opensource.com and so blog post here we did a four four kind of series of it I think the third or fourth one is out at the moment and so we'll drop a link to that as well in the presentation and that that explains a lot of background it goes into more detail than what we what we said here today as well also on your movie stops and the QR code will bring you to the openpracticelibrary.com link all of the concepts that we took for our planning process we took from open source library which is free and it's completely like go you can even put in your own templates as well we probably could do that with this one but yeah feel free to scan it and there's loads of resources if you're trying to just better your planning planning processes and team norms and all that all that kind of boring admin but could do it really fun way so I would highly recommend using this resource it's excellent I think that's it thanks a lot everyone also don't forget to pick up your party tickets from the registration desk are they doing any like the not to clip on oh it was just because I walk a little bit and but it's like okay and so this one's gonna be on yeah hello everyone um let's welcome Kali um for our next talk around community data it's can people hear me I don't know I'd not really um let's see I can just okay I can also just like talk loudly I'm good at that um so hi my name is Kelly Dolphy and I am a data scientist at in the open source program office at Red Hat um and today we're gonna be talking about community metrics and starting to make decisions around what we want to measure and why we want to measure them so let's break down a little bit what we'll be discussing today so first we're gonna be starting at the value of community metrics why do we want to do this and then the methodology behind generating community metrics and then going into the analysis on lessons learned this third one is really where what I hope we're all to leave with and the type of things you want to think about as you go forward and do your own community analysis so first question here is why do we even want to do community metrics why do we want to spend time doing that and first I'm not going to tell you that data analysis and optimization is going to be the one thing that informs your community decisions it's actually the opposite this is to build off of your own open source community knowledge incorporate others and uncover potential biases or perspectives not like already considered you can think about some of your colleagues that might have a specialization really understand how community meetups work and the impacts that can come here or maybe somebody else is really great with PRs and understanding what balance needs to come there to be helpful towards the community those type of specialties can be integrated into your visualization so everyone can benefit from that type of knowledge and so the next thing is that we all have a thousand and one things to keep up with and it always feels like there is never enough time there's never there's always one more email and if getting an answer about your community is going to take 10 15 hours you're not going to do it multiple times if at all once is really going to be getting there so how can we start making this to be a regular process and a realistic process and also there's just so much data available and there's such a pressure to take okay I have so much repository day I have IRC day I have all this stuff and wanting the pressure to use it all and really capitalize on it so how do we go through with just all the piles of data and pick out the little pieces that can actually bring you insights and be helpful towards your everyday work with your community so first we're going to start thinking about what perspective you want to gain from your analysis and what do you want to receive what do you want to give and there's a few things you want to consider first is your main goal to gain information or influence an action that will be made is there an area of your community that is not understood or are you trying to take a first or you're trying to take that first step of getting to understand that or are you starting to try to influence some actions is there initiative that you want to establish is there a new meetup is there a new processes that you want to do and trying to measure that initiative that's already in place another thing you want to consider is are you wanting to expose an area for improvement or trying to highlight strengths there are going to be times when you're just trying to hype up your community and show how great it is this can be to show business impact advocate for your community all of the above but when it comes to informing yourself about in your community about things that you might want to change or improve on just focusing in on the things that your best at is not going to get you any farther than you already are and identifying these short comes is where you can go and see most of the value from your metrics there is not a problem with highlighting strings there's actually a great time and place from it but you don't want to just have a yes man to just tell you how great your community is and stay stagnant and never change but you may want to show this for outsiders or use this to boost morale and really highlight some community members that have been doing awesome work there's just a balance and really what comes from this is determining what that perspective is for you can keep that separated lastly you want to consider whether you're looking for community impact or business impact the languages that many businesses speak in is numbers and data that can make it incredibly it can make it incredibly difficult for you to advocate for your community and have people listen to you and this can be a way to speak the language that they are used to seeing putting some weight behind your words the language and the message that you're getting across might never change from the analysis that you're showing but you're putting weight to try to get people to hear you and or like I said that's the business side of it you might want to just be looking at your community impacts how is it impacting the open-source ecosystem as a whole how is your community impacting the people inside of it those are some of the things you want to consider and these are not always like an either or situation but the framing of it makes it to be where it's a more deliberate metric and we can hone in on what your actual goal is anything you get beyond that is just a plus so when talking about general data science or machine learning work you're going to see some version of this workflow and so we're going to kind of take a step back to look at the problem of open-source community analysis from the lens of data science analysis as a whole and we're going to be focusing in on this very first step the codifying problems and metrics and we'll be looking at a splash of second but mainly this first one and this entire presentation can actually be viewed as a case study of this first step and this step is often overlooked if you've talked to me you know that I will hype on this often that everyone wants to go to all of the fancy feature engineering the model validation all of the big fancy words and tooling and not actually focusing in on the thing that you want to get out of this what questions are you trying to answer and so you just don't wake up one day and be like I know exactly what I look at I want to look into and I know that this is realistic it's not really how that works so we really want to focus in on this codifying problems and metrics and so now we're going to hone into our goal of our time here and surprise it's codifying problems and metrics the tooling debate is one for another time this is one that I will always talk about if you want to have this debate come find me afterwards and we can argue which debate which tool is better um we're going to start with truly figuring out what we want to know looking at what data that we have and take us to the goal of thoughtful execution of data analysis so we're going to go into some different analysis angles and these are not the only ones they're just the main examples for this talk but they can be generally applied throughout so let's look at our first