 So, I would love to talk to you all about this new tool that you may have seen floating around our various networks called Lassie. And essentially, what Lassie is, is a retrieval tool. That is the bottom line. It is a thing to retrieve stuff from our networks. It's a new IPFS client that can fetch content from pretty much anywhere. It's written in Go. In that respect, it's a new Go IPFS implementation like Pupo. We've started from the ground up and we have a single goal, which is downloading your data in IPFS and Filecoin. It should just work. We basically say, if you want your data, just tell Lassie to fetch it. That's our motto. And essentially, the origin of Lassie is that as our networks have grown, we start to see a proliferation of some transfer protocols. Well, BitSwap is our bread butter when we see a lot of graph sync on Filecoin. We're starting to see a lot of folks interested in HTTP. And this is all very interesting from a programmer perspective, but from a regular person perspective, no one really thinks what transfer protocol would I like to use to get my data? That's not something that people care about. What they want is they want to get their data. And so Lassie is a multi-protocol retrieval client. So it can figure out what data is available on what protocols and figure out the best way to get it. We already support BitSwap and graph sync. We're working on a trustless HTTP transfer protocol. And then in the future, maybe we will also add some cool stuff like this new thing that I wrote is working on. What is it called? Bow or if you've heard of carpool or car sync. And we have another sort of like approach, another sort of aspect of Lassie, which is in addition to not worrying about how you, what protocol used to transfer, we don't want you to worry about how you find your data. So Lassie can find data through both the IPFS DHT. We use IPNI, the network indexer to talk to the IPFS DHT. And we can find content on the Filecoin network. We don't have to know where your data is. We'll just track it down. We can even track down some providers who don't put their content in the DHT through BitSwap. And as new content networks appear, we intend to add them to Lassie. So goodness, though, if you haven't heard of the network indexer, I think that's the best way to advertise your content. It's super fast and super awesome. So how do you use Lassie? So we've designed Lassie from the beginning with basically three main ways to use it. The first is as a command line executable tool. You can download Lassie already compiled and ready to go and just run it immediately to fetch data. It's designed like we try to design it like a Unix command so you can pipe it and compose it with other things. And I'll show you how that works in a moment. The second way you can use Lassie is we have built essentially a mini HTTP server that exposes a trustless HTTP gateway that serves car files. It is much like the spec for the IPFS trustless gateway, but we do some other cool stuff. You can find data at a half beneath a SID. And there's a bunch of other things that go a little bit beyond what the current gateway spec is for trustless data, though we are actually working on an IPIP proposal to extend it or for trustless gateways for other IPFS implementations. And finally, we've designed Lassie from the start to work easily as a library so you can easily incorporate it into your Go application to seamlessly add retrieval from IPFS and Filecoin into your Go program. We see this as a sort of long-term Lassie superpower and we'd love to partner with other organizations, other teams who want to integrate Lassie into their systems. Cool, so there's a couple of things that Lassie will not do. We've built some design constraints into it and we put this in for a specific reason, which is that we want to stay focused on a goal, which is retrieving data. As a result, Lassie does not store data permanently on your machine. We return a car file to you and we think that's up to you to figure out what you want to do with it. We also will not provide records to the DHT because again, we're not holding data permanently. We are not a way to advertise to the IPFS DHT or other content indexing systems. Essentially, Lassie is not designed to be a full-fledged IPFS server node. We are largely stateless. When you run us as an HTTP server, we hold onto a little bit of temporary state just to make some optimizations, but one of our goals from the start is there shouldn't be a config file or any other file that lives permanently on your system. As a result of Lassie, we are essentially a stateless tool for the minute you stop the program running. And we think this somewhat artificial design constraint is necessary as a way to keep our program simple, focused and easy to use for other folks in their programs. Cool, so let's see what Lassie can do. Essentially, as I said, Lassie is a simple command line tool is one of the ways you can use it. You can essentially tell it to fetch a given SID and then it will return that to you and it can return that to you as a data stream that you can part pipe to another program. In this demo, we actually fetch a SID and pipe that output directly to the go-car command line tool and then we use the go-car command line tool to convert it to the flat original data file and then we pipe that into FFMPEG, the video player and this is what you get. Again, we'll see if this actually works. Cool, yes, so here we are, we're typing it in. We're getting it, we're putting in a provider, we're getting a SID, then we're passing it to car, we're passing it to an FF player and you magically get a video playing on your screen. It's kind of cool that you can just go from a SID to a video playing. All right, cool, I'm gonna stop that because it's super loud. So another thing is so while Lassie's pretty new and it's definitely still in development, Lassie's not really a prototype project. We are already in production. We are the primary retrieval tool for the Saturn network to do cache misses. So when the Saturn network gets a request for data and they don't have it, they just turn over to Lassie and say, find this data and through the REIT project, which is the decentralized gateway project, that translates to we're downloading about 140 million SIDs a week through Lassie. So we've got some heavy volume use case and we're working on optimizing, we've got a whole team working on it. And Lassie is also increasingly one of the easiest ways to download data through Filecoin. So we're starting to recommend it to enterprise clients who wanna put data on Filecoin and get it back. And we have, as I said, we have a team working on Lassie. We're continuing to optimize, we're continuing to optimize how we select protocols, we're continuing to optimize each protocol implementation. And our goal is to make it faster and better over time. So if you wanna integrate Lassie into your project, I'd say go for it. We do have some work to do on documentation but we're also working on that. Yeah, so if you wanna know here more about Lassie and you happen to be at an IPFS thing, you can come by the data transfer track where we're gonna be doing the deep dive into the architecture. You can also just hit me up on Slack. It's probably the easiest way. I'm also happy to meet with folks one-on-one. If you start to use Lassie, give us feedback, file a GitHub issue. I didn't actually include the repo in here but talk to the Bedrock team, that's the team that's developing Lassie or complain to us on Twitter that we do not have a Twitter account. So you'll have to figure that one out yourself. Yeah, that's the tool. That's what we got. Yay, thanks.