 to see just how much interest there is. Andrew, do you wanna kick us off? Okay, I've just hit record. I'm assuming you've seen a prompt saying I've hit record. Good. All right, so thank you for joining us for this webinar. As Tom mentioned, had I known Tom was gonna say all of that, I would have hit record 30 seconds earlier, but there we go. This is an opportunity to hear about open source program officers from Said Chowdhury, who is here, well, for a variety of reasons, but in part because he's director of the open source program's office at Carnegie Mellon. I've known Said for 20 years now, I reckon. We first bumped into each other in the institutional repository space when David Gronowhegan was also involved. David seems to have vanished, but never mind. So Said has been working in the intersection between scholarship and technology for a very long time and brings deep experience and deep knowledge to bear on the space. But in this particular context is most relevant because of his direct work with open source program officers. I've given Said a bit of context so he understands where we're coming from. We've got 11 participants on the webinar, so I'm very happy to keep it fairly interactive, but maybe if we give Said a chance to provide some introductory material and then we can dive into a conversation about particular issues. Over to you, Said. Thank you, Andrew. Thank you for the opportunity to be here. It's great to chat with colleagues in Australia. Good to see Andrew and Tom and David and some new faces. So hopefully we get to know you a little bit better as well. I do hope this is interactive. The last thing I want to do is talk at you for 45, 50 minutes. So please do jump in with questions either raising your hand or going off audio or in chat as I move along and I'll do my best to address them. As Andrew said, I think some introductory or contextual remarks may be useful in terms of the Ospo landscape in the United States, how it's come to be and the questions around open source software. So I often begin by mentioning thinking about all software, not just open source, depending on which recent reports you believe, anywhere between 95 to 98% of all software is reliant on open source software in some way or another, either consuming open source modules, for example, or being released as open source of software or both. So when a lot of people come to me and say, I don't know if my university's producing open source software, using open source software, I take a step back and say, does your organization have any software whatsoever? And if it does, chances are there's at least open source embedded in it or it can potentially be released as open source. And that tends to make people think, oh wait, yes, of course we're producing software or using software in some way or another. Another important dimension to that is the assertion that software, open source software is a primary research object. So a lot of the policies that have been developed in the United States of the last decade or so, particularly at the federal government level, which of course influences the funders both that are on private, have moved in some sense from articles to data to software. So the first set of public access, my most of you wish to call them that focused on open access to articles and then moved progressively toward the data that are cited by those articles or that are necessary for re-producibility and so on. And then now are getting around to thinking about software itself. So the Office of Science and Technology Policy in the White House issued a memo last year that is an update to the public access memo that came out over a decade ago. And I spoke with one of the authors of that memo and he said in the earlier versions they had mentioned open source software specifically but getting feedback from some of the funders and some of the universities, everybody said, yes, of course it's important, but we are not ready. We do not know what an appropriate set of policies would be around open source software. We don't know what the appropriate infrastructure might be. We don't know what those social norms are and so on and so on. And the last thing we wanna do is create unfunded mandates or set up policies that may get difficult for funders and universities to respond. And by the way, can you help? Can you help us in fact to inform the update to this policy as we think about open source software? So at least in the U.S. there's recognition of its importance, its role in supporting open science more broadly and there's an understanding that universities funders need some time and some dedicated effort to become better organized around it and that an OSPO or an open source program office or a network of them can be a wonderful way to do that. One description I've used for an OSPO is an organizational API. No matter how large or small the organization might be, if someone is interested within your organization about knowing what kind of open source activity it's happening, where does one go to ask that question? And for those of anyone who's outside of your organization that wishes to interact with your organization, again, can the OSPO become that organizational interface that allows them not to have to deal with all the complexities of the internal workings of your university, your museum or whatever it might be, but rather just say, I'm interested in open source, can you help me navigate that landscape? And the last thing I'll say about context and background is I've only been at Carnegie Mellon since last August. Andrew and I met when I was at Johns Hopkins where I worked for many, many years. And I was a member, I led the library team that supported the Johns Hopkins Global dashboard, COVID-19 dashboard. And it is cited rightly so as a real success story and celebration of open science. Proud to say that I was the initially lone voice that said the data should be