WEBVTT 00:00.000 --> 00:08.000 That's right. Hello, people. 00:08.000 --> 00:18.000 I am Christine Lumberwebber, executive director of the Sprayley Institute. This is Jessica Talon, 00:18.000 --> 00:22.000 founding technologist at the Sprayley Institute. I hear a lot of people talking about the 00:22.000 --> 00:29.000 company pub thing in the room. People seem pretty excited about it. Anybody explain to me what 00:29.000 --> 00:37.000 activity pub is? It's like email. Anyway, sometimes it's a little bit hard to get some talking 00:37.000 --> 00:43.000 and with Jessica, but I'm going to do my best. I just remembered, actually, we spent a few 00:43.000 --> 00:48.000 years working on activity pubs. We're co-authors, co-editors. We're excited about activity 00:48.000 --> 00:54.000 and we're glad that people are clearly excited about the Fediverse as a few users online. 00:54.000 --> 01:03.000 But I think we're happy with that. There might be a little bit more to do. 01:03.000 --> 01:13.000 Maybe it's done. Maybe what else could we do? People worry about their sights going down. 01:13.000 --> 01:17.000 They might have to move somewhere to appear. That's one clear one. There's this whole thing about 01:17.000 --> 01:21.000 if you read the activity pubs back, there's this authentication, not the 01:21.000 --> 01:25.000 authorization thing and it kind of goes like, it just kind of stops there. 01:25.000 --> 01:31.000 The W3C at the time was like, don't do that. It's not ready. 01:31.000 --> 01:37.000 We did our best. We spent a lot of time thinking since then about what we might do. 01:37.000 --> 01:43.000 There's things like moderation. It's tough for people who work on things. 01:43.000 --> 01:48.000 You know, like instance administrators get tired. They're like, I've walked this whole instance. 01:48.000 --> 01:51.000 But we also want people to self-host as much as possible. 01:51.000 --> 01:55.000 But if people stare at self-host as much as possible, then I guess you're just 01:55.000 --> 01:59.000 walking individuals at that point. That might be a problem. 01:59.000 --> 02:03.000 At one point, I wrote up this thing called OCab pub. 02:03.000 --> 02:07.000 One of the things I think we really loved about activity pub is that it's an 02:07.000 --> 02:11.000 actor model system. Who here knows about the actor model? 02:11.000 --> 02:17.000 Okay, about 30%. It's great. It's all about sending messages like email. 02:17.000 --> 02:21.000 But it turns out you can do all of computing this way. 02:21.000 --> 02:23.000 It works great for asynchronous systems. 02:23.000 --> 02:27.000 One of the things that Jessica and I loved about activity pub is that it's an 02:27.000 --> 02:30.000 actor model system. There's so much you could do with it. 02:30.000 --> 02:36.000 We're kind of doing the decentralized version of this major web 2.0 site. 02:36.000 --> 02:39.000 The centralized version of that major web 2.0 site. 02:39.000 --> 02:41.000 That's about good enough. 02:41.000 --> 02:46.000 Maybe there's some more rich interactions. 02:46.000 --> 02:50.000 What is social? What does it mean to be social? 02:50.000 --> 02:53.000 We think about social as these kind of constraints. 02:53.000 --> 02:56.000 You're sending a message to somebody. 02:56.000 --> 02:59.000 The actor model, you can do things. 02:59.000 --> 03:04.000 The turns out that people really want software where you can do things. 03:04.000 --> 03:07.000 Maybe why can't all software be social? 03:07.000 --> 03:13.000 And in fact, once something like Google Docs appears, it's really hard to convince people 03:13.000 --> 03:15.000 to use something like LibreOffice. 03:15.000 --> 03:19.000 Once you end up getting to the point where you can make something 03:19.000 --> 03:23.000 collaborative, then it's really great. 03:23.000 --> 03:25.000 We do more of it. We make it more peer-to-peer. 03:25.000 --> 03:28.000 It's an actor model system. 