 My name is Stephanie Kittes, I work at the Ansmerben office and I'm also a member of the Secretary of the Research Data Alliance and I would like to introduce it to today's speaker, Mark Parsons, the Secretary-General of the Research Data Alliance. Mark, before he became the Secretary-General of the Research Data Alliance, he was the Managing Director of the US Component of RDA or the Research Data Alliance. And he also is the Director, I think, of the Rensselaar Center for the Digital Society. And prior to joining Rensselaar, Mark was a Senior Associate Scientist and Lead Project Manager at the National Snow and Ice Data Center, which sounds pretty chilly, quite cool. He was there, he implemented the overall data management process and let the data management effort for the International Polar Year 2007-2008. And we look forward to Mark talking about the Research Data Alliance. Okay, thank you and thank you everybody for calling in. So as Stephanie said, I'm Mark Parsons, I'm the Secretary-General of the Research Data Alliance and just to give you a little bit more background than what Stephanie provided. Yeah, I was at the National Snow and Ice Data Center for a long time and where I managed data ranging from remote sensing of the environment to indigenous knowledge of the Arctic to the qualities of Keraloo poop. So it was very diverse was my point. And so I'm trained as a geographer, so a lot of what I will say comes sort of from that context. So trained as a geographer, practically as a data manager, but now as an administrator. So let me just jump in. So we take it as read within the RDA that all of society's grand challenges, be it curing cancer, understanding and responding to climate change, feeding the 7 plus billion people of us that are crawling all over the planet. All of these grand challenges required diverse, sometimes large data to be shared and integrated across cultures, scales and technologies. And so by cultures, I mean national cultures, but also disciplinary cultures, cultures of stewardship versus cultures of compute, if you will, libraries versus repositories and all those different sort of cultures. Scales, I mean in terms of physical scales, temporal scales, but also conceptual scales. And then technologies, interfaces, communications, visualizations, analysis, etc. So we take that as red. To do this, to guess these grand challenges, we need to integrate data across many different things. So correspondingly, RDA's vision is researchers and innovators openly sharing data across technologies, disciplines and countries to address the grand challenges of society. Our mission, more targeted, more specific is that we build the social and technical bridges that enable open sharing of data. I think that's an important metaphor, this bridge metaphor. And I'll talk more about it as we go through, but at one level you can see that what we're talking about is building infrastructure. And when we're talking about infrastructure, I don't just mean the pipes and the wires. Infrastructure is a much more complex thing than that. And I'll argue that it's really more like a body of relationships. And so the first question we asked is what is infrastructure? But at a simple level, what we want to do is go from something that looks like this to something that looks like this. Except that we're here down under, so we want it to look like this. So what we really need are things like this. And if you think about it, that's an example of a socio-technical bridge. I like the XKCV version that's a bit broader in its context. But that's essentially what we're trying to do is build bridges. And a plug adapter is one sort of example. And this is not without rationale. There's actually a whole discipline called infrastructure studies that looks at how infrastructures evolve as opposed to being developed. And if you were to read one thing on infrastructure studies, I recommend this report by Edwards et al. Understanding Infrastructure, Dynamics, Tensions, and Design. It really sums up the field well, and it highlights the way infrastructures mature. And then ultimately, an infrastructure is mature when it becomes ubiquitous, accessible, reliable, and transparent. In other words, you only recognize infrastructure when it's broken. You don't think about electricity until the lights go out. And that's what we ultimately want to achieve with the data infrastructures, that same level of ubiquity and invisibility. But if we look at how historical infrastructures have evolved, and this is everything from the rail network to the internet, and they go through this phase development, starting with the system-building phase, where it's characterized by deliberate development of specific technologies to develop new services. And then it goes to this sort of networking phase, where you start to get some transfer of technology across domains, across locations. But this also results in variation on the original design, and sometimes with emerging competing systems. It's a competing phase. And then finally, there's this process of consolidation, where these networks start interconnecting. And as Edwards et al. call it, is gateways. But those gateways, you could also think of as bridges. They allow these dissimilar systems to be linked into broader networks. And so I would argue that we might be at that phase now, where we're starting to move from this network to the internet work phase with the data infrastructure. So really it's not a question of what is infrastructure, but when is infrastructure. And to which we might also add who is infrastructure. And that is, I think, a critical component as well, because if we think about these bridges or these gateways, they're not just technical solutions. They're not just technologies. And it's not simply just either or a technical versus social. Often it's a combination of a technical solution in combination with a social choice. So, you know, a standard only works if people adopt it. I'm thinking about