licensed using a creative license. And some people said, what is that? But we moved fairly quickly toward using that kind of license, which I think was important. But I've also heard from a lot of the medical researchers at Hopkins in terms of the vaccine development that it was still difficult to share data with each other. There were still patient records and so on, even if the federal government said everybody be open and access to articles was made more freely. It still wasn't easy to share data. But by sharing code, they were able to infer what kinds of data they were using and what kind of analysis was being used. So it was a clever and more rapid kind of workaround from trying to share the data. And I say that to point out that for all the things I've talked about in terms of the federal landscape and universities thinking about research objects. Open science is a good thing to do period, right? And what we've heard is, well, of course, we wanna do that for COVID. But what we're also hearing now is, well, we wanna do that for cancer and heart disease and diabetes and all sorts of other medical issues. We wanna do that for climate change, recruiting security or electrification, de-governmentization. Why can't universities do this with all these other global problems that we have yet to address? And I don't think universities at least in the US will have the luggage saying, well, that was COVID. And we did all that for COVID, but we don't really want to be open for all these other types of things. So I think this is a one-way street. We will find more and more funders assisting on this, more and more policies around this. And I think that open source software is a key component in any of these kinds of conversations moving forward. So let me take a pause there in terms of context background. I obviously don't know the Australian context background certainly nearly as any of you. So be happy to take questions or perhaps, hear your comments about that before I move more into sort of hospital, setting up an hospital. Maybe I can comment to kick us off. I don't think that we have a stronger policy environment, but it's always important to remember, I think with this particular problem. There's a few different types of policy and it seems unlikely to me that a kind of mandating particular behavior, that kind of policy is going to happen in the Australian landscape all that soon. But there is a type of policy that captures norms as they emerge. And the important thing to keep in the back of my mind as we do this is that this is already happening. Software production and use as part of research is huge. It's enormous and it's not visible, but it is happening. And really where the change actually is, is an understanding characterizing and amplifying the behavior that's already happening basically. And that is a precursor to the emergence of normative policy or an environment that supports it. So, yeah. Okay, that's helpful to know. Just picking up on the point about norms, one other observation I've made is that the norms around open source software are much more conducive to openness than they are around data. Or certainly around articles, if you think about the four publishers in the world that they play in that ecosystem. Faculty members, again, at least in the US, are quite comfortable putting their code out on GitHub or GitLab or take your pick. They're quite comfortable basically saying, sure, I would like people to see the code and to adopt it and to use it. That isn't true in a way around data. Sometimes for very good reasons. And certainly isn't true around articles, if you sign obviously the publisher agreements that the four proper publishers wish for. So, the other aspect of thinking about things like reproducibility, which is a key aspect of the policy development in the US, the code is critical for the reproducibility. I can read an article and basically gain no insight into how to reproduce your results. Even if you give me the data, it's probably not enough. So having the code is really key for the reproducibility aspect, but it also lends itself to the kinds of norms that make it easier to share. And to distribute and manage the software in a collective way. So let me shift a little bit to the OSPO concept, which is something that's actually existed in the private sector for many years, at least a decade, depending on who you ask. So we're learning and borrowing in some sense from that context. I've been very fortunate to work directly with some of the people who started earlier and even more recent OSPO's in the corporate setting. But even in a very early set of conversations, I made it clear to them that whatever work in the corporate environment isn't going to work in the university or academic sector. We don't have a specific set of licenses that we can compel faculty members to use. We don't tell faculty members how to produce their software. It's just a very different environment. And for most people, that was clear, but as we started going down this path when they literally said things like, well, what do you mean? What do you mean you can't tell them? This is the DevOps practice that they have to use. Well, what do you mean you can't say these are the licenses? I said, you just can't trust me. It would be a very short unpleasant conversation if you try to tell somebody with 10-year academic freedom whatever term you want to use that this is the way you will do it. So very quickly shifted to best practices, right? It's that the private sector has developed a set of ways in which they develop software. And specifically toward more industrial scale usage, so large numbers of people and to do so in a way that fosters this sort of community development and use. One person I've worked with for about three and a half years at this point, Steven Wally at Microsoft, he's in the Azure office. He has this wonderful metaphor to cooking. He says, if