We can do more together: sharing a collection management system

Since the mid-1990s, six museums in Western Massachusetts have shared one CMS. The arrangement enables significant collaboration on research and exhibitions, reduces certain costs and regional collection redundancies, and allows the public to search all six repositories’ collections at once. Using the example of a current project to correct metadata about and enhance access to Indigenous American collections, this session will detail how the shared CMS promotes collaborative data creation, management, and cost-sharing. Track:Big Ideas


Unknown Speaker 02:01
115 We will get started. We'll introduce ourselves in a few minutes. First I want to introduce the session this session is we can do more together sharing a collection management system and we are all colleagues at the museum collection management Commons project seated at the Five Colleges incorporated which is an academic consortium and we are located in Amherst, Massachusetts. And I'm gonna hand it over to Sydney for our land acknowledgement.

Unknown Speaker 02:36
Hi, my name is Sydney when Sorry, I'm lucky Piku UI Kuya and lasagna woman from what is currently known as San Diego in Southern California. We are currently streaming from what is commonly referred to as Amherst, Massachusetts. I'd like to begin this event by acknowledging that we stand on unseeded than it took land. I'd also like to acknowledge our neighboring indigenous nations the Nipmuc and the Waupun off to the east, the Mohegan and the peak was to the south, the Mohegan and the Will he came to the west and the Abenaki to the north. Thank you. Thanks, Sydney.

Unknown Speaker 03:13
So these as I mentioned, we are all from the museum collection management Commons project. I'm Erin Richardson. I'm the product manager for this product. I also run a web server or consultancy called Frank and glory

Unknown Speaker 03:33
oh and I forgot to identify, describe myself I am a cisgender, middle aged white woman with short hair and glasses.

Unknown Speaker 03:46
Hi, my name is Carrie Evans. I am the data specialist for the museum collections management Commons project and I am a cisgendered white woman with short hair in her early 30s. And in my office, and yeah, I'm also a maker a gardener plant. person and a loving cat

Unknown Speaker 04:11
Mom. Hi, my name Sydney when I am one of two assistants for indigenous American collections for the SAM collections management Commons project. I'm a cisgendered Asian and indigenous woman with black hair glasses and blue earrings. I'm also a dancer.

Unknown Speaker 04:39
Hi, excuse me getting over a cold. I'm Isabel Cordova. I'm the other assistant the second assistant for indigenous American collections in this project. I'm also an artist and I met this woman of color in my mid 20s With a dark wavy hair and a black T shirt on

Unknown Speaker 05:00
and hello I'm Hilary causal it I'm an IT analyst at five colleges Inc and part of my job is running maintain the back end of Mimsy also a booker person I'm in MANY and coordinate two of them. I am a sis white woman in my mid 50s I have bangs and I'm wearing a headset.

Unknown Speaker 05:23
Thanks everybody for introducing yourselves the picture on your screen is an example of a record from our public database portal that will mention later. It's a drawing by an American artist ally who Vetter was created about 1888 and I'm using this as a specific example of some things we're going to talk about later. So the database description for this work is a cup that is sort of a complex phrase close up of five sad women in veils under a dome. The picture also has some Latin text under the image which is not translated in the database record. The actual image is of five women looking out toward the left of the viewer all the women are wearing headscarves and the dome is actually an arched upper margin of the image created by the artist. The translation of the saying is leave all hope you who enter and I've made a stab at some alt texts for this image which is five women with light, light skin tone are crowded together, shown from the shoulders up all are wearing head coverings and appear distressed. So I use this as an example of the kinds of art artwork, data and information that we're dealing with as part of this project. And just a note about illustrations and artwork that are part of our slides. We're using them for illustration and interest purposes only. They're not directly related necessarily to the content that we're going to be discussing in the presentation. For those with visual impairments, we will read the alt text but not otherwise describe or discuss the image except for this one as part of the presentation. If you have questions during our talk, you can add them to the chat or you we will have time for questions at the end and we'll go by show of hands or if no one's feeling brave people can just unmute themselves. So let's go so here's the plan for today. We are going to share an overview of the shared database history and a little bit of its context to talk about some of the functional and technical details of how it works. discuss some pros and cons of the arrangement. Talk about some improvements that are underway or are on the horizon. Give you a concrete example of some shared database benefits and have time for questions and discussion. And the the picture for this slide is a word called telephone and it's an abstract geometric black and white diagram of a telephone system with wires and connections. So a little bit of history and context. Six different museums share one database way back in 1995 These museums got together to share a database. The Mead Art Museum at Amherst College, the Smith College Museum of Art, Holyoke College, Art Museum and Skinner Museum, University Museum of Contemporary Art at a mass Hampshire College Art Gallery and historic Deerfield. These museums do not share a common parent entity, but they do participate in the consortium. The campus libraries at Amherst College, Smith College, Mount Holyoke College, UMass Amherst and Hampshire College had all been coordinating library services with an iOS and OPAC arrangement since the 1980s. So by the mid 1990s, the museums were interested in seeing if something like what the libraries would do were doing would work for them as well. The collections of these six organizations are vastly different

