Conference. Um so hi everyone, I am Brian. I am on the planning committee and I am here to introduce the first keynote. Um when I was a younger librarian, there was this sort of this species of webinar that was like future tech what's happening so this is in the before time before covet before ai before cryptocurrency maybe even before Instagram right there were these sort of webinars about this and maybe they still happen now And um I think these were kind of low-key toxic. It just had like librarians kind of chasing uh the the ghost dragons of like Silicon Valley press releases, right? But of these webinars that were happening, there was one librarian that sort of could see the big picture. was willing to be critical and uh most importantly understand the value of copyleft and open source uh for libraries and that was Jason Griffey um And then of course I'm okay who is this guy? He's insightful and he's also funny and charismatic. And I come to learn that like, okay, he uh actually lives and works here in Tennessee. So he started his career at the universit University of Tennessee, Chattanooga. And of course, soon enough, other people, not just me, started to notice that he was really in insightful. And soon enough, he became a fellow and affiliate at the Burke McKlein Center for Internet and Society at Harvard University. Never heard of it. And then he goes on to become. uh the director of strategic initiatives at the Nash at the National Information Standards Organization, NISO is how we normally uh call it, his latest book Wait for it. It's titled Standards, which came out this year, co-authored with Jeffrey Pomerance, and was published by MIT Press. And uh on your blog, Jason, you said you were very proud that you were published under MIT Press. Rightfully so, brother. Someone who's been a personal inspiration to me in my career, I'm proud to introduce uh Jason Griffey to AspenCon. We'll probably go to the next one.

