 All righty, we'll go ahead and get started. Afternoon everybody. My name is Bruce Hedrick. I'm the Senior Vice President for Open Collections and Infrastructure at Ithaca. Cliff Anderson from Vanderbilt is joining me up here today to talk a little bit about Vanderbilt's experience with the Open Community Collections initiative which JSTOR launched, right? I guess really about this time in 2020, right? As the pandemic was getting started. So it's been an interesting couple of years. We're going to talk about a few things and then leave plenty of time at the end for questions that people might have. I wanna give people a sense of kind of how we got started doing this and Cliff's gonna talk about Vanderbilt's experience and it's thus far and I'll talk a little bit about some of the lessons we've learned over the past two years. So most of the folks know JSTOR from the standpoint of the work it started 25 years ago and the digitization of the back files and journals. So we've been at this for some time for a couple of decades. Actually this is the 25th anniversary of JSTOR being launched. And the interesting thing over that time is we've sort of constantly been in this process of bringing together content in the humanities and social sciences. Started with journals, moved on to books and over this time we've also been trying to move more and getting more and more open content onto the platform. So beyond what we started with the back files of journals, everything was paywalled books, everything was paywalled. Now we're seeing more and more books come on the platform that are open, journals that are coming on the platform that are open. We put research reports on the platform a few years ago. Some of you are familiar for Veal Digital which is really about finding marginalized content and bringing that on. They came into the fold in 2018 and that content's been on the JSTOR platform. We have over the last couple of years been moving all the art store content onto JSTOR and a lot of that content is now open. And of course what we've started in 2020 with the open community collections. So the idea here is we're trying to broaden the content that researchers want to use when they're connecting secondary and primary literature and trying to get more open content onto the JSTOR platform. And that has been an interesting experience both with doing things like research reports and books where we actually found that having that stuff on the platform really started to expose it to a much broader audience and started to get quite a bit of use. And we're starting to see that with images on the JSTOR platform as well which has been really interesting. All of this is underpinned with our mission as a not for profit really to be the long-term preservation home for this content. So all the things that we're putting onto the JSTOR platform, we're also doing long-term preservation for that content with Portico including the community collections which Cliff will be talking about. So why did we start to get into the primary sources? Well, we did a lot of work with libraries really in 2017 and 18 to find out things that they were struggling with that a small organization like JSTOR might be able to put some time and energy into. And one of the things that came across over and over again was that people were struggling with making their primary source content, their special collections, their distinctive collections, getting those really into the research workflow. And we were hearing constantly that faculty were wanting to teach more of these resources and that students were wanting to do research with these resources, particularly early career researchers. So this sounded a lot like the issues we started to hear back in the early mid 90s with the back files of journals. They were hard to find, they were expensive to maintain, they weren't in the research workflow. And that sounded like something that we might be able to actually put, have some value add into that space. And so we decided, let's do some experimentation and see what happens when we get these sort of primary source special collections on the platform, how do they get used? Can we connect them to the secondary literature that's already on the platform and things of that nature? And this also began to bring out something for us that we wanted to build out, which is JSTOR has traditionally been a research platform. And our colleagues at ArtStore, when ArtStore was built, ArtStore was really built more as a teaching platform. So one of the things we were trying to bring is the teaching aspect of ArtStore, the tool sets and things of that nature onto the JSTOR platform and see if we can build JSTOR out to be a teaching platform as well as a research platform. So that's really where this experiment has begun. The things that we thought we could actually have some impact on, there were many problems that were brought to our attention, but there were things that JSTOR has a bit of a unique position where we thought that could really add value in. Number one is in many disciplines, JSTOR already is in the research workflow. It is a starting place for a lot of people, sometimes wrongly, but it is a starting place for a lot of people in different disciplines. So this idea of these primary sources being difficult to discover in the silos that they're currently in, we might be able to do