WEBVTT 00:00.000 --> 00:13.000 Right, we're going to kick off this afternoon with talk about ties by Rafael, take it away. 00:13.000 --> 00:20.000 Alright, can you hear me? Nice. Thanks for the introduction. Hi everybody, I'm Rafael. 00:20.000 --> 00:25.000 And today I want to show you a project I've been working on for the past few years. 00:25.000 --> 00:30.000 And it is for Federated Bookmark Sharing. 00:30.000 --> 00:35.000 Federated Bookmark Sharing, what does it mean? What can ties even do? 00:35.000 --> 00:42.000 So let me start off by saying that I think there are good websites on the web today. 00:42.000 --> 00:48.000 But I think sometimes it feels like there aren't any good websites because it's really how to find them. 00:48.000 --> 00:58.000 So my inspiration for this was to create a better way to kind of explore the small web, the good web, the indie web, whatever you want to call it. 00:58.000 --> 01:06.000 So basically in ties you can, if you find a good website somewhere, doesn't matter where you can save it, 01:06.000 --> 01:16.000 kind of organize it in a structure, then you can share it with others, and you can discover good websites that other people found on ties. 01:17.000 --> 01:22.000 There are two things that ties us a little bit differently than other normal web apps. 01:22.000 --> 01:25.000 So the first one is the approach to sharing. 01:25.000 --> 01:33.000 We don't want to have a global timeline, so like a big stream where all the posts are in there. 01:33.000 --> 01:40.000 That kind of opens up lots of problems with abuse, spam, and also use the experience in general. 01:40.000 --> 01:45.000 So we don't want to have content from strangers pushed at you. 01:45.000 --> 01:50.000 There's still means that you can publish the bookmarks you find publicly. 01:50.000 --> 01:54.000 So we want this to be a normal website accessible without an account. 01:54.000 --> 02:00.000 We've had a rate with other federal services and we want to support our S feeds. 02:00.000 --> 02:06.000 But the feed you see when you log in that contains things from people you follow only. 02:06.000 --> 02:13.000 So you might see where this goes, this is built for more of a small scale interaction. 02:13.000 --> 02:25.000 For example, like your friends in the team at work, your colleagues in the team at work, or your friends, or some kind of niche topic group on the internet that you met. 02:25.000 --> 02:31.000 The second thing ties us a bit differently is how it organizes bookmarks. 02:31.000 --> 02:38.000 Usually bookmark managers use folders or directories or tags, which have some problems with folders. 02:38.000 --> 02:42.000 You end up with a really complex tree and hierarchy of folders with tags. 02:42.000 --> 02:48.000 You end up with a kind of a huge list of tags and you don't know which tags to apply to any bookmark you save. 02:48.000 --> 02:57.000 So with ties we're going for an approach that works like Obsidian or Rome Research or Arena, a few part of these services. 02:57.000 --> 03:06.000 You have basically lists that can be connected to other lists to form this flexible graph that you can organize however you want. 03:06.000 --> 03:11.000 And then you can connect bookmarks to as many of these lists as you like. 03:11.000 --> 03:18.000 And then they form kind of this beautiful structure that they can use however you like. 03:18.000 --> 03:23.000 A little bit about the philosophy behind the project. 03:23.000 --> 03:35.000 I'm very inspired by the ideas behind the web, such so when I say a link as a fundamental unit, how I've written it down here, that means that this is a very powerful concept. 03:35.000 --> 03:41.000 I think baked into HTML that we can use in other forms as well. 03:41.000 --> 03:45.000 For example, for the graph structure, I just showed you. 03:45.000 --> 03:53.000 And I think this is something that will be a very influential for the way that I built the UI of of ties. 03:53.000 --> 04:01.000 Also the ideas behind the web for me are building on existing standards being interoperable with as many services as possible. 04:01.000 --> 04:11.000 Back what's comfortable with as many browsers as possible, which also means working without JS. 04:11.000 --> 04:25.000 Also this means that when you run a ties instance that is a public instance, like a public website that can get spotted by search engines, it can be used without an account. 04:25.000 --> 04:29.000 And I think this is really important to be kind of a good citizen of the web. 04:29.000 --> 04:37.000 We also want to poison AI scrapers and bots that don't respect resource limits, this is something that is killing many small websites. 