scenario here this is current data analysis say you kind of already started to go down this path you've gotten your data you know what you're looking what you generally want to look into and now want to start making it better the idea here is to build off of the common traditional open source community analysis commits overtime is cool but what does that actually tell you besides a straight up value we can start taking increment steps of just having a value to having insight so let's first look at number of contributors over time say at this point you know that there has been 150 people who have contributed to in your community at some point okay that's a great number and you can put it on a slide but you're not going to change any actions within your own community on it so why don't we take it a step further and look at active versus drifting contributors um out of those 150 how many of them have actively committed in the last three months has that value changed significantly in the like in the last period of time maybe a year ago about 50 of those people were actively contributing versus now maybe only 10 what is causing that decrease and we can look at it by repeat or fly by contributors so when you have that person come in and do that very first contribution how many of them are actually becoming active contributors in your community and staying around another example we can look at this is commits overtime say this month we have had 100 commits and this next month we see that there's 40 those are units great for a presentation but you're not going to change your decisions so what about looking at the depths of your commits so say those 100 commits there was maybe a single line of code changed or a period in documentation was switched out versus maybe in those 40 there was a huge refactor in the code base an entire series hundreds and thousands of lines of codes moved around those are two very different scenarios that you would not get from just a straight 100 commits versus 40 another way of looking into this is commits by a subset of your contributors is there a part of your code base that is heavily dependent on one two three people is that manageable and how vulnerable are you if something happens in one of those contributors can no longer contribute to your community and so now we'll go into the second scenario which is community campaign impact measurement that's a lot of words let's figure out what that actually means let's think meet up conferences or a community initiative how do you start to view the impact of those goals and there's two steps to this and they actually figure they actually feed into one another once you establish your campaign goals you can go from there to determine what can be measured to detect the impact of these actions it's really easy to fall into the trap of being a little hand wavy and non-concrete with your goals when the campaign comes up and at the end of it kind of feeling a little bit empty and not really knowing what happened with those actions in that effort did something actually come out of it and one example here that I think really shows this scenario is some work that I've been doing with the fedora community they had came and mentioned a goal of doubling the contributors by 2025 that seems like a pretty simple straightforward question but what are we considering a contribution is it just code is it documentation is it working on the website doing designs where do those contributions need to happen to count and so those it might seem so simple in that one sentence but there's really a lot of dynamic to it and once you determine what those where you want to look what counts then you can actually change the campaign and how you go about implementing your goals and what are you going to put your effort into and so yeah we can go into our third scenario here and we're going to spend the majority of our time here as the prior examples can be viewed as different parts of this workflow this is a living cycle and improvements or extensions can always be made you have to kind of determine if it's worth that time to continue putting into it but it's something that you must consider I don't know about y'all but I'm more of an examples person we will walk through these steps at a theoretical level and then go into an in-depth example so this first step here is breaking down your focus area and taking in those perspectives that we just discussed in the earlier slides I like to think about this in a three-part process first let's think about a magic eight ball you know that toy that I think a lot of us had as a kid where you could go and ask it any question made by question today is is the green line running and most likely and it's probably gonna say definitely not but I probably could look out the window and tell you that for myself but you get the point so let's think about your analysis area if you could get any answer right away no like no limitations what would it be you have that higher arching goal now we can look into the data that is available when you look at this focus area in that magic eight ball question what are the data sources that you have available to you that could have anything to do potentially with the question or more generally that focus area now that we have the data we have our goals what questions could be actually answered with the data that you have to bring you closer to that magic eight ball question you can start taking and breaking it down into sub components that's actually realistic with the data that you have one thing to note here is that it might seem tempting to be like okay I have four different sub parts of this question let's bring it all together and it's going to give me the exact answer that I want from that big overarching that can start to build on top of a lot of assumptions if you try to bring it together like that so you really have to be careful with it but those sub questions is still going to bring you closer to what you wanted to learn in the first place so now we're going to go into that second step of converting those sub questions into a metric and this is a process that you would repeat every single time for those sub parts determined in step one so taking one of those sub questions we want to first select those specific data points needed maybe originally you said okay github data do you want to look at issues are you trying to look at contributors what is really honing into the specific data points that you're going to be using next you're going to want to look into what visualization is actually going to give you the information you want here is it a heat map is it actually just one metric point do you need to do some form of like linear regression whatever that might be you have to kind of spend some time being like okay I have my data I have my question what is the tool that's going to get me there and then the next step with this is starting to hypothesize and look into what the impacts of this information might be what are some insights maybe is like what is your goal here and that can feed into the next portion of this process so now that you have your initial work in progress metric this is in my opinion where the real magic happens and this is where you have the collaboratory person portion and this is when you want to bring in your biggest skeptic the person who is going to try to break down and tell you every single reason why you're wrong in a constructive manner but still trying to tell ask all of the questions for you can start to iterate on the first thing that you developed many times the best ideas from my work that I've done here has come from bringing forth a singular visualization or an idea that I have and that person is taking it in a direction I had never thought of before multiple different conferences I've been to I have left being really excited because of the advancements that happen just by having new eyes on the work that I have been doing so this third step now is taking this analysis to go into action so we have three different steps on things that can happen here with every cycle of this sometimes you'll definitely be doing the the alignment with prior knowledge but these second two steps you kind of have to see what you want to do with this so first you want to see if your metric follows