WEBVTT 00:00.000 --> 00:14.000 What's cool to see some references to what we're doing already and the power of open source 00:14.000 --> 00:20.000 so that's really cool to be in a very collaborative room. 00:20.000 --> 00:27.000 So we're trying to build like a modern blockbase editing experience for the open web. 00:27.000 --> 00:30.000 So quick run through. 00:30.000 --> 00:36.000 I want to share like okay what we've been up to past year mostly and look a little bit 00:36.000 --> 00:38.000 forward to next year. 00:38.000 --> 00:40.000 First a little bit of an introduction to block now. 00:40.000 --> 00:45.000 We'll do a quick demo to show you how it works for those not familiar with it. 00:45.000 --> 00:56.000 We'll show what we've done mostly in 2025 and kind of like our ideas of okay what's ahead. 00:56.000 --> 00:59.000 So for those who are not familiar. 00:59.000 --> 01:05.000 Here's like a quick 10 second overview of what you can do with our editor. 01:05.000 --> 01:07.000 So it's up. 01:07.000 --> 01:13.000 We started building this a couple of years ago when we did an editor like this for ourselves 01:13.000 --> 01:17.000 that we noticed okay this is actually a lot more work than expected. 01:17.000 --> 01:24.000 It was hoping there'd be something existing that we could really easily reuse. 01:24.000 --> 01:29.000 And though there was great low level open source libraries it was still a lot of work 01:29.000 --> 01:34.000 to build something with a good UX and a good developer experience on top of it. 01:34.000 --> 01:40.000 So we started with block now and the idea is okay we have a good UX included. 01:40.000 --> 01:47.000 So notion style blockbase UX including all the UI components that you might need. 01:48.000 --> 01:57.000 Is your APIs so no low level position juggling for those who are familiar with lower level text editors. 01:57.000 --> 02:05.000 Blockbase design and internally we're proud to build really on some high quality open source libraries like pros mirror for 02:05.000 --> 02:15.000 managing the DOM content editable and YJS to power collaboration with CRDTs. 02:15.000 --> 02:21.000 We started kind of spitting out of a research project sponsored by AnnelNet. 02:21.000 --> 02:24.000 And then the last one and a half years. 02:24.000 --> 02:29.000 So Alex mentioned it and I think Bill and also mentioned it. 02:29.000 --> 02:37.000 We've been really proud to unfortunate to collaborate with Zemdus and Dinem from France. 02:37.000 --> 02:44.000 They're building a, they reached out one a half year ago like hey we're building last week ducks in our collaboration, 02:44.000 --> 02:50.000 which is like a notion style collaboration documents collaboration products. 02:50.000 --> 02:53.000 And we're using block for not for it. It works great. 02:53.000 --> 02:55.000 Can we help you take it to the next level? 02:55.000 --> 03:05.000 We also need this and this and this. So that's also really accelerated our product and organization last year. 03:05.000 --> 03:08.000 Besides that, we're also supported by industry. 03:08.000 --> 03:12.000 So now you see more and more sponsors of the last year. 03:12.000 --> 03:19.000 I think 2025 for us was really a year where we went from something that started as a side project to a small team. 03:19.000 --> 03:29.000 So it's like four of us now with a mix of different sponsorship models. 03:29.000 --> 03:33.000 One thing that's core to block now is blocks. 03:33.000 --> 03:39.000 Nothing to do with block chain but about the cool little pieces in the editor just picked out here. 03:39.000 --> 03:42.000 Why blocks and why this kind of editor? 03:42.000 --> 03:47.000 If you look at different formats existing, so let's position yourself a little bit like what we're trying to be. 03:47.000 --> 03:53.000 We're not like a pure markdown editor. It's a little bit too limited if you want advanced formatting or colors or something like that. 03:53.000 --> 03:57.000 That's not easily possible with simple markdown editor. 03:57.000 --> 04:06.000 And we also definitely don't want to be a word style editor because all of us have seen these kind of like word documents. 04:06.000 --> 04:14.000 That have a super inconsistent layout and design because you can change the font here and everything there. 