 Thank you very much, everyone, for your patience with me getting us started. My name is Stephanie Butland. I'm the chair for this session. In my day job, I'm the community manager for our OpenSci, and I'm really excited for today's final session on community and outreach. I will let me see. I'd like to remind people that there is a code of conduct that applies to this session, and that can be found on the USAR site with information on how to make a report and guidelines for enforcement. For the first session, we have, for the first talk, we do have English captions on a Spanish language YouTube video, but I'll mention that for subsequent sessions, there will be closed captioning available for the other videos if you would like to use that. I will also let you know that we have a Q&A session. Many people will be comfortable with this already for previous sessions, but there's a Q&A button, and I'll invite you to ask your questions there. And when the session is finished, there is a channel community and outreach in Slack that the speakers will be happy to answer or have discussions there after the session. So, folks, can let me know if I've forgotten anything, but with that, I think it's safe for us to get started. I'd like to introduce the first presentation called Using R in Latin America, the Great, the Good, the Bad, and the Ugly. This is work by Virginia Garcia Alonso, Paula Corrales, Claudia Raya, Andrea Gomez Vargas, Jocelyn Chavez, and Denise Fierro-Arcos. Hi, everyone. Well, thanks for being here and wanting to hear about our project. My name is Virginia, and I'm here along with my partners who worked in this project, and now the presentation was recorded. So you will see the slides ahead with the transcript in English. My name is Andrea, and I'm here with Paula, Denise, Virginia, Claudia, and Jocelyn. We're going to present the work carried out on the use of R in Latin America, strengths, challenges, and weaknesses. Next slide. What is this initiative about? We know that the R environment is used globally for various purposes. However, how widely is it used in peripheral countries such as Latin America, where English is not the official language and where infrastructure and access to various resources are not guaranteed, among others? In August of 2020, a group of people belonging to communities of users and users of R in Latin America proposed to us to generate a survey to meet the people who use R in Latin America, trying to find out what are our possible strengths and what challenges we face. Next slide. Based on the goals proposed, we designed a survey of 31 questions which addressed five main topics, their interest in R, demographic information, their academic level, their relationship with the community of R, and the area in which they use R. With the aim of promoting inclusivity, the survey was made up of three languages, Spanish, Portuguese, and English, so that people would answer the language with which they felt more comfortable. The initiative was spread on numerous networks such as Twitter, Slack, and Telegram of groups that work with R. For this survey, it was invited to participate in those who were born and are residents of Latin America. This small project grew thanks to the great work and collaboration of many people who added questions, helped correct mistakes, and generated the different translations so that it is available in Spanish, Portuguese, and English, and take into account the diversity of realities in the region. But, in addition, it became a project of R Forward, the group of work of the R Foundation that seeks to improve inclusion and diversity in the world of R. Proxima Slite. Who completed the survey? Proxima Slite. More than 900 people completed the survey. The same people were born in 26 different countries, most of them Latin Americans. A small proportion of those who answered the survey were born in countries outside of Latin America but currently reside in this region. Of the Latin Americans who answered, approximately 3% currently reside outside of Latin America, in different countries of North America, Europe, and Oceania. Proxima Slite. The survey was answered by people between 19 and 69 years old, in most of them under 40 years old. Of the people who were perceived by some gender and the percentage between the feminine and masculine gender was very equitative. And there were also 2% of people who identified with a different gender, who did not identify with any gender or preferred not to answer. At the same time, the majority of the surveyed people had a higher educational degree. The 87% had a master's degree, a university degree and a PhD degree, which surpassed the average of people in the region. Only 13% had a third degree, a technical degree or a secondary degree. There were no answers to people with a lower degree in secondary. Proxima Slite. The majority of the surveyed people used R mainly in the field of research and development, presenting almost 60% of the surveyed people followed by their use for the educational field. The other people declared to use R for the private sector of the industry, for the private sector, for the public sector as government entities and in other areas that include activities of various disciplines such as design, finance, journalism, among others. At the same time, we analyzed the application that the surveyed people gave to R, finding that in all areas, the data analysis was the activity carried out in greater proportion, followed by data visualization. In this same line of questions, we asked the people who responded how many years ago they used R. 40% of the people declared to use R five years ago or more, which is in accordance with the large percentage of people with high academic degree studies. However, more than half of the surveyed people used R relatively recently and the fifth part used it two years ago or less. Proxima Slite. We already know a little about the surveyed people and their use of R. Now, the living or alternation in Latin America carries challenges and challenges. Proxima Slite. Due to the fact that the environment of R is built and is used in English language, we asked ourselves if the language would be one of the main barriers for their use in Latin America, where Spanish and Portuguese are the most used languages in the region. In fact, Spanish was the native language of approximately 85% of the people who completed the survey, followed by Portuguese, used by 13%, and in the last place in English. One percent of the surveyed people speak Quechua, Dutch, and other official Latin American languages. However, the people who speak Spanish and Portuguese declared a high level of English in their majority, being few of those considered beginners and less than one percent who have no knowledge of English. Moreover, when they were asked if English had been a barrier to learn and solve mistakes, less than 25% indicated that it had been the case. Due to the structure of our survey and the group of people who finally answered we cannot identify if the surveyed people use R because they know English, if they learned the language themselves to be able to use R, or if simply the surveyed people had a high level of English due to their academic preparation. Next slide. Another possible challenge that we ask ourselves if the people of Latin America use R are related to participation in conferences, since the majority of international events are carried out in the United States and Europe. In the survey, we asked if they knew or had they participated in any of the 10 international events related to R. Most people, of course, did not know the mentioned events, unless they were going to attend the same in the last five years. Even the Congress of Latin R, which we expected