what is currently known about your community so in the scenario that it does align is there some assumptions that you've made to cater the results you want to take a little bit of a step there to make sure that you didn't honestly put into the assumptions that you wanted to see and it outputted with the data and if it does not investigate a little bit further is there a data or calculation issue if not is this just something you've discovered to be a privately misunderstood part of your community and something that you can build off from there and then once you have a better understanding of how this aligns with your prior knowledge community initiatives can now be implemented and these these community initiatives should be informed by data analysis and we want to gear them towards being measurable where we can see what those impacts come and so once we get to the point of observing these initiatives we want to see okay when we're looking at these metrics or the things that we thought we're going to be looking at are we measuring the right thing if you don't see the results or you're not seeing a real change in the activity is there an impact happening and maybe you're looking in a different space maybe you expected to see it in the commits and it's actually around the activity in your IRC channel there's more people asking questions more people getting involved or is the initiative does the initiative strategy need to change you thought that something would have an impact in a way and it didn't you kind of need to tweak what you thought to be doing before so I'm gonna get a quick drink and we can go into an example of all of the words that I had just said so let's look at this concrete example let's say I want to analyze new contributors for the first time so I go with my magic April and I figure out what I want to ask what do I wish I knew if I could just get a straight answer and with this one I'm I want to know if people are having an experience that converts them to become a consistent contributor that's my overarching goal and the question I'd like to have answered next what data do I have that can go towards this analysis area and the magic eight ball question and this is going to for me in this case it's going to be an individual contributor activity and so now that we have our individual contributor activity on repos and getting those timestamps and now that you have the data and the magic eight ball question let's break this down into sub parts and see them all the way to the end if you're thinking about this in the prior steps that we went through we're going to take three sub questions from that first step and bring them all the way to analysis and action so this first sub question that would come up with is how are people coming into our community looking at new contributors let's see what the first action that they do so from a data perspective we're going to look at contributions whether that be issues prs commit whatever that may be by the contributions over time and we're only looking at the very first one that they're doing and then we're going to be looking at the visualization the visualization I want to see here is the first time contributions broke broken down by quarter and so once I have this visualization and I will show an example of this at the end now I'm going to talk to some people and try to see what we can learn further how can we take this farther and let's see from this we can see that maybe we want to break down whether that first contribution would like be a signal of the person staying and being an active member of the community is a pr a better sign or is an issue and as a community are we supporting our new contributors to be able to do a pr are we labeling first new issues well is that something that you want to add to your new contributor documentation maybe getting that first pr merge is very overwhelming and having a pr buddy of some sort might help might help them figure out through those steps as well as having a connection point within a community and so this is kind of that first sub question all the way into an actual action another example of this is what is the conversion rate from a first time contributor to an active or repeat contributor and this is going to be looking at the same data as we were before and I want to look at a metric of the percent or number that was converted to an active contributor versus not and some questions and actions that could come from this is looking at is this number or percent going down is there something that we were doing differently in the past that helps support people to become active contributors that we need to start doing again or what is that change coming from and then we can also look at is there some trends of the ones that are staying around are they getting more direct communications from other active members is there issues getting comments or their pr is getting more activity than they were before so now with all of this we're going to go into our analysis lessons learned and the things that I really want you to leave today with so metrics they do have limitations but there's a lot of opportunities that can come with it numbers and data analysis are not facts you can make them say anything you want and your internal skeptic needs to be alive and well the iterative process of really breaking down what you've made and trying to build off of it is what's going to bring you value you don't want your analysis to just be a yes man of the things that you've already known you want to take time to take a step back and evaluate the assumptions that you have made also if a metric just points you to a new direction to investigate that's a huge win you can't think and look into everything right off the bat with no with no kind of triggers when you get sent down a rabbit hole that can be a really great thing for your community that one email that you didn't get to is not that big of a deal and that can be a conversation starter that brings you to a new place and sometimes exactly what you want to measure is not there but you might be able to get valuable pieces of the puzzle with that you can't just assume that you have all of the pieces put together if you try to force an answer a solution like that really pesky puzzle where you think this should fit and it just doesn't you can lead yourself down a dangerous path by assumptions leaving room for the goal or analysis to change can lead you to a better place than where you were before and next the analysis this is the start each scenario that we went through today is starting at different points of the analysis process if we look at that first example that's the second go around of your data analysis where you're starting to build the second example of looking at your community initiatives your goals and scope have been you have been established now you're trying to observe your initiatives and loop back around and seeing what you can change and that third one is a new idea just building straight from scratch you should never stop asking yourself if more context is needed or if you're truly answering the questions that you want and there's so much going on whether that be data other community responsibilities if we can cut down the amount of time it takes to get this information and make it to an easy check at a regular cadence that's a huge win think about how much we can use this information to guide ourselves in way you think about your community even if it's not in a direct application that you have thought about you want to make this process sustainable and if it takes only five or ten minutes a week to check into something that's much more realistic than the 10 or 15 hours that we were starting with before so now I'm going to leave you with my closing thoughts and that's data is a tool but it's not the answer it can bring together analysis and information that would not have been accessible otherwise and the methodology is vital to the success and the value of your analysis you need to get comfortable with the process of breaking down what you want to know and to manageable chunks and then building off of that and taking a step back open source data analysis is a great example of the care that needs to be taken with all data science work you must take into account the nuance of