you cook for yourself that should be relatively easy. As soon as you add one other person, you at least have to know what that person likes or their dietary preferences and so on. If you have a dinner for a friend's family, same thing. You should probably ask, but you're starting to now harmonize the menu in a way that isn't specifically for one person at a time. If you're a caterer, you don't go to every person at a hundred person event and say, what would you like? How would you like me to prepare this just for you? And then if you're preparing food for a thousand people or 10,000 people, you certainly don't do that. So what are the metaphors in terms of software development? And if you've ever been involved in software growth, you've probably heard that phrase, it works on my laptop, right? What do you mean it doesn't work when you put it onto the production server? What do you mean when a thousand people use it, it didn't work? So again, not going to researchers, academics can say, you must do things this way, but rather, are you interested in taking this software to this more industrial scale? Are you interested in other people being able to interact with the code that you produce? The answer may be no. There is a large amount of software produced according to Mellon that is used to train students that is used to experiment on something that should never see the light of day. That's totally fine. But there are some subsets where the faculty actually say, yes, I do, I actually do want this to be used, I want this to be adopted. And if I can do things further upstream to make that more minimal, then sure, I'm willing to listen, I'm willing to work with you. So the OSPO's fundamental sort of mapping of that corporate role into the academic sector we can help advise you around best practices, we can connect you to tools that are being used by thousands of developers throughout the world that will make it easier once you're ready, should you be ready to share that software, it'll be easier to make that happen. Same thing around license choices. So we do work very closely with what's called our technology transfer office in the US, that is the commercialization type of thing in US universities. And I will tell you both at Hopkins and at Carnegie Mellon, the Tech Transfer Office immediately came to me and said, are you willing to tell everyone to produce everything open source and what licenses they're going to use? The answer is no. I don't feel the need to tell everyone should do everything open source software. That's a choice that someone needs to make, particularly because it involves some effort, right? So telling someone to do this and then having telling them that all these other things you need to do, it's not a good recipe. So working with Tech Transfer, what we've basically argued is ultimately what we're about is impact. And I'm happy to say that the head of Tech Transfer at CMU, when I kept saying translation, which is the term US universities use to talk about research being translated beyond the Wilson University, I kept saying commercialization is only one form of translation, right? And he finally interrupted me and said, I don't actually use the word translation, I use the word dissemination, right? That is ultimately the goal of the Tech Transfer Office. It's a different kind of dissemination, but open source software is a perfectly valid kind of dissemination as well. So I'm not saying this is the case at all US universities, but it's a good discussion to have with a Tech Transfer Office is that we're not here to compete with the things you're asked to do, the metrics you're asked to work toward, but rather here's a different set of metrics that relate to open source software and its impact. And mind you, this is not all just altruistic or sort of high principle or however you wanna think about it. One conversation I had within CMU, the president of CMU was present when the CHIPS Act was signed at the White House. This is a major act of funding in the United States to focus on next generation manufacturing, semiconductors, so on. And the point I made to him was, Carnegie Mellon can produce software in this realm that's relevant and we can spin off a company and make some money. Or we can produce open source software that impacts the entire sector of manufacturing for city semiconductors or metal refining or whatever you wish to call it. I had this phrase that I didn't really think it was that useful, what people seem to think it is, is that if data is the new oil then software is the refinery, right? So if you produce software that an entire sector of the United States, maybe even the world uses, isn't that valuable to Carnegie Mellon as well? And he said, well, of course it is. That's the kind of thing I would take to a funder. That's the kind of thing I would take to the representative we have in Congress and the Senate and so on. So it's a different kind of conversation about impact, social impact and so on. So when that kind of conversation happens you can start to see this inflection towards, oh, okay, maybe we should manage our software more intentionally and more strategically. And curiously enough, what I've found is that faculty are actually quite open to this. I think many faculty have understood that they don't know a lot about private sector industrial software development. Some do. I met a couple of faculty members early on in Carnegie Mellon who had taken their software through the Apache Foundation process. They don't need any help from me, right? They don't need to talk to me. They know what they're doing. They went to Apache, figured that out. That's wonderful. But they are the exception. They're not the norm. The vast majority of even other places that Carnegie Mellon faculty will say, I don't actually know, right? If I've run this, you know, developed the software using this practice, I don't actually know the best