it. But you can also think about how certain technologies that they work best when they don't specify exactly how the work is to be done. An example, I think that Edwards uses is email versus Lotus Notes. People use email for absolutely everything from their to do list to, you know, basic communication. Lotus Notes, which was actually a pretty clever program, was very specifically targeted, though, to how you work. So it never took off. Email remains, you know, ubiquitous to a fault. So with that sort of concept in mind in terms of this phase development, what is the problem that we're trying to deal with? Well, this notion of a data deluge has really become somewhat of a cliche. We're drowning in data and so forth. And so, but to me, it's not so much the simple zeros and ones falling from the sky. Instead of thinking of a data deluge, maybe we should be thinking of a data blizzard where data is flying everywhere and it's chaotic and we're just trying to escape to somewhere sensible. And then if you think about, more importantly, each one of those chaotically flying snowflakes is unique. And so it's that diversity and the chaotic behavior that really needs attention as much as the volume, if not more so. So that's what I really want to focus on is this notion of diversity and the diversity of the data that we're trying to deal with is a central challenge. And so looking over the list, I see there's a lot of data scientists in the audience you all are probably familiar with this concept of the long tail of science. This is now a somewhat classic figure from Hydorn some years ago, just showing a distribution of awards by the National Science Foundation in the U.S. But I think you see a similar distribution of awards by almost any research granting agency. And the idea is that there's a few people that get lots of money and there's many people that get a little bit of money. And those guys are the individual investigators, small teams, groups of graduate students that are developing these smaller research collections that are incredibly diverse. And that's a central challenge that we're trying to deal with here. So when we're talking about diversity, it's not just the diversity of the data and the cultural diversity that I alluded to, but we also have to think about the humans. And George Grain brings in this concept of surface level diversity versus deep level diversity. And so when you think about this with people, the surface level diversity of those things that are obvious and immediate, race, age, gender, versus the deeper level diversity, which can often be more challenging in terms of their values, the conceptual metaphors, their personalities. And if you think about that diversity across scientific disciplines, we have that same sort of notion. So I think that's important to recognize is this deeper level diversity. So then when we're dealing with diversity, it's nice to go back to this guy, William Ashby. And it's sort of the quintessential nerd there, but a very smart guy. And he worked in this field called cybernetics. And it's not something you hear about so much for anymore, but it was a big thing back in the day. And the term cybernetics comes from the Greek word to govern, and it's to study or control information and communication systems. And his, so he has this law, if you will, of requisite variety that is simplified as only variety absorbs variety. To state it more prosaically is that the minimum number of states necessary for a controller to control a system of a given number of states. In other words, the larger the variety of actions available to a control system, the larger variety of perturbations is able to compensate. So in a sense, this is what you have to sort of understand the complexity of what you're doing to have sufficient complexity to address. If you think about it, this is why humans still control machines. And no individual can address this complexity alone. So only organizations with enough variety and enough diversity can adapt to the changing realities that come out of this diversity and really thrive in them. And only that way can be prepared for the scene as well as the unforeseen contingencies. You know, the organizations that are too homogenous will crash when the wrong wind blows and unable to adapt. So this concept of variety absorbs variety I think is very central. We add that to Metcalf's law, which says that the value of a network increases as the square of the number of nodes in that network. And you can argue whether it's a square or whatever it is, but the point is it's not linear. It's exponential. And here I think was originally looking at the cell phone network. And so you can see simply that, you know, the value of the cell phone network increases as the number of devices are out there and interconnected. It's not a linear thing. It grows exponentially. And I think you can see the same thing with other technical networks, but also with social networks. Social networks grow in value as exponentially compared to the number of nodes involved, number of people involved. And then further we explore that work theory a little further and we look at how networks form. This is a map of the piece of the internet from actually some years back now. But the point is you can see how it forms into these sort of these nodes, these sort of super nodes. And that's sort of a natural way that they connect. But as Barbara Rossi and others have pointed out that these networks, or if you think about them as sort of ecosystems, rely on these weak links or these weak ties. So this person right here that connects these two big nodes becomes critically important. They're not broadly connected. They're only like two connections, but that weak tie becomes really important. So it's these people who play in multiple spaces that are most valuable. And this and these ways. And so I think if we think in the Australian context, I think Australia is both a node in terms of that there's a lot of activity