the topic area and as we all know open source community is about as nuanced as it is the process of working through what to ask an answer is often overlooked in every application of data science and it can be the hardest part to come up with that knowing what to ask and building off that to make something insightful and innovative but the time spent there is much more important than whatever tool you choose to use so if you're a community member with no data science experience just looking for a place to start I hope this can show you how important and valuable you can be to this process you bring the insights and the understanding of the data and the needs of your community and I hope you can see the places where you can fit into this process without touching any data or line of code and if you're a data scientist or somebody implementing the metrics or visualizations you have to listen to the voices around you and other people in your community even if you're also actively involved in that community and that is what I got for y'all I will show some of the visualizations that I was referencing to earlier and leave room for questions thank you so much I think we have time for a couple of questions they're highly specific that's so fun you can come up and talk to me afterwards it's also like 4 p.m. on a Friday I understand people are mentally dead so I know I am but yeah if anybody would like the links to like seeing this analysis in action and looking at your own communities this is also available through Project San Diego and I can provide the link as well oh yeah thanks hey great talk thank you um I don't know how much you're gonna be able to answer this right now but if I want to get into it do you have like RSS feeds I should follow to like read up maybe get it into my life more because this is really exciting and I never get to it yes um I would say that the first steps that I would follow would be like kind of following the prescribed route of like determining what you want to ask and what data is necessary to do that and then from there it's actually a project that we're building on now of how to get from those questions to actually seeing that analysis without having as much of a barrier of knowing all the data science work on the background so my initial would be get to that point of the question knowing what data you need and reach out to me or hopefully in the next couple of months this will be a little bit more streamlined and I can reach out and give more information about the project that we have going right last one hi yes so you mentioned turning your the metrics that you collect into action actionable insights how do you avoid the turning those actions into yes man questions once you've gathered metrics on what it is you want to do so I would say that you wanted to start like where your baseline is so if you're looking at you're trying to change like how many contributors are in your community you want to look at the baseline of what it looks like before that initiative has started and then if you start to look at the data and you start to see a trend upwards you could you want to also take into account okay we have established this initiative is there anything else going on that might be impacting this as well and kind of take that into account but it's really important to have that baseline and knowing where your community was at before you started implementing your initiative to make sure there's nothing else weird going on thank you yeah for sure thank you so much next up is the lightning talk and don't forget to pick up your party ticket from the registration desk thank you pretty much of the example i was talking about is like okay we really need to break down where okay what is the contribution I think fedora is a little bit the good example that is a lot more complicated because it's like you don't want to just look at certain repositories it's also like for the repositories which ones matter yeah and and so and then whenever it comes to stuff involving like design or website like where can we see that activity and so I think we have a really great opportunity here but that's the first step where we can actually get a list of those then we can start going one by one being like okay this is how we can look at this this is how we can look at this and this is how we can go there yeah exactly and so then it's like okay which are irc channels and gain those links because I think that we can get like a chart like a document going of being like okay here's the repose that matters here is with links the irc chat that matters here the website like kind of like not only like what but where and having those links then we can really start moving okay and there's probably some more work on one exact question yeah because I think that if we can because I would say that's good to be working on in the background but if I would really want us to start because if we can figure out where to go measure it going from okay we want to look at doubling contributors versus like looking at something else it's so much but it's so easy I mean it could be anybody's first message but maybe even if you could specifically like identify messages and how do I get involved how do I contribute and find those kinds of messages I'm getting good is it going to be a turf war it is it's standoff it's lightning talk man I need Tommy it's um tumbleweed themed all right I'm just taking the earpiece off all right y'all here for lightning talks do we have more people for lightning talks I saw flying a room all right so the way we're going to do this is we are going to go in order of popularity of vote and we'll see how many we can get to in the amount of time we have then we will immediately be following with egg karaoke so right after this you want to stay it'll be fun we're kind of thinking like five or six minutes like five to seven minutes so maybe you know two for five minutes and then maybe you can take a question does that work all right let me make sure the stick might work test test all right uh I think uh mr. Dawson I think you are first I'm holding them up right now yeah oh yes it's both fire uh sorry the the carefully manufactured link process that I did did not actually work um one sec no no it's 36 though right 36 is current I can never remember there we go all right did anybody else send me links to wherever their talks are for the lightning talks so I was going to try to run them all off of here because that will be way less painful than switching computers um so you can either do that or you can just type it in when you get here whichever you prefer all right and I don't know probably you want odp or pdf it doesn't matter to me all right let's do the odp it's probably easier exactly we're going to do word star I keep the computer's on the wrong side sorry computer was on the wrong side which kept confusing there's also a fly at the start of your notes does is there like a present mode in firefox though or do I have to do 11 well yeah but will it do it as pages no print it I'll hand them out exactly that works all right uh I'm going to stick with stick mic you will use the podium mic if you run too long uh this is Troy Dawson with his open source shirts talk how do you open source the shirt uh I've always loved Hawaiian shirts here I am at 20 30 40 and 50 all wearing Hawaii shirts I love Hawaiian shirts I love penguins and Linux uh that penguin is on the that's a vector based has the creative funds attributions license which means I can use it as long as I give attribution and so what I did was gave it a Hawaiian shirt and I kept the same license of Hawaiian created commons attributions and then I made more penguins with Hawaiian shirts and then of course you've got to give color variations and that's there there we go you have to have color variations of all the penguins wearing Hawaiian shirts now I have a brother that actually was making Hawaiian shirts and his graphics artist was nice enough to create this graphics artist and so he gave me he made me a Hawaiian shirt not only am I here with this I'm showing you my dirty laundry this is not that shirt because that one is uh like thing is eight years old and it's faded and this is like I remade it so there's a Hawaiian shirt and that got me thinking you know he made