03:28.000 --> 03:31.000 We can send messages. It doesn't matter where they live. 03:32.000 --> 03:34.000 This is really great. 03:34.000 --> 03:36.000 When I was writing up OKPub, I'm like, yeah. 03:36.000 --> 03:39.000 There's this whole great security model. 03:39.000 --> 03:42.000 We left out that authentication and authorization. 03:42.000 --> 03:45.000 There's a way to do things that we can understand. 03:45.000 --> 03:48.000 It's called capabilities security. 03:48.000 --> 03:52.000 I started writing this document called OKPub towards networks of consent. 03:52.000 --> 03:54.000 People read it. They're like, yeah. 03:54.000 --> 03:56.000 It's a problem domain. 03:56.000 --> 03:57.000 You're explaining the things. 03:57.000 --> 03:58.000 It's really exciting. 03:58.000 --> 04:00.000 It's like, it gets to the part where it's like, how do you do this? 04:00.000 --> 04:01.000 It just kind of stops. 04:01.000 --> 04:06.000 And the reason why it just kind of stops is that I tried talking to a bunch of people who 04:06.000 --> 04:07.000 were doing whatever its implementation. 04:07.000 --> 04:09.000 So like, yeah, this sounds great. 04:09.000 --> 04:11.000 How do I do it on Ruby on Rails? 04:11.000 --> 04:15.000 And actually, Ruby on Rails is a pretty interesting thing. 04:15.000 --> 04:17.000 Ruby on Rails in Django and stuff like that. 04:17.000 --> 04:19.000 They did something wonderful, right? 04:19.000 --> 04:23.000 Like, in kind of the web 2.0 era, they kind of provided. 04:23.000 --> 04:26.000 They didn't just make it easy to write these kind of web 2.0 applications. 04:26.000 --> 04:29.000 They kind of taught you how to do it, right? 04:29.000 --> 04:31.000 They kind of taught you how to do it. 04:31.000 --> 04:34.000 But right now, writing peer to peer applications, 04:34.000 --> 04:36.000 writing really distributed secure applications. 04:36.000 --> 04:38.000 We can do things. 04:38.000 --> 04:39.000 We can live. 04:39.000 --> 04:40.000 We can interact. 04:40.000 --> 04:41.000 It's pretty hard. 04:41.000 --> 04:42.000 You kind of have to be an expert. 04:42.000 --> 04:44.000 It's like photography is complicated. 04:44.000 --> 04:46.000 The security stuff that's difficult to think about. 04:46.000 --> 04:48.000 How do you know whether or not a person can do it? 04:48.000 --> 04:50.000 Think that just capability stuff. 04:50.000 --> 04:52.000 It seems like it can do it. 04:52.000 --> 04:53.000 It seems like it could do it. 04:53.000 --> 04:55.000 But it's just really hard to write this, right? 04:55.000 --> 04:57.000 But wait, wait, wait, wait, wait. 04:57.000 --> 04:58.000 Wait, wait, wait, wait. 04:58.000 --> 04:59.000 Wait, wait. 04:59.000 --> 05:02.000 What if we did the thing, the thing that happened with Django and 05:02.000 --> 05:07.000 Rails and stuff like that, we make a new software foundation that 05:07.000 --> 05:09.000 makes it so that's a default thing you get. 05:09.000 --> 05:10.000 And it walks you. 05:10.000 --> 05:12.000 It doesn't just, it doesn't just make it easy to think. 05:12.000 --> 05:16.000 It teaches you the default way to be able to write peer to peer applications. 05:16.000 --> 05:17.000 And you're doing it. 05:17.000 --> 05:19.000 We can do that. 05:19.000 --> 05:21.000 Or we should even do that for a few years. 05:21.000 --> 05:23.000 Oh, it's called goblins. 05:24.000 --> 05:26.000 You're doing that for a few years now. 05:26.000 --> 05:27.000 It's called goblins. 05:27.000 --> 05:31.000 And it's after model based in the teaching, it makes peer to peer 05:31.000 --> 05:32.000 to distribute applications. 05:32.000 --> 05:34.000 It's the default thing you get. 05:34.000 --> 05:36.000 Well, I forgot about these things sometimes. 05:36.000 --> 05:37.000 Oh, we're also doing a fundraiser. 05:37.000 --> 05:39.000 You can donate to us by the way. 05:39.000 --> 05:40.000 But also. 05:40.000 --> 05:43.000 But also, we have this other protocol. 