All right. So now we get to do the uh we get to do the tech live. I do live here in Tennessee. Hi everybody, I'm Jason. Hey. Um I do live here in Tennessee. I live in Suwannee. How many people know where that is? Yeah, awesome. Um love, love living there. And normally from Swanee to downtown Nashville is about an hour and a half. Except this morning, when I left my house at 6, thinking this will be easy. I'll get there early. I'll have some time to do a tech check. I'll have coffee. And uh that was not to be because there was you know normal traffic around the airport and normal traffic around everywhere else, and I literally just ran in the room. So um we are going to do our best to see if this works um before I get started. Sorry. Apologies for doing this live. Normally we would, you know. Have done it a little earlier. Alright. So ah, there we are. Awesome. We did find Now let me see if we've got the slides in the right place. Look at that, that even worked. Hey, cool. All right. So um Hi, I'm Jason. I am a librarian, but I I usually call myself a librarian and a uh technologist. I've been working in technology technically longer than I've been working in libraries, but you know, when you show up in a library and you you tell your dean on the first day you really want to be in reference and instruction and all of that technology, you don't want to deal with that anymore. She looks at you and smiles and grins and says, sure, Jason. And uh so I I ended up in technology whether I wanted to or not, frankly. Thank you for the incredibly kind um introduction. Uh I it it is always uh nice to know that people pay attention to your work, even um even if it is mildly embarrassing. Um but I want to tell you a little bit about myself and sort of where I am and what I do and where I come from. before I talk about the actual topic of the talk today, which is all of the ways I've failed in open source. And why I still think open source is valuable, to be fair. Let's be clear. So yes, indeed, I I am a librarian. I have uh an MLS from UNC Chapel Hill. Where I did think I wanted to be a reference instruction librarian, but again the call of technology was too strong. Um I ended up uh being a fellow at the Berkman-Klein Center for Internet and Society at Harvard for a few years. Which was a fantastic opportunity to sort of expand my vision of how the world works because it turns out they know a lot of people there at Harvard. I did move on to the National Information Standards Organization. My career track has basically been help an organization, help a bigger organization, help a bigger organization. So I I've tried to sort of expand my uh my scope a little bit. With NISO, I was the director of strategic initiatives, which basically meant that I was the one looking for emerging technologies. That's my gig, emerging tech is sort of what I do. And trying to figure out where that works. Sadly, because of Wave's hands at the world, I was let go from NISO back in May. So uh if anybody knows anybody who likes uh wants someone to work on emerging technologies, hi, come talk to me. Um I have published a bunch of things. The two that I'm going to point out just because I think they're interesting is I was the editor and co-author for Artificial Intelligence and Machine Learning in Libraries, which was written in 2017 and 2018. So, ha, emerging tech. Um and is only partially dated with you know the explosion in the last uh six years. But This was written before uh before uh Chat GPT was even a thing. So pretty cool. Um and then uh uh as uh was pointed out my uh current book, the most recent thing that I've published is um standards. with Jeffrey Pomerance uh via MIT Press. This is their um sort of Cliff Notes for Smart People uh imprint, which is uh a really kind of fun, breezy introduction to why standards are important in the world. And I think it's something that people would or should be interested in, I hope. But what I am here to talk to you today about. Excuse me. is uh my two open source open hardware projects that uh ended up failing and talk to you a little bit about what they are, what they were, why they failed, and what I learned from that. So I did two open source projects designed specifically around libraries. One of them was a library box, and one of them was called Measure the Future. Has anyone ever heard of either of these things? A couple of them, okay. Cool. Yay. Um if you want a copy of my slides, there is a library box in the room that is currently projecting Wi-Fi that you could connect to. and uh and grab my slides if you would like to do that. I'll obviously share them in the in the traditional way in the conference portal and everything. But, you know, gotta do what I can. So open source and libraries and uh my failures. I just want to talk a little bit about the projects themselves. I'm gonna give you sort of an idea of the history of them, the rise and the fall. And then uh at the end of the talk, I'll talk about obviously the lessons learned from that. All right. So we're going to start with Library Box, because that was my first open source project. It was the first thing that I did. that I worked on that was kind of fully transparently out in the open uh for library. So what was Library Box for those of you who don't know? It was a hardware project. Just set that right up, just like that. That was uh a form of portable private digital distribution. This is basically a battery-powered web server. That you can connect to the Wi-Fi and you can grab things off of at will. So It was designed to be a place where the right this was uh I'll talk again, I'll talk about the history and such, but this was sort of Uh 2012 when I uh worked on this when I when I sort of had the idea for it. So this was like Very early in the mobile phone explosion, very early in sort of broad-based uh cellular connectivity, especially uh data, uh high-speed data. through cell and in rural parts of the US and in uh developing parts of the world, uh the ability to sort of grab educational materials