it for me I was like wait a sec what if whoop I started making my own so I well I took that pattern of of his from his graphics designers I made my own because this looks familiar uh I made others this one didn't turn out as well it it looks wonderful on a woman's dress not as a Hawaiian shirt for at least for a guy anyway and I did them all as creative commons attributions for those of you who are wondering uh I went over to my my brother's friend and I paid him he he actually does that things I paid him and then I open sourced that this pattern so it's it's not encumbered by somebody that that wasn't doing it okay now open sourcing is a shirt this in the fabric industry is called a design okay just to keep make things clear the graphics is the design the shirt pattern that says hey your sleeve is this long your pocket goes here and all that is called a shirt pattern now we have more dirty laundry this is this is uh you guys might have recognized this from yesterday this is a pattern I got from walmart and it's my calls or whatever and it's really nice and it fits me really good but I can't open source that so this pattern which is like I said is not the one my brother did it I went to a professional pattern making company they uh made a pattern I am now able to open source this pattern so step two of open sourcing here just so you know it is expensive oh what's my next slide sure patterns oh even better if you don't I sort of wish I'd done this if you don't know this free sewing dot org is a wonderful place for sewing patterns for shirts one minute shirts anyway go there what do I got next um okay here's a red shirt this is not the red shirt I'm talking about he has a red shirt this one uh some of you who are here with the the centos thing saw me wearing this one I am making shirts I will be selling them in a couple months I will be making the shirts and uh that's how you open socials open sourcing shirts all right uh next up we have uh next up we have uh something to do with git with sally oh malin it doesn't matter you're all fancy which one did you send it to this mic wasn't working test test it's working it's just close what's that sure yeah I just need a mirror oh this is going well we gotta hurry right I'll be I'll be super fast mic working like can you guys hear it through the mic it went black this is one of the reasons we didn't want to switch I know what the heck why isn't it is it okay well do you have us be like like I don't need this oh I do switch box oh on the wall like this guy that's a black one that I'm like oh wait it's different now damn it that was a box here that I know the word box is it coming up oh there we go there we go now I need to mirror it can you come here real quick because you'll be able to do this at wire life re-showing that was a very good question uh I just want to do you should I know close it close it yeah I think it's confused let's do oh you don't have to click no wonder I can't get anything nope that's not helpful display I just want somebody if you can just set it to anything and I can just move it over too it's fine I don't know why it's doing this I don't know why it's doing it either uh you should have a prompt up here that says mirror but it doesn't I know built-in display so that one's two and so it's open it's pass three put two over there two over here yeah keep going now hit apply apply now no matter what happens you just hit enter well now we have to move our now we got to like move the okay okay okay gosh no that's good that's good okay so what do you want to put over there um this my terminal and yeah so just drag it over it's a lot does this count as far for a minute no well it can it can where is my pointer okay bring that over come on other way other way other way now where did it go uh wait oh my hands are all sweaty I'm nervous uh is that chrome selected uh yeah you can do chrome no I'm trying to get it selected then alt have seven and then and this is the reason why no one else can give their lightning yes all right anyone's actually someone else can go next I feel bad um yeah yeah yeah yeah well we can do that for the next one too uh hold on alt f if I hit it okay just go on to the next one my talk was on six store and I was gonna point out that six store offers less encrypt for signing software and I was going to redeem myself from my uh talk yesterday because I didn't get to show you the output of my get sign um verify from fetch it that's all I was gonna do so you'll have to trust me that it works it's not giving the option it's doing something really weird oh I just switched the function key all right here we go really all right I'll drive your slides okay you don't have a maximizer though uh okay get verified commit if you go to the if you go to the repo um six store slash get sign it gives you a very easy way to install get sign on your system and um when you do you can sign every commit that you create and now it's a little bit of a pain in the butt because it's keyless signing which means every time you sign a commit up up pops an OIDC UI gooey thing and you have to click if you want to sign with Microsoft GitHub or whatever it's OIDC and then your your your your commits are signed keyless it gets uploaded to a public recore log you can serve your own recore log that you don't have to use the public one signed by the sig store um root uh ca but with managed by full ceo and then um if you run fetch it and you opt into the verify get sign you'll get this nice information about your commits and you can see that uh I just want to scroll up yeah that's not what's being yes fine oh really no that's a projection of the screen of the that's fine you can see it's a validated get signature shows you the the subject and the issuer of the key and the validated recore entry means that that the content of that commit was verified from uh signed by my email that's it yay next yeah so who's up next uh let's see uh yeah thanks Eric you're an analysis for your data all right did you send me something are we doing okay do you want to drive projection we can just turn that one off if you want to use this one uh you want to type the url sure it's going to take a zoom i'm probably going to use google what i know this over here so it's a little less confusing and uh what's the url yeah and then just throw the window off the left thanks okay yeah you know now so this is the window yeah and slide it to the left oh my god freaking out okay okay so I know right um this all started because like you know anybody's like done data analysis of any kind it's like you know most numbers in real life have units right it's like you know it's 100 kilograms or 50 miles and when we do like matrix math and data science like the first thing we do is like throw that right out the window right we just start treating them as you know numbers right and they got me to thinking it's like well is that necessary um and I don't really know um but I started I started trying to I started trying to like say well what what if I like you know what if you like kept the units and try to do the math what would happen and so um I don't know if I can scroll this uh yeah just like down here so and then you can just two two fingers scroll oh okay yeah okay so like you know here I did a bunch of proofs I like actually showed you can like take the inverse of a matrix it's got units um and I'm going to skip all that because we have now three minutes or something but like you know here's a matrix x and I'm going to keep the units um and my y I'm going to keep the units and I'm going to say well what if we did so the oldest data science in the book right like linear regression um so let me go here I'm going to see if I can find the formula that I've been the formula you know so like here's your linear regression formula right and it's using the more penrose inverse applied to the y so it's like solving your equation but if you have like more data than columns which mostly we do that's how you one of the ways you do it right the other way is like singular related decomposition um so I did some proofs um and I invented this concept of unit signature so like I did the reverse it's like what if we just throw the numbers out and keep the units um because uh think about this units are like data types and so if you can