04:14.000 --> 04:21.000 And so you kind of end up with a little bit of a messy thing. So that's also not our goal. 04:21.000 --> 04:26.000 What we want to do is have a better UX and for us the blocks are the answer to this. 04:26.000 --> 04:32.000 So it helps blocks help users to create structured content naturally. 04:32.000 --> 04:37.000 And yeah, we call that prioritizing content over form. 04:37.000 --> 04:40.000 So you end up with nice and structured documents. 04:40.000 --> 04:44.000 And this same principle also really helps in the developer experience. 04:44.000 --> 04:49.000 So because everything is a block you can easily address it by an ID. 04:49.000 --> 04:57.000 You have a clear schema of what kind of blocks are allowed in the document and it makes it easier to work with as a developer. 04:58.000 --> 05:07.000 On to Nick for a quick demo about that and I'll need to try to hold this with you. 05:07.000 --> 05:09.000 Okay. 05:09.000 --> 05:18.000 Yeah, so what I want to go over here is a little bit about how you actually use block now in terms of the API and such. 05:18.000 --> 05:25.000 So over the past year we've put a lot of effort into the ability to customize the editor. 05:25.000 --> 05:36.000 So you'll see here that I have a very simple document editor so you can see that there's a bunch of elements that you can add to the editor. 05:36.000 --> 05:42.000 So all of these are blocks that are available by default and built into the system. 05:42.000 --> 05:49.000 And on the left hand side you can see all of the code that it actually took to get that to happen. 05:49.000 --> 05:54.000 So what we're really trying to go for here is a really easy to use out of the box experience. 05:54.000 --> 06:03.000 Whereas many other lower level editors tend to force you to build out all of this content. 06:03.000 --> 06:08.000 Yourself rather than it being provided by default. 06:08.000 --> 06:15.000 But what we've done on top of this is, well actually I should go over the code a little bit. 06:15.000 --> 06:21.000 So you'll see here that we block notice based on we're using React as the framework. 06:21.000 --> 06:25.000 And you'll see that there's this used create block note. 06:25.000 --> 06:32.000 Hook that actually gives us an editor instance and then we can render it out. 06:32.000 --> 06:40.000 So what I'm going to show here is actually adding to the schema. 06:40.000 --> 06:50.000 So what you'll see here is that we can extend and add additional features into the editor by extending its schema. 06:50.000 --> 06:55.000 So you can choose what blocks you actually want inside of your editor. 06:55.000 --> 07:05.000 So what I've done here is created a very simple alert block, which we've we've spent over the past year. 07:05.000 --> 07:15.000 One of the major refactors that we've done is worked on the actual API for defining these custom blocks and adding them into the editor. 07:16.000 --> 07:23.000 So block note in general we heavily focus on type safety structure schema. 07:23.000 --> 07:30.000 Like that's kind of our whole bit is that everything in the editor it's content is completely type safe. 07:30.000 --> 07:35.000 It's always valid content. 07:35.000 --> 07:44.000 So what you'll see here is that there is this configuration that describes what an alert block has as. 07:44.000 --> 07:53.000 So it's the blocks type so it's an alert block and the properties that it's available to have. 07:53.000 --> 07:59.000 So this is defining a way that it can be validated at runtime. 07:59.000 --> 08:04.000 And the content that the actual alert block supports. 08:04.000 --> 08:10.000 So in this case it's it's allowed to have content inside of it. 08:10.000 --> 08:25.000 So you'll see that there's a very simple function for describing how to map that block, which is just a json object at this point and how to map that into something that actually renders into the editor. 08:25.000 --> 08:35.000 So what we can do here is go ahead and actually add that block to the editor. 08:36.000 --> 08:42.000 So again, very simple example so you can actually see how it actually renders out. 08:42.000 --> 08:51.000 So you'll see here that we have defined this severity property for the alert block. 08:51.000 --> 08:56.000 And defined that it should change its background color based on that severity. 08:56.000 --> 09:08.000 And I had added a button that actually will insert that new element, the new block into the editor. 09:08.000 --> 09:15.000 So what I would really like to stress here and show off is that everything is type safe. 09:15.000 --> 09:21.000 So you can see that the actual properties of the alert block are being validated. 