locally rather than having to use very expensive data connections was a big deal. And so this was a project that was driven basically open source. uh software and then commodity hardware. We we effectively targeted the cheapest hardware we possibly could for any given area so that it was the most affordable for the people in that part of the world. So we had a sort of a small range of of um commercial uh hardware that we used. This was about a 40 at the time, 2012, 2013, about a $40 portable travel router. that we um that we used as part of our platform for the for the uh software. Um so this was Um originally a project called Pirate Box that was designed in 2011 by an artist actually at New York University called David Darts, named David Darts. And David's idea was effectively a um uh an artistic statement on copyright, let's say He developed a portable router that fit inside a lunchbox. This was like lead acid batteries inside. It was like again, 2011. I know it's many years ago, but try to you know, try to put your brain back. Um and his idea was wouldn't it be cool if there was a little web server that you could just carry around with you that people could connect to and they could put things there and other people could take things away, right? So this like portable pirate radio digital thing. So it was it was an an art project. So I um saw the project and then thought, well, this is kind of cool. I could You do some changes and I could uh make this an interesting thing for libraries to use out in the world. So in the spring of 2012. I built one, sort of a proof of concept, and cheekily put it inside a book and took the book with me to computers and libraries and talked about it. And that garnered some interest. People were like, this is cool. We could put this out on bookmobiles and we could put this, you know, we could take this with us out into the world. And so I uh kept working on the code and got to sort of a 1. 5 version which was much easier to install and much better. It's designed to be sort of a much easier less of a heavy lift. You didn't actually have to code much to do it. You could sort of just flash some firmware and be done. And then We I started getting like support requests. Oh, oh, open source, the wonders of open source. Hey, can you do this for me? Sure. So libraries started contacting me and saying, like, well, here are some things that we could use, that we we need some things for this to be a useful product for us. And I was like, all right, well, I need some help. Like I I need help. I had a full-time job. I was the head of IT at a academic library. Like I I couldn't really do all of this. So I needed to raise some money. And again, this is 2013 at this point. Very early in the days of Kickstarter. But I thought, let's try it. Let's just see. Let's just see what happens if I put this up as a Kickstarter project. and um see if anybody's interested. So I uh launched it officially at uh ALA uh annual in Chicago in 2013. Uh had a little like you know, printed some business cards and like, you know, walked around and told people about it. I was like, hey, go check out our Kickstarter project. Um that after I explained to them what Kickstarter was in 2013. It started sort of organically getting a little bit of um a little bit of uh traction. I had set, uh for those of you who haven't used Kickstarter, don't Don't use it regularly, um I set a goal of three thousand dollars. I thought, what do I need? You know, I need a few hours of a developer's time. I don't need hours a lot of time because I can do a lot of it, but I need a few hours of somebody who can sort of dig in a little more deeply than me. You know, if I uh okay, so let's let's say twenty hours of their time, I'll raise three thousand dollars and I'll figure out how to hire somebody. I'll get them to do it, right? Cool. So I set a goal of $3,000. And then this guy, who knows who this guy is? This is Corey Dochtrow, the author, sci-fi author, slash sociologist, slash digital guru. Um Corey was at ALA that year and he saw me talking about library box and um after he saw that he posted his on his at the time most popular blog in the world, Boing Boing. uh about it and then he uh before twitter was a Nazi hellscape um he uh posted on his personal site that or on his personal Twitter that uh he backed it and then things got real weird. And uh so I ended up with this open source project where I thought I was going to have maybe 10 people interested and I wanted $3,000. Um I ended up selling almost 500 of them over the course of it and you know 10xing the ask uh for it. So It was extremely successful, and that was extremely uh stressful because success is its own form of headache. But it did give me funds to actually pay a developer, right? I could hire someone to help me with this open source project, and we could open the we could do more work with the code and we could get it out. 2. 0 incredibly successful. So uh then we kept getting requests because support requests never die in open source And a lot of those had to do with the international nature of it. I'm going to talk a little bit about the sort of spread of the project um here in just a second. But we had a lot of requests for like, come on, you need languages other than English. Duh. Um we need to support more hardware because the hardware we were supporting was very localized to the US and Europe, and we needed to expand into hardware that was more available in other parts of the world. We needed to do more work with our responsiveness on it, et cetera, et cetera. And so I was able to find some more funding. It was great. Things were going fantastic. In twenty fourteen we received a Knight Foundation prototype fund to keep working. On the 2. 1 release for Library Box. This allowed us to expand into a bunch of different languages, which was fantastic. And the 2. 