do this you could actually write software and I have um that keeps the units along with your stuff so it becomes like a data type thing if you're checking type checking um and so I started following the formula and just applying my new unit signature laws that I proved in a previous post and if I do that I'll all the way through this is my unit signature operator I invented an operator um the uh and if you do it you end up with this and so like my my weights like the linear weights that you solve for have these units um and that's actually cool because it all works um if you it's like if you actually do this you know if you multiply it out you find out that like oh look at these you know centimeters and years cancel and what's left is kilograms which is exactly the unit of the thing you were solving for so like I followed this all the way through and did not lose faith and sure enough you can do data science with unit analysis questions all right oh like has a question right exactly um so by the way part of the reason I started doing this is because I actually have um type I have like type level unit analysis and scale where like if you screw the units up it's actually a type error it'll fail to compile and I wanted these kind of laws so that I could actually you know apply that and so you could give it you know matrices with units as types and it comes out with the correct types I care all right thank you next up open source with uh sorry the importance of d and i in open source which originally I thought was p and i and took a while for me to figure out that was actually a d on the whiteboard and not a p and I am unaware of what p and i is pets and inclusivity right uh so do you have slides you have no slides y'all have me so who here is not a computer science major great this is good so I put this up originally as the importance of d and i and I'm glad to see so many different majors because we all bring something different to the table I was an animal science major and extension education I don't palpate cows in open source but you learn how to ask questions you learn how to look at things differently and the importance of d and I within open source and in computers and anything else is everyone brings something different to the table how you ask a question how you see things and as a career in computers I started off as tech support answering phones then I moved into development software management operations and again how you look at a problem how you solve a problem how you investigate that problem is so different based on your background because someone else might go right to the log someone else might go to the screen you know taking something out of the equation to see how that changes things so these are the importance of diversity in computers and in open source and one thing I went to a imposter syndrome talk from major Hayden if anyone knows major and you would think he would never have imposter syndrome at rack space he had his picture on the wall he had his own little badge sticker I mean major was a big deal major do everything if you had a question you went to major and to hear him talk that he had doubts makes you wonder well if major has doubts it's okay to have doubts this is kind of going to Pat's talk but so you also realize that when you go to a conference you are the only female in the room half the time I'm the only female on a team the only one in the room we joked at open source summit this past year that we actually had to wait five seconds at the bathroom you know you're a grace hopper when you have to wait in line for the bathroom in fact in houston I kind of knocked on the men's door and said incoming and let a group in we would have been there for 30 minutes but those are not our normal problems our normal problem is we are the only one there so we do analysis of our communities and I hope everyone does that we send out diversity and inclusion surveys to see where we are and it's not just a gender issue how you word your questions do you feel like you're a minority in the community do you feel like you're a minority in your country because in your country you may not be a minority in your community you might be so asking questions that aren't just a or b are important to find out what's actually going on in your community because it's how those people look at the problems have the discussions that is what is important about diversity and inclusion within open source and we don't think about that you know it's like well everyone in the room has a computer science degree and we all think the same and we all you know so that was why I wanted to come up here and talk and the fact is if you notice you're the only one in the room put in a conference talk speak run for a leader position leadership position remember I started in tech support I am now on three board of directors you know if you put your heart and soul in these open source projects you can do anything you want and I think that is the lesson so don't think because you don't look like the person next to you that you can't be something and that's my talk all righty uh next up we have uh an unknown talk it's one of two I don't know which choice they're going to make okay not to know things do you have slides I do not have slides but I have a timer okay it's okay not to know things they tell me in a presentation you open with the joke here's the joke on the first day of your job you know how to do it that's the joke those of you who have not had jobs in the tech industry before this is how it works they hire you because they think eventually you can learn how to do this job they do not hire you because you know how to do it they do not pay you because you know how to do it they did not get you there because you know how to do it it's okay not to know things this is what your team members are for this is where you have an opportunity to learn and ask questions because a lot of the time the things that we're doing are just that's the process we ended up with why uh yeah it's like it's just it's just what we do like steve when he was here 35 years ago picked it well steve retired can we do it differently we have this thing called like web servers now and so it's okay not to know things ask questions if you're new in your industry ask them all the time because it's okay not to know things we need people to ask what's going on because half the time the people who have been there forever don't actually know what's going on they just know what they're doing and that's a big difference so ask the questions don't be afraid when you get a new job and you don't know how to do literally any single thing they've assigned you and you don't actually know what they've assigned you um oh wait on so day one oh where's the ticket system oh yeah someone should probably tell you that yeah yeah they really should but you don't know that and you know to ask about it well what about how we had this weird grub boot loader problem that like exploded the box yeah if you don't know how to fix it you don't know how to fix it and that's fine it's okay not to know things as long as you're willing to learn the answers and that's the other piece of it um don't be that person that always doesn't know but don't be afraid to not know it's fine it's okay when i started at my current job i did not know how to do it i had i had built zero rpms in my entire life when i became one of the developers for scientific linux that builds rpm packages for a linux distribution i did not know how to do this but i knew to ask the questions and learn the answers and that's what this industry is about that's what we're here for that's what your peers are for if you've seen someone up here giving a presentation they genuinely meaningfully actually want you to ask them your questions this q and a section at the end of everything isn't a joke that's just there to like make them feel better they want your questions and i want you to ask them your questions i'm going to beg you i'm going to plead you with you don't be that person that sits in the back and thinks well i don't know what's going on so i shouldn't keep like it'll be like i'll just sit in the back here none of us know what's going on there's too much in the tech industry to keep track of one interface in the kernel um there's what