something to help remedy that. Oftentimes too, these primary sources are found out of context. So they're found in one place, the secondary literatures in another. The idea was that if we could connect these things together in an effective way that we could be able to actually make the research process a lot easier. We all know these things are expensive to maintain. Libraries spend a lot of money on that and they're getting more challenging to maintain, particularly when you have to start to work in things like are they optimized for preservation? Are they optimized for machine learning? Are they optimized for accessibility? Are they optimized for inclusiveness? There's a whole set of dimensions that people are trying to work through on how to actually keep these collections in the research context, the way researchers are using things today. And that's difficult. And of course, JSTOR has a very broad audience. So we have 12,000 institutions around the world that use JSTOR and we, I think are like the 980th most used website on the web. So we have a big reach out there for institutions. And of course, one of the underpinnings of what we've done all along is preservation. And that's really important with these materials. So our two working hypotheses were these, that if we could make the content available in the JSTOR channel, and this is not about building another institutional repository, this is about providing another channel for this content to be made available, that we would be able to potentially improve the usage footprint of these materials and make them more widely available across the globe. The second piece was that if we can co-locate the primary source contents, these special collections with the secondary literature, that we potentially can make the research process more seamless by putting those things in context. So those are the hypotheses that we started with in 2020. And that's what we've been working on over the last two years as institutions have been putting things on the platform. I'll talk a little bit about what we found in just a bit, but I'll turn it over to Cliff and let him talk a little bit about Vanderbilt's experience. Let's see if I can get this to fit here. I wanna smush your computer. All right, nice to see you all and you can all hear me okay? Is that all right? Okay, great. So thank you, Bruce, and I just wanna reiterate this has been a really exciting partnership that started at a strange time. So Vanderbilt's experience with JSTOR community collections began at the CNI meeting in December of 2019. And so I attended Bruce Hedrick's presentation, it was called Unlocking Opportunity. Using the JSTOR platform to get library special collections into research workflows without a paywall. In my role as digital strategist at the Vanderbilt University Libraries, I'm always looking out for ways to reduce the complexity of our systems while broadening access to our digital collections. So Bruce's plan, as he put it, to take the infrastructure we've developed at JSTOR and make it openly available to libraries intrigued me. Of course, JSTOR is immediately popular and known to students and faculty alike. Our library liaisons know how to teach with JSTOR since Vanderbilt also subscribes to ArtStore, the product formerly known and as well as the product formerly known as ArtStore Assured Self, now JSTOR Form and we'll talk about branding a little bit in a minute. Members of our technical staff were also acquainted with the metadata tools. So it seemed like consolidating our digital collections within this popular digital ecosystem by adding JSTOR community collections would be a good option to explore. So the Vanderbilt University Libraries began its pilot in May of 2020 and the timing actually proved fortuitous. At the beginning of the COVID-19 pandemic and probably a lot of you had this experience, the library leadership had to quickly equip staff members to work from home that were really not used to doing that. And the transition proved challenging, more challenging for some units than others. For example, in our special collections units, the curators worked so closely with the collections that it made it difficult to identify work that could be done effectively remotely. As it happened, we had a large backlog of digitized items from special collections and other units in the library that needed to be described and shared. So we commissioned what we termed a summer project team to experiment with putting those backlog materials online with JSTOR community collections. And Nancy Godleski, who is in the room, I think. Nancy was my partner in that and so you can ask her questions about that as well. So over the course of the summer, staff members from IT, cataloging and metadata as well as our distinctive collections units collaborated on building trial collections in JSTOR community collections. We also migrated a few existing collections from Islandora to community collections to ascertain how easy it would be to transfer our already existing special collections in those content areas. So the pilot proved a smashing success. Staff members learned to use the JSTOR form interface with minimal, if any, training. Coordinating over Microsoft Teams, staff members contributed several collections over the course of that summer and even more during the past year and a half, including student