1 was released in public beta in October 2015, which is, I am extremely aware, exactly 10 years ago now. And uh the official release was in December of 2015, which was fantastic. And that was the last release. We'll talk about why in a minute. But while this ride was going, if you've got the timeline roughly in your head right, I started in about 2012 and then end of 2015. was the last time we released any uh code for it. So it's about a three-year run. For a typical open source project, that's not bad. That's for the sort of hobbyist-level open source project. It's not too bad. But we had a lot of users. We actually had quite an impact with this particular project around the world. we had uh libraries actively using it in their um in their work, right, in outreach to the communities. We had uh all around the world. This is the State Library of Queensland getting ready to build a bunch of library boxes and take them out into the remote areas of their service area. We add things like the Beaverton Library in Oregon doing uh work with it. And then it started popping up in places like France. This is a street library kiosk in France where um they have information about their library and then a library box serving files to the um to the community. It showed up somehow at South by Southwest. They were using it to distribute um books and media by walking around South by Southwest. It showed up again, oddly, uh the Monterey Bay Aquarium was using it to take scholarly work out to researchers who were doing research for them on remote Pacific islands where there were no other connectivity. They would load a library box on a boat, take it out to the island, hook it up to a uh solar panel, and then the researchers could get to their research um, you know, that way, because they were in the middle of the Pacific. It was being used on army bases to bypass the government uh restrictions on what could and couldn't be seen on the internal uh computers. The librarians could preload it with things and take it um uh take it to uh the base so that people could get access to things that were otherwise uh otherwise blocked. And then companies started using it. SparkFun, if you're familiar, if you do any sort of hardware work, SparkFun is sort of like Adafruit or one of the other. open hardware vendors. Spark fund started using it for their class materials in their classes they were teaching. The American International Health Alliance used it, and this is one of my favorites, uh, for Zambia Rising, an educational uh endeavor for children. in uh Zambia and 26 of them in South Africa were used for AIDS education out in the remote areas of South Africa. And for some reason IBM started using them, and I don't even know what happened here. This was just like you're IBM, what are you doing? Um but uh I've had somebody sent me this from uh a conference where IBM was presenting and um they were using library boxes to give uh access to manuals for hardware, which again, weird. So uh this is to say, right? We had three years of run. It actually had some traction in the world it was being used. We um one of the disadvantages of developing a uh piece of software and hardware that is designed to be off of the internet is that it is off of the internet, which means that you can't tell when it's being used or where it's being used. I don't know, because it's not connected to anything. Um so I I I had a little open Google map where I let people tell me that they were adding one. And so this is sort of the slow motion growth. The last update that I had for the map was in 2017, two years after we sort of released our last release, but this is uh sort of the spread of the uh project around the world. At its sort of top, we were on uh uh 40 US states, 46 different countries, and on all seven continents, I had someone when I inquired on our uh discussion board, um I said, you know, I we're on six continents. I would love like, you know, go go Antarctica. And I had a researcher contact me and be like, oh, I'm going there in a month or two. Do you want me to take one? I was like, yes, here, please. Here is some hardware for you. Please take one to Antarctica. Anyway. So we ended up on all seven continents. So um so the question, right? The question for open sources, like what happened? Why did it fail? And part of that answer, I would like to believe, is not my fault. Right? Nobody wants to believe it's their fault. And partially, partially, it wasn't entirely my fault, but um in 2015 The FCC proposed a modification prohibition to anything that had a software-defined radio That includes routers, that includes travel routers, and anything that has uh 2. 4 or 5 gigahertz um you know band Wi-Fi chips in them. In um so the the FCC prohibition in 2015 basically said the the rule change was um in order to protect Wi-Fi. The FCC said anything that has a software-defined radio chip on the 5 gigahertz band has to have a lock on it to prevent the power from being modified by a user. So no other restrictions. That was the FCC's rule. It just said, hey, you're not allowed to crank the power on these radios up and down because it can mess with things, right? It can blow out the ability to connect to Wi-Fi. But that meant that the easiest way to comply was for manufacturers to just lock their firmware and prevent any changes at all. So instead of putting some like subtle restriction in, TP Link, who is the primary vendor that we were using the hardware for. Um just said we're just gonna encrypt our firmware, no key, no access, good luck, goodbye, thanks. We're you know, use it the way it's supposed to be used, and we we're not interested in you hacking it anymore. And so by 20, late 2016, early 2017, um they had effectively locked us out from being able to modify their devices anymore. So to one one degree, yeah. That that that's not my fault. Like it's not my fault that the FCC changed the rules on me. It's also not my fault that uh a lot of you are probably thinking, yeah, but why didn't you just like use a Raspberry Pi or something because like those are around, right? Um Uh why why the insistence on using sort of this commercial hardware? Just use something open. For those of you who are thinking that and weren't