Unknown Speaker 09:18
thanks so I'm gonna explain in a little bit more detail about how this works. So this slide, just to get us started shows a picture of a painted leather fire bucket from the collections of historic Deerfield. And now we often explain our instance of Mimsy x g. In terms of a bucket, and which is to say that there is one primary Oracle database that's hosted on a server that lives at UMass Amherst. And each organization that participates enters data through the Minzy interface into the same bucket. And so on the user end, the database is partitioned by museum view, and we'll get into that in a minute. But effectively all museums enter data into the same database and some of that data is then published to the shared public portal museums dot five colleges.edu. And I think it's also a little important to explain where the MC MC project, the museum collections management Commons project stands in relation to the data and so five colleges incorporated both stewards the data and facilitates communication and collaboration across the different museums. So that's who we work for. That's the educational nonprofit that administers this grant project. And then museum staff at each Museum, create edit and use the data so we're not creating editing or using the data we're stewarding it and planning for the future and of those staff. There's also a group of registrar's or collections managers that meet monthly to discuss things like implementing controlled vocabularies and finding and cleaning up bad or non standard and data skill sharing about different database functionalities, how to create and generate reports and also what data to provide to the public portal. And the museum's primarily engaged with the database through their own records. So the initial intention of this setup was really to facilitate information sharing, and so certain modules within the database are accessible and viewable to all parties. But some are partitioned. And so the FCI staff specifically Hillary, alrighty analyst can see everything and communicate with Axiell and z's parent company, and so I will pass it to Hillary to better explain the complexity of the data partitions.

Unknown Speaker 12:11
So we have a diagram here, sort of illustrating in shapes what Carrie just said in words, which is that everything is hosted on the server at UMass. So that's both the Mimsy XG instance of the software and the public portal, which is run through a PHP MySQL. Each of the six organizations feeds their data into an MC XG and can then edit and manipulate it. One thing to know is that this is custom code that was built by whoever owned the memzy software back in the day. We have been through MANY versions of the software and MANY holding companies owning it. And I have been here 10 years and I don't know the details of why which which things were chosen to be shared and which weren't. So the partitioning is based on as Carrie said on the record view of each institution. And so, for example, somebody at the meed cannot see Hampshire colleges records that relate to their objects, or their people or their facilities, but they can share and also mutually edit if they have permissions to edit, other shared authority files like Places Events, exhibitions and thesaurus. And then there are some others that are that are viewable like media publication records. And then there's a whole raft of Mimzy functionality that isn't necessarily being used, some of which is shared, some of which is not. So the the, the partitioning controls both viewing and editing editing is, is allowed not by institution, but by roles. So there's this sort of complex interaction between who can see and edit what and who can edit, at what level, delete all that all that kind of stuff. The connection between the Oracle database, which is the Mimsy records, and the public portal is handled by a daily push from an Oracle database to my SQL database. That means if something goes the data is not live and if something goes wrong with that connection, it can get stale. And only selected information is is even pushed over to that database and then not all of it is is displayed. So it is pretty complicated. And if there were, you know, a decision to share different things or or change how it's set up that would require more custom code. So just to be closed is not something you can just like buy off the shelf. On the other hand it it works for us and now I will pass it along. Back to Aaron.