seven eight hundred packages in rel that are just binary type enumeration you can't keep track of that no one knows all of it no one knows half of it no one knows a third of it don't be intimidated by your ignorance be intimidated by the fact that you're afraid to ask the question and chase that and don't be afraid of it and with that i have 42 seconds left all right next up we have academic study and community matters and do we do we have a vkp oh okay that might be coming up all right and so next up we also have cross-disciplinary inspiration for information systems which you can give if you prepared it but they were last on the list and weren't sure they were actually going to want to present it but i don't see i don't know what the guy's name is but i don't see it uh no that's yeah it's one of Alex yeah yeah i think you're right um so is he hiding in the back over there okay he just came up he knew he was coming up and said nope i'm out uh all right well then we will transition to what is currently scheduled which is um so if you're unaware of this devconf us normally we do an easter egg talk so something that on the face of it sounds like a good idea at an interesting talk then you get there and you find out it isn't sick so this year we're doing a small variant which is uh adam has had uh some good luck with a concept called slide show karaoke we will need participants from the crowd uh and uh i didn't see if you wrote me back do you want to use your machine or mine okay um we'll exactly uh so what we would love though is um because the point of this exercise is it is you don't know what exactly you'll be presenting so if we can get some volunteers to get ready to present all right uh maybe we should let's say uh how about uh guess a number between one and ten um and uh and then i'll figure out the winner let's see i have ten fingers let's see all right so hold up a number of fingers how many fingers is that all right uh let's see we got none of you have gotten it all right let's try again i will do it another number apparently i pick odd numbers all right one more time numbers or i can't count one of the two uh it looks like jen is up all right uh then we will go in order of i don't know first name we'll figure it out all right so these are the rules and we should have titled it would you like to give a lighting top i have nothing nothing to say um this is your opportunity um anyway sorry that was funny um all right so you can use this is awesome you can use the stick mic or you can use this one let's not have them both okay let me just use that okay how do you mic okay and it's autoplay right it's autoplay 15 seconds a slide five slides and you go so this guy has a big idea in the form of a light bulb look how lit up it is it means you too can hold a light bulb in your hands and have great ideas just know you too can do this and look your brain is inside this multicolored skull you too can have a multicolored skull look at that the light full green you have in our occipital region isn't it wonderful and then you can evolve into a wooden owl and the green really is symbolized by the moss just know that anything is possible because it began with a light bulb in your hands and it looks like the character from cars has come into the slide deck he's become a planner maybe he was a car once but now he's a planner and i have no idea what this means i think the brain may have exploded inside the multicolored skull and i'm afraid that these two raccoons had rabies and are playing basketball the end well done well done live streaming i pray to god okay awesome um all right so who's next you now need to hold up uh let's do let's do five fingers maybe we'll figure it out faster i'll put it in my pocket uh all right so hold them up uh niche you're up i can see the slides much easier we can hide it for me oh that would be interesting okay okay oh yeah all right um have you ever been pensive enough that you don't into uh moose you're out in the hills and we're really going animal theme this time i'm sorry yeah i like i think that's what amy was thinking right now when i called it a moose that's a go oh the wood goods um oh so we're going animal theme this time yeah so um you know you're in the hills you're thinking what do you want to do next um and someone suggests horse racing jockeying and that's what you end up doing um you're racing off the dev cons and then you see jen present and she's hilarious so you just lose your mind there's a sheep are you on the list would you like to present uh i was actually thinking about running you the same one uh let's give it a no wait let's wait so you forget some of it all right uh let's see uh hold up some numbers uh back left corner i can't quite see who you're that i'm actually being honest about the numbers i'm not picking favorites the paintbrush creativity spark your mind all this and more welcome to the b u school of fine arts a little known degree program here at the b u school of fine arts is our parkour school our parkour program is one of the strongest in the nation the art of motion is our most popular class here at b u we go from complete beginners all the way to flips and in the really advanced courses of course you actually blend different disciplines you learn how to paint while flipping the way that works is actually more blood goes to your head so you can be even more creative and this will really give you wings and let you take off and let your creativity really soar so here at the b u school of fine arts we only use the finest duck feathers and uh we actually pull them right from the fence you might see them on the way back and uh we also have a culinary program which not a lot of people know about we developed this one in collaboration with uh the disney company of course and uh our chefs are in resorts all around the world we call them pancake artists that's just a journey like he is anyway who's next hold up some numbers you gotta hold up different numbers do we just have the two people here is there anybody else holding up numbers come on you did it every other time what what do you got all right uh who's closest yeah let's let's Troy's closest how many oh sorry it started Troy so he's up Troy's up once upon a time I was walking through the forest just acting oh just acting like the chicken I was because I was scared of everything I saw this giant fly flying around I thought it was a bee but in reality it was a storm trooper so I had the duck behind the trees they were shooting around shooting around and I didn't know what to do so I hid behind my desk pretended I was another storm trooper oh no I didn't but I said I set off my my fireworks so that so that my uh rebel friends could come find me and my fireworks went up and they went bang oh wow and they came zooming in at light speed and they rescued me and then we had we had a party and my my best pug friend Frodo came and licked me all over and that was how I got out of indoor of course it was thank you all right a test hello all right who's up next let's see are you gonna go no all right anybody else well then I guess you win you're up exactly actually also had the right number all right space when I'm ready the fact is penguins are the single most important creature on the face of the earth they control the universe through the power of their mind for example this emperor penguin is currently making me give this presentation and so we have to talk to ourselves about woodworking and the epic advances that penguins are making within this important field of climate engineering woodworking penguins are at the forefront of modern practices for architecture and design and so we have to work within them as we talk about things like how do we send penguin woodworkers into space these questions are in national importance and perhaps the greatest ideas of mankind can ever have is truly to look at our penguin woodworkers and say can they fly helicopters the answer this of course being that the helicopters are not terribly handicap accessible and we're working into this project it's important that we increase our disability access because these penguin woodworkers are controlling my mind and making me think about gummy bears all I want in life are gummy bears I don't know why I don't have a choice in the matter but