in the middle of hardware. The Raspberry Pi didn't actually have wireless. It didn't actually have Wi-Fi built in until 2016. So the entire time we're building this, like it didn't even have a Wi-Fi chip on it. And the little USB dongles and everything you could buy were wonderfully stable. Not at all. Um and if you bought The board and the adapter was way more expensive than just buying the little piece of hardware that we had that we were using. So we had sort of this failure to see where things needed to go, like when the FCC made the rule in 2015. Me and my team, the little couple of developers that were working with me at the time, we probably should have seen where the road was going, and we didn't. And that's probably our fault. And then this is something that might sound familiar to anyone who's ever done open source anywhere ever. So this is the team that built PirateBox. And the gentleman in the center there is Matthias Strubel. Who um ended up, he was the primary developer actually for uh for Pirate Box with David, uh who's the guy in the hat. Um I actually after the Kickstarter uh campaign uh got Matthias to work with me directly on Library Box because he was the he knew the code better than anybody, right? He had written all of the original code. So I just went to him and said like hey, would you just you know, can I pay you to help me with this? Um and life happened and he couldn't do it anymore. Right? Uh imagine in your head that XKCD uh cartoon with the one little you know big tower of blocks and the one little block holding it all up, right? Insert that here. Um uh he just had things that had to happen. And while I like originated the thing, I am not the developer he is. I I don't write the things he writes and I just I did not have the skills to keep up with it. So we failed. And we failed because partially FCC threw a wrench into our hardware supply, right? We had enough ongoing revenue in the project to continue as we were, but not enough to pivot to a whole new hardware stack. Like we didn't have sort of the runway necessary. Again, probably mostly my fault. The limited developer pool meant that when we needed to find options, we just didn't have them. That was, again, sort of our fault, right? And then also sort of out of our hands, but the change in the overall mobile landscape reduced the need as well, right? Like things got better. Like coverage got better, coverage got cheaper, data stopped being metered. in a lot of places. They just it got a it got to be a less attractive product. And all of those came together To mean that by 2015 we weren't writing code and by 2016, 2017, we basically shut the whole thing down and said, this is a dead product at this point. Like, you know, we're done. So, that's failure one. Failure two. Thank you, thank you. No, no. You know, everything, you know, nothing ends except that it ends badly, right? Otherwise, why would it end? So our our second failure, my second failure in open source, uh was a product a project called Measure the Future. Um measure of the future was in my head a fantastic idea. Aren't they all, right? In your head, all ideas are awesome. This was 2015. So just as library box is sort of falling apart, I'm like, there's gotta be something else. We need something else. And the idea for this project, for this uh bit of library code, was Uh a sort of Google Analytics for the building. We this again, 2015, 10 years ago. So, you know, all of the All of the things you know about now as far as like measuring uh uh door counts and little sensors that tell whether you're seated at your table and cameras that watch you and all that, like try to go back to 2015. when this was actually still a cool idea, cause now it sounds kind of silly, but trust me, in twenty fifteen it was uh it was a little more interesting. So the idea for Measure the Future was to use open hardware. We were going to concentrate on making sure that all of our hardware was open so that we did not get trapped by the vendor sort of TP link. Problem that we had with the library box. And of course, open source software, right? And we were going to combine a bunch of those two things, and we were going to use computer vision to do some attention analysis in library spaces. So we were going to take uh a Raspberry Pi and a camera and we were going to put it in a library space and we were going to use computer vision to analyze how people paid attention. Where did they stop? Where did they browse? Where did they just move through quickly? How much time did they spend in there? All of those sorts of things, right? And the idea was like you pointed at a new book display, or you pointed at a shelf, or you pointed at a computer lab, or you were wherever, wherever you wanted sort of those sort of metrics. You could build one of these, throw it on a shelf, collect some data, make some decisions about how to change things, right? You could A-B test your spaces. Yeah, that was the idea, at least, right? We very quickly iterated through a whole bunch of things. We just we uh you know I built built built built built lots of hardware The development stack was basically, you know, Linux, Go, Open Computer Vision, React, and Postgres. Super straightforward, trying to make sure everything was open and easy and trying to make sure that you know stuff that people knew, trying not to get caught by the lessons I learned with Library Box. So what did it do? How did it work? I'm just going to show a very, very quick short video here because I think it is pretty cool what we did. So this is how the the software worked, right? Use some computer vision to just identify individuals Identify people, not individuals, sorry. And then uh count, you know, put points where they were, center of mass, and um build that up over time in order to produce heat maps and things like that. advantages were things like we weren't taking pictures of people. All we were doing was recording the door the dots. We weren't like videoing, saving video. Very, very, very privacy forward. I'll talk