 All right, it's the last talk, so I'll be quicker. So what I want to talk about is open science with open data on the open web using open source. But that's just because I knew that the real title would never get accepted. Spreadsheets. So what I want to present today is the collaboration between three really cool projects, some of which you've heard before in this conference and that's stencil and substance and one that I'll present a new and that's PubSuite. So we'll go through each of these individually to just explain what they are and how we work together as a team. No, no, no, it's about Spreadsheets. So Spreadsheet is an interactive peer application for organization analysis and storage of data. Should I go on? So what is Coco? Oh, did you miss that one? I'll have the slides after the talk so if you can ask questions. So Coco, Coco is the organization that I'm presenting here today and that's the Collaborative Knowledge Foundation. We call it Coco for short, even though it doesn't make any sense. But it's Coco for us because that was the easiest to pronounce. Coco Foundation is a new foundation. It's about Adam Six, seven months old. Yeah. And we just recently got a fine amount of funding from the Moore Foundation and we've been initially sponsored by the Shuttleworth Foundation. So we really want to thank them for making Coco possible. I really like how Jeremy before gave pictures of people. I think that's important, like to connect the people to the projects. So in the last minute I added that. I didn't have the foresight to do it before, but I added it after Jeremy's talk. So Coco was founded by Adam Hyde and Kristen Ratan. And Adam Hyde had several projects in the open source space before. And then the activism around knowledge production. So things like Book Sprints is a company that he founded where you basically write a book with a group of people in a week. And Kristen Ratan was a publisher for PLOS as her last position. And then decided to find Coco. So Coco is working on this project called PubSuite. And what is PubSuite? PubSuite is this project for... It's a framework where you put together components for knowledge production. So say for example you wanted to have a science blogger. Like nothing like that exists. That's really funny. Like there is no blogging application out there where you would be able to add equations, add references, add figures and present that in a scientific way. It just doesn't exist. That's funny. So that's one of the things you could do with PubSuite. For example you take the text editor, you take the post editor, you take the users manager components and you bring them together to form a science blogger. And then there are other applications that you can build with PubSuite. It's for a lot of the applications in the knowledge production sphere that we want you as the community for Coco to build. And it will help you along the way and will present the best examples. So for example the first thing that we're building is a science blogger. And we're building a book production system. All of this is based in open source. We are very open to the way we run our meetings, to where our source code is, where we chat. Everything is open. So anyone who feels the need to form some kind of knowledge production system just come talk to us and we'll walk you through it and we'll help you along the way. Because I think there's a lot of people who want specific things to work in a specific way and they don't have any support in doing that. So if you want that, come to us. We'll help you. This is how PubSuite looks right now. And it's just a mock-up. We have an actual demo and it's nowhere near as pretty as this. This is just pretty. And it's based on the work that our UX designer has started working on since about two weeks ago. So she's very new on the project. So things will improve in UX-wise. Right now, this is how we want PubSuite to look. This, for example, is a science blog. It's very simple. That's the whole thing. Why doesn't this exist yet? You have a list of posts. You can edit them. You have a list of users. You can add permissions to them. You can say, hey, userX, you're a contributor. You can create a new blog post. Hey, userY, you're an admin. You control everything. That's it. There's no magic there. But all of these components are individually developed, and they come together in this whole. So the user's manager is developed individually. The post manager is developed individually. And you bring these together to form a whole system. So I haven't talked about the other parts of the project. That's substance. So what is substance? The two developers of substance are sitting there in the back. If you could raise your hand. I'm Oliver Superstars. They're working on a web-based text editor. And they've been doing that for the last four years. For six years. So it went through a lot of the initial pains that a text editor on the web would have. And they've gone through that. So if you want to use Prozmer, don't. Just use substance. Substance went through all of this. If you know a bit about how text editing works in the browser, it's very fraught with differences between browsers and cross-browser compatibility. And some things work weirdly in its own browser and the other. Substance went through that. And they produced a lot of very cool, final, production-ready text editor-based products. So you can follow them on substance. And there's substance on the web page. Oh, there they are. Michael and Oliver. Putting the people to the project again. And then the last part of the project, and the collaboration is Stensilla. And Stensilla is data-driven documents. Some of you were at the previous talk where Mike and Oliver gave a demonstration of what Stensilla is. It's basically this templating language where you can include R code and run it in a browser context. And then it has two parts. It has a sheet and a stencil. And I'll demo all of that a bit later. So you get a good idea of what Stensilla is. And it's made by this guy from New Zealand. This is a common thing. We have a few people from New Zealand, actually. I don't know what it is about that place, but it inspires something in people. So Nakomi is working on Stensilla. And he's been working on this alone. And he, I think, just recently got a lot of attention by being on top of