place to deposit it. I don't know the best license choices and so on. So it's fundamentally about a consultation in terms of what are your ultimate goals, right? Where are you trying to get? And what early decisions can we make that sort of make that easier or make it harder or maybe open up pathways you didn't realize or constrained pathways you didn't realize? So one interesting nuance that I've learned recently is people who want to open up their software even though it's been closed or vice versa, something that's closed and open it up. license choices have profound implications for your ability to do that. And they don't know that. They just think, why can't I open this up? Or why can't I close this? So having those kinds of conversations I think has been really critical to helping them to think of the auspice as center of competency, right? Not a center of excellence. That's a somewhat loaded term and I don't necessarily believe the auspice is going to be a center of excellence. But a center of competency implies you may have questions, we may be able to answer them directly or we can connect you to other people who can answer those questions directly or we can help you find resources that can be absolutely questions. So it's very much a, I'd like to do the following thing. Can you help me? And then just leave the choices to me. As long as I understand the choices and the implications, that's fine. So help me understand the implications so I can make better informed choices. Now one related aspect to this is the organizational home for an auspice. I've been at two institutions where it's based out of the libraries. So I have that bias, I guess, if you wanna call it that. But the reason I have that is it's fairly rare in my opinion and my experience that a faculty member will show up and say, I need help with software only. And I don't wanna talk to you about the data or the articles or the open science or any of those other things or mentoring my students or so on. Typically it's, I'm trying to advance my research or I'm trying to teach a certain way and software is a piece of that. So libraries in the U.S. have developed data management services, have broken relationship with publishers around open access or it can provide consultation about looking at author claims and so on. So I've made the argument that aligning that software kind of consultation with existing consultation around data or articles, best practices for science is a natural alignment. That doesn't have to be the case. There's another round of funding. So the first six auspices in the U.S. have been funded by the Apple Creek Sloan Foundation and they are going to fund another round next month. I don't know how many but let's just say for the sake of argument, they double it. My understanding is they are trying to look at some diversity or because national models as well. So not necessarily looking for all of them to be in libraries, all of them being computer science or research computing. It's early days for university auspices. So there's not a blueprint. There are guides. There are recommendations and there are observations but there's no, this is the way it needs to be done. So again, let me take a pause there and see if there are questions or comments. So I'm interested in your last comments there about the organizational position, I guess. That's what you're saying. What sort of relationship do you have or do you see having with the university's central IT function? Because I guess one of the things I guess I've seen a lot more in recent years is university IT is increasingly risk averse for all sorts of good reasons, right? Because lots of open doors into the university create lots of problems, right? So the things that make them particularly nervous is open doors that they don't understand. And also I think in my time of universities a push towards their main purpose in life is keeping the corporate governance systems going, right? So if it comes down to the email server's gonna work or Professor X's little bit of code is gonna work they're gonna focus on the email server, right? Because if the email goes down the vice chancellor hears about it or the president hears about it. If Professor X goes down, yeah, that's a problem but unless the professor has a really good line to the president, the president's probably gonna hear about. So I'm sort of curious about what the relationship you see there from in that it is and whether they have that works for you and cause nervousness. Yeah, yeah. So it's interesting to know that this is one way in which the US and Austria are similar. So I do think there's another key relationship to have for an hospital is with the Central IT Group that will vary depending on the nature of the institution. I used to be at Hopkins which has a very significant medical and public health enterprise which gone email and does not. So that has an impact in terms of the culture in terms of what Central IT is worried about and so on. There are some broad general things you're right. There's an increase in emphasis on security, on compliance and all sorts of things that you know, CIOs have to think about. I actually remember talking to the CIO at Hopkins in her NCI at the time and your job is really difficult. He goes, yes, it really is. So I'm sensitive to this. I'm not trying to be unfair in any way but I do agree with your general characterization that Central IT in some ways has had the proportion of academic kind of focus and interest and more corporate administrative because of Hopkins medical types of concerns has grown over time, right? And I think that's just a reality. I think that's a fair statement. So in that context, it's not to say a Central IT group couldn't be the organization owned for Ospo but I think my own opinion is having an Ospo outside of the Central IT group that works closely with them is a better model and the way that, what are