about that in a little bit. The system was designed to have what we called a scout, which was our little camera unit, and then a mothership, which was a little um server that that would talk to it, and then you would interact with the mothership. um through a browser in order to download your um statistics, your images, your whatever. Why did we not just put it on a network? Why didn't we like Hook it to the Wi-Fi in the building. Why didn't we? Well, because IT departments don't like it when you walk in with a with a random little thing and say, hey, can we put this on your network, please? That doesn't go very well, right? Uh all of you have probably had that happen. Like, no, no, no, you can't. You can't actually put that thing we don't understand, didn't build, and don't have a warranty for on our Wi-Fi. And so we built it as this little self-contained thing. Um so that uh it would be easier for libraries to roll out. It also meant that if it's not on the network, it can't be hacked as far as uh like trying to get in and take over the cameras or anything like that, right? It's closed systems. Anyway, um so we ran some um demos and some beta sort of uh testing and things. Um And uh I was lucky enough to do so. Does anybody recognize the library? I'll be impressed if you can tell me the library. No, cool. This is the New York Public Library, actually. This is the Rose Room in the New York Public Library. So I was able to like test it, right? Like New York Public Library, Rose Room. This was fantastic. We were able to run uh several weeks worth of um of tests. We put in a system for them to be able to use At the time, uh this was oh gosh, I'm gonna forget the year actually because um they all blend together at this point. But the uh the reading rooms in the New York Library had been uh New York public had been closed for Eighteen months, two years at this point, and they were just about to reopen, and they were very interested in sort of seeing how they were used when they were reopened. Um and so that was sort of our little time period to be able to get in and and um and try our project out. So we, you know, did some uh did some data gathering. Um the design of it, right? I'm going to libraries and I'm saying, hey, I've got a camera. Can I put it in your library? That goes real well with uh with librarians. And so we were just aggressively privacy focused, right? We we designed the system so that there would be no photos or video of patrons at all, ever in any way as saved at all. you know, sort of period. All of the communication between everything is secure, uh nothing's on the net and open to random uh random attacks. Um we worked on thinking through sort of how to avoid identification of individuals in spaces at all. So we Anytime we saw a single person, we sort of rounded that into the next group so that they weren't identified as individuals in case they there were other sorts of cameras in the space. We averaged 15-minute reporting blocks so that we didn't have uh instances of being able to tell when only two people were in a space. We did all sorts of work to try. and make it as comfortable for library use as possible. There's a lot of thought in the project, right? Uh we had about eight beta sites that we were able to um to install. Uh these beta sites were happy to pay us to be able to move forward. Um and we still failed. It still didn't take off, right? Um so failure number two. What failed? Um part of it we couldn't grow fast enough. Um part of it was just revenue. Um we needed a lot of development, but we didn't have the revenue to fund it, and we were stretching our small pool way thinner than it should have been. Asking people to do things more quickly than could be done and disjoint between what we needed and what we could actually do was very heavy in the project. We had a lot of librarians actually in um libraries who were super interested in hacking away at our code. There were actually quite a few that pulled our code and forked it for one reason or another and did a lot of work on it. Um and then never rolled the code back to uh us. They did like legit improvements. And then for one reason or another, whether it was library policy, state policy, right? Lots of reasons. But they forked it and they never integrated. They never pushed it back to um to core. So um That was a problem. And I'm uh you know, this was a small project. It's not like we could like go after the state of Iowa with a GPL uh you know logger. I mean like They're they're they're being mean to our code. That doesn't go, you know, the state of Iowa is gonna look at us and be like, go away, kid. Um part of the reason we failed was a refusal to be creepy. We built it again. This was 2015-2016-ish when we were putting it all together. And um limited data collection, the m the method by which we put all of our data together made it look less useful than the commercial products that were out there. We made decisions about how we valued library privacy and then we had libraries tell us, oh yeah, but it can't do that. This product can do that. Well yeah, but they're collecting things you don't want them to collect. Yeah, but it does that. Oh my god. Okay. Alright. Cool. Cool. Cool. Cool. Alright. Nice. So um our our refusal to be creepy and our sort of um limiting our development by our uh ethical stances did hurt us, like legitimately. And then the lack of our sort we you know the in the the need for it to be local in order for it to be private, um we had a huge amount of people who when they came to us and we were explaining the system and like doing uh doing a look at it, they said, yeah, but I want to be able to sort of, you know, I want to be able to get to it from anywhere. I'm like, well, but it's not online. We don't it's not in the cloud. This is not a thing that we're doing. And he's like, they're like, well that's what we want. Okay, again, refusal to be creepy. Sure. Okay. Cool. Um so It fell apart. It just fell apart. We we had a product, um, it worked. We had cool stuff going on