Hacker News. And people are recognizing the effort that he put into making this project. So I think things will pick up in terms of who the contributors are. If you feel like you could contribute, do check him out and reach out to him. He's very friendly. So demo time. Hopefully my everything is, wow. And that's never happened before. It's a good sign. It's gonna go well. Okay, so first I want to demo an initial configuration of Popsuite and that's the science blogger. So just quickly, you create a new blog post. So this would be a separate component. What you see now is the post manager. You create it, ah, demos. I love demos. So you create it, you go into the editor and enter some abstract, enter a title. And then the fun part begins when you want to add references. So you can search Crossref and you can say cats and dogs. I think that's a common theme here. You search for cats and dogs. And now it goes out to the Crossref API and if Internet works, hopefully. Come on, Internet. Okay. That's a bummer. I'll try that again. There we go. Supplemental information. That's pretty okay for the query that I put in. CSV, that makes sense. So you add the reference and then in the paper or in the blog post that you're writing, you can insert a citation. You say I'm citing this supplemental information and the references build automatically and the whole site rock is in there, like the metadata that is, the reference is in the document. And then you can go further. So you can say, ah, I want to insert a figure. So you say insert figure. Hieroglyphic pepper. That's an interesting one. So spreadsheet, 4,000 years old. You title it. Wow, that's a great typo. And then when you, again, when you're entering the text, you can say referencing. And you say cite the figure and you cite the figure. And again, this figure is cited. And then you save this, go to posts, you publish it. You go to the main page, to the landing page. And you see this post, and I say this. We just started working with a UX person. So please forgive the brutalist look. But if you click read more, you get the entire paper in an elife lens kind of reader. So this is a substance reader. And this is basically what you need for scientific blogging. That's it. There's nothing more that you would want. So this is version 0.2. And in 0.3, I think the one thing that people still miss is collaborative editing. So two people can write on the same blog post at the same time. And that's courtesy of substance who worked on the back end, the OT transformation that made that possible. So that's the science blogger demo. And then we have the sheets demo. So sheets would be the stenciler sheets. I will just bear with me. So you create a new sheet. You edit it. And that's the spreadsheet. So you go... This is backed by R. So I can do things like a1, a3, b1, b3, boom. And I have the plot that's coming from R. And what is more, I can then take... I can take this address. Right now it's a bit rudimentary, but I can take this address. I can create a new stencil, which would be the blog post scientific document. I say, create it, edit it. Well, hello there. And I say, paste the address, select B4. Sorry that the interface broke with the low resolution, but here we go. Ah. Okay, it's not gonna work. Because the interface broke because the resolution is too small. But you would select this and you would say insert and it would go to the stenciler to the stenciler sheets and get that figure. And if you change any of the parameters when the figure was generated, it would update here. So you can imagine this. You would have some simple analysis going in the sheets. Spreadsheets are something that people are really used to working with. So you would have something simple in there, do some R statistics, and then create a figure included in the blog post. And if you change anything in the analysis, the figure will automatically update. So that's the sheets demo. And there are some. If you go to stenciler.la, you have the real spreadsheets with real analysis with this platform. So that's the demos. But then I had a really good idea yesterday. It was amazing. So imagine that you could follow the feed of DOIs minted in real time. And Joe's smiling because he knows we had that discussion. But minted DOIs or newly created DOIs is a signal of where papers are published. So if you could look at that and geo-locate the affiliations, you could know where science is coming from. And if you plot that on the globe, you'll have this real-time map. I think DOIs are like 20,000 a day or something like that. So you'll have a real-time map every few seconds of where science is happening. I was like, that's so cool. I should do that. And I tried. And I came up with this. It spins. And I think there's... There's maybe... not somewhere. So imagine that this is a real-time view of the scientific output of the human race. But with DOIs. But if they were, if they were, when somebody publishes a paper, that's usually the end of life in terms of how the paper will change. Usually the publication of a paper is the final step. There's nothing beyond the publication of a paper. Compare that to open source. When you publish your project, there's the dot somewhere in Russia, when you publish an open source project, that's when its life begins. And this is something that we at Calco want to bring into science. We want it to be okay to send a typo correction to one of the leading experts in neuroscience. That should be fine. If I look at my open source contributions, a lot of them is like hey, there's a typo in your reading and then the person says thank you. And it's amazing. We should do that in science. And this is something that we at Calco at least hope to build the tools for. Because it's also a social problem. And we want the world to look like this. Connected. So the world of science, not just like dots, isolated dots on the globe. But connected dots. Things that people improving other people's paper. Even if that's just a single line or a typo, or just like oh, you misspelled your name. Or some silly thing like that. It should be a collaborative effort that doesn't stop with just an isolated publication somewhere in Russia. That's it. And this is just a quote from Marcus which I think resonates with how open source works. Why do we contribute to open source projects? Those of us