within Australia. But then it's also a link to other areas as well. So this concept of weak links I think is really important. And also if we think about that person in the middle, if you will, that person who knows people and is connecting things, you can think about that as mediators. And Chris Borgman and others wrote on this increasing complexity of mediation and that over time mediation has gotten much more complex. And I don't want to dwell on the details of this figure except to disagree with it in large sense. But also to point out that as we go from each one of these levels, there's also an increase in sort of standardization. You can see it really clearly as we went from symbol mediated, as she calls it, of basic books and letters and numbers to when we started getting into the telecommunication. So that required a whole growth of new standards and so forth. Same with what as we went into the next phase and as we move into this phase and for there yet. And so it's in that context that we're trying to develop things. And so it's this notion of you're trying to build some level of unity. And if you look at work on how organizations come together, particularly volunteer organizations, particularly looking at like open source movement, John O'Bacon writes a lot on this is that there's something that drives that is this notion of I think what he calls an axiom of unity, but something that unifies people towards a common goal, you know, a dream, if you will, that we have a shared dream. So that makes sense. We want to move forward on something we need to have this sort of shared dream. But Anna Singh, who writes on collaboration and fascinating ethnography of the biodiversity movement and the destruction of rainforests in the 90s, notes that these universalisms, as she's calls them, are actually hybrid transient and constantly being reformed through dialogue. They work out through friction that we are not as united as we think we are. And I think, you know, a good example of that is say open data. So I would think that most of the people on the call here would support the general principle of open data. That's a nice unifying principle. But when we start to get into what in actually implementing what that means, we run into all sorts of frictions. But some of those frictions actually reveal interesting things that lead to new solutions, such as the development of robust citation schemes. I think it's something that's emerged out of the friction that comes from this notion of open data. And so thinking along those same lines, there's no reason to think that collaborators really have common goals. They may have this unified dream, but they're all working towards their particular local objectives. And so it's a form of coalition politics, if you will. So we have to recognize all that. And then this last point I think is really important, but it's kind of hard to get your head around. And this notion that unity and diversity mask each other. They cover each other up. And the example she gives that I think works really well is looking at early biological classifications. So early on, people would go to new exotic places and they would see new plants and animals they'd never seen before. And so they would rely on locals to tell them, what is this plant? How does it relate to this? What is this animal? Later on, we get Linnaean classification and so forth, and they come in and they're trying to impose this unified view and that masks the local knowledge. But then if you use the local knowledge in the sense you're covering up that systemized view of taxonomy. So the point is you have to work in both places and you need to remember the local. Another book, and I'll get out of the theory here soon, and this is a much more popular, easy to read book. And I think there's actually a TED talk from this guy, Steven Jobs, where good ideas come from. And he introduces this concept of the adjacent possible, and I think it relates to that notion of the local, where it's this connection between people that really generates these ideas. And the ideas often are not a eureka moment, the sudden insight, but this sort of slow hunch that sort of fades into view over time. He gives a nice description of how Darwin's concept of survival of the fittest and adaptation was not a sudden realization that we actually developed over decades. But then he argues for this that these hunches, if you will, need to bump into other hunches. And so, and we see this, you know, you have an idea that you've been milling around in your head and you talk to somebody about it and they give you some insight and something boom, oh yeah, now I get this connection. And so he argues, you know, we shouldn't be protecting IP, we should be sharing it. He has this little nice little phrase connecting versus protecting. And that means sharing of failures as well. And so when we're trying to build a global data infrastructure, we need to learn from what we didn't work as well as what worked. And so we want to create spaces for that to happen. Sort of coffee shops, if you will. Johnson illustrates how during the Enlightenment, the growth of coffee shops and bringing intellectuals together really helped foster that intellectual new youth. And so just to close with his nice little quote here, the chance favors the connected mind. And it goes back to that Winkley concept again, that connection makes a difference. So in a sense, it's all about relationships. That's a theme that I'm getting at. And I know this as an introvert and Amirah brings sense and that, you know, interactions and relationships exhaust me. But they're central and they're necessary. And so to sort of summarize so far, we have this central challenge of diversity. We address diversity through diversity and a variety of mirror interfaces, connections, relationships. But then fostering those relationships is not only central to community and building the community that we're trying to build. But I would argue that it's also