most Latin American people to know, was not the session. This was the most popular conference among all the options, but it was only known by 14% of the surveyed people. By directly asking about possible barriers found to participate in an event, more than 30% of the surveyed people indicated that the event was expensive. And in fact, only half of those people pointed out not having attended any conference. Among the main barriers pointed out, it was also found the high level of time to participate, the lack of comfort, the lack of language, the lack of representative in the event, and the lack of technological resources such as staff, among others. Next slide. We wanted to explore the role of the communities in Latin America. Next slide. Approximately 40% of the surveyed people belong to at least one community. Of them, at least a quarter of them belong to two or more communities, with some even forming part of up to five communities. The Arlayis Group chapters best represented our survey, since around 43% of the people who are part of a community are members of the chapter. This proportion of members among the surveyed people is almost the same as the members of the Arlayis Groups and the Special ES group. But Latin American users are not only members of the local group, but around 11% of the people who belong to a community are also part of international groups. It is important to note that these results are conservative, since some people who answered that they do not consider themselves members of a community despite participating in events in these communities because they did not have an active participation in the organization of the same. When we focused on the composition of the communities in general, we found that women represent a little more than 60% of the people who responded to be belonging to communities. That caught our attention because they represent a greater proportion in total of the women who answered the survey which was almost the same as the men, approximately 48%. Mr. Mango, it is clear if this difference is that an important number of Arlayis responded to the survey since many chapters actively distributed this initiative or if this is representative of the community of users and users of error. Additionally, around 10% of the people identified as part of the LGBTQI community which is similar to the global results of the survey. These results are advantageous since it does not suggest that the community of R is offering spaces in which women and other groups are represented and feel safe to learn and share their knowledge with other members of the community. Next slide. Additionally, we are interested in knowing which social network is most used by the people who use error. In 91% of the people who are on the side responded that they use some social network to communicate with the community of R or to keep some of the news. Most people, 70%, use Twitter, feel the most popular network followed by Facebook, Miran and Slack, among others. Next slide. Up to here we have presented some of the results of this first Latin American survey on the use of error, describing to the people who use it, identifying some of the barriers found and recognizing their participation in the communities and networks. We are very happy that the initiative was able to take place and that we had a great number of responses. We did not want to stop thanking so many people who participated in their construction like those who completed it. But what do we want to highlight? Next slide. In the first place, we wanted to share what we learned when doing the survey. In the process of doing it, we realized some aspects that we did not consider at the beginning of the initiative. The selection of questions was very complex and it is possible that the final version with 33 questions has been very extensive. On the way, we learned a lot about how to ask the questions taking into account the diversity of each country. But we are sure that some questions could be improved to get better-directed answers. The organization and data manipulation of a survey of this magnitude takes a long time. This usually happens with most of the data sets. But many of us found ourselves for the first time with the analysis of a large number of categorical data among other factors that made this analysis a long-term procedure. We also found points to improve in the design of the survey that could facilitate the generation of correlations and new analysis among some responses. We recommend to devote even more time to planning the analysis that is expected to be done before the confection of the same. Next slide. And what did we learn about the surveyed people? It impacted us, although it did not surprise us, the important role of the communities and networks. More than half of the surveyed people indicated belonging to at least to a community and that that community helped them solve problems. The networks also play a key role, even acting as communities in itself. In that context, we also noticed that many people are part of some of the communities mentioned without knowing it, since participating in meet-ups, webinars and other activities also implies being part of a community. For sure, we can make greater emphasis on that. Women represent, most of the people who are part of communities, at least according to the group of people who responded to the survey. Although it is positive that women are well represented in them and that they are possibly creating safe spaces for other women and minorities like in the relays chapters, we must consider that this voluntary work falls over minorities. We found studies that show that when there are projects to improve diversity, the work not remunerated and not recognized usually falls on people who are part of minority groups. And this additional work can put you in disadvantage because this means that they cannot dedicate at the same time the development of their careers to minority groups. As for the challenges faced by the surveyed people, we corroborate that there are challenges in terms of resources and infrastructure that condition the use of ERR and participation in conferences. It is important to undertake and continue actions to promote participation of the Latin American community in international events. For example, allowing to present talks in other languages that are not as English as the USR 2021. Sending a job or summarizing the conference of ERR can be an important challenge for many people. For that reason, we suggest implementing and promoting initiatives such as chat clinic within the Latin American Slack where people can share ideas about their proposals or also the ERR ladies review system. We consider that there is still enough information to answer some of the initial questions. In the future, we say we can identify and survey people who know the existence of ERR but for various reasons do not use it to identify the aspects to improve from the ERR community to improve the experience of those who want to learn to use the language. Next slide. In this same context we consider that it is still necessary to deepen certain key topics. How do we facilitate the integration of the missing population to the use of ERR in their communities? For that, we must first determine what are the reasons for which people have not done so far. But we believe that presenting the benefits of using ERR and belonging to these communities, solving problems, accessing free capacitations and even working opportunities can be a good strategy to expand inclusion. Following with the goal of improving inclusion we would also be interested in deepening about what we can improve so that minorities feel better represented in the ERR communities and events. It is the challenge of expanding participation within and outside the ERR communities towards a more heterogeneous and inclusive composition in the region that includes a greater participation of people with disabilities, people of the LGBTQI community, Afro-descendants, Afro-Latin Americans of the Caribbean and indigenous descendants who are interested in ERR in particular or in general data science. One last topic to go deeper is related to the educational level. 