publications, diaries from the Civil War and just recently Dizzy Gillespie's personal photographs. Last year, special collections held a symposium about James Robertson. He was an early settler in the Cumberland Valley who is sometimes called the father of Tennessee and among the scholars on the panel was the state historian of Tennessee. As our associate university librarian who organized the panel puts it, the panelists were quote, blown away that the material was available to them. She recalls the state historian waxing enthusiastically about the usefulness of this collection for teaching the history of the state in K through 12 classrooms. I think that's because he was familiar with JSTOR already, so he kind of knew how to use it. From her perspective, the capacity to reach audiences beyond the traditional academic ones that we normally reach with our digital collections underscores the community in JSTOR community collections. Yeah, I was just gonna say, there we go. There's James Robertson. Okay, I think I'll pause here. Okay, so let's talk about a few of the key advantages that make JSTOR community collections a good fit. The primary selling point for us was, as I've mentioned, ease of use. When using platforms such as Islandora or DSpace, a skills gap divides curatorial staff and technical staff. For curatorial staff, using the command line feels foreign and they're unsure how to upload files from network shares into those systems with the accompanying metadata. Community collections seems to have eliminated that gap almost completely. During the period of our pilot, I didn't receive any feedback about technical impediments. The relative simplicity and ease of use of JSTOR forum allows the curators to focus on the description of their collections rather than on technical questions about deploying them in a digital content system. Another advantage of community collections is the interface. With community collections, you get a great user experience right out of the box. In the past, digitizing materials and making them available in online content management systems proved only half the battle. After digitizing products, we then had to engage our web community committee in the library to build websites or create in a Mecca exhibit to showcase them. We also had to publicize them to the relevant audiences. With community collections, we no longer have to take these additional design and communication steps. In fact, we discovered that audiences are finding their way to us through the JSTOR ecosystem without having to draft press releases or create social media posts, which we still do anyway. But you can see, here's the usage statistics. There's a simple but nice tracking mechanism that shows that there are indeed being used. And this connects to a third critical advantage for us. The JSTOR platform drives traffic to our collections in measurable ways. After moving existing collections into community collections, we've been pleased to see that they've attracted new and larger audiences. As Bruce has demonstrated, community collections comes with this simple tracking tool for showing web metrics and engagements. This tool lacks the bells and whistles of, say, Google Analytics, but I imagine better protects user privacy. And since curators are also the ones that have direct access to these metrics, which they don't normally do, it's encouraging for them to check the level of user engagement and to also use them to gauge public interest in their collections, which in turn incentivize them to bring more collections onto the web. So that's been a good virtuous circle. Let me talk a little bit now about Portico and the integration of Portico with community collections. So in August of 2021, we joined the so-called Preservation Early Adopter Program for the purpose of, quote, testing the long-term preservation of collections published to community collections on the JSTOR platform in Portico. The integration of community collections with Portico held forth the tantalizing possibility of a comprehensive system for describing, providing access and preserving digital collections. As with community collections, the onboarding process for our staff proved straightforward. Portico really does offer something close to push button preservation with a few caveats. The introduction of Portico raised questions about our existing procedures for digital preservation. On the one hand, staff have been accustomed to uploading low-resolution access images, typically the JPEGs or the PNG files, to JSTOR forum, while maintaining full-resolution access copies, in most cases for us, TIF copies, elsewhere. And ideally that elsewhere is the library's Fedora digital asset management system. But in practice, many high-resolution images remained on network shares due to bottlenecks in our systems, including lack of metadata description and inexperience with loading digital assets into Fedora. So this pattern of uploading access images onto JSTOR forum into the JSTOR platform, while preserving the original files off the platform emerges a problem when we were trying to integrate with Portico. Because we wound up preserving a raft of lower-resolution derivative images in Portico rather than the full-resolution ones. So that was a mistake. In the Visual Resources Center, which serves as our center that assists with art histories, the traditional slide lab