central to data science. And I'm putting that in because I'm just looking at the attendance list. It looks like that's something that a lot of you guys are working on. And so part of it is that, you know, those relationships build social capital. You know, you give you professional growth, it increases your influence, your distinction, your gravitas as the more connected you are. This notion of, you know, the more you give of yourself, the more successful you are. And that's why volunteerism works. But also I think these relationships and fostering these relationships uncover tacit knowledge. And I think that's in managing data that's often a really central challenge, especially when trying to share data across disciplines, because this embedded knowledge is not necessarily explicit. So just to give an example, I was working with a modeler who was modeling snowpack, and I was trying to understand his needs, his data needs for what he was bringing in and understanding the data that he was bringing in. And so part of the data he was bringing in was from these telemetry stations that record basic meteorological data and snow data. And he just knew that you don't use the wind direction and wind speed from those telemetry stations, because they're often in these valleys, they're small forest clearings, and the wind speed is just not representative. And so he's just like, everybody knows that you don't use the wind direction and wind speed from snow tele stations. I'm like, oh, everybody knows that? And so it wasn't until that relationship was formed between him and I that tacit knowledge becomes exposed. And you understand then why the wind speed is not included. You don't do stupid things like using that wind speed to try and improve his model. I'd also argue that focusing on these relationships inform our methods as data managers and scientists. And so I would argue that we're all into user driven design and developing systems, but it's not just the end user. We need to be thinking about the data providers, the funders, the entire ecosystem of people that are involved when we're designing systems. And sort of in a similar sense, we need not just use cases, but case studies. We need to not only ask what people want, but how do they work? We need to understand how they work and how data fits into their processes. And so we really need to take in some cases an ethnographic sort of approach. Participant observation is one definition of ethnography. And to really focus on those relationships, because data is often at the center of those relationships. It's what we call a boundary object. And the boundary is not putting a barrier in between, but a boundary upon where communication goes across. This last point is, it only works if you're into agile development. And so agile is a philosophy of software development. And there's all sorts of methods around it, but ultimately it's centered around a philosophy. And Bruce Keran, a colleague of mine, translated this philosophy into agile organizations. And the idea is that we focus on individuals and interactions over processes and tools. Not that processes and tools aren't important, we need processes and tools, but we've given more precedence to the individuals and the interactions. We focus on the working volunteers over detailed documentation. We focus on the member collaboration over detailed contract negotiations. We focus on responding to change over following a plan. So this isn't to say that we don't follow plans or negotiate contracts. It's just where we put our emphasis and where we put our priorities. I just thought that was kind of fun. So if you're in the agile world that resonates, it probably doesn't resonate too well with you. Okay, so that was all sort of the theory dump, as an old professor of mine used to say. So I'm a geographer, tend to think in sort of conceptual ways, particularly in sense of place, space, and scale, but then bring it down to something real. So given all of that theory, what does it all have to do with RDA? Well, first of all, RDA focuses on developing these gateways. So they talked about it at the beginning, these bridges. And we don't do architecture in the classic sense. We do not design the entire system up front because that will not work. It has never worked for any infrastructure ever developed. And that we need to recognize that it's more of an organic-evolving process through that phase development that I talked about. So how are we explicitly doing that? Well, we have this sort of mantra of create, adopt, use. And it's focus on short-term deliverables. These are these little bridges. And so they might be codes, a particular specification, implementation of a specific standard. The devil is always in the details of implementation. And these practices that enable data sharing. They're reusable or harvestable efforts that can be done in 12 to 18 months and eliminate some sort of roadblock. Get over some sort of barrier, bridge across some sort of gap. And they're efforts that have broad applicability, but don't necessarily solve all problems. I mean, if you have a generic solution, that's great, but those are very rare. But they need to have substantive applicability to a group within the data community. And things that can start today. So we want these short-term bridges. And while we don't have an architecture, we do have that unifying vision. And that's encapsulated. Captured. Sorry. And that vision is captured in our principles, if you will. So we have that I've summarized in six little terms. So openness, and by not just open data, but in terms of open in our processes, open in how we negotiate across technologies, et cetera. Similarly, we're a consensus-based organization. We want consensus within the community in terms of how to do things. By balance, I mean balance across different perspectives. The data practitioner perspective, the data user perspective, the library perspective, the repository perspective, but also balance across