Unknown Speaker 15:05
Thanks. So here's some things that are really great about this database. Some costs are shared and therefore reduced. So those costs include the annual maintenance contract with axial, the, you know, expensive Hillary. The original cost of the system was funded by an implementation grant back in 1995. system upgrades over time training so those are things that you know, you can imagine splitting six ways instead of paying for yourself. The public portal also allows a single access point to all the regional collections in the Connecticut River Valley. that are you know, how is it these organizations so museums, can they radically reduce redundancy, regionally. campuses, you know, campus art museums are usually lousy with works on paper, and works on paper are also you know, come in MANY editions. So it's really helpful for any of these database partners to be able to look if they're thinking about purchasing print. If one of the collaborating organizations already owns it, then they don't need to make that purchase. They can museums can share information with each other. And it also really helps in facilitating inter organizational loans for exhibition. There's a very long tradition of all the campuses borrowing from each other for exhibition and sharing a database makes that a lot easier. The shared tables, also theoretically reduced data entry time and effort. We're not exactly using that to the best of best of our ability. And we can talk more about that in the q&a if you have specific, specific questions about that and the image on the slide is called Geronimo's Cadillac. It is a hand colored photograph of three men in Native American attire, sitting in an antique car in the foreground and a big $5 bill with a Native American man in the center fills the background.

Unknown Speaker 17:31
So I'm going to talk a little bit about what could be better about the current arrangement. And so just to start off the two images in this slide and show two sides of the same coin from the mead art museum collections, and Amherst College, it's a Greek coin. It's made of metal, and one side of the coin shows a bird and relief, while the other side shows a crab in relief. Yes, and so two sides of the same coin. So with everyone entering data into one bucket, it's very important that museums that all the participating museums are on the same page about things like the definitions of data elements, whether or not to use certain data tables, what control vocabularies they're using, as well as protocols for certain data entry. Because the data published to the portal the public portal is ideally should ideally be consistent but data governance has historically been very limited. So the fact that we have six different organizations funneling data into one database, but the six organizations have very different structures, and parent organizational structures, so that makes decision making about data very complex. And right now, the monthly meeting of the registers is the only sort of proto data governance body for the museum's for the consortium and there are very few formal procedures documented. So this also means that representation of different museum roles and cultural perspectives has necessarily sort of been limited in decisions about data until this point. And because the governance is minimal, there's almost no formal controlled vocabularies that have been adopted. So most of the data are a lot of the data is entered in free text, and this creates barriers to access. So for example, this coin of the 1000s of coins in the entire five colleges, collections ecosystem, and some can only be retrieved in the public portal by searching for coins. Others have to be searched on numismatics, others on money, etc, depending on which vocabulary the individual Museum is using. And so linking data from controlled vocabularies either within, you know, elsewhere on the web was wasn't necessarily a possible possibility for museums in the 90s when this was implemented, and also the Getty 80 was kind of new at that point as well. So that is, all of which is to say, governance is a huge sort of step for us going forward in the planning grant as well. We talked a lot about the data partitions you know, I tend to think that those are a little bit of a scapegoat for governance historically. So there's also not a lot of validation that's going on in the database at this point. And that is part of our work here on the planning grant and the partitions are However, at this point limited limiting for things that you want to use a shared database for so copy cataloging, for example, from other collections, and sharing artist and maker and places records, collections management activities, like collection, development, or deaccessioning and exhibition development as well. So during the planning grant, we really learned about how these limitations can complicate future projects, like facilitated bold cleanup efforts or migrations or sharing consortium New Museum data with other disciplines. So Aaron will tell us about some of the improvements we are making as part of this planning grant and that we hope to make in the future.

Unknown Speaker 21:46
Thanks, actually, I think yeah, yeah, Hillary's gonna take it away with ControlLogix everywhere.

Unknown Speaker 21:54
Yeah, I'm gonna I'm gonna start with a little bit of a deeper dive into this issue of controlled vocabularies, which is a very rich one. Basically, early in the process. There was basically only free text entry and limited picklists and over the years the meeting of the the monthly meeting was devoted to at least agreeing on what should be available in picklist. So that there's a limited set of basically approved words to be used in various contacts, but they weren't actually locked down. So if people chose to think that they remembered what the right term was, and they made a typo, it would just be in there and we would find all kinds of interesting variations on permanent collection, for example, which you can you can omit all kinds of letters and make all kinds of different different things. So that's something that has been being worked on over time. But thanks to COVID and lots of time at home, just dealing with computers. We've actually made some significant strides there. We have come up thanks to Aaron and Carrie with a workflow for approving new keywords. That section is now locked down so non approved terms cannot be typed in. And the really exciting thing was importing a fresh, complete version of a current Getty Arden architectures to Saurus into the system. This was a module that had not really been used, and we are now working on linking objects to proper terms based on what's been entered in their object names and their materials and their descriptions. So there there's a lot of exciting work to be done there. So we went from kind of no control to pseudo control in some areas to now actually full fledged, controlled vocabulary in certain small areas, which we hope to expand.