that's become a digital lab for art history and beyond, our staff have worked with technical partners at JSTOR to develop a method for substituting those source files for the derivatives. This method is a little clunky. If anyone's used it, it involves downloading a spreadsheet of the metadata, swapping the names of files, and then uploading the revised spreadsheet to amend the records. But the process works reasonably efficiently. In the future, we anticipate that JSTOR, we hope will release an API for editing metadata, which would allow us to script that substitution process. In special collections, by contrast, staff members have held off substituting the preservation for access images. Why? Well, the reservation arises from a difference in the way that community collections handles preservation formats like TIFFs. While TIFFs display just fine when you view them directly within community collections, when you try to embed links from community collections to other web-based systems, of course, if you're using a TIFF, the links will behave differently than a JPEG or a PNG. For example, browsers do not render TIFFs, and so users receive a prompt asking them to download the TIFF to view it. And you can see here that this is an archive space. There's a PNG that's embedded and that link will go to community collections. And that's what our special collections librarians like to see inside of archive space, which is not possible now when you're using TIFFs as the image that you upload into community collections. This difference makes it problematic to embed images safe from archive space, as I've said, to community collections. We're in context, in conversation with JSTOR now about a fix to the issue. And it seems to me at least that the ready at hand solution would be to have a more robust implementation of the TIFF image API, which will allow staff members to provide links from community collections to any desired format and resolution. So these quibbles serve only to highlight how unusually powerful the combination of JSTOR community collections in Portico is. In our production workflows, we've grown accustomed to maintaining distinct workflows for access and preservation. With this integration, we can pull those processes together into a single stream. And to be honest, we are starting to wonder whether we need to keep the copies of objects in community collections in Fedora and S3 Deep Glacier at all. Or should we just learn to rely on Portico? As with nearly everything in library administration, I think that financing and staffing may eventually dictate where we come down on that decision. Oops, I think that's jumping out, okay. So let me just talk about a few other challenges. If our pilot of JSTOR community collections has proved so successful at Vanderbilt, why aren't we using the platform for all our digital collections? There are a few reasons that prevent us from scaling community collections more than we have already. Copyright questions remain a concern among staff. Initially, the staff worried about sharing materials publicly in community collections if the collections contain material that was potentially under copyright. Because when hosting collections in our campus infrastructure, we can respond quickly to takedown notices if a copyright holder objects to the display of particular materials. But how would we handle a complaint about copyrighted materials on community collections? Would we even receive it? Now that community collections allows a finer-grained control of user access levels, providing the option to restrict collections to campus users only, our curators are growing more comfortable of sharing in copyrighted items in community collections. Still, there are certain classes of digital objects, such as donor restricted materials that we really have to maintain elsewhere because we provide access to them only on request to a handful of authorized users and for legal reasons can't store them with third parties. A second challenge arises from the variety of formats in digital collections. When digitizing archival collections, for instance, we may produce digital versions of photographs, maps, manuscripts, codices, articles, along with audio and video recordings. And at present, community collections offers excellent support for documents, but audio-visual support is lacking. The lack of features for audio and visual formats means that we need to maintain a separate workflow for those objects, even if they belong from a special collection perspective, to the same series. Open source content management systems like Avalon or proprietary platforms like Aviary exist to meet the specialized needs of those type of audio-visual formats. But sorting out AV from other objects feels artificial to our curators. They use finding aids in archival space to link these collections back together, but users do not always start their searches with finding aids or necessarily even understand how to use them. Another issue relates to the discoverability of content in the JSTOR community collections. Of course, items and collections are easy to find when you're working within the JSTOR ecosystem, but what about through library discovery systems? In our library, for example, we use XLievers' Primo VE as our primary platform for discovery. But at present, there's no easy way to federate item-level metadata from community