But we couldn't grow fast enough to support the needs that we had, and we were getting asked for things from our clients that we refused to do because we knew they were wrong. And we fell apart. So, why is open source and libraries so hard? Why is it so hard? It's so hard for a lot of reasons. I'm gonna give you a few you that in as I reflect on our work, as I reflect on the work um that I put into these products, projects, why is it so hard? So part of it is hard. Because libraries, by their very nature, and this is public, academic, sort of the US library structure, is hyper, hyper, hyper local. If you go to a library and you say, hey, we're building this thing for libraries everywhere, they will say to you, yeah, but I need this for my patrons. Okay, cool. Yeah, but you understand we're building something for everyone. It's like I don't I need this for my pe my patron. Right. And that's right. Like their job is to be the people that do things for their patrons. That's literally, you know, that is the job. But that hyper-locality, that like need to have mission, vision, funding all tied to the place rather than the idea. is uh is limiting in how projects get done. It's limiting on how open source projects get done with libraries. It's limiting on how technology writ large often gets done in libraries. Um so the hyperlocality is very challenging sometimes to get people to coordinate and do things with a bigger vision. Um funding is always a problem, right? Libraries um are underfunded top to bottom everywhere all the time. And uh nothing in a library is more challenging than ongoer ongoing forever funding. You might get, you know, a one-time thing and you can implement a thing, but You know, uh infrastructure, actually getting like a permanent line on a budget for infrastructure? Ooh, good luck. It's tough. And um that's been that it's been that way forever. Funding in libraries is just always a problem. Um and that is a it's a thing that keeps Especially smaller projects from growing, but even big projects, right? Even uh even uh even things like Aspen Discovery, like it's it's hard. Um organizational structures. I um talked about this just a a little bit earlier when I was talking about you know walking into a library and going to their IT and saying, hey, can I put this on your network? But libraries themselves are parts of bigger structures, and often they are beholden to those bigger structures, especially for technologies. Right? Academic libraries are beholden to their university IT for their infrastructure. Public libraries are beholden to their local governments usually for their infrastructure. And so those organizational structures have limiters on what they can and can't do and make it difficult. um especially sometimes for open projects, open source, open hardware, open whatever. Because there's nothing there's no Um there's no like there there, right? There's no like uh they can't go to Microsoft. What is the what is the nobody got fired for buying IBM, right? Like it's uh there's none of that behind it, or at least in the minds of the people. People that are often pulling the purse strings. And sometimes there are incentive mismatches. The incentive of a specific organization may or may not be the incentive of their cooperative or their consortia. spending, support, like there are often mismatches in incentives that are not obvious. And so that can also be a problem for open source projects. Sometimes these challenges are sometimes benefits. For instance, hyper-locality, especially for funding. makes it difficult for there to be a top-down imposition of things. So sometimes they're benefits, but often they are are just challenges. So patterns of failure, things I learned. This is basically just a reiteration of what I just said. But funding versus sustainability, often there's seed money, but no plan for maintenance. Often there are individual champions, but no institutional sport. One person cannot keep a project alive forever. That person will leave. Community enthusiasm versus community ownership. Just because a community is interested, that does not mean they are going to contribute. Applause is not the same as contribution. And innovation versus infrastructure what excites us isn't always exactly the same as the day-to-day need. So I'm gonna just whip through this one real quick because I know I'm running all on time. Thank you. Um patterns of success, uh alignment of mission, things like transparency. I mean open source is great for libraries because of these reasons. Transparency, shared responsibility, adaptability of the um of the of the work. And um open source aligns with library's mission for public good. Like all of these things are reasons for open source to be important in libraries. So I do think it's worth doing. Oh I think it's worth fighting for, that is. So important lessons learned, and this is the last uh the last slide with real content. Um governance These are three things I would like to sort of impart to you. One, governance is growth. If people are arguing about your project, it probably means the community is invested in it. So don't always take argument as a bad thing. Passion, passionate feelings mean people care. So remember that when there are arguments inside your community? Ownership is power. Moving beyond a single maintainer? Uh painful, but you have to. It's hard, but you have to do it. And finally, conflict is collaboration in disguise. This is related to the top one. But if people fight, it means they care enough to fight, and it means that they care enough to help make changes. So that's what I have learned about open source and open source development. Thank you so so much for having me here today. I'm going to say this is me. If you want to check out any of the projects, they're there on GitHub and you can look at them. Yay. If you want to get in touch, there's a couple of ways to do that. And I'm I'm I know I maybe got two minutes left or something. But I'm happy to uh take questions if that is the thing you want to do.