Unknown Speaker 23:51
So I just realized that we did not explain. Everybody keeps talking about the planning group. I didn't really explain that. So I'm going to take a second to explain it. So you're seeing five people from five colleges Inc. on the screen here. Four of us are part of a Mellon, Andrew Mellon Foundation funded planning grant to essentially find out what's been happening in the database since 1995 and see what we need to do for the future. Prior to that, the only person seated at five colleges that was in like responsible for the database was a few hours a week of Hillary. So you're seeing what you're seeing on the screen belies what is is real about the support of this database. So back to the slide, that any so some other things that we've been improving upon, is including more voices in determining what exactly is the content of the records of the records describing art and artifacts in the museum's collections. So we applied to NH, this has not been awarded. I'm just you know, I'm very excited about it. Maybe we'll get the money, maybe we won't who knows. But we want to take what we learned from the keywords and object names, improvement projects and use that to create some intentional governance around artist identities, so that we can ingest some linked open data and contribute back to linked open data. And the intentional governance is is really about understanding that the museums and their staffs, our museum staff are very small and I've only gotten smaller during COVID and we don't have all of the knowledge to accurately describe works in the collection in a culturally appropriate way. So we want to bring in other expertise that resides on this campuses just not in terms of faculty and lived student experience that doesn't actually exist in the museums and is not part of the current decision making process about what our information is made public and open in public portal. We've also had some preliminary discussions about sharing a public portal with archives. So five college compass is an archives discovery platform that runs on Windows seven and or eight, depending on who you're talking to and what day it is. The museum's public portal is, as mentioned runs on Mobius and one is it possible to have one solution that replaces the two aging backends of these two public discovery portals? We've also been piloting. Sure, I'm not sure what that means at the end. That's not That's not what I wanted. To say. But Smith College Museum or in conjunction with the Botanical Gardens, archives of library is piloting shared common data model so that all of their different types of collections can be discoverable more more easily. And so we're testing that common data model on museums data more, more widely. The image on your slide is a silhouette of a rabbit standing on hind legs behind an overall psychedelic pattern in orange, blue and green, and this is called the White Rabbit in Wonderland. By Joe McHugh from the

Unknown Speaker 27:22
Smith College Museum of Art collection.

Unknown Speaker 27:27
So for a real life example of how we can actually work together and use the power of this database, I'm going to hand it over to Isabel and Sydney.

Unknown Speaker 27:39
Hi, I'm Isabelle I'm one of the postback just as a reminder, one of the postback assistants randonneurs Dungeness American collections along sides anyone as a model for how this collaboration might work, we were hired in a year long project. And we're currently in the middle of that project of evaluating some of the records for the indigenous materials across the museum's collections I wrote aka settler colonial collections because that's kind of a framework shift that we've had to develop as we're in our early stages of our project. And I just on that note, wanted to reiterate what Sydney acknowledged in the beginning of the presentation that these institutions all have really complicated histories, or colonial legacies that have been destructive for MANY of the indigenous communities in the area. So MANY of the museums and collections that we are researching are a result of those damaging practices. So I think it just shows how it's the need for this restorative work and things behind the scenes like, like the Metadata database for proving access. And research is one of those things that we're attempting to address. So most of our work thus far as an overview before I get into how we broke down the project, is just trying to evaluate and get familiar with what information is available on Mimsy exchange, the database that feeds these public portals used by all these institutions, and we're trying to understand whether or not the database and objects are stewarded well are accessible to the relevant audiences are you even utilize those tools for curriculum development or tools for education through these educational institutions, so and as was mentioned earlier, as five college employees and the consortium employees bear the only ones who can see all the records on menzi from each institution which is actually refining a unique position that allows us to see all these variations in the institution's collections. function and how that is in there, how that manifests in their Metadata. So as an overview for our project, it's shorter than we thought it would. It feels shorter than we thought, but we're organizing it into three parts so far. So first, we are un has been working to compile a comprehensive list of objects and render records or with indigenous or settler colonial the themes so either made by or made about or inspired by indigenous communities and those are across the Five Colleges collections and historic do your fields collection and the ones that are on the database. So we're compiling that. And after we compile that list, we have been developing rubrics to evaluate the Metadata for those object records. And so we're in that stage now and we're, once we make those observations, we're going to summarize those findings in a status report and develop recommendations in early 2022. And then finally, in that last point, we would like to re emphasize collaboration that isn't always best facilitated through the database by making connections with community members, museum staff faculties are starting those connections to because it's going to be an ongoing process with students and the public, and as well as other museums, libraries, archives and special collections in the area. So I'll pass it to Sydney.