88% of the people involved have a high degree of education which equates approximately to the double of what was observed in Latin America. This made us wonder if we should promote the teaching of ERR from basic or intermediate academic degrees given that the results suggest that there are inequities from the people who have access to learning programming in ERR and potentially in other languages. The same is probably related to the technological gap in Latin America where in average, less than 50% of the population has access to the internet. It can include programming in the basic education curriculum or offering opportunities for potential training to help close that gap. Next slide. There is still much to be done to investigate but until here we come to our presentation to help use ERR 2021. Thank you very much for your attention and we are at your disposal for any questions. Thank you very much for that. I think we should take time for one question and there's a question here from Linda in Spanish so I'm wondering if you could address that one and then the others from Shani you could answer in the chat. Yes, while Linda was asking if we will do another Latin American survey to gather more results and the answer is we want to do it. We believe we still have more information to analyze with this first Latin American survey. We had a lot of responses and there were so many questions we wanted to answer but definitely something we want to pursue in the future and actually learn from this experience. So if you want the next questions we respond them on the chat so everyone can hear about them. Thank you. As I give you a specific thank you I want to ask Rick who's giving the next presentation to start sharing your screen and get yourself ready make sure that works but I wanted to give like the biggest possible shout out to this group of people and everyone who's working in the Latin American art community because this has been such a grassroots effort and it's incredible the progress you've made and I want to congratulate you because you're having an impact not just in your own community but on other international communities seeing how they can do this, so fantastic. Yes, this was, I want to also say that to the users who are organizing this conference thank you very much, it's exactly what you said, it's their job. So, Janie was asking if the slides are available and they have just shared the links to slides which are both in Spanish and in English for those who speak in English and I want to follow with exactly what is being said there maybe it was too fast but you have these slides there to see whenever you want and of course if you want to reach out to any of us we are six in this presentation but we are many people who are participating constructing this survey so we are a lot of people who can answer questions so and I don't know if there's more time I think, you know what I think we'll move on because thank you so much anyway and thank you for Paola to paste the information in the chat it's already in the slacks thanks again everybody for your patience good thing it's a nice, it's a community presentation session so I know everybody's understanding the next presentation is ourcommunity.org a central community infrastructure for R this is being presented by Rick Pack on behalf of the team including Ben Uba, Meet Bagnagar and João Vitor Cavalcante take it away Rick thank you Stephanie I will attempt to move a little faster so that we can make up time thank you all for your patience it is my pleasure to be able to represent lead developer Ben Uba in a sort of debut of a website that documents our activity both in terms of users and the software ecosystem called R-community.org this work was also presented during a video recorded for last year's user 2020 presentation among participants for the new developments that will be shown along with some former are Meet Bagnagar and João Vitor Cavalcante who are both Google Summer of Code 2021 mentees with Ben and I having the opportunity to serve as mentors I will say that this website it's continuing development of the website continues and so in your browser you will need to go to r-community.org slash user groups at this time and there is another fun surprise that I will show soon so there are quite a few pages that are available through this dashboard some of these pages were present before and some are new among the pages are a calendar featuring the many R events that are available on the planet CRAN packages, the R-Studio Community website including activity among members of that website the Saturdays events, Twitter, Stack Overflow R-Bloggers I think it's .com sorry r-bloggers.org perhaps and more many pages Ben's Twitter handle is R-Central with two R's and two L's please consider reaching out to at R-Central with two R's and two L's by following him to show your support he also will document new developments of this dashboard as well as surprises so the URL at the top of this page will take you to a new dashboard that Ben and team have debuted which features activity about USAR 2021 as acquired through Twitter and I am happy to show here Yanina being recognized with the most liked tweet today you may know that she is a very hard working organizer of this event that we are attending and we are all very grateful so why r-community.org? Well, it helps users connect with each other it shows trends concerning the R language which could help those that want to train others as well as new users to know what is popular it also motivates targeted assistance in the sense of where geographically R might need a little bit more energy in order to become popular I think about the presentation that just ended concerning Latin America and I look forward to learning more about what is happening in the world of R there and of course we want to boost attendance of R events speaking of R events here is a calendar and this what is on screen may be small so I'll note that at the bottom right I've zoomed in some on a date so that you can see every day there are almost every day there are R events if you click on the filter at the top left you can look on the calendar at particular geographic regions including worldwide and worldwide we see R is popular as may have been communicated in the prior presentation there are opportunities for growth in Latin America, Africa and some other locations but we can still celebrate during this global conference the worldwide activity of R Saturdays are free R conferences that have been going on since 2016 and we see popularity has grown beginning about in 2018 when you scroll down on the page of any of the dashboard pages more information will be revealed many elements are common between the pages as far as interactivity such as scrolling and clicking and when you scroll down the Saturdays page you can see that R studio and jumping rivers are among the top sponsors along with R consortium the Twitter page helps to see what individuals are discussing in the R universe you may know that the hashtag R stats so that's hashtag R stats is a frequently used Twitter hashtag to denote that a tweet is about R and everyone probably knows but tweets are short messages that are communicated through Twitter.com we can see that thousands of tweets are generated about R