04:37.000 --> 04:43.000 This is super close to my heart, this will be like a core part of the philosophy. 04:43.000 --> 04:53.000 And the last point is, we don't want a single big flagship instance, like one big domain or server where every other body signs up. 04:53.000 --> 04:59.000 We want to foster small network of instances that really talk to each other. 04:59.000 --> 05:05.000 And as part of that, we want to make it as easy as possible to self host the service. 05:05.000 --> 05:07.000 What's the current state? 05:07.000 --> 05:15.000 The basics are there, you can save bookmarks, you can nicely organize them, you can federate and publish bookmarks to mess it on. 05:15.000 --> 05:18.000 But there's lots of more stuff we want to build. 05:18.000 --> 05:34.000 First off, some full text search through the contents of bookmarked sites that we want to actually fetch the content and make it searchable, also make it possible to search the bookmarks of people you follow to discover new sites. 05:34.000 --> 05:39.000 If you want, we also want to add more publishing with other services. 05:39.000 --> 05:52.000 For example, with TICE instances, so following people on other TICE instances, we want to support RSS and we want to support collaborative editing and this shared spaces for communities. 05:52.000 --> 05:57.000 And of course, there's always more we have tons of accessibility things on the roadmap. 05:57.000 --> 06:03.000 We want to include that AI poisoning mechanism I talked of and so forth. 06:04.000 --> 06:15.000 It's really easy to install TICE. You can get the container or just download a single binary, or you need an addition to that as a Postgres scale database, then you set three simple options and you're ready to go. 06:15.000 --> 06:21.000 We have a guide for that. We also provided Docker Compulse file and a guide for compiling it yourself if you want. 06:22.000 --> 06:36.000 The stack is heavily focused around rust and around the experience of preventing as many errors and mistakes at compile time, so it's heavily based around back and template rendering. 06:36.000 --> 06:43.000 And we only use HTML to sprinkle some activity on top, which also makes it really nice to work with our JavaScript. 06:43.000 --> 06:47.000 And we lean heavily on Postgres scale for things like the full text search. 06:47.000 --> 06:55.000 That sounds interesting to you and you like the philosophy I outlined, come talk to me, look through the HTML tracker, we are looking for contributors. 06:55.000 --> 07:00.000 We want to make this as easy as possible to contribute and make this into a community. 07:00.000 --> 07:08.000 If you want to contribute in another way, you can book a short call with me and do a user interview. 07:08.000 --> 07:14.000 We're really interested in getting feedback and future requests and seeing how this works for people. 07:14.000 --> 07:27.000 So that, thanks for your attention. I will be waiting outside, come talk to me, message me on the very first sketch of stickers and thanks. 07:27.000 --> 07:30.000 Thank you, Raphael. 07:30.000 --> 07:34.000 So this is another one of the shorter 10 minutes, we have time for one question. 07:34.000 --> 07:41.000 If anybody would like to ask Raphael a question here, otherwise he said he'll be outside and we have to admit you and chat with you. 07:42.000 --> 07:44.000 I'm excited, I'm really excited to see this. 07:44.000 --> 07:49.000 As I said at the beginning, if you were here, folks were here earlier on this morning when we did the intro. 07:49.000 --> 07:51.000 I think it seems like a question. 07:51.000 --> 07:52.000 Oh, I'm so sorry. 07:52.000 --> 07:54.000 Oh, there we are, I'm sorry. 07:54.000 --> 08:00.000 I'm really excited to see all of this innovation progress in the Fediburse. 08:00.000 --> 08:08.000 Will there be extension with a browser sink or a folder with a file box where you automatically see 08:08.000 --> 08:12.000 this size or something like that? 08:12.000 --> 08:20.000 So the question was whether it be an extension for sinking the Firefox bookmarks, for example, the folder. 08:20.000 --> 08:23.000 Like a local folder with files. 08:23.000 --> 08:40.000 So you basically want to import stuff from Firefox. 08:40.000 --> 08:48.000 So we will add lots of importers for other bookmarking services as well as definitely planned and won't be that far in the future. 08:48.000 --> 08:52.000 So that's basically a very important part of the server setting. 08:52.000 --> 08:53.000 Yes. 08:53.000 --> 08:54.000 Thanks. 08:54.000 --> 08:55.000 Great. 08:55.000 --> 08:56.000 Excellent. 08:56.000 --> 08:57.000 Thank you, Rafael.