collections into the central discovery index, CDI, that forms the shared metadata store for Primo and related tools like Summon. We're an active conversation with XLievers about adding content from community collections to the CDI, but for the feedback we received, that project will require development effort and has not yet been prioritized by XLievers. So for the present, we rely on metadata from other systems of record, primarily from Archive Space for pushing information about JSTOR community collections into our discovery system. And because of the way that Archive Space is set up, this work around means that users cannot discover objects at the item level in our discovery systems only at the collection level. A final sort of challenge, as I mentioned earlier, relates to branding. As Bruce has emphasized, JSTOR community collections are open access collections. But for many users, JSTOR designates licensed and paywall products like the JSTOR digital library of books and journals, and to a lesser extent, the ArtStore digital library. Understanding what is open and what remains closed takes nuanced understanding. And frankly, not everyone gets it. As a case in point, an anonymous peer reviewer recently questioned our proposal in a grant application to share assets on JSTOR community collections because the reviewer mistakenly thought that that decision would prevent the public from engaging and accessing with our digital archive. So summing up here, when I asked a curator who works extensively with community collections to share his opinion of the product, he responded, I love it. Like I just love it. I love it so much that I want to use it as much as possible. And literally like he kept me on the call to tell me how much he loved it. So his main concern was that we moved beyond the pilot period because he and his fellow curators are already using the system so extensively. And our associate university librarian for distinctive collections, engagements, and assessments responded with similar words of praise. For her, community collections has become the default option for providing access to newly digitized collections. She now routinely asks whether curators would be able to put their collections into community collections. And if not, then why not? For me, JSTOR community collections helps us again to refocus the attention of our curators on the content of their collections rather than on the technology, saving both staff time and budget. And from my perspective as digital strategist, I want to say what more could I ask for? So let me turn it back to Bruce. Thanks Cliff for that honest assessment. I appreciate that. Whenever you get into these things, particularly in sort of a pilot period, you learn a lot. We've learned a lot over the last two years. We've learned a lot of what we had no idea about in some instances. So we have a long way to go still but I think we're moving in the right direction. So where are we today? We have about 850 collections on the platform. One of the real ideas here was that this really would start to, its value would start to show itself as we got more scale. So we need to get, we're looking for thousands of collections on the platform. And so we're working with institutions to have them dip their toe in the water and see if they want to learn like Cliff has learned at Vanderbilt. One of the really interesting things is that, how much of the content, now again, this was during a pandemic. So, and a lot of people were not logging in to their university credentials when they came in. But this stuff is available openly on JSTOR. And what we're found is that about 62% of the item requests coming in have actually come in from IP ranges that we don't recognize from those 12,000 institutions. So it feels like the open part of this idea of allowing just people to find it is working. The other thing I'll mention, and I think I'll get to this in a minute is that the amount of Google traffic that's driven in this content is much higher than what we see in the JSTOR Journal and Bookside. So Google is playing a much larger part and playing such a large, much larger part. In fact, that we really for the first time as an organization are having to think about our SEO with Google, which is we didn't really have to do before. And so that's been really interesting. So collections and a number of institutions who are participating continues to grow. The usage keeps escalating through the blue lines 2020, the red lines 2021 and 2020. Obviously we're adding more content over that period of time, but it's really starting to escalate. These are very similar things to what we saw with the journals in JSTOR early on, this type of growth. Cumulative usage is continuing to go up. And we have things on the, as Cliff was saying, we haven't gotten into audio and video yet. We've really been focused on text and images and really focused on things that obviously have already been digitized. And most of what we're seeing on the platform is stuff that we're harvesting from other repositories, whether it's Internet Archive or Content DM or dSpace or Islandora. Cliff mentioned JSTOR form, which is sort of our content management system we use. A few people use that, but most of the collections are coming in, exist in other repositories and we're simply harvesting them and putting them on the platform. Okay, so what are