every single month and so what Ben and team have done when we scroll down on that screen is help us more clearly see what is happening with R tweets and so this world world excuse me word cloud shows some popular terms among those tweets and I noted CASO CASOS about in the middle of your page so I am not familiar with that usage in our world and that presents a potential opportunity for learning more our user groups are documented through the meetup.com API we see again the international activity and then there's all when we scroll down there's a screen that depicts every country and the number of groups in each of those countries this is also a zoomable interface and you can zoom by clicking and dragging which is denoted by the arrow so that you can zoom in on some countries and zoom out it's quite interesting in general the suggestion I would give is to click and explore so to talk a little bit more about navigating the website this is the RStudio community page on the left there is a bar depicting the various pages available to use our 2021 page may not be present yet I will show that URL again at the end the button at the top for RStudio community the button at the top that I've noted may not yet be active when you click an arrow that you see on any of the pages it will sort so that is what I am indicating at the bottom left I mentioned click and discover so on the r-community.org page on the left you will see the users if you click popular R community member and Twitter advocate Mara Averick's name click it will take you automatically to her r-community profile on which you can see information such as when she joined and how many posts she has created and this is an opportunity to celebrate just how active Mara has been with over 2000 posts created by her on r-community.org and thousands of tweets as well thank you Mara so how does r-community.org come to us how was it developed well RStudio's flex dashboard now drives those dashboards this is a conversion from the previous I just blanked I'm sorry but this new flex dashboard usage is what has been implemented this year during Google Summer of Code 2021 on screen is an R mark pardon me R mark down page that shows the invocation of flex dashboard with this being an open source website all of the activity all of the code can be viewed through github.com and those links will be provided in the notes that will be attached to this presentation to just look at the code a little bit we also see this is Joao's one of Joao's commits we see the usage of e-charts for R sorry one second the e-charts for our package as well as the HTML widgets package so as you think about potential displays of your own or how you might get involved there are frequent usage frequent usage is a flex dashboard but also other R packages to enhance those displays we want to thank sponsors that include Syncra the R consulting company RStudio and Google Summer of Code with the logo being in yellow at the far right one second so a few notes issues are welcome you can use this URL to communicate potential enhancements for the group such as if you see that your user group is not among those featured on the R user groups page please let us know also regarding the calendar you may see an event that needs to be added please let us know this is a community website and it will thrive because of the activity that we know our users express through various channels we are a vibrant community so this page is going to be a great resource for us also Ben and team are looking for sponsors so please do reach out if you would like to further this work Ben has committed an enormous amount of time for this work so please do consider sponsorship so thank you for listening I hope that this more rapid version saves us time our hyphen community.org as I said is going to be a vibrant central resource for us and your involvement will be key please do communicate through GitHub any thoughts you have about its enhancement a reminder that you need to use the slash user groups so that's our hyphen community.org slash user groups at this time as an entry point through which you can acquire the other pages although as I said there's a fun surprise when you go to this URL you will be able to see the user 2021 page I will say that out loud for accessibility that page is our hyphen pardon me HTTPS our hyphen community.github.io slash user hyphen tweets thank you all for your time are there any questions thank you very much Rick you've done us a wonderful service by jamming all that information in and we have three minutes before we have to start the next one I will invite people to ask a question in the Q&A if they like while we're waiting in case as someone might have a question in there I wanted to share with folks a couple of comments that people have made you know one it's a super amazing dashboard and someone was saying that they like shared this in every one of the presentations they've given to other people and someone else said Ben is a dashboard wizard so congratulations for that there's a lot of appreciation in the community that's beautiful to hear and I will say Ben is online now should that invite any questions or well I just want to celebrate him so I'm glad that Ben can join us today indeed that's good that he's here and that you're presenting on his behalf yeah happy to be here and happy to take any questions yeah Ben is there anything you wanted to share yourself at this point I think there's a great future for this project and we hope that we find sponsors and they help us put free land put sponsors in code and maybe financial and infrastructure one thing we don't have yet is automation we're thinking of making this fully automated some parts are automated not all so we hope that in the future we could be able to do that and now there is one question thank you for bringing that up there is one question here from Gabriel thank you for the presentation it would be nice to have this dashboard in other languages are you planning to translate it and of course that's the kind of thing that probably requires sponsorship and participation and learning from others so yes any consideration or translation yeah sure we don't have that in the near term but if we have contributors that can help with that we're happy all right thank you all I will also say in the chat there was a thank you for the out loud more accessible URL fantastic that's much appreciated so the next presentation is called COVIDR an RMD gallery a streamlined process for collecting community contributions in a gallery website and this work is by Ricardo Perica and Francesca Vitellini please go ahead when you're ready okay thank you so the work I'm presenting today was done as part of Iran 2020 one year ago so I would like to thank a lot my fellow organizing committee members for the amazing journey of turning Iran into the first large virtual art conference and if you want to know more about this journey I link here in the slides some nice reading about the process of making Iran going live as a virtual event in particular one of the things we organized was wanted to engage the community around the topic of the COVID-19 pandemic and for this we organized COVIDR which was a contest of open source contributions built around an open source repository to collect and ultimately showcase contributions in an RMD gallery website this COVIDR also featured a pre-conference event when participants were invited to present their work and eventually granted some prizes and awards so let me quickly navigate to the website so this is the home page where you can see the list of contributions already with the result of the contest from one year ago you can expand the abstract of each of them and in particular each of them has a