 Hello everyone, I am Anikit Pathan and today we will be talking about Neuro Fedora. So, talking about Neuro Fedora, it's a special interest group in Fedora that we have been working on. So, before we talk about Neuro Fedora, so let's just talk about some other things that we have in our slide deck over here that is basically starting with, we are starting, we are going to start off with open science. So, basically like what open science is and so open science, it's very similar to what open source is, open software that is like open, basically that science should be transparent and knowledge should be accessible and shared. And it should be developed through collaborative networks and everything so everyone should have the freedom to study, modify and share the scientific material. So, basically it applies to all the scientists, academics, students, researchers, and basically like almost everyone that have access to the scientific information. So, irrespective of the social status, age, location, so that's what we call opens, that's what we refer to what open science is. So, and that's so we know what open science is now but like what is neuroscience, what is Neuro Fedora based on so neuroscience as you have guessed probably it's the study of the brain. So, we study different aspects about it that is like basically how the brain functions, how it is structured it's the anatomy of the brain and everything right. So, the chemicals inside the brain, there are a lot of chemicals that falls on their domains of oncology and biochemistry and how it processes information so that goes to the computational neuroscience bit over there. So, about the behaviors and cognition that is the behavioral and cognitive neuroscience. And that's it. So, this is what neuroscience is we study all these different aspects about the brain. So, basically now let's just talk about how we actually do that so so to do that we have a lot of tools and there's a whole pipeline and workflow. So we all start with like data analysis then we go to theory then modeling then experimental So, it's actually a very simplified diagram but it's very complex because you probably just go from data analysis to modeling directly and skip the theory been between so. I did not obviously skip that 30 but like with the with the availability of tools and softwares you can actually just go to modeling part. So it's basically a general workflow for almost like all resource based work. So, yeah, so that's it. So, what are what are the tools of the trade or in neuroscience so it's basically like we have experimental experimental softwares that we include like and I come or image viewers we have some FSL tools. We have the softwares to drive the big machines like heavy machines that do the computational work. Other than that we have some data analysis softwares as well that is that just ranges from like simple from like very simple to very complex libraries like from lumpy sci-fi and cycle cancer flow and that's like one of the basics like we talked about the neural networks but there are a lot of data analysis libraries as well. We have simulators we have simulators to model how our brain is going to work. So for that we have Neon or and these were one of the biggest the biggest computational neuroscience simulators that we have, but there are a lot more as I said. And so and for theory and now for theory and modeling, rather than that we have lots of hardware and softwares obviously required. So let's just like summarize what we I have talked about right now so we have free and open science that everyone should have the freedom to study share and modify the scientific material. Similarly, we have force that's free and open source software that is everyone should have the freedom to share study and modify software. That's it. So free and open neuroscience free and open science includes and relies heavily upon free and open software because open source software basically because if you don't have the tools and how are we going to like how are you going to share a scientific material right so that's why the force is really necessary for free and open science. So with the help of your federal year basically just want to consolidate the two moments that is a force free and open science that we have. So, regarding that we have a very, very interesting paper that we have it's a fun paper to be upon that's basically based upon like open science and open software. It discusses various aspects of why open software is required and why opens open science is also required. Basically discusses about reproducibility crisis where people are unable to reproduce the data or results. Also, it also benefits also also has like contains like why we need, why, what does what are the benefits of open code like it helps the community and people can reuse your stuff and they can build upon and improve. So your publication becomes an ad word for the code. So, yeah, if you have any questions. Okay, so I can see a few questions. So can we work on a cloud image of the spin. Oh, yes, David, you can actually so that bit is going to come in the slides in a few minutes. So, yep, I would definitely not on the cloud you can work it on work. You can for now like use it as a container. So, I will just tell you on how you can do on what ways you can use the neuro Fedora spin. Yeah. So, so then your science community is multidisciplinary. There are people from all the all like various fields like. And so, basically we are discussing why we need, why, why do we need a federal so basically only a small proportion of them are trained software developers and not all of them have the required experience to use or just create a software right. So, this is a basic flow chart on how software is distributed in general, basically we have the developers and which provide the code to the end users and the end users they use it and they provide the feedback to the developers. Right. So, this is not always true because there's some issues with the flow chart if you can see. So we are assuming that the dev the developer we are assuming that the developer knows that the end users are knowledgeable. So, we are just assuming that the end users are knowledgeable and they know how to use the tool. They know how to build it they know how to test it and how to install it. But we are also assuming that then users would be providing regular feedback to developers over here. Right. So, based upon like anecdotal evidence over here we have like software using research is not generally the best quality. Right. So, it may or may not meet the development standards that we have the code may not have an instruction set on how to install the software. The users may also suffer from dissolving dependencies and