we seeing so far? Well, in the first, you know, 850 collections, we see that image content is still dominating what's in repositories. If you look at what's in the community collections, over 60% of it is image content. And that's what people have been putting in. We're seeing more text content come in now. One of the things I was really surprised by early on is we didn't see a lot of electronic theses and dissertations put in, which I was really surprised by. I really thought that would be the most common thing that people would put in because that connects to the secondary literature so directly. We're starting to see more of that now. I think rights issues kind of fed around that one a little bit, but I think we'll see more of that coming in. As I mentioned, Google referrals have grown and are growing significantly. And, you know, stuff that is already indexed by Google, the canonical version at your repository, because of JSTOR's just high use in Google, it helps boost those rankings as well, even the canonical version. I did mention it looks like open is working. And as Cliff mentioned, we also have the ability to put things on the platform that institutions can only serve back to their own institution. They haven't cleared the rights to make them open. And we have more people who are putting things on in that space. Preservation is an immediate priority. So many of these institutions have said, we may or may not host things on your platform, depending on what the evidence sends to us, but we definitely want to preserve things with you through Portico. So we're hearing a lot more of that. So to Cliff's point, we have work to do to make that. Portico takes this content in, and then we're kind of used to serving it back to publishers or showing it back to publishers. So we need to have a better way of showing it into the library space for the needs they have. One of the really curious things that has come up during these last two years is this almost universal cry for metadata enhancement and enrichment services that JSTOR might be able to provide. This has been pretty much with almost any institution we've spoken with, no matter how big or how small. And this is particularly interesting. We've sort of begun to develop a tool where we can feed metadata in and kind of score it on different dimensions of discoverability, preservation, inclusiveness, it's accessibility. Trying to think about a tool like that that would really help institutions understand things they might be doing to help their content score better on accessibility, for instance, or be better in the preservation side. And then libraries want to understand that and decide, okay, these are pieces we can take on ourselves, but maybe it'd be interesting if JSTOR could take this on and give this content back to us. So this gets beyond transliteration and adding all text and things of this nature and gets into some artificial intelligence, sort of machine learning ways of understanding with it. So we have a lot to do in that space if we're going to provide a service like that, but we're looking at what the possibilities, where we might get started might be interesting. Finally, I'll just mention that we're still in a charter period. So institutions who want to dip their toe in the water and try a couple of collections out, we're still inviting institutions to do that. Through the end of this year, there's no fees involved. You can get stuff on the platform. I think Cliff would attest there's not much effort at the institutional level that has to be done, particularly if we can just harvest it from your repository or you can send us a spreadsheet. We take care of mapping the thing to the JStore metadata schema and getting stuff published onto the platform. We're looking at providing the preservation service later this year for folks who want it. There's a lot of work that is going on the platform right now. We probably spent the first year working on search, trying to make search work effectively across these different content types. Obviously that's an ongoing thing that never stops, but we're really focused now on how to connect the content more conspicuously. So making sure that when someone comes in and finds something, a journal article or a book chapter, how do we connect them to stuff in the community collections that might be related to that and vice versa. Ithaca has launched a text and data mining service called Constellate. And there's been a lot of interest in pushing this content into Constellate. So folks can do text analytics on it. And we're looking at that. And of course, the metadata enhancement. I'll just say one other thing before we start taking questions. The audio video piece is a really important thing for us. It's on our roadmap. It's just we are trying to make sure we don't put things on the platform until we have the tool sets built to actually use them. And so audio and video are definitely things we're looking at and doing discovery work on right now. But I imagine it'll be a little bit before we actually get that content on the platform because we want to make sure the tools are there first so people can use it for research and teaching. So I'll stop there. And Cliff now I'll take questions. Thank you. I was wondering if you could unpack a little bit more about the LIF because you were mentioning challenges around that but I