related gallery page like the one you see here and from this menu you can browse to the different contributions from the gallery page besides the actual content submitted by the participants we also featured the like button where basically the community could like the contributions they consider particularly interesting and we also featured as part of the contest a dynamic badge that participants would be able to find the code in the website itself to include in their repository or website like you see in the case here so let me go back to the slides and here yes so I have a few screenshots just for the records here but actually for each contribution in the gallery page you would see that we would feature the title of the contribution the authors of course the open source repository very important the abstract and ultimately the actual content for the gallery page so how did we collect this information actually we asked participants to provide these in the form of a YAML metadata file where besides the usual title author abstract information in particular we would ask them to provide us the content we would display in the gallery page along with the specific type like Shania, a web page or a GitHub GIS for more complex and heterogeneous content and all this was collected in a GitHub repo via pull request and GitHub issues using community features like PR and issue templates and if you want to know more about the submission process itself we had prepared a nice video tutorial that you can rewatch on YouTube to see basically how submitting a contribution would feel like so for the rest of the talk I will be showing how we actually developed a package called RMD gallery for kind of turning these YAML metadata into actual gallery pages and then I will show you something more about the badges and the light mechanism and last but not least how we have automated pretty much everything with a bit of actions so let me start with the RMD gallery which is a package providing a custom armor downside generator which would include in the rendered website a gallery of pages generated from a YAML or JSON metadata you can see this in action in our companion example repository or and website RMD gallery examples you will find the link in the slides and beyond Covidar here is a link to our gallery of online resources that was sort of a POC even before Covidar was put in place so the idea behind RMD gallery isn't new and in fact here I would like to mention Xbin, the SVing project which was based on JSON files also done GitHub and most relevantly Jekyll which is a very popular static site generator based on markdown templates and YAML data and this is in fact what we use for our company website gallery and this is something I have covered as a topic in a presentation two years ago at the user so RMD gallery allows metadata in both YAML or JSON format that can come in the form of individual metadata files or as a metadata collection files where each metadata would be under a key entry so in both cases however the provided content would be rendered by RMD gallery based on several alternative templates into an HTML gallery page and a gallery menu entry and possibly one could also specify in the metadata a menu icon for the gallery so how can you use RMD gallery you can simply specify the provided gallery site generator as the armor down site generator for your website and you can define configuration and customization in the usual site YAML file of an armor down website and in particular I can mention here the fact that you can define templates for custom page types like shiny applications and also highly customizable additional content to be rendered above or below the main content in the gallery page in the form of HTML or R glue expressions based on individual pages metadata and this is actually how RMD gallery renders this very customized and ad hoc top heading showing the repository the abstract and all these fancy buttons in a similar way to this concept of including before and after actually the parsed metadata can be retrieved and used to create arbitrary content in any other location of your armor down website here I have a simple example that I'm probably going to skip in the interest of time but this is in fact how the homepage of RMD gallery is put together with this index of contributions by contest result including collapsibles models and this was done with a combination of HTML tools and BS plus in R so the next thing I want to talk about is how we came about with dynamic badges so dynamic badges were created using the budgen.net generator service which is able to render dynamic badges using an API basically and the API we implemented was based on a YAML file that we would generate and update along with the live website as the Kovidar contest would evolve to all the stages for the actual API we were using a tool called an instant API ranking notebook which is a very simple way of writing a simple Node.js function that would be offered as an HTTPS and point out of the box so this was something pretty cool to explore and you can find some further information if you want in the slides the other aspect I can give some more details about are the community likes and if you can see here this looks pretty much like GitHub issue comments and this is not a coincidence because we were implementing these using utterances which are lightweight comments widgets really built on GitHub issues and this was a perfect fit for our GitHub centered project in particular this we wouldn't care about comments but we would create a predefined comment and collect thumbs up reactions on this specific comment so the last thing I want to talk about is automation automation automation and this is just a funny quote from one of the enthusiastic attendees of our workshop we gave a theorem one year ago so the process behind Kovidar was as I mentioned a number of times built around GitHub with a poor request and issues and in particular we put in place all the nice GitHub community profile features in particular around issues and poor request templates and to ensure a nice experience both on the GitHub repository and the website with a lot of repeated content among all these and on the website and the approach we use is something I can label as don't repeat yourself but let R write everything twice for you so indeed we centralize the definition of the content that would be repeated and we would let R do the hard work of rendering this content in different locations now talking about automation of course a key part here where a GitHub actions workflows and in particular upon every new poor request for a new contribution we would make sure that the website could be built with the new contribution producing an artifact to preview the contributions as it would be integrated in the gallery and then it was just a matter of merging the poor request on the main branch to see the contribution live something a bit more special we did was also managing as part of the CICD workflows the creation of these voting utterances issues with a standard comment inviting visitors to give a thumbs up vote for the contribution and since this is all GitHub based we could use the GitHub API and we implemented just a small functionality for doing that now with all this machinery in place actually the evolution of the context and making the website co-evolve with the various steps of the context was just a matter of defining poor request emerging poor request from each individual submission up to announcing the constant the contest winners and our this live during the pre-conference event these are these pictures are actually contains actual links to the corresponding poor request for those who would be interested in seeing what was behind these