 So I suppose the third problem as we start is that I have 35 minutes with you and we want to leave at least five minutes for questions. So it gives me 30 minutes to present what it ended up as 67 slides because because today I am so I'm here from the Linux Foundation research team and I the team has been publishing a lot of research in the last two years 40 reports actually and there's so many things I want to share with you and I just could not decide in the end I decided to summarize four reports so that's 10% of our output and that itself is a lot so I'll at some points just skip over some things I think are maybe not that interesting or maybe yeah are not that important so let me start by saying it's nice to meet you my name is Kailin Osborn and today I'll be presenting on these four research reports so I'll introduce them in a second so the agenda quickly introduce the team and myself then very very quickly introduce some highlights from four reports so the first one is the one I wrote was anyone here at Anna's presentation on Ospose okay so Anna was a co-author and shame she cannot be here she's at the Ospo happy hour having fun but anyway I'll present some this came out last week so this is fresh off the press and then here are three other studies which I did not write firstly it's a study by professor at UC Berkeley on measuring the economic value of open-source software for companies next one is a study I came out two months ago about maintainers and the fourth study is about global collaboration and challenges that are caused by fragmentation and then we'll leave five minutes for questions and answers okay let's get started so what is the Linux Foundation research team the kind of the mission is to investigate the impact of open-source software project technologies and standards the team was founded two years ago if this QR code will take you to the website for the foundation's research the research team and you'll see all our reports there before the team was established there were two big research studies conducted one was just a history of development of the Linux kernel and in 2020 was a contributor survey and then after this there was a lot of people commented that they found these reports interesting and useful and they wanted to see more of that so then the foundation set up a research team dedicated to do doing this kind of research the kind of three broad research methodologies are used our interviews surveys and also kind of big data analytics you know get up analytics software as SCA and so on and as I said 40 reports have been published this team is led by Hilary Carter who's a senior vice president Linux Foundation highly prolific I myself have only written two of these so I don't want to over claim what I have done here this is really much Hilary and and the rest of the team in going work and here just some pictures of some of the other reports and so today I'll talk about this one this one and this one so five minutes in let's move on okay who am I well I joined last summer I've been with the team for about 16 months I'm also a PhD candidate at the University of Oxford where I research collaborative commercial participation open-source software projects and the reason I'm in in China at the moment is not just for KubeCon but I'm visiting the open-source software data analytics lab run by professor Joe Minghui at beta and yeah this is if you're on Twitter that's my username okay so let's start with this this is a port that I wrote it's on Europe so to show me your hands if you're interested in what's happening in Europe okay three hands or five hands okay I won't spend too much time on this most of the slides are about this so we're gonna if we skip over it it will be able to cover the whole presentation so this is the report and the QR code if you want to read it you'll also find it on the website which the previous QR code took you to so no worries background there's a lot of political interest in Europe about in digital sovereignty and this goes back a few years the French government in particular was the kind of driving force for this the covert pandemic made this much more gave much more fuel to this debate because governments realize how dependent they were on a few platforms schools across Europe were running on Microsoft teams and just you know the services and applications provide very very few companies who are not European companies mostly just US American companies so European governments have been since then talking more and more about the need to invest in the digital commons which right which includes open source software okay there's interest in a lot of talk by politicians but little action there are lots of challenges many of you will be familiar with these you know vendor lock-in governments are in 20 30-year contracts with with service providers there's limited awareness amongst bureaucrats about the benefits of open-source software there's limited training amongst bureaucrats in development of open-source software there's also a contribution gap this is a report that we published a year before which shows I mean this was a survey about a thousand practitioners in Europe and the public sector as you see alongside education and finance and insurance is one of the sectors that encourages contribution the least what does that mean well developers during working hours are not are not very much encouraged to contribute to open source so it's difficult to realize the political interests that politicians talk about when this is the reality so our objectives here were to understand the state of open source in the public sector across Europe identify barriers and enablers for shifting to open source or increasing open source activity and also identify priorities so I'll just quickly over this we spoke to 30 people from 14 countries in Europe and yeah key findings so with all this political talk about open source and digital commons there's an increasing recognition of the value you know cost savings control localization and reducing vendor lock-in but a lot of the people we spoke to who well how many people know what an Ospo is so of course you will know because you came to Ospo talk but how put your hand up if you know what Ospo is open source program offices these are dedicated offices or teams within organizations not just industry but increasingly governments which are dead yes focus on open-source software and I think when we did this research this summer either 12 12 countries in Europe out of 27 