some of the ways that we can work together? So security is one aspect of that. I don't think a lot of academics are thinking about security throughout their software development process. I'll just put it that way. Whether they do it at all or not, it's a fair question. So a real conversation with a faculty member recently where I said, are you aware of the US NIST, National Institute Standards and Technology Framework for Secure Software Development? He said, I have no idea what you're talking about. Being able to say these things exist, you may want to look at them, you may want to consider them, you may want to think about the implications before you release your software to the world. And you do so without a license, by the way. It's very interesting implications on you if something goes wrong. So Central IT groups tend to be happy when they hear, oh look, there's another group of the university that's reminding the faculty that this is an important issue and this is something we need to think about. I'll also say that, again, I don't know if this is common in Australia, but in the US, some Central IT groups also run the research computing groups within the university. So that's another area where I think there's a lot of potential common interest and overlap between an OSPO and the Central IT group. And we're finding that this sort of rise of research software engineering as a role in the US is becoming a much more prominent type of development. And I will say that I've spoken to the deputy CAO at CMU and had a very good conversation about what kind of services are we going to provide around research computing and open source computing and do that together. And where we initially spoke about it, and I'm slated to speak to a group in July about this, the research computing advisory committee, is the library is a good place to do sort of narrow, thin, large numbers of engagement, right? So training around Python, R, Git, so on, where undergrads, grads who know nothing about it can come to the library and learn. As you go sort of deeper down, or I said it was a set of concentric circles, right? Is as you get to these outer circles and Carnegie Mellon has a relationship with the University of Pittsburgh Supercomputing Center. Supercomputing Center saying, hey, we want to develop these open source tools for orchestration of the workflows here. That's not something the OSPO is going to do. So as you get to these very specialized kind of needs, then it's probably more appropriate for a research computing central IT someone to say, we're going to develop that, but in concert with the OSPO, right? And they're looking to us for help around things like Git training and so on. They don't need to reinvent that wheel. So it is very much trying to find at least some common points of intersection. And again, trying to make it clear that we're all trying to move in the same direction. We just have different levers and mechanisms to do so. Yeah. Oh, sorry. Just as I agree with you absolutely on that model of where the libraries will stops, because I think there's a point at which you say, well, we're good at volume, but we're never going to have the expertise for that really complicated stuff. And it becomes very messy very quickly. But there's a meta point there as well. The host within, if it's a university, and that's a big if, because we're also considering organizations that are not universities, the nature of the service or the capability is going to match what's actually possible with that unit as well. Sorry, Nick, I saw you had your hand up though. Yeah, I'd be interested to know how these OSPOs are progressing in the UK because they are much more advanced and there's much greater deployment of research software engineering groups within the universities and where, so does it align more in the UK with those groups rather than the library? Yeah, it's a good question. And, you know, just probably not surprising, I don't know the UK landscape nearly as well. I do have had some conversations with people at a couple of universities and one of them being leads and with someone in JISC, you know, which is sort of the shared services group. I don't know if that's the right way of describing it in the UK. And I think that they're interested in thinking about OSPOs. My sense is the US is definitely a place where a few universities are going to throw something at the wall and see how it goes. And then everyone will learn from that and we'll come up with an approach. Whereas I think the UK and my limited understanding tends to be more let's plan this, let's design it, and let's roll out a national type of service. I don't think they're looking for each individual university to go off and experiment and prod around this. I don't think it would stop them and the people in these actually said, you know, if you're ever in the UK, please let us know. We'd love to hear you. Have you talked to our faculty in here and so on. So there's interest, but I suspect that interest will probably bubble up and just might say, let's create a national type of an open source programs office or a network of that. Which again, I think that's a perfectly fine model. It's impossible to do in the United States. I remind everyone, the US is made up of 50 sovereign entities who increasingly don't agree on a lot of things. So it will be very hard to say, here's a national, you know what? So universities are a good place to do that kind of experimentation in the US, but that might not be true in the UK. So I see a question in the chat. I'm sorry, was there more? No, that's right. I was just gonna direct you to Kathy. If Kathy, you wanna take the microphone? Sure, Tom, thank you. And thank you so much for the presentation. To date, I'm getting a lot out of it. I wanna be clear here. I'm here in a personal capacity. I'm not representing the Institute for Long Involvement with Open Source in Australia. One