05:43.000 --> 05:44.000 Oh, shit. 05:44.000 --> 05:45.000 I'm going to get in trouble. 05:45.000 --> 05:47.000 It's probably talking about activity pub. 05:47.000 --> 05:49.000 I'm going to talk about this other protocol called O'Captain. 05:49.000 --> 05:50.000 Well, it's pretty cool. 05:51.000 --> 05:53.000 It's distributed garbage collection. 05:53.000 --> 05:54.000 Why? 05:54.000 --> 05:55.000 You can leave that. 05:55.000 --> 05:57.000 It's like really fast and efficient. 05:57.000 --> 05:59.000 It, like, has this capabilities. 05:59.000 --> 06:00.000 The security can some antics. 06:00.000 --> 06:02.000 And it's pretty cool, right? 06:02.000 --> 06:07.000 Actually, Jessica wrote the specs documents for this new stuff. 06:07.000 --> 06:10.000 If you go to O'Captain.org, you can check it out. 06:10.000 --> 06:13.000 So are we abandoning activity pub? 06:13.000 --> 06:14.000 Okay. 06:14.000 --> 06:15.000 No. 06:15.000 --> 06:16.000 Okay. 06:16.000 --> 06:17.000 We're not abandoning the air pay to be. 06:17.000 --> 06:21.000 I kicked out from this room for abandoning activity pub. 06:21.000 --> 06:22.000 No. 06:22.000 --> 06:23.000 We're not going to do it. 06:23.000 --> 06:24.000 We're not going to do it. 06:24.000 --> 06:26.000 Because activity pubs after model. 06:26.000 --> 06:29.000 The same way that we can do O'Captain. 06:29.000 --> 06:32.000 As, like, this stuff can do the same thing where at the perimeter, 06:32.000 --> 06:34.000 we're sending messages back and forth between actors. 06:34.000 --> 06:36.000 Except instead of just doing social media things, 06:36.000 --> 06:38.000 they'll be able to do all sorts of things. 06:38.000 --> 06:39.000 But we can use O'Captain. 06:39.000 --> 06:41.000 And the other things we want to direct communication. 06:41.000 --> 06:43.000 We want to have distributed garbage collection. 06:43.000 --> 06:45.000 I think we can do this. 06:45.000 --> 06:46.000 I think we can do this. 06:46.000 --> 06:49.000 And in fact, actually, I think we are doing this. 06:49.000 --> 06:52.000 And we've actually been building quite a lot of stuff. 06:52.000 --> 06:54.000 And things are actually starting to get to the point where, 06:54.000 --> 06:55.000 you know, activity pub. 06:55.000 --> 06:59.000 There's about four, like, longer than four years. 06:59.000 --> 07:02.000 We're like, I would try to explain to people what we're doing. 07:02.000 --> 07:04.000 And people were just kind of like, and then, you know, 07:04.000 --> 07:07.000 when people use it, they're kind of like, whoa, this is amazing. 07:07.000 --> 07:08.000 Right? 07:08.000 --> 07:10.000 So I think we're starting. 07:10.000 --> 07:12.000 There's just baby steps. 07:12.000 --> 07:14.000 Instead of territory. 07:14.000 --> 07:18.000 Or maybe people can start to experience it and kind of explore it. 07:18.000 --> 07:20.000 But I'm getting ahead of myself. 07:20.000 --> 07:24.000 Because, because this is a lightning talk. 07:24.000 --> 07:25.000 There's two minutes left. 07:25.000 --> 07:26.000 And probably less than that. 07:26.000 --> 07:29.000 Because actually, it doesn't tell you to be passive perimeter. 07:29.000 --> 07:32.000 So I think I'm going to stop right here. 07:32.000 --> 07:36.000 But I think the real exciting thing about all of this is that 07:36.000 --> 07:41.000 it's not only possible, but actually there's all this research. 07:41.000 --> 07:45.000 And at the end of activity pub, you know, we spent a few years. 07:45.000 --> 07:48.000 This was a solo project of like, what's next for the Fediverse? 07:48.000 --> 07:51.000 But then we started this nonprofit when it got to the point where, you know, 07:51.000 --> 07:52.000 it's actually possible. 07:52.000 --> 07:54.000 And we've been doing it. 07:54.000 --> 07:56.000 And it's happening. 07:56.000 --> 07:57.000 Fuck yeah. 07:57.000 --> 07:58.000 Yeah.