who do. What drives you to make that typo change? What drives you to update the package JSON of some module somewhere? I think you get something from it and we're made to help each other. That seems weird, but if you do a lot of it, you kind of get into that mode of helping others. And it's really nice. If we could do that in science, that'd be great. And that's really the end. Thanks. Datasets do not come kind of due to help you with that? We don't. We only query the Crossref API right now. And then Crossref's API doesn't include Datasets, obviously. But I mean, it's not technologically it's not a big issue. The bigger issue is what is the usefulness of citing Datasets and what can we do to improve that? Yeah. We can do it. Technologically, it's possible. One of these PubSuite documents actually written now to get a window or something so that if I send a correction to your email where does that come out? Or is it running a PubSuite instance for you somewhere? So this is, so PubSuite would be, all of our code is hosted on GitLab and our self-hosted GitLab instance. But a PubSuite document right now is an XML that Substance produces and it's not something that I would consider interchangeable with other platforms. But just two weeks ago Substance worked on a Jax editor and those of you who know Jax is a format that's used widely in science and when we have the Jax editor the next step is is very simple. If you put that anywhere even if it's Git or it's wherever people can then edit that and that's the interchange format that we think will support at least initially though we have other ideas like the HTML is one of them and so forth. But right now Substance has a working Jax editor which will include as a component in this. A condensed question maybe but why didn't you just build a WordPress plugin? And I don't need to like create your work and to sound like why did you but yeah I wanted to elaborate on because it takes a lot of effort to get people to do platform or they don't have a blogging as other platforms I think it's a very different group of people that we're addressing WordPress is a great CMS system but it will be hard to get other people to contribute to WordPress. We could start with WordPress we could make a plugin for WordPress but the inertia that WordPress has or the I wouldn't not inertia I would call it a rebounding force that it has when people talk about WordPress is something we consider the barrier if you want to sell people on WordPress it's a much harder sell than if you want to sell them on something that is new that has their intentions in its scope from the initial conception that is not kind of hacked on like we thought about what is useful for knowledge production and we built it into the platform from the get go so there's not work arounds that we need to do in order to support those that we think people will need built into the platform What is your audience do you much as all scholars? It's scholars yes but also people from other knowledge production industries such as book production or Wikipedia like organizations or anyone who deals with a collaborative generation of knowledge which I think is almost the entire web like one of the cool things we imagined is XKCD the comic there's an XKCD editor that you can create your own comic and if we plug that into as a component into PubSuite we can have everyone can have an XKCD comic that would be great and it's not entirely related to science it's not related to book production it's just something that generates knowledge and it's something that we'd like to support I wonder what your target audience is science bloggers they often use for workshops so how do you get them to use your tool because they have to then they publish on WordPress and would go back and have to somehow import again to WordPress so how do you manage that? So I think the back and forth between WordPress is not in our scope right now what we would be more concerned with is how do you get people out of WordPress into PubSuite and this is something we haven't handled yet actually we haven't talked to a lot of people that would be using WordPress so if you know people who are using WordPress to publish scientific content that would be great if you could connect us this is the user that we're looking for to ask them what they need Mark I spent some time writing plugins for WordPress to do things like citations but I think that's just something you can do but I think having a fresh start doing things is a way forward WordPress is a because a huge platformer can many things it's possible to do that but it's not the only way to think logging needs WordPress I think it's really short-circuit especially because like the next step will lead you short of the shortness that you have sort of little programming you have code that's executed when figures aren't generated when you publish the blog post it's just you can't get there to get through the first step but then it's just messy Oliver? I think it's not misleading just to put the login use case into the center also there is a full-fledged publishing system where editorial processes writing processes everything management it's way more complicated than just logging and logging is one easy to start on. One of the use cases for PubSuite is a journal platform you would have a component for workflow management you would have a component for editors you would put this together and form a journal platform that's basically our main use case that's maybe why I didn't mention it because this is the thing that we really want to do this is maybe a follow-up to Oliver's comment but I see a huge potential to take this to the classroom and use classrooms as a test bed especially students are submitting where they have to make references if things are automatically formatted it sort of takes away that complexity of really bad formatted submissions of papers you have like Cincilla where you can actually run our code for in a spreadsheet for people who are both from who are or just using spreadsheet I would say if there's training documentation or anything that educators could use and bring that into the classroom I think that would be a huge way to get usership that's an awesome idea I'll start that early and convince them that this is the whole way to go cool, that's a cool idea, thanks alright, thank you