technologies that we don't want, you know, someone to push their particular chosen approach. And so related harmonization. We're trying to harmonize, and that's, again, in keeping with that consolidation phase, we're trying to harmonize across different systems, different communities. We're community-driven, and we're nonprofit. So further to this, building these gateways and just create, adopt, use, this is our overall organizational structure, and I'll come back to this in more detail later. But the basic thing is we have this membership. That's all the people that are involved in RDA, that sign up to be members of RDA. And a particularly important group within that, you know, these working groups, and these are these groups that exist for only 12 to 18 months to produce these little ridges. And here are the ones that we currently have, the ones in red are ones that are just wrapping up right now and are starting to deliver, and I'll talk more about their deliverables in a little bit. But I think if you look through this list, and I won't go through them all in great detail, you'll see that there's probably things here that are relevant to what you're working on, that you're trying to address some of these issues as well. The next point I want to make is that RDA plays in both the global space and the local. So you can think global. And global is, it's kind of a created word, but you actually see it in the literature some. And the idea is not to just think global act locally, but to simultaneously play in both spaces at the same time. And so, again, RDA is a global organization. These stats are about a month old, so I'm sure it's grown since then. So it is a global organization, and we have members from 96 countries pushing 2,500 members. Yes, we're a little biased to the North Atlantic right now. And so we're trying to spread to a more global situation. So maybe we're international, not quite global. Just to point out the Australopacific piece, that's primarily Australia, some from New Zealand. And just looking at the map, we've got some people from New Guinea and Indonesia, but it's primarily Australia. And so that 5% in terms of population is actually a pretty heavy contribution. Also, just to note that we're primarily academics, a lot of government folks, but the private sector is growing. And I'll get back to that in a little bit, but I think that's particularly important. And so not just growing in numbers, but also growing relatively, so as a piece of the pie. But further to this local and global at the same time, we have what we call regional RDAs. And part of this is just an artifact of the funding that funding agencies want to fund things in their own country. And currently that's what's called RDA US, RDA Europe, and then ANZ is somewhat the representative of RDA Australia, if you will. But it's more than just an artifact of the funding because with this focus on implementation, we recognize that implementation is inherently a local activity. And the objectives in Australia are going to be different from the objectives in the US and that the Australians are going to be best equipped to recognize those objectives and how to implement in a local level because the devil is always in those details. But at the same time, it's ensuring that those regional issues or national issues get echoed up to a global level so that you're not working on things in isolation. And then finally, and I'll get back to this a bit more later, is that we have plenaries meetings every six months. And to date, these three sponsoring organizations have been the sponsors of these plenaries locally. Okay, and then so points four and five. RDA fosters relationships, interfaces, and connections. Remember, I focused on relationships as critically important. And we provide a neutral place, if you will, to identify these issues that we need to work out identify the friction and work through the friction. So going back to this, part of where that happens is in the working groups, as I discussed, but also part of where that happens is in these interest groups. And interest groups were not originally conceived when RDA was first organized, but we found that it turns out that they're quite important. And it turns out that we also have many more of them than we have working groups. And again, if you skim through this, you might see some areas that are of particular interest to you. And interest groups originally were conceived as incubators for working groups. We recognize that having that tight, focused deliverable that you can deliver in 18 months requires maybe some setup and some discussion to get going. So that was the original concept of interest groups, but it's broadened since then. In some cases, it's a way to bring communities together. For example, there's one on here, yeah, number 24, the Preservation E-Infrastructure Intergroup. People are working on preservation all over the place in the library community, in the data community, in Europe and the U.S., in Australia. And this was simply a mechanism to bring those groups together to start sharing information. So that's sort of the opposite spectrum. But then there's many interest groups in the middle that are still delivering things. They're not necessarily delivering those little bridges, but they're developing things, delivering things that we hope will work towards those little bridges. So they might be, for example, there's a group on here, legal interoperability that is doing some intensive studies of the actual success and failures of some of these tools, like CC licenses and CC0 waivers, and to develop some recommendations specifically on how to address them. So interest groups are a way to foster that connection, foster that community, give that chance to the connected mind. But then it really works best when you come together in place. So as I mentioned, we have these planaries every six months. The next one will be in San Diego in March. Our last one was in Amsterdam. We will have one