several steps of the of the contest from a technical viewpoint I have one additional slide about GitHub actions we went kind of all in with with GitHub actions here so we defined a specific workflow for locking the issues that we use for voting because once the contest was over we wanted to basically stop voting and this was done again with the GitHub API through our GitHub actions workflow and the other thing was having an automated way of sort of moderating the comments to the voting issues to make sure that we would only keep this predefined comment and ensure that participants to the event would actually just use thumbs up reactions so I'm probably going to wrap up maybe it was a bit fast but I'm happy to have more time for questions perhaps so what I've shown is how we have put together an open source contest of open source community contributions which was centered around GitHub using a combination of open source code tools and services and relying on a newly developed package called the IMD gallery something I am particularly grateful for is the fact that this was developed with good development practices in mind so this is helping me being my future self today one year down the road being able to still know the details of what I did how I did as part of this project one year ago so the slides are available at the link which is a bit maybe too complex to spell out but I will add it to the Slack channel to the session and I'm sure there is a lot to unpack from the presentation a lot of details but the idea is also that these slides we serve as a potential starting point with a lot of links and pointers to explore several aspects more in details I can mention here also some more resources you can browse through the already mentioned submission in YouTube registration and tutorial you can actually rewatch the full pre-conference event and also it was a kind of a many of the features that we were using to support this project were actually also part and covered by a workshop we gave at ERAM and again within the context of ERAM we were using a similar approach to actually collect the conference material of ERAM in a collaborative way on a GitHub repository and then finally just some contact information so thank you all and I'm happy to take any questions that was fantastic and I love the fact that you were doing this like bringing in community contributions but facilitating this in a way using tools that people in the community are already familiar with working with it because then it sort of lowers the barriers for the people who were already participating that was fantastic yes absolutely and this was also the reason why we actually didn't only have a poor request because we knew that even poor requests might be kind of a slightly higher entry barrier for some members of the community who are not really used to GitHub but opening an issue is something that we believed was really easily to approach and really the community features of GitHub with templates were very very useful here that was such that's such a thoughtful approach to things thank you for that certainly I will invite people to add a question using the Q&A tool we have three minutes for questions or discussion pause for a second to give people a chance to think about that one thing I wanted to ask you is I know there was an incubator session that just took place about organizing conferences and not reinventing the wheel have you been in touch with that group or connected with them in any way yes I was following it was an interesting discussion there was a lot of reflections about the current user but of course I believe that this type of tools and especially open source tools will be the future for conferences like user and I'm sure that also the organizing committee obviously this year is fully onboarded into this based on what I heard in the incubator so looking forward to the future and to more things like this this is such a good contribution and I'm reflecting back to like Rick's presentation of Ben and the group's work too yes because I don't see another question popping up one thing I can ask you is let's say you're going to do this again typical kind of question let's say you're going to do this again what would be the next thing that you would add or something that you really feel is a gap that if someone wanted to add that for their conference to build on what you've done what would that be yeah difficult question of course I am fully aware of the fact that what was developed was really tailored for this particular COVID-R although we tried and this was one of the reasons for developing a package to do the hard work because I think when you start writing functions and packages you also have a bit this fresh mind of abstracting from your specific use case and try to already generalize a bit but of course I'm aware that there is a potential for a lot more also because even COVID-R is something that was organized a bit on the side of a large event so for large events there is definitely something where there might be some shortcomings of this approach which would be very interesting to explore with future developments there's a something in the Q&A here from Francesca but it might be more of a comment but Francesca feel free to build on that there saying this is basically a great example of being an R developer even in the organization of community events like using these R developers skills to build community is a brilliant thing and actually this was also the chance for me as a developer to explore a lot of the tools so actually having these events oriented to the community is also a chance for me it was a chance for me definitely but it's a chance for the community to build more and more knowledge around these topics wonderful thank you so much for doing the work and presenting it I think with that we will move on to the final talk of this session um Eric because he's so experienced with this kind of thing is ready and on board and so our final talk in this session is how open source and R enables state-of-the-art media production and this is Eric Nance hello thank you Stephanie and this has been an excellent panel our excellent session I have tough acts to follow but I'm going to be excited to talk to you about how open source technology is enabling my production of R media content and how I think it can supercharge your workflows too if you want to go on this journey so there are lots of ways as we've heard already to interact with the R community I'm going to highlight a couple of ways for both interacting and putting your own content out there one of the awesome groups that I've been a part of for a while now is the R for Data Science Slack community great place to go back and forth and learn about the best practices of R in a safe environment and that channel has a ton of great dialogues a ton of great collaboration there and of course some of the tools that you've heard about throughout use R and beyond you can share your own insights via blogs often powered by our markdown like what Thomas Maka has done in his popular blog on highlighting lots of cool things that he explores such as when he was looking at the GT package and these are excellent ways to collaborate and to interact with the community I think there are ways you can add even more to that if you want to go on a journey because sometimes when you see some of these additional sources that I highlighted you look at what might be called the destination but you're not as much seeing the journey to get there so to speak and recently for the shiny contest this year ran by R studio I had a good fortune of serving actually as a judge for that but I wanted to get in on the fund so to speak so I added a submission of a fun application