had a dedicated Ospo they're not always called Ospo they have different names and like but you know they come in different shapes and sizes and we spoke to we representatives of these teams and they said what they want from their politicians and what they want from their senior leadership is systematic endorsements and resources so money going towards the things not just comments so as I said digital sovereignty is very important events like log for shell also drew attention to the importance of cybersecurity Germany did a very the German government did very I'm German so I'm not seeing this because I'm German but I think German government did very something very what awesome in October last year they set up a fund called the sovereign tech fund has anyone heard of it oh so in its first year of running it has a budget of 11 million euros and it's dedicated to giving basically funding open-source software projects that it sees as critical dependencies for the German government and how it works is open-source software projects or foundations can apply they have to for basically a grant and they get it so for example open JS which is part of the Linux foundation I believe we see 500,000 euros for one years of work to hire maintainers to work on critic on cybersecurity within the product ecosystem so that's really cool so that but that's an exception that's the only government that has set up something like that spoke to Fiona from the sovereign tech fund she said we need more actors because funding is the maintenance is very important but we also need more coordination because it's great to see other organizations getting involved in open-source software funding this includes the private sector this includes philanthropies and now governments but everyone is pushing in different directions there's very little coordination between them so she said from her perspective it'd be great if there could be more coordination okay I'm moving very fast let's all the information so I'm sorry for that I thought this was a really interesting point from Boris he's based in Amsterdam the Netherlands he said Europe is in a very lucky position most countries in Europe have policies and laws about open-source software use that's not the problem the problem is actually implementing these laws one of his compatriots from the Interior Ministry of the Dutch government told me for many years they've had this law called open-source unless what does this more mean well if the government develops IT systems or software systems they have to open-source it unless they have justifiable reasons not to and hit in his opinion this is the least followed policy in the whole country and why is that well he said well people just don't know how to implement it and often that's the justifiable reason I mean they don't know like you know all the legal or even technical procedures are to open-source software from the government so many people emphasize and many were from Ospo so it's a biased sample Ospo's are a key mechanism for facilitating contributions to the open-source software ecosystem including you know actually implementing the open-source and less policy or funding open-source software projects or just turning up to events like this this is a view from Leonardo who used to be head of open-source for the Italian government he says you know most organizations just consume and use open-source software and very few give back and he sees Ospo as the key way of contributing back one of the things they do and so he talked about his experience of the Italian government is create catalogs which basically help with the discoverability problem so basically if you're a developer working in one of the ministries of your government knowing okay what other projects have other developers within the government apparatus in a different ministry or maybe at a regional level or a local level what have they been building and can I reuse any of that so that's very useful Francis also created one and Francis done something very interesting it's called the sill they've created basically a social network around their catalog so basically let's say you focus on the Linux kernel right and you are a pytorch expert you are not just using the catalog but you can share that you are an expert so you become a referent as they call it and say you come along and you want to learn about both Linux kernel and pytorch and then it will point you in the direction of these two developers within the government so you'll never have met them but you know okay they have expertise and the technology that you want to work on and they are work within the same organization as you and so he said through that so the person I spoke to said through that a lot of developers were starting to get to know each other and share resources and build capacity so that's been very useful guidance is of course very useful as well this is an interesting point from the CTO of Estonia so he said building something this is about the discoverability problem it's great if you build something open-source it but that doesn't necessarily guarantee that people come and use it this is probably quite familiar to all of you and so he said he emphasized in these catalogs and kind of guidance documents are useful for getting people to come to your project and contribute to it okay so I'm sorry that was a lot of information about a report that maybe is not that interesting to you because it's about what's happening in Europe so now I'm gonna talk about more generalizable findings and how are we doing for time with 17 more minutes okay so this study is on measuring the economic value of open source here's the QR code and this was run by professor Henry Chesbro at Berkeley I'll give you a second to take the QR code so it's based on US companies but I think it I assume is generalizable across the US across the world and I'll give you my reasons why so the survey was sent to CIOs and IT managers at Fortune 500 companies as well as companies that could be reached through Linux Foundation as well as members of the Berkeley Innovation Forum who are also Fortune 500 companies and this resulted in 439 reusable responses so because of the mixture of the distribution of the of the survey we cannot say that all these 439 responses correspond to 439 Fortune 500 companies but roughly and here you see the diversity of the sizes of the companies so what did he find okay here's a lot of information but we see a lot of the kind of key benefits and key costs or risks of