of the key challenges that I see particularly for a country which has a very small research footprint, we only have 40 universities. We have a scattering of research institutions. There are several efforts in this space already in the Australian landscape. We have carpentries. We have the open data, GovHack movement. We have several sort of HPC communities of practice that are doing excellent things. My, I guess the challenge that I see here is an OSPO is another actor in this ecosystem. And how might that streamline rather than sort of duplicate or proliferate some of the efforts? Yeah, it's a very good question. I think it comes back to that first, or one of the first points I made about the organizational API in that if you're an individual in any one of these institutions that are being represented here and you have to figure out how to deal with your open source software, you'll do your best faith effort and you'll go off and do it. But if one talks to, hopefully if one talks to an OSPO you can hear, oh, you don't need to take two weeks of your class teaching your students about Python. There are these carpentries that will help you learn about Python. There are all these other resources you can use rather than doing it all yourself. And maybe this is the US mindset, but most researchers will poke around or try to find some help. And then eventually conclude, I'll just do it myself. I'll just figure it out. I'll help the students do this. So fundamentally, what we're hoping people start to realize is the OSPO. And then, you know, I guess I'm, you know, I said it was last August, I'm coming up on a year, Tony Mellon. But I'm already encouraged to see a number of faculty who reached out and said, hey, I heard about this OSPO thing and we've been using open source and can you help me? And, you know, the answer is always yes, we'll do something to help you. But again, the answer may very simply be this is external research you really shouldn't go look at. You don't need to do it on the ground. And quite frankly, neither do we, right? But I will also say that it's not one directional or sort of passive. One thing Cardinal Mellon Ospo is doing is working with somebody in the private sector in the U.S. named Duane O'Brien is a very well-respected leader in the open source tech world, at least in the U.S. About looking at the carpentry's current training for Git and evaluating it and either seeing are there things we can add to it or is there a second, you know, stage or training and so on. So we're also contributing back, right? We're hearing that students take this and they don't necessarily feel like they can actually work with it. You know, how do we improve that? So I think it's a two-way exchange. And I think particularly a network of Ospos could be a really powerful two-way exchange. It's one thing for CMU faculty or students to say we're struggling with this. It's another for 10 universities and you have to say we're struggling with this. And then I'll speak for the carpentries or any organization for that matter outside of CMU. But I would think they would be much more compelled by hearing, oh, 10 of you care about this. Sure, let's do something together. Saeed, I've just noticed we've got 15 minutes left. Did you have other slides you wanted to present or are you just happy to let the conversation run? I'm happy to let the conversation run. It really would be best to hear from all of you about what would be most useful. Yeah, I was hoping to kind of... It's a little bit Dorothy Dixer. Actually, that's a very Australian term. It's a question in parliament from the same side of the government. But... Please tell us about your latest wonderful initiative at length, that kind of thing. So to do one of those, I wanted to go a little bit from the strategic and tactical down to operational, to give people a sense in the room. For instance, how many people are usually located in an Ospo? And bearing in mind, it would be different at a university versus, say, some of the other institutions that are coming here today. Sure, I can tell you that in the corporate landscape, the number I've typically heard of is six to eight. But I will tell you that many of those are technical roles so that there are individuals who will work with code directly, who will set up software infrastructure and maintain it and support it. That hasn't been the role to date of the Ospos within universities. So I'm not saying that could not eventually become the case, but to David's earlier question, Central IT might say, why are you going off and building this other kind of team to do these things, we already have people and so on. So not to rule it out, but I don't think that that's another case where I don't think the corporate Ospo maps directly to the university one. What I will say is in these early stages, we're finding that the Ospos typically have tended to be one to two people. Now, we're all very fortunate in that we're getting funding from the Sloan Foundation, at least initially, as I'm sure is true with any funder anywhere in the world. They are also asking, what's your long-term plan? And the answer should not be you keep coming back to us, I'll break you down for money. So part of these proposals, part of the planning, I'm happy to say as of July 1, I'm being moved off of soft money onto combination of soft money and hard money at Carnegie Mellon. So this is the trajectory that the Sloans of the world are hoping for, that you prove there's value here and that your institution will step up and start to support them. The roles have tended to be sort of that consultative advocacy, connection type of roles. We aren't going in and propping up student projects, although we are supporting hackathons, right? If we wanted to get into that direction, then that would obviously entail more people. But whether that's people existing already within the organization or it's new