what I do know is that if we don't put the penguin woodworkers into space I won't ever get off the podium very well done I am curious is uh is the penguin woodworking another division that at boston university is there a collaboration there with the department of fine arts all right all right uh so yeah yeah right exactly all right so do we have any other takers we have several more slides we don't want to waste them that's the point of the exercise I guess it will be sally sometimes known as Molly that's my sister's name Molly I'm really curious you know what's going to come out of my mouth because like I never know so that now this is making me feel calm I'm going inside I'm not worried about my children minimal with no next okay why am I getting the strangest ass slides in the mouth and where I really thought the opening was the best it was a very good mimic it was really that was really really well done who else we have any other takers there will be no reference to operating systems I can talk about other things sure sure you there's one thing you should learn from this conference it is that dogs are the master race of the world and they are going to take over everything but they're disguising it with their cuteness and they their their enemies are the little tiger parrots who are flitting around and they use a field of deception so that you think that they're just pigeons when they are in Boston but like the fearsome braying horse that comes through the city in the guise of the tea all these animals are around you at all times in disguise and ready to attack like a weasel or stote or ferret or something just leaping leaping probably to their own doom because they don't really judge distance very well I think it's probably going to happen here and I can't even see this picture very well I have to look here we've got a bear in a compromising position here and that's probably how things are going to end for us all with the animals attacking the world and there were no operating systems all right do we have any other takers we have several more slides for your enjoyment come on we've got to have somebody all right good afternoon everyone thank you for joining me here today it's a very important discussion we'll uh drive in too so when it comes to the serengeti the main thing is to leap leap as big and as high as you can because you need to get out of the way of the sand as it's coming after you in this photo you'll see the sand is chasing the quadcopter thing right and and the most important part of doing this is to make sure that your actions your maneuvers are following the careful martial arts in this case the swan maneuver is a good way to keep yourself out of the way of that sand whoa like that right just keep that in mind as you're going through the serengeti and in this particular case you see we we were able to capture in stop motion a 100 000 racers all doing the whoo motion at the same time and and the the the way that they moved caused the light to reflect to make them all appear the same color now there is a caution this is a cautionary tale whatever you do don't fall in love with that which is toxic to you okay whatever it is don't forget it will not turn its head away if you really it will learn it will just you know it'll buy two need you up so that's all um and the last thing is as your the key thing is as you're as you're contemplating your changes in life and whether you're going to become you know one of these serengeti racers and actually do these maneuvers just remember eat carefully chew your food before you swallow thank you thank you thank you very well done very well done uh we have one more set two two more sets so do we have any takers we have two more set don't yeah we could go back to the animal one that's no that is impossible never mind all right there will probably if you've noticed there's actually been a lot of animals in general so this one of course all of the different animal this of course will not have any it was a ferret how do you know the difference the markings on the face of course i have no idea so even though we're not talking about animals we apparently have dolphin bananas who are eating grapes so the dolphin banana can also be used as a gun so you just turn it around so that the face is in your hand and then you can shoot everybody and you have to make ugly faces while you're doing ah but after you have used all your banana bullets you can also get your enemies by dropping the peel on the ground and then as they're walking they will squish and slide to their doom but certain types of bananas can be turned into tents and give you safety and warmth in the dark of night while you're on the mountain under the stars are there any more but if you alienate your dolphin banana he turns into the devil notice his horns notice his fangs the dolphin banana can be your friend or your enemy thank you for the educational moment on the dolphin banana uh i've i've actually been curious for a long time about the the various you know kind of life expectancy etc you know life path of the dolphin banana i know adam was too so we really appreciate the education all right who's who's up next me all right hopefully i get a good set we'll see we'll see all right should i have uh i liked your start of the once upon a time uh so there was a student of basically like how does art come together right so you you want to bring the pieces together and they're trying to understand at the b u school fine arts of course um how to merge these animals and bring together an amorphic picture and then sometimes when you make a mistake uh you end up with things like this so this is where we cross those pictures together and it just it it really is not bringing out the mood that you really intended so you really when you're at the view of school fine arts you got to be careful that however as you can see when you get truly good from the parkour school you actually are able to uh you know like run with the hyenas right i mean this is really important and you know and then of course ultimately when you can kind of bring it all together you can actually uh you know be a lovely calm goat maybe um llama uh and so i'm also not good at animal identification um but there's always a danger right you can you can get too close to the sun this is this is the problem you don't want to have too large an impact on the overall world because you can just just destroy it and i mean what could happen the penguins will take over all right thanks everybody uh one of these days we will also try to do the slideshow karaoke that adam and i have been debating about for several years which is where you see the set of slides first and then you have to get the talk but you have multiple people give the same like from the same set of slides so we will not be doing that today because it is two minutes of five and uh we don't want you here anymore oh sorry we think you should be able to have a chance to go and have dinner and all those kinds of things before the party at seven p.m uh do not miss it uh it will be a lot of fun the theme is space exploration uh so who knows what that could mean uh it could be the penguins yes oh yeah so the party is at seven p.m uh it is at the photonics building uh which is kind of like down the street a little bit um but it's across the street it's very very obvious because it's a tall silvery building um and it does have a beam of light on the side that's right and it is on the ninth floor uh in a big huge open blasting area so it's like got great views of the city especially at night so we strongly encourage you to come and don't forget you will need a ticket which you can collect from registration or if we have leftover tickets we will have them at the door so if you don't have a ticket we probably have some extras uh follow us on twitter and we will definitely announce if we run out okay yes there's food and drinks yes exactly and prizes maybe prizes to definitely games and a dj unfortunately we couldn't make Keetar bear happen this year sadly it was all on me i apologize um but hopefully next time so but we do have a dj uh who is not only a professional dj but also a professional redhatter so uh with i have been assured space exploration themed music so we'll see what it goes all right thanks everybody