open source are pretty much standard having kind of repeated over and confirmed over the last 20 years so sorry information overload but here are the kind of main benefits draw your attention to these three so the three main benefits according to Fortune 500 companies in the US are faster development speeds cost savings and open standards and interoperability I assume that doesn't surprise anybody when it comes to the to the costs but if you want to take a photo of that you can take a photo quickly otherwise you'll see in the report when it comes to the costs the three main ones are hidden support costs cost you to security gaps and costs for training did anyone expect any other kind of major costs to come up that you don't see here yeah I think it's pretty straightforward right but I think that the value of summarizing the costs and the benefits in this way is that you can put them in front of decision-makers whether it's in governments as I was just talking about or in companies that are not active in open-source software and it kind of gives you a very useful summary of what the main benefits and costs are so hopefully stakeholders who are not already believers and supporters of open-source will find this useful okay the third report this is about maintainers so here's the QR code if anyone wants to take it quickly the methodology here was to interview 32 super maintainers from various projects hosted by the Linux Foundation so I'll get more I'll talk more about the methodology in a second but this is like the reason why you know we all know that maintainer burnout is real everyone recognizes this diagram right is this the first time for anyone seeing this okay so everyone knows that right so maintainer burnout is real so the aim of this study was to examine the practices of super maintainers and I'll define what super maintainers means in second and identify key sustainability factors and then provide insights into how we can support maintainers so how did we select which super maintainers we spoke to well so this is a I'm just showing another idea I wanted to include this in the presentation but I already had four so this is a previous study by researchers at Harvard University it was for the Linux Foundation where they analyze identified the most you know the most you know the most common dependencies doing software composition analysis and this ended up with a big ranked list of software opens for software projects and then so the methodology here was to select super maintainers of these the most you know the highest ranked dependencies so that's what super maintainer means and to cover a range of different domains so front-end libraries operating systems infrastructure package managers databases and storage DevOps tools and data analysis and AI and yeah lower-level languages so we got a kind of a range of projects in the sample the key findings here's a lot for me this was the most striking one that only 35 percent of these super maintainers have a contributor pipeline you know who here has heard of the bus factor the bus factor is this is a new concept oh so the bus factor measures how many developers or contributors make up at least 50% of contributions and it goes back to the 90s when people talked about the role of Lino's Torvalds in the Linux kernel and someone said well what happens if Lino's gets hit by a bus he will not be able to contribute anymore so since then people or researchers talk about the bus factor which is quite morbid but basically is a measure of dependence on individuals and so if some of these make super maintainers decide that they cannot continue well they're in a they have a problem if they don't have a contributor pipeline for training and mentoring contributors who can then step up and become maintainers so they shared a bunch of best practices I'm just gonna highlight six categories of best practices the first one is about is anyway sorry is anyone here a maintainer do any of these resonate with you do you do any of these yourself we can focus on this box because then we'll move on you have what's your project it's a graph database so it's a very large product we have more than nice on the stars in GitHub so we have greeting mechanisms we have the good first issues a list of simple issues or bugs for new commerce to do how effective do you find that it used to be effective but we do have a serious pipeline issue we don't have many new contributors coming so used to be effective are you considering any other tools or ways of reaching out to new contributors we are waiting for our incoming GA of our new version which we rely on it to boost our open source contributions after that we may rebuild our contributor pipelines and and we will reinstates it on these measures we will hope that will help the contributions we should bring out a second version of this report with this these insights so then yeah community governance is important establishing code of conduct I think the second one is interesting design yourself out of the job this relates to the bus factor so basically ensuring that your project can continue without you so we have to put your pride to the side and think in terms of the interest of the longevity the sustainability of your project okay and yet distributing power and decision-making is very very important of course documentation wait do we have any other maintainers I don't want to just pick on you and ask you your retainer to different project or same project different okay do you want to comment on documentation or any of the I'm from kublox it is an operator actually for multi database management multiple cloud and you can check how we're both in the first floor and last month I spent a lot of my time on documentation because many users we complain how to contribute to your project or how to use a project better and I'm also doing the second second one recruiting some doc document writers from our community to contribute to their ideas or the experience or how they can successfully run our projects yeah I think that's very useful and they experience when they share the experience out it will help more like our users or potential users yeah so moving on time is running out here are three other categories of best practices so the first one is diversity you know mentoring is really important for bringing in new people and training them up a major problem that was brought up is well of course there's a certain geek culture but also a cultural dynamic here a lot of the projects are us