I did as part of a community group I've been a part of which I'll talk about shortly and we were required to write a blog post on the community site blog post or post in general to talk about the app in more detail but I wanted to take it a step further I also augmented the actual post with development streams of me working on the application code and sharing it live through all the errors that could happen and all the winds along the way as a way to get inside that to that destination of the shiny app and that's what I'm most excited about the share of you today is the tools that I've used for the past year or so to make that happen that are all completely open source and to take a step back on when I thought about going around this journey because those of you that might have heard my content before I've been mostly an audio kind of person I do a podcast about R I've you know done things like that but it was a couple years ago that I ran into Jim Hester a very talented RStudio software engineer at one of the RStudio conferences and he had started producing some great screencasts on some of the packages he's developed and using awesome like lower level R functions and he posted those on YouTube and I came up to him and said hey Jim these are great tutorials how did you make that you have such a distinctive look to it and he said guess what Eric I'm using OBS I have heard of OBS before but I did not think it would be the right fit for sharing our content but I saw Jim do it so this kind of was a little bit in my head to maybe pursue later and then late last year and early this year I became part of a very fun community my other kind of open source itches so to speak of Linux and software development Martin Wimpress the architect of a Linux distribution called Ubuntu Monta decided that since the era of the pandemic to try actually live streaming his development of those Linux distribution all the utilities inside and on top of sharing the code he was developing which of course wasn't R but that's not the point he shared how he was using OBS to produce these streams at that point I am hooked as I just alluded to each of these very separate people have producer content with OBS now what is OBS OBS studio is a cross platform almost jack of all trades if you will for audio and video production it has actually become very quickly that the fact those standard for streaming video content on YouTube and Twitch it has lots of flexibility via its plugin system now one thing to note is that up until recently most of the tutorials on using this platform of OBS for media production were always targeted to arguably the biggest audience that's using the platform the gaming community many many gamers out there live streamed their gaming adventures no matter what Janair or anything like that on Twitch and sometimes YouTube as well but what about people like me what about data science well first before we get to that let's talk about what makes up OBS how do you actually start building your own kind of digital garden so to speak for video content there are two principle principles I want to highlight for OBS everything that go is basically a in two categories one is called sources think of these as some components that have individual attributes that are going to be displayed in your production you might have a camera feed audio inputs for microphones capturing screens media files and perhaps much more in fact there is a lot more but all those are enveloped into what's called a scene it's basically a collection of one or more sources and any type of source can be included including other scenes now wasn't until recently that I took a step back and thought about these two principles and I thought you know that sounds quite familiar with another thing I use a lot let's talk about how this is related to R itself in R one of the fundamental object types is a vector it's basically individual components that have their own unique attributes that have to be of the same type you might have a string of characters might have some numbers booleans of true and false and then you have the list object the list can be a collection of one or more objects and any type of object can be included they can be completely different and yes including other lists now this is probably the only time other than me as somebody saying that R it has direct relations to OBS but to kind of learn the nuts and bolts of how to use it effectively it was great to translate that into terms that I as a data scientist has been using R since 2005 could understand and let's look at a high level one of my we might call scenes in a lot of the streaming setups and video content I do I have a scene called welcome scene which ironically you're looking at in the video of this talk right now it's composed of a backdrop it's a nice image of our hex logos it has a footer at the bottom to share my various ways of connecting to me on social media and other media content I produce and then it has me from the camera that I'm using to show my video and my microphone to share my audio and these are actually those four kind of main categories are actually other scenes within a bigger scene and because this is an R talk of course you can actually export the data from OBS and actually interact with it as a JSON file so I can show you quickly these are all the scenes and components that are nested in my quote welcome scene it's pretty cool I dare say that I could take that data and actually play with it in R itself again to help make some of these connections and how I can use OBS more effectively now OBS is admittedly a very complex tool that was why I was pretty intimidated about even when I first heard about it a few years ago and I'm sharing a couple quick tips but these are not the only tips I have but if there's one thing you take away about your effective use of OBS it's take advantage of nesting of scenes because that is going to help you organize things immensely easier there are a couple quick tips here well but I want to basically mention that I completed my first ever live stream of inside my OBS setup earlier this week on YouTube and that's hyperlinked right here in the slide and I plan on doing more video tutorials of using OBS but tailored to the R and data science crowd because again this is an audience that is now starting to use this tool a lot more and I want to make sure that I pass along the things I've learned mostly the hard way to all of you if you're interested on going on this journey there's another tool that has changed my production of another effort that I've had going for a couple years now and this tool is called VDO Ninja it is a free web service for audio and video sharing sharing meaning like your camera and a microphone much like you might do on the platform like zoom which we're using right now for use R it is actually powered by WebRTC and peer-to-peer technology some of the Stadia things for sharing video and multimedia content between two parties if you will this is easily imported into OBS as what's called a browser source and what you're seeing on the screen here is a capture of my latest shiny developer series episode where I connect with a guest in this case Nick Strayer from the RStudio team via video Ninja and I import that into OBS along with his screen share of the shiny app that he demonstrated and this has just changed my production of this important effort and it's improved it 110 percent I am so thrilled with it I'm still learning the ropes so to speak I want to use it effectively but it's been a lot of fun to learn and to give you a real kind of aha moment I can actually bring in video Ninja in the slides I'm making right now that's me hello hello that is really cool to me now I'm going to leave with