You have 12 minutes for questions.

Awesome. Beautiful. That is some awesome timing then. Cool. If anyone has questions, I'm happy. About about this or or literally anything at all. I'll opine on anything if you want.

Surely someone has a question.

Um so With a dependency on streaming and companies revoking previous quote unquote ownership of media like iTunes, do you see a renewed interest in something like Library Box, similar to the increase

I do see that. There um is a fair amount of um uh chatter online. I I do I keep my finger in a bunch of different online communities and um there is a a fair amount of interest in sort of a rising C copyright indifferent uh media acquisition strategy. Um so uh Uh I don't always piracy is a weird word and it has a lot of loaded um a lot of loaded connotation uh to it, but yes, I do think that um there are communities out there that are uh extremely afraid of the lack of ownership of media writ large and that um the yeah streaming is this the world of everything streams is a world where everything can be taken away. And that should worry all of us, but I think there are definitely people for whom something like a a private network of um of uh Copyright in different media might be uh something that they're interested in, yeah. Whether that's enough to bring the project back to life, I don't know. That's a whole different yeah, maybe.

Way in the back corner. Um while I walk over there, I'll tell you that uh I I bought a library box and Jason loaded it up with like totally like public domain copyleft Nashville specific media. And I accidentally deleted all the and I Oh no And I never and I never told him until today

Yeah, I was curious, do you have any tips on how to grow you said an individual champion to making it an institutional thing? Is there anything you've seen that helps move any project towards that?

That's a great question. Um the things that I have seen that have moved the needle on that are sort of the the the the boring work of building community. honestly like uh a lot of developers how many how many people in here identify themselves as developers i'm i'm guessing that's most okay how many people in here identify themselves as librarians primarily like for number one Okay, all right, so a good mix. Um so I will say that um that building community is harder than building code. And that um maintaining community is harder than maintaining code. And the stronger your community is the better chance you have of moving past that sort of XKCD cartoon um of the single developer ruining everything or the single project, you know, single library, whatever, ruining everything. Um the broader the base of support. And you only get that broad base of support through community actions like the one that we're having over the next couple of days, right? This is the way that you change that. Like legitimately. It's the way I've seen that work in other areas. So another question?

And if you're watching us online, uh please remember you can use the dedicated thread on Slack to ask questions about it.

Oh wait, no, I'm just kidding.

Not only online, but you can also answer those questions after the conference.

Oh great.

All right. Going once. Going once. Going twice. This is my gone.

I absolutely will, yes. I absolutely will. Yes, I love I love taking I mean, you know, it's a bit self-serving, but one of the uh if you connect to this, uh Let me see. It's uh 192. 168. 11, right? Uh-huh. If you connect to the Wi-Fi and then load up a browser. Um 192. 168. 11. It'll load uh and one of the one of the uh folders is Jason Griffey and it just has a bunch of my publications in it. So if you want anything I've written, it's probably on there.

Jason, can you uh excuse me?

No, go ahead.

Can you let us know can you let us know which question you just answered? Just for the Oh, oh, sorry, sorry.

The slides are actually already on here if you wanted to get them. There's a folder called AspenCon, and the slides are there as a PDF if you wanted to grab them now. You can grab them later too. It'll be fine. All right. Thank you all. Thank you so much. Sorry I came in a little hot there from the drive, but I appreciate all your time and I'm gonna be around for a little while, so feel free to grab me if you wanna say anything. Thanks.