Unknown Speaker 31:55
The larger questions that guy each phase of our research are, one, what are the current Metadata standards or lack thereof between the institutions to how can each institution lean into their own strengths and each other's and three, how can this assessment of records be a model for the larger project goals? As you've been working to try and answer these questions, we notice a lot of things but these these particular ones stand out. The first is that each institution has their own mission and their own style of managing their data. The collections exists in a variety of contexts. They serve different purposes and they function in different ways. Some are geared towards geared more towards display student towards tours and public programming, while others focus more on research coursework and or community engagement. The goal is to develop standards that in practices that accommodate all of this next is that there are variables and collections scale history, and collection policies. For example, Smith College has a small collection of indigenous objects, but it's focused on acquiring more art from contemporary Indigenous artists. While Mount Holyoke already has a large collection including a large number of objects inherited from the Skinner museum, but both primarily focus on display object study and classroom education. In contrast, historic Deerfield is a collection of historical houses and early New England material culture. All of these have vastly different histories. of collecting or actually not collecting indigenous materials. And their data illustrates this. There is a more diverse representation of objects and materials when looking at the collections in relation to one another. And there were a wider variety of what one can do with the objects when they have access to the entirety of the Five Colleges collections as opposed to just one institutions. Third, is that developing these frameworks are complicated and difficult but not impossible. In the final stage of our project, we hope to address first address the ongoing legacies of extractive colonial museum practices by forming relationships with the indigenous communities of the objects at the museum steward. Develop a standard from an agenda that accommodates the variability and change throughout and within the consortium and outline a framework that can be adapted to adapted to new systems and projects beyond to swim, the scope of what we're looking at. Kind of our thinking right now is that it'd be good to institute some kind of government system governance system that oversees the Metadata and ties together the colleges and institutions. This we think would contribute to a more robust database and consortium as a whole.

Unknown Speaker 35:06
Thanks, so thanks. For listening. I'm going to stop sharing my screen that's the end of our of our talking and we would love to hear questions from you. Carrie, I think or Hillary has been monitoring the chat. So if you put a question in the chat, we can get to that. And if you want to just raise your hand or just unmute yourself, there's not a ton of us in here. Feel free to throw your question out for the group of the group and we'll see if we have an answer.

Unknown Speaker 35:41
So reporting back on the chat no actual questions, but a fascinating backstory that apparently the origin of the MC n is actually tied to this idea of a shared system for museums. So this is this is really, really cool. And I'd love I mean, if we have time, you know, instead of questions to us, I would love to hear more about that. But let's let's have questions from you to us first.

Unknown Speaker 36:07
So I did post the link to the public portal in the chat if that's something that you're interested in looking at. You can that the public portal has not undergone any evaluation or user interface upgrades since the early 2000s. That right 2012 2012

Unknown Speaker 36:27
There again, there have been tweaks around the edges but we are well aware of the MANY aspects that are not modern.

Unknown Speaker 36:37
Hence are interested in potentially working with archives to improve.

Unknown Speaker 36:45
Claire Yeah,

Unknown Speaker 36:49
I as someone who has implemented a lot of dams in various places, I'm interested to know how you get this kind of anarchist structure of every museum in the consortium, an overarching, like person that says yes, let's go and do this. Let's go and do that. How do you actually do find that you've been able to get people to agree on making changes and additions? Is that particularly hard? Do you have any advice on how to make that happen?

Unknown Speaker 37:17
So I'll hand it over to Hillary in one second. But I want to say that the group of registers is incredibly collegial. They all get along really well together. It's a super fun meeting, which someone impedes the ability to actually decide something and I think the big challenge is how do we get to more decision without losing the friendly banter?