people, that's yet to be determined. I think a good metaphor I can give you is that in the US academic landscape, the data management services groups, they span all the way from multiple individuals who will actually take data and curate it and add metadata and then process it, so on, to single individuals or maybe even a part of an individual who does nothing but the consultative piece because they simply don't have bandwidth or more. I suspect we'll see something similar with hospitals in the US, that there will be some full-fledged ones that have multiple people and then there will be some universities where it just isn't necessary to have that kind of full team. And they might just say, we just need someone who is keeping an eye on the open-source landscape, can let us know what's happening and can sort of see what's happening with the institution and do that mapping. Got another hand up. Yes, I just wanted to ask, and it's probably a question for everyone else. It's more about funding given that Australia doesn't really have philanthropic donors or anything like that. How would we look at being able to fund something, especially in the library space where budgets are quite thin? Funding one or two people can be quite an expensive task for something that may not seem in the IT space for IT departments to pay for because they're very open-source adverse. So how is everyone going to be funding this if it was just set it up? Perhaps, you know, as the code designer of a scheme to, in our words, co-invest into this space, I can jump in there. We're not quite doing it at the scale or in the way that the Sloan Foundation is doing it. We want to kind of ease our way into this and we also want to consider a broader array of partners than the Sloan Foundation has focused on universities, I believe, entirely in doing this, but OSPOs are actually present in public sector organizations and we're specifically interested in medical research institutes, glam sector, so that's galleries, libraries, archives, museums, producing software, and increased facilities in this round as well as universities and really will be guided by where the weight of the interest is, but we see value in proceeding through the initial steps as laid out in the guide written by Sayed and also in forming a network of people noting again that Sayed said that part of the real value in this is being able to talk to other OSPOs about what they're facing and it means that you distribute the knowledge and the experience around that. And so we are looking for people who are interested in taking on this initial work and to make that a little bit easier, we're offering a co-investment in that effort to reduce the cost. And in order to do that, I might just jump in with the bit that I would have wrapped up with. We are seeking expressions of interest in order to gauge whether there is sufficient level of interest out there. The expressions of interest are due on the 26th, which is Monday by midday and it is a very low bar, let me say, for an expression of interest. It is the digital version of raising your hand. So we do need people to raise their hand digitally, please, in order to proceed. Yeah, and if I can just add, I was involved in setting up one of the first data management services in the US University of Hopkins and I will say that it took a lot longer to persuade the researchers to include the data management services in their grant proposals than it has with software. That has happened, I've been included in three different proposals at this point to represent the Ospo, include the Ospo in their proposals. So there's a much quicker path I've found where researchers are saying, I need this involved in what I'm doing and I'm willing to put in some funding into my grant application to support it. And that's actually the signal that led Carnegie Mellon to say, well, okay, I mean, that's great, but number one, you can't be involved in 50 different proposals, that's simply not reasonable. So we're taking this as a signal that there's value in this and that the researchers need it and we need to sort of free you from having to sort of engage with all these different proposals one at a time. I will also mention something else within the expression of interest mechanism, sorry, request is that we're willing to consider possible partnerships in achieving this as well. So if resourcing is a problem, there are some structures out there where this does make sense. There are alliances of universities, there are clusters of increased facilities, there are peak bodies that sit over the top of Glam Sector Institutions that might be a better place to host this out of. And if you have ideas that you'd like to put forward, we'd love to hear them. David, go for it. Just follow up on what Sayed said a minute ago about the sort of sense that he has that it's gotten some traction more quickly than data management stuff. Do I, I guess when you're establishing this, how much overlap has there been with your existing data management services? So are you able to build on that brand, I guess, by having some of the same people or, you know, having that existing, here are some nice people who helped you before. Here's another service that's got a slightly different name. You think that's been, that's an important factor? Definitely. So Carnegie Mellon Libraries has spent over a decade thinking about open science and has built a fairly substantial program around it. I know my colleagues would say, we need more people and we need more capacity. But I think it's still a fairly impressive set of both tools, services and capacity. The open source software services is considered part of that, right? So faculty at CMU already started to think about what I care about open science. I know the library has some help. Maybe I don't know exactly what it is, but there's some help. And now they hear, do you have software as part of what you're