to install the software so that's a lot of things that that's a lot of issues over here like how can general how can a general audience use the highly sophisticated software that the developers are creating. Right. So, for that we, as Neuro Fedora are going to come in between, we are going to liaison between the developers and the end users over here. We will take the code from the developers just package it in a binary and then provide it to the end user directly so then they can install the tool with just one click. Right. So, on other, on the other hand, like, if the end users are providing us feedback that's great, but we will provide our feedback and give that to the upstream developers so that they can follow the best practices that they have they can test their codes. They can resolve the bugs that if we find any. So, basically what we're doing is we're basically simplifying the installation and the usage of the tools for the end users and providing a regular feedback system for the developers over here as well. Right. So, distributions like Fedora over here we are actually in a unique position because we have the infrastructure over here. We have high end servers we have multiple mirrors across the globe. We have a phone packaging guideline we go through a heavy duty review process and we test the software before packaging it for the users. So that's a good thing that that we are testing the softwares and many contributors we are from in Neuro Fedora they hear from different backgrounds and they have a lot to learn as well. So that's a good thing and for the movie also providing help to the users if they are facing any difficulty installing or using the software we are there to help them. Why, because a lot of us from the neuroscience community and they have they have the experience on how to use the tools and everything so if there's anything then obviously we can help. So, about you know Fedora so what's our primary goal it's basically to provide ready to use integrated force platform for neuroscientists. That's, yeah, that's pretty much a primary goal over here. And our secondary goal is like basically like helping improve the standards of, like, of the tools that is there, like so we are improving the standards on how to maintain them how you test them how you use them. We also help users develop their software development skills, and we are basically making neuroscience accessible to everyone like not just a specialist but not specialists over here as well. Right. So Neuro Fedora what are we offering so let's just talk about that David, I think I will be answering your question and right now right so what we as a cigar offering to the community. Right, so the company or lab that we have it's a spin of the Fedora OS that helps us enable neuroscience. Right, it has all the near tools being installed and so it's just easy to use you just have to install and start using the tools and we just like ready to play right. So it's public analysis and general productivity tools that are using computational neuroscience. The few examples of them would be the nest or the neuron simulator that we have it also has Brian so Brian is another similar that we have. It's heavily using computational neuroscience so we have all the big softwares over here, not all the big I mean like we are still in the process but yeah. It's integrated with no. So no man is like known for like the, it's very simple to use desktop environment and it's intuitive and it's simple so that's why we are integrating it with no and so that like everyone could use it. And the lab that we are offering we are also offering a computer container so for people who are using containerize environments. They can just download or pull the container from the Docker hub and just start their development on it they can deploy clusters on containers or maybe clusters and they can deploy images and clusters so that they could just launch a sort of hub for the for the people for the people who will be accessing it so for example like I work in a lab I could just like get I could just pull the container over here I could just deploy it and I could use Jupiter lab or I could provide a Jupiter lab entry point for all the people inside the lab so they can they can use nest that's the simulator, they can just access this without any difficulty like so you can also use it at podman and Docker and that's great. So I see another question that is your plan on making the image available on other registries like quay or gx here. That I'm not sure I mean like it is as I tested it is available on the Docker hub with the Fedora repository with the Fedora images, we will have to check if it is available on quay or gx here right so wait for a few minutes if anyone has any questions. Wait, so, right, so I guess I should move on. So, neuro Fedora so how can you help so I'm assuming like all of the most of the audience most of the people in the audience from technical backgrounds and they could help us with our little group that we have. So, they could help us with packaging and maintenance right so since our main work is revolves around like maintaining and maintaining packages and creating packages, creating software packages for people. Then, yeah, I mean, definitely if there are people out there that can help us with that, we would obviously be very glad to help. And other than that we have the QA test that is the testing bit that we do. Right, so packages that we provide must like they go through a very heavy duty Fedora QA process right so you are supposed to pass a review and then they have to have automated tests that we have. So, either you can help us with the reviews or you could just simply enable the updates testing repository and just test the tools and like as the updates and if you see anything that is worthy of reporting then you could report it to bugzilla and that brings us to the third point that is basically you can file bugs, right you can file bugs you can report issues that you are facing with the packages right so bugs can be reported on to the bugzilla that we have that is basically the bugzilla is basically an issue tracker that we have for Fedora packages. So, any packages that related to any issues related to the packages can be reported just add over there. Furthermore, any bugs related to the packages the users can directly go upstream and help them out so that's one of the big plus that we are working on. So, other than that you can also help us improve the user documentation so it's an important resource and you could just like help us by you can help us like by contributing or improving our documentation because there might be some human over there and it's not always perfect so yeah. And that you can help other users