in Australia, but it's probably a little ways out. But I will note that this is as close as you can get to Australia in the U.S. without going to the body. It's just an ocean between us. I can tell you, it's just a 14-plus hour point. But anyway, I encourage you to attend planaries. And actually, we're hoping to have one in Japan soon, so that's also kind of in your neighborhood. But the planaries are a lot of fun. They advance our work substantially. You get Stephanie Snickered at fun. They're really dynamic. Let me put it that way. But I at least get energized by them. The community, they're hands-on working meetings. Sure, we have planary sessions, keynote speakers, all that good stuff. But the vast majority of the meeting is devoted to work and the working groups and industry groups coming together and actually working together. And people really like that. The members really like that. Oh, and so then finally, just to cover all the pieces of this figure, all these yellow boxes up at the top, that's our governance structure. So our council, our senior statespeople, they're currently appointed by our funders. Ultimately, they will be elected by the membership. And they're responsible for the overall mission of the organization. They're to ensure that they're to detect consensus within the community. The technical advisory board, I think is a really critical group. And a big focus of them is on that balance piece and that harmonization piece. So they're there to make sure that people aren't pushing their own specific agendas and providing technical advice to the council on the approach. The organizational advisory board, I'll talk about it a little bit more in a minute, but is to recognize that organizations have a key role in this as adopters. After all, it's organizations really that adopt, not individuals. And then in the middle is the secretary that's the office that runs things that Stephanie and I are part of. So just to give you some of the names, here's the council. Ross Wilkinson, director of ANS, is the Australian representative. Stephanie is the representative on the secretary. And you'll notice the technical advisory board also has good Australian representation with Simon Cox and Andrew Trelore. Andrew Trelore actually was just re-elected. He was the chair. I would imagine he'll be elected chair again, but technically he is not at the moment. He'd has to be willing, I think. And so, but going back to these organizations, as I mentioned that organizations play this essential role as adopters. We have this focus on implementation. We want to actually have the little bridges that we build be adopted. And it's all well and good to be adopted by an academic in a lab, but to really have an impact, you have to be adopted by an organization, be it a private organization, a government organization, a nonprofit organization, a library, what have you. And so organizations actually join as members as well wearing their organizational hat. What they do is to do so. Individual membership is free as long as you agree with our principles. And so the organizational members come in together in what we call the organizational assembly, who in turn elect an organizational advisory board that represent organizational issues to counsel. The organizational advisory board, so currently we only have about 25, 30 organizational members. So the advisory board is acting as a committee for now. Once we get more members, we'll probably have a smaller group. And as I mentioned, they pay modest dues, but they have a special voice within RDA to ensure that what we're developing aren't just academic little exercises, but they're actually relevant to real organizational problems. Finally, oh yeah. Okay, so here, I think this is a rather dated slide, but here are the organizational members. And just to note that we also have affiliates. They have the same rights as organizational members. They're a little different. They don't pay dues. They're like-minded organizations with international scope, also working on data sharing issues. So the world data system and co-data are sort of classic examples. They were our first affiliates. They're working on similar issues. We want to make sure that we're working with them and not against them. That we're not overlapping, that we're collaborating. And so you can see that our organizations are pretty diverse, but they are leaning towards the nonprofit and research type and not the private sector. And that's the area that we want to focus in in the near future. Returning to this figure for the last time, this bottom box, the RDA colloquium. And you'll notice that it is, this is RDA up here. This is the colloquium. It is a separate entity. And that is the body of funders. Currently governments or government agencies, but also could be nonprofit foundations, for example, Sloan and Wellcome Trust are expressing some interest. But they, and the point of this, it's an informal way for the funding agencies to gather, to share their plans around supporting data interoperability. But they don't have formal relationships. There's no MOUs. There's no really, any sort of binding agreements to collaborate, a casual collaboration to share their ideas so they can amplify their impact. So it's related to, but distinct from, RDA, sort of a parallel informal organization. And they did that intentionally to keep sort of a hands-off approach to say, we want better data sharing. We need this vision-solve community figure it out. And so they're intentionally keeping that hands-off rule other than appointing the council to make sure that we have these sort of senior states people keeping an eye on things. But they're not presuming that they have the solution. They're leaving a community to determine what the solution is. So what have we got? We've actually, so we're 18 months old, well, I guess 19 months old now. And we've actually already delivered some initial prototype deliverables. They're not many international organizations