the few minutes we have left with a few of the wins that I've seen by producing media content in particular this video content with the R community first of which a very talented R user that I've been watching since about a couple months ago his name is Tan Ho he is a very much an expert in R and especially shiny development he was on that aforementioned R for Data Science Slack community and there's a channel about shiny questions and somebody asked for some help with a complicated issue they had with a shiny app that was using leaflet well Tan actually took some time in one of his his afternoons and put up a live stream on the spot to debug the application that was strictly amazing and it was a great way to hear what Tan's solution was to solve that problem and you can definitely check that out on Tan's Twitch channel but that is just a great way to take advantage of these skills to actually help others in the community and speaking of community the thing I want to leave you with is that we are having a very much a growing community of R and Data Science Streamers Jesse Mostapak who works at RStudio as a content creator has written an excellent blog post that is being regularly updated that showcases who she has been watching and who else is involved with streaming data science in the community we don't have time to go over all the names here but I've been actually collaborating with quite a few of them sharing knowledge, learning from them and I think we're really becoming a much more closer knit group and it's just a fun journey to be on and I would not have been able to meet all these people if I hadn't just tried putting myself out there trying something new but you never know what can happen when you use these the power of open source to connect with communities at large and I've been so fortunate to connect with the R community even more than I even had before and also my friends from the Linux community who guided me on this journey of learning this new software to make my media content even more and just to show you that I eat my own dog food so to speak all this presentation you've seen has been using OBS to actually stream it which is pretty cool so that's what I've got I really appreciate all of you for tuning in happy to answer questions along the way thank you again your talk was perfectly timed thank you Eric again that doesn't surprise me there's one question I'll relay to you from the Q&A and I'll also invite people we have five minutes till we have to end this session so please add your questions for Eric Question from Elio besides the technical tips do you have any advice on how to make a compelling stream of our development and debugging? Great question everybody kind of has their own you need take on it I think my biggest recommendation is to be true to yourself and what I mean by that is even when I first started some of this I kind of felt pressure to kind of prepare for them so to speak and try to think of all the issues I might see in like this Shiny app I'm making for the Shiny contest and trying to narrate through what could happen but in the end I think the best way that people learn is seeing how you solve problems and that's where it's actually kind of good to I don't want to say challenge yourself but to put yourself in a situation where you have to solve maybe a small issue or maybe a bigger issue and that just not being afraid to show that has boosted my confidence quite a bit and many of the streamers I watch also have that same approach and the chat will also jump in and help each other out in fact Tan was helping me out on a couple of my streams when I was dealing with an esoteric Shiny issue he knew what to look for and I plugged it in and it worked so I guess don't be afraid to put yourself out there and you don't have to be super polished to do it just have fun with it you never know what can happen that was a great question there are lots of people putting like multiple pluses in the chat which may not be visible to all attendees I think but just letting everybody know there's a lot of enthusiasm about this and people commenting on the power of open source here I'm wondering if there's a good way to do this it sounds like a lot of people are curious about using OBS they're really interested in when you have this tutorial out I wonder if people can even add in the Q and A just leave a statement like are you have you already tried it out are you thinking of trying it it would be nice because that way community wise we can expose ourselves to each other there's a question here from Garrick do you use peripherals or extra equipment to transition between scenes so kind of yes I have a piece of equipment in fact I'll stop the screen share here if I can because I want to show this on the video here see if I can do that I have a piece of equipment called a stream deck which I'm not sure if you'll be able to see on there that has a way to program different functions in OBS to basically transition to different places and to be able to click a button to aid instead of clicking in OBS directly so that was something I bought online a few months ago and that's been helping quite a bit otherwise most of it is done in software and scripting but that's a great question that was a very cool looking thing Eric and I'm wondering for the benefit of some people here could you describe what that looked like yeah so if you weren't able to see that it's basically like a mini keyboard that has a set of buttons that are initially blank but through software you can program them to perform like a keyboard shortcut or actually run a script on your system so I have a bunch of buttons that are mapped to different operations in OBS via its API to kind of change things hide things switch around so that I can just have a nice icon and just push it like I would just typing some on a keyboard and that's been a great great purchase for me lately that's fantastic thank you for sharing that and I love how in asking you to describe it we all got a little bit more information about how that works so accessibility for building on what we all learn from the process let me see I will note here that Rick mentioned that he recorded his use our 2020 presentation 20 about Ben Uba's dashboard using OBS and Rick has shared a link to YouTube perhaps Rick you could also put that in the Slack channel for this session this is it for our time brilliant first of all that for a partly rough start partly owing to me starting with the wrong link this went really smoothly so thank you so much to all the presenters and they're supporting collaborators I love how everybody's giving shout outs and encouraging everyone it's like I love a good community session maybe all the sessions are like that but I really enjoyed this one let me see so I guess to wrap up I also want to thank our Zoom host Yuya who is like totally working behind the scenes and managing all of this stuff and the screen sharing and all this so thank you very much to Yuya thank you to the person who's been doing the captions for the session please do continue the conversation in Slack this has been a really neat session and I would like to invite you now we don't get extra time for Q&A at the end of the session today because in nine minutes we're all invited to and encouraged to go to the closing session that starts it's going to be the conference closing let's all sort of give a pat on the back to all of the incredibly hard working organizers there will also be some awards announced so just a little bit more time of community spirit and let's all hang up on each other now thank you all for joining and like really appreciated the participation here and I will see you all over in the other Zoom channel for the closing ceremonies