Unknown Speaker 37:40
So I think, you know, we were tremendously helped by two factors. One is just simply longevity. This was set up so long before anybody who's currently employed, knew anything about it, that it's got this real weight of history behind it. It's like we work together because we've always worked together. So that's tremendously helpful. And the other factor that's helpful is being part of this academic consortium where you know, to varying extents, different types of department staff roles, they're all different ways in which the Five Colleges work together. And the library's especially have been a super tight, active coordinated bunch since, as Erin mentioned, the 1980s So that kind of set a standard so I feel like everyone is everyone is totally willing and eager to share. But where it gets really sticky is actually moving forward actually getting everyone on the same page at the same time and especially any decisions that need to be moved higher up the chain. If the registrar's feel like well, this is something we need the directors input for on then you know, getting that bandwidth can be difficult. So I feel like the obstacles are more structural and procedural than anything and if there were more resources, and this is what this grant has been absolutely fabulous for is to have actual people with time who are able to wrangle the agreement. So it's it's not that anybody doesn't want to agree. It's that figuring out what a sensible way forward is and saying yes to it. It's hard to get to that end point. And this is really helped tremendously. Yeah,

Unknown Speaker 39:30
I would think Hillary that was a really, yeah, really lovely way to talk about the grant. But I also just wanted to share a little bit as part of the grant about some of the things that we found sort of helpful in getting a an understanding, sort of piloting, or scoping the work that it would take to to arrive at a formalized governance structure and at this point, you know, I think just some things I don't know if it'll be helpful to you and your work. But some things that we've noticed are, there's sort of an unspoken agreement. Or maybe it's spoken somewhere that, you know, a unanimous vote is not necessarily required. So if three or more museums agree, if over half of the museums agree on a certain topic or to implement you know, a certain use of a certain field that the other three museums aren't necessarily interested or, you know, in using at that point. So so that's the collegiality that Eric mentioned really helps with the non unanimous agreement. And also one of the things that's been really helpful is we've created almost like a, not necessarily an agenda for each meeting, but almost like a decision log. So at each meeting, where decisions are made, we are not only tracking the discussion, but also any agreements that were made mutually and then also, like action items for the next meeting, but with the agreements. We are sort of internally testing making those a little bit more accessible to the wider museum communities, in an effort to be a little bit more transparent about how decisions are made. So those, you know, tracking your decisions is a really good, good way to do it. I don't know if that's helpful, but yeah.

Unknown Speaker 41:45
Other questions? We've got about five minutes

Unknown Speaker 41:53
so a question in the chat. Is there any interest in switching to new software you can tell with what you have? Is there interest? Yes, yes, yes. Yes. We we would love to. We would love to have the ability to move to something that was more flexible. more in line with modern standards. You, the group of registrar's and museums, has been speaking for forever about the possibilities of change and that is an extremely slow process, both time consuming and resource consuming. So I think that's been one of the big hitches but um, Aaron, do you do you have anything to add to that?

Unknown Speaker 42:44
Yeah. So as part of this Mellon funded planning grant, we are engaging to external research studies, one of which is a user study, to understand perceptions and expectations of the public portal to really learn what students and faculty who are the primary audience for these collections want in a discovery layer. They Yeah, so the results of that will be available sometime in January. And then the second part, which is going to start by the end of the month is a gap analysis, which will really dig into what what do staff need? What how are they using this tool? What else are they using? That's not Mimsy because there's lots of stuff takes too long to learn how to use something and Mimsy. I'm just gonna make this Excel sheet. So there's lots of data that lives outside of NZ and lots of tools that are in use outside of NZ to understand that and then so the the sort of end result of that entire process will be an RFI or RFP for a successor system.

Unknown Speaker 44:03
Yeah, hashtag rogue databases. Indeed, competing we war war of the databases or the pseudo databases

Unknown Speaker 44:14
or just the paper files in a stack on someone's desk. Other questions? Okay, if there are not any, I'm good. We're gonna post our slides and the unscared in the schedule for for this event, and I may also post them on, on Slack. And yeah, thanks everybody, for your time and attention today.

Unknown Speaker 44:52
Thanks for coming. You did it. Yay. See you later at the at the wrap up the wrap up. Thank you. I have guys called TK.

Unknown Speaker 45:19
Thanks, Aaron. In the background. Yes. And Aaron's dog. All right. Have a good rest of the day,

Unknown Speaker 45:27
too. Bye, guys. Thank you.