that deliver in 18 months, if any. So they're pretty small, little building blocks, but I think they're important and very foundational. So one is simply a foundational terminology. The first thing you need to do is figure out what the heck you're talking about. And that's, you know, you can argue forever about terms. So this is only a handful of terms that we anticipate growing and then a query tool. But it includes basic terms like digital objects, for example, things like that. And as the chairs put it in, it's part of building the culture of RDA is to, you know, having the terminology so that we know what we're talking about. Another thing is a data type model and registry. You might think about this as sort of like mind types for data. So it's a model of how you would describe a data type. What are in the columns? What are the units? That sort of thing. And then a registry for registering those data types. The vision is that this wouldn't be a global registry, per se, that people would use this within their local context. So, I gave this talk, a very similar talk to Geoscience Australia yesterday. So I can see that Geoscience Australia might want to have their little registry of all their Geoscience data types. But if they follow this model and then say the bio group, a biodiversity group is following a similar model, over time it can start to have interconnections across these registries. So that's the idea. Very closely related notion is persistent identifier type registry. So persistent identifiers can identify, locate many different things. Articles, data sets, individuals, organizations, instruments. So for a machine to understand what that thing is identifying can help a lot in terms of you know, oh, I am pointing to a data set that tells me something other than I am pointing to an article. And it's interesting. So I imagine a number of you are familiar with digital object identifiers and using those in data citation. And one of the things that's emerging now that is people are getting confused is when is the DOI pointing to a data set? When is it pointing to an article? Wouldn't it be great if the machine could figure that out? So some other things coming soon. A basic set of what they call practical policies. I think policies kind of are loaded words. I call them rules. But the idea is that these are some of these basic functions you know, a fixity check say or a simple ingest process that could be repeated and used in multiple contexts and the idea ultimately is to enhance trust. If we are following the same basic rules then we can ensure greater trust of our repositories. A metadata standards directory this is a rather grand notion I wish them success but the idea is that you would have a directory that you could decide what is the appropriate metadata standard to use for the type of data that I have. So you can think, I got a thumbs up from the data library next to me and you say you think that would be great for working with a researcher who's trying to write a data management plan and doesn't know what metadata, you say use standard metadata, that doesn't help. Another one is a dynamic data citation methodology I know Andreas Rauber gave a talk here at Ann's a few weeks ago, he's leading that group and essentially so what it is is a way that we can reference precise subsets within a ever changing data set. So it's a reproducibility issue so if someone does science with a particular subset of a particular version of a particular time series that we can reference that again and essentially what he's trying to do is put a query ID of a persistent identifier onto a query so that you can capture that. It's an intriguing idea and a lot of use cases coming together, I think it has some good hope. And another one that's coming up soon that one of my personal favorite groups is the wheat interoperability group and they're taking a, basically building a small ontology to describe a basic issue, things around wheat data harvests and so forth so we can share that information around the world and if you think about it, that's just fundamentally critical it's talking about feeding the world. So these are some of the deliverables that we have coming forward these are some of our initial little bridges like I say, adopt one today feel free to talk to me or anybody in the secretariat about what that might entail. So I encourage you all to get involved as I mentioned joining RDA is free, you just need to agree to our principles. If you're part of an organization that would like to join there is a fee but it's fairly modest we encourage you to join or start up an interest group or a working group and come to our plenaries. So just to sort of summarize infrastructure is created in phases with this final consolidation phase relying on gateways and bridges and I think we're in that consolidation phase now just the beginnings of it. Diversity is a central problem but we only address diversity with more diversity and networking and interconnection are a way to handle that and to solve these complex problems and we're in a more global and more democratic world but we're also in a more local world we have more local concerns as well as we're working in this global space and so this notion of coalition politics with all kinds of new coalitions and new kinds of identity really as to what and this is the human geographer in me coming out that the identity question I think is central and we're creating new forms of identity all the time with new professions for example and that data science as a discipline I think needs to focus on these relationships, these connections, these interfaces not just human to human but machine to machine human to machine etc and you need to participate both globally and locally at the same time at 16 and that RDA provides the mechanism to do all of that so thank you very much Wonderful, thank you Mark Thank you everyone, thank you Stephanie for chairing it down in Melbourne and thank you everyone for coming