centric and many contributors who are not American might find it difficult to to climb the ranks and become a maintainer in a project that is technically is technically global in nature but might feel like an American project and this is can include like you know the language that is used and do we have something about the language no but later on in the presentation I'll talk about this but yeah just basically making it feel like an inclusive global community rather than just you know the culture of one place in the world avoiding burnout is this something that the two of you can relate to have you burned out from wait so let me step back are you paid to do your maintenance work as part of your job okay so you're not doing on the weekend excellent so your your company or your organization supports you okay so would you agree that that's an important reason why you have not burned out yes so I think this is a problem amongst maintainers who are not doing this as part of their jobs who are just doing it on a voluntary basis after work on the weekends so on and then funding an interest of time I'll move on this is the last study so thank you you've made you've done three out of four so this one came out in February and yeah as it as the title says it's about we basically the study interviewed open-source software leaders from projects and foundations what they saw as the biggest challenges and priorities for the global open-source software ecosystem and this project is probably maybe the one that is most relevant to China or to a Chinese audience so yeah we spoke to a bunch of people I won't focus on methodology key findings so open source has changed a lot in the last 30 years as I just mentioned before you know it has its roots in Western Europe and North America but it's truly a global community and that brings with it language culture and geopolitics barriers it's not part of this presentation because it's not from the next foundation but my favorite favorite book I'll share two things with you my favorite book which was written by a Russian scholar at Stanford in the US about developers of Lua in Brazil in Rio so put your hand up if you know Lua okay some of you so some of the problems that these guys in in Rio de Janeiro had is that you know they speak Portuguese with each other but the lingua franca of the kind of the global open-source community is English and they were finding it very difficult to to contribute and at the beginning he spent five years researching these guys and at the beginning he he could speak Russian he could speak English but he could not speak Portuguese so he was speaking with them in English and then maybe say let's say you speak very good English and you don't speak a good English if I spoke if when he spoke to you privately you would speak very well and when but when you spoke in front of sorry what is your name when he spoke in front of you he would reduce the quality of his English because basically your ability to speak English reflects your social status you know what your education background is how wealthy you are and so on so he in this book he this is not the focus but this is just one chapter he talks about the the the politics of language ability for developers open source software developers who don't come from English speaking countries very interesting the second one is a report I should have these written down I'm sorry is by a woman from Bangladesh who was based at Berkeley and she had a grant from the Ford Foundation to research open source communities in Bangladesh and she looked at the communities of Rasta and Brave and no sorry Mozilla and Brave and a lot of the volunteers are high school students or university students who are you know contributing why well they want to basically build up a portfolio so that they can get a job and you know just studying computer science or electrical engineering or mathematics is not enough and their dream she says is to work for you know a big international technology company but by just doing very local community work in a city in Bangladesh anyway so those are very interesting studies about the kind of experiences of developers who are not from English speaking countries and maybe some of those challenges resonate with you as as developer Chinese speakers in English speaking environments another challenge is techno nationalism you saw this definitely during the Trump era with the trade wars and so that's a major concern here's the last set of findings so what do I focus on here so this is I think the one I will want to focus on here which is there are so many foundations you know Linux Foundation is one with lots of daughter foundations it's not the only one open Adam Foundation Eclipse Foundation and so on so forth and so an outcome of this was oh let me get I'll come back to this an outcome of this was in July this Congress took place with foundations from around the world to talk about kind of shared priorities and looking at ways how they can coordinate to make sure they're not just duplicating the same efforts I just want to share some insights from three different people we interviewed first one is from Mexico and this is I talked earlier about the problem that people from outside the USA experience when they want to contribute to open-source software projects he said in Mexico a lot of developers struggle with this in Mexico their ability to speak English is quite good but of course they're Spanish speakers so he said the language barrier can be difficult the perspective from Huawei that your political conflicts were fragmenting open-source we see this with you know the last couple of years and then also in Europe digital sovereignty is very important and European governments want to kind of cut their dependence on US tech companies so I think I've just managed to go through 67 slides in 30 minutes so thank you for suffering through this session with me I hope it was interesting and we've left some time for some questions so please ask me questions we have nine minutes yeah since you're visiting I think you're visiting Peking University right so so since you're visiting China are you comparing the open-source communities in China with that in Europe and I think to some extent the European community and the Chinese open-source community may share some common problems like a lot of the projects and their co-contributors are from the United States and we do share this so problem that we want to have more homegrown independent open-source is that are