that we have in our communication channels so we could basically help them troubleshoot issues and if they're asking for help then you can obviously help them out. So, there are a lot there are communication channels like on the on IRC there's a lot on telegram and matrix. So, I will get on to that bit in the future slides. And the last thing that you could help us by help us by is by spreading the word because right about you can write about your federal you can share your opinions about your federal on social media. You could refer us to your friends or colleagues who are working in the neuroscience domain so that would help us help people to know that there's there's this thing known as your federal and they can actually get involved with it as well. Right. So, about you know fedora so it with our current metrics as basic are basically like as follows we will be three years old in September 2021 that's like next month. We have around 30 voluntary contributors right now. And that is a good thing. And most of them are not but we would like to definitely like to get that number up. We have to move we would love to have more contributors with us. And for software we have 190 tools packaged, and they're ready to install. And so the recent ones that we have packaged a new normal like the pioneer and all these tools are like heavily known in the neuroscience community, and a lot of people use them as well. So, it's a, it's a good thing that we are actually having a lot of people that are using them so we can get like reports and everything as well. So, other than that we have over 200 softwares in the queue already. Right. So, we have tools like even the pie, the neuro minora, the fly brain lab, the gen tools. Right. So, these are all again the heavily known softwares and would love to have help from a volunteers on this right. So, how can you reach reach out to us and we have basically a mailing list that is neurosecatalyst.fedoraproject.org. We have an IRC channel on liberal chat that's Fedora Neuro. We have a telegram channel as well. And that's the link over here t.me slash neuro Fedora. So currently the IRC and the telegram channel bot is not working the link about that. So, we usually hang on around the bot as a channel so we would be able to help you on either one of them. Our documentation is available at neuro.fedoraproject.org. Please do visit that there are lots of users over there you would love to go through what all softwares we have offer and how you can contribute. We have a blog over here neuroblog.fedoraproject.org and the published regular updates about our project like what we are doing and what is the current state of the whole SIG so it's all over there as well. And our pager IO that is the first getforge that we have it's over there neuro SIG slash neuro Fedora right so that's a small group neuro SIG so you can just search it over there. And that is the end of slides if we have any more questions any comments or anything we we would definitely love to have them over here. If there's something I could clarify upon. I'd like to just say thank you for attending the session it was amazing. Thanks a lot for all the people who attended the session. It was really great having you all. And if there's any question that we can answer. Please do let us know. Thank you. Thank you David. Thank you Shane for attending. If you have any questions regarding neuro Fedora if you want to ask anything literally anything if what are what are our plans for future. How are we going about with the containers and with anything literally please feel free to ask. You will be taking the containers as well. You want deep learning as well. Don't be all. That would be amazing like having a deep learning image as well. I am pretty sure the science team the science or the robotics know there was some Sega forward the name show definitely David I would love to talk to you as well regarding this. Yes the machine learning sick my bad yes the machine learning Sega is actually working upon something like they were trying to packet some open source tools that are used in for GPU computing and everything. If I remember correctly there were some tools known as hip and everything basically for AMD graphics and everything so that would be cool to work upon as well. Yes, exactly. Yes, there's a machine learning Sega as well David that you can probably look up. It's similar to what we are doing over here. They're also trying to import all the tools are not packaged for Linux and for Fedora basically for Fedora they're trying to create a they're trying to package everything as well. So I guess that's the end of the session. If anyone has any questions or anything. Please. Obviously I mean. No worries like us. We would wish we would have done the same if I had known this that we could. Oh wait I skip on to the chat plan so they're so talking about deep learning is they can be issues with CUDA packages right so any heads up on how and if okay so basically our cars deep so we don't we don't necessarily use packages that use we don't actually use neuroscience we don't specifically use packages or to software to do that. We don't have any tools that rely on the GPU right so we don't have a hard dependency upon the CUDA libraries as well right and as uncle said it's a no go because the drivers are not open source right so that's another issue that we have. On the other hand, as I said, about the machine learning so they were they were trying on packaging the hip libraries that could work with the AMD graphics card series that they have but then again it's not optimized as the Nvidia one size so. Again, it's a big issue that we have. But we could probably clear the code poll on like what desktop environment do you use because I was talking about normally over here but I don't specifically use normal I use I use the plasma desktop so we could probably have a neuro Fedora. A spin off neuro Fedora with the plasma desktop so that makes things more amazing. What do you guys feel about it, please, if you have any comments on please do leave them at the chat. See, yeah, I can't agree is that could be fun. I'm talking about the plasma desktop, even then, even a desktop with I three, that would be amazing. So that's worth looking into as well. So I guess that's the time. It's my 25 minutes are done. So if you guys have any other sessions to go do. As I caused the recommended the session that I'm the WME session. I don't know what that is. I'll jump into it as well. So please do it and it. All right, so I'll leave the room. Thank you all for attending and it's the websites and applications we would love to attend it. Thanks a lot everyone for attending and that is time so I'll have to close the room as well. See you. Bye bye.