didn't see that in the next step. So I was wondering if we could connect the dots. Is this on here? Can you, is this working? Yeah, it's on. Oh, it's on? Okay. Yeah, that's a great question. And I almost want to just defer that question or ask Bruce the question because, so it seems to me that there are some technologies from the triple F standards that are at work already in the platform. You can see them. You can sort of zoom in very seamlessly on items that are in community collections. The problem is that there's no way that I know of and I don't think it exists, correct me if I'm wrong, to like serve out a triple IF image link so that you could say I want to see this in like a P&G format at this resolution and then put that into like our archive space link. That's what the curators would like to have. And at the present, I don't believe that's possible. That's correct. We've only implemented part of the triple IF framework and that is on the roadmap to do, to try to more fully incorporate the framework in there, but we've only done part of that up to this stage. To be honest, I was thinking about that in relation to your right issue too because you've been carrying some of the right things and then you have to flip all of that. Right. Yeah, that's a great point. And yeah, so I think as you think about integrating those standards, there are probably some solutions that would work well for our curators that we could think about implementing. Yep. So if we don't have someone speaking right away, if there is, that's great. Otherwise, I'll invite my colleague Nancy to maybe share her thoughts too. Just curious what the screen you showed us with the A on link comment. Where is the Robertson thing? Which system was that coming from? What this was, archive space? I think it was, I think this is archive space if I'm not mistaken. I was just curious whether it was Portico or JSTOR. Yeah, so curious then how the A on link was working. But it's local, so it's just a local link. Right, so they're putting in the links then directly into the collection so that people can go from the finding aid into the collection itself. But you wouldn't have that link in JSTOR. Well, the link goes to JSTOR? Oh, it does. Right. So if you clicked on that image, then you would go into community collections to that image in its context in the collection. But the A on request would come back to your A on request. Oh, right, right. If you're making an A on request, I'm not actually sure how we've done that to be honest. That's a good question for me to go back and see. Yeah, I was just curious because the A on integration is obviously interesting. And then also wondered, I mean, Bruce, you had your two hypotheses. I'm wondering how well you're matching against the JSTOR materials. Whether that hypothesis is... Yeah, so I'll say that we have just started that work, honestly, Lorke, and the connecting of the secondary literature and the primary literature, I mean, secondary literature with the primary source content on the platform, that platform work on our platform really just began last quarter. So I don't think we'll start to see the results of that probably until the fall when that gets completed. And that scope of work is quite large. So we'll be starting, I think, with small things and growing out as you all probably know very well at OCLC. Hi, Bruce, Jason Price from Skelk. How are you? Have you done any comparisons on the... So the first hypothesis, the making it open and it's gonna reach more folks part, have you done any comparisons with local usage of the same content and what does that look like? We have. It is, in some instances, it's really been a lot higher in some instances, it has not been. And some of that comes into the way... Did we get embedded text, things of this nature to make things findable? Some of it is just us getting better at being able to surface things in the search. So it's kind of in the eye of the holder. I think one of the things I wouldn't... What I wanna see is happening is that usage jumps off the page at people the way that it did with when books, when we started like chunking books in the chapters and the usage that jumped off the page at people like, holy smokes, look at what's going on. We're not seeing that yet. But I see it sort of trending up into that direction. But I think we still have a little ways to go on that. Nancy, do you wanna share any thoughts? Well, thank you. You covered it so well, Cliff. I don't know what it is. Bruce, I'm still waiting to see pricing after the pilot here. And actually, Cliff knows, we've been in a lot of discussion about the preservation side too. So I think that's where we're seeing a lot of promise to really work these together. So... Yeah, I was telling Cliff before, I think we're gonna have the fees understood probably before the end of June. Yeah, I'm the budget person. So yeah. Again, the premise here is we're trying to make it as low barriers as possible for people to put stuff on the platform. So that hosting fee has got to be, in some sense, kind of a no-brainer for people. And then the preservation will be tied into with kind of what it costs us to preserve that. And I imagine that will have a function of storage in it that we'll have to work through. All right, I think we're at time. Thank you all very much. Enjoy the rest of the year. Afternoon. Thanks everyone.