doing? Yes, oh, there's this other group you should talk to as well. You know, from a research perspective, there isn't a hard boundary between data and software, right? It's not, again, getting back to that day of the life of an academic. It's not like they wake up and say, today I will only deal with my data and tomorrow I will only deal with my software. So the fact that that line isn't so sharp and clearly delineated helps. They just need help moving forward with whatever they're doing. And it probably includes data and software. But another factor is not tied to any particular university. And that is that there are a set of canonical licenses for open source software, right? The open source initiative manages a set of canonical licenses that everybody understands, maybe not everybody understands, but everybody knows. And you can ask someone to interpret what it means from a liability perspective, from a sharing perspective, from a commercial perspective and so on. And they're sort of organized into permissive and not permissive and so on. We don't have that in the United States for data, right? We just don't have that for data. So I was involved in very early conversations around data and there was this, this is obviously my own editorial view. There was a bizarre dance where the funders kept saying universities, you go figure out your data policies. And universities said funders, go tell us what the data policies are. And nobody put the proverbial stake in the ground and said, this is the data sharing policy. This is the data set of licenses we need to use. But that conversation I had with the Office of Science and Technology Policy, one of the first things I said to them was whatever you do, point to the OSI licenses, do not let universities go off and develop their own open source licenses. So I think that makes a huge difference, right? That we have that framework in place, it's tied to the fluid boundary between data and software. And it comes back to that question of norms, right? Faculty seem much more willing to get help or share a code in a way that's not true in data. I think it's worth commenting too that data sharing or developing a culture around data sharing is actually relatively young, but really actually the movement of open source software is much more mature. And the nature of the object being considered is also different. I mean, they are different objects under the eyes of copyright law in Australia and the US. And so one thing to be cognizant of as you leverage what you understand about data sharing and how that fits into a university is that you're coming from a different set of cultural norms that have emerged under this open source software banner. I think actually, I'm sorry, Myra, but we're actually pretty much on time. I'm okay with staying past the hour if others are as well. So I know people might leave, but if there's a question, I'm happy to address it even. All right, good. Well, we might do that, but I would also give permission to people if they want to leave at this point. This is being recorded. The recording will be available. There is no FOMO to be considered here. We won't hide the link to the recording, I promise. Yeah, sure. Thanks very much. Just very quickly, my question is about what was put into the chat by Kathy Reid. So it's a quote, software defined era. I mean, it's this question for anyone. So what's the reference for it and can you please provide the context? Sure, I'm happy to take that question. And firstly, I apologize for it not being clear and not having a lot of context around it. So my argument is that we're in the era of software defined everything. So the asterisk in that comment is standing for everything. So we see things like software defined networking, software defined infrastructure. When we start to think about research engineering in the university and research institution context, we're seeing a huge growth in software supported, HPC supported research, hence the need for growing research engineering capabilities. So my comment there is basically that open source underpinsed a lot of that research software. And so in a world of software defined everything, there's an increasing need for open source capability within the research sector. Yeah, in the US people use this term critical infrastructure and software open source software in particular is mentioned everywhere in that context. Thank you. Yeah. We might wrap it up there, John, I think. That's okay. Unless there are any more questions in the room, we can go over time. Or if you do need to run off to the next meeting, then feel free to do so. I'm just filling the air by talking now in case you've got that question. And otherwise I'm going to basically launch in to say thank you very much, Said, for joining us today. It's been great to actually hear how this works and how it's worked and is working in the US. And so thank you for sharing your time with us at whatever hour it is there in the US. Presumably it's not too bad. It's not bad. Okay. Well, thank you very much. And I will have to wrap up by saying, please, the Australians on the call, the clock is ticking. We need those expressions of interest to proceed with this, even if it's a half a hand, half a digital hand, whatever that is, something between zero and one. We need that in by for the recording by the 26th of June, Monday, midday in your time zone. And in order to proceed with this activity, which would be actually that we'd be seeking proposals based on what we get in terms of expressions of interest. So please, even if you're just tentatively interested, let us know so that we can bring this into reality in Australia. Right. Well, in the case, thanks all. We will send out the recording and my thanks again to Saeed for staying up after 10 o'clock at night to talk to us. This is my pleasure. Thanks for your time. Thanks. Thanks everyone. Thanks everyone. Bye. Bye.