not that American so what's your opinion on this are you doing this comparative study in China so it's not the focus of my own research and I should say when I'm at beta I'm not while I'm at beta for the next six months I'm not part of Linux Foundation research it's just academic research as a PhD student my research focuses on commercial participation and I'm interested in how different types of company managed open-source software projects how collaboration takes place between companies in those projects because academics so far have only studied how companies collaborate together in community-led projects but we haven't we basically company-led projects which are very common which come in different types of different types of community structures have never been studied so that's what I'm studying however I've had three weeks there now and there's a lot of interest for my lab about doing these kind of studies and one thing that we might look at is are you familiar with scikit-learn it's a machine Python library for machine learning developed the core maintainers are mostly in France at the National French Computer Science Research Center called INRIA and for my PhD I did a study on their funding model because I think it's interesting the field of AI which is you know where most popular with the most popular open source software libraries for machine learning or AI are developed by big tech companies but they are just a few academics in Paris you know they don't earn millions and there are a few of them but they have a really strong community so I'm interested how were they able to sustain themselves and so I was speaking visited the many times and the community manager François showed me a dashboard of their contributors so he showed me a map of stargazers and about 35% of stargazers are in China and then followed by I think USA Germany UK France and so on and then when he changed to issues and pull requests and then even worse commits China goes from like you know over 30 percent to less than 3% and the USA France UK Germany go up and I asked him you know what is happening here and he said there's a fork where a lot of Chinese developers are working in Chinese and then there are a few who contribute back who can speak who are bilingual and so one of the studies where my colleague Beda and I I want to look at is the process the role process of collaboration between the fork and the actual project and then and the role of these I don't know how many I think let's say two or three bilingual developers who speak English and Chinese the role they're playing to diffuse information in two ways you know so that's something another thing is looking at I know one of my colleagues is interested in international collaboration on kitty so not just China but yeah anyway that's hopefully that answers your question any more questions we have five more minutes yeah so from your slide that some of the like most popular open source projects are maintained by a few handful of super maintainers yes so is there any study around what keeps motivates these these super maintainers in my opinion I don't think like free open source project the maintainers are doing it for free so what are those like major motivations for them to keep working on these projects thank you yeah thank you for your question here this is the so studies about maintainers my favorite study about maintainers that kind of covers this it doesn't specifically answer your question but it covers this is a book called working in open in the in public the making and maintenance of open-source software and written by Nadia eggbell and she was a researcher at github for two years where she just spoke to hundreds of maintainers that was her job and she published this book which is also I think it's open access but don't quote me on this where she basically talks about the why maintainers maintain open-source software and as you might imagine the reasons are many you know there are many reasons and the lots of factors to consider firstly is are they being paid to do this or not she has a statistic in her book which I can't remember but many are you know let's say around 50% maybe more there are many different types of projects so she gives four examples of projects she has names for them like a stadium a stadium for example is where you have a few maintainers and core developers and then lots of users basically lots of people watching it another one is like you know basically with the two attack two axes are developer contributors and users basically so they're like four quadrants and she finds in each of those quadrants so say like low use low development or low use high development high development lots of use whatever the maintainers of those different types of projects have different motivations and resources so hopefully that's a good question but basically there's no simple answer what's the name of the book again I wish I could write it down it's called working in public the making and maintenance of open-source software thank you you're welcome I think we have time for one more question okay if there are questions I just want to ask you one question you know I'm coming from Europe where everyone is talking about digital so everyone but digital sovereignty and sometimes when you're you know so deep into something it's hard to see outside so how common is this term to you are you familiar did you have you heard of this term digital sovereignty before today okay it's completely new that's fascinating yeah okay thanks that's the answer is my question what is the concept in China wait do you want to term is to do cocoon just you you you have the ability capability to control the software independently so I think that that's ideal I think it's similar to digital sovereignty it's like you own the intellectual property yeah if there are security risks and you know it you know how to fix it and you have the power to control its development and commercialization and so on yeah that sounds I think that's a Chinese term for digital sovereignty okay thank you the other term that we see being used in Europe is strategic autonomy the strategic autonomy so digital sovereignty is a subset of strategic autonomy yeah yes okay I think we're at time so I'll wrap up here thank you all very very much please if you found the reports interesting check them out on our website please feel free to ask me any questions if you see me and please forgive me for only having 18 stickers next time I'll come with more thank you