09:22.000 --> 09:27.000 So you can actually also see the different types of blocks. 09:27.000 --> 09:36.000 Since we've extended the block note schema here, there was a number of blocks that were already built into the editor. 09:36.000 --> 09:41.000 And so we were just adding on the alert block to the list essentially. 09:41.000 --> 09:48.000 So you can see that there's a number of other. 09:48.000 --> 09:58.000 You see that there's a list of all the blocks that are available in the editor and all of this is defined in a type safe way. 09:58.000 --> 10:03.000 So yeah, I think that's the majority of what I wanted to show off. 10:03.000 --> 10:17.000 We spent a good amount of time over the year to work and refine the API for defining these blocks and allowing for more customization into the editor. 10:18.000 --> 10:19.000 Yeah. 10:25.000 --> 10:30.000 Nice and I think it was important to keep in mind with this. 10:30.000 --> 10:33.000 So thanks for the demo. It's like okay because we're. 10:33.000 --> 10:35.000 Thank you. 10:35.000 --> 10:38.000 Because it's an editable service right. 10:38.000 --> 10:42.000 It's a little bit different than when you're just making a web app. 10:42.000 --> 10:45.000 So having this type safety. 10:46.000 --> 10:48.000 How are you doing? 10:48.000 --> 10:49.000 Yeah. 10:49.000 --> 10:51.000 Okay. 10:51.000 --> 10:53.000 Cool. 10:53.000 --> 10:56.000 Because an editor is a dynamic service. 10:56.000 --> 10:59.000 So these kind of things are a little bit more complex. 10:59.000 --> 11:03.000 And when you're building a regular app so hard. 11:03.000 --> 11:06.000 So looking back at 20 25. 11:06.000 --> 11:11.000 So what we've been focusing on and some key numbers again. 11:11.000 --> 11:15.000 Yeah, I think a project is really taking off now and that's cool to see. 11:15.000 --> 11:17.000 And I think you can see it in these numbers. 11:17.000 --> 11:19.000 So we've done over 60 releases. 11:19.000 --> 11:24.000 Lots of PRs and we've grown from I think it was like 30k. 11:24.000 --> 11:25.000 This is like mpm downloads, right? 11:25.000 --> 11:27.000 It's the best set that we have. 11:27.000 --> 11:32.000 But the question marks, but at least it was like 30,000 about a year ago. 11:32.000 --> 11:37.000 And now we're like 120 or 130,000 weekly downloads. 11:37.000 --> 11:40.000 It's really nice to see the adoption. 11:40.000 --> 11:46.000 And apparently that to see our project really resonates with the community. 11:46.000 --> 11:48.000 Yep. 11:48.000 --> 11:49.000 We worked on many different things. 11:49.000 --> 11:54.000 So definitely a lot of features going from export system. 11:54.000 --> 11:57.000 I'll go over some of these in a bit. 11:57.000 --> 12:00.000 A lot of like API enhancements. 12:00.000 --> 12:06.000 But I'm even even more proud that we were also able to also really invest in the core. 12:07.000 --> 12:09.000 That has been around for a couple of years. 12:09.000 --> 12:12.000 But you really make sure that the core is also sustainable. 12:12.000 --> 12:15.000 And you're being able to re-factor parts of this. 12:15.000 --> 12:19.000 So for example, this custom block API that Nick just showed, 12:19.000 --> 12:23.000 like with which you can make your own alert block in the demo. 12:23.000 --> 12:28.000 Last year we migrated all our default blocks to actually also use the external API. 12:28.000 --> 12:31.000 First version of our AI integration and it's been designed. 12:31.000 --> 12:37.000 So you can have an agent understand the documents and then hand in hand with an AI. 12:37.000 --> 12:39.000 You can understand the documents. 12:39.000 --> 12:40.000 So you can ask questions. 12:40.000 --> 12:42.000 You can redact the documents. 12:42.000 --> 12:45.000 So ask it to make changes or write new content. 12:45.000 --> 12:46.000 Hand in hand. 12:46.000 --> 12:48.000 And you can really integrate this. 12:48.000 --> 12:50.000 Of course with your own language model. 12:50.000 --> 12:56.000 So you're not bound to any surface or whatsoever that we build. 12:56.000 --> 12:59.000 And it's done in a way. 12:59.000 --> 13:01.000 Yeah, what we call like human in the loop. 13:01.000 --> 13:04.000 So the AI, you really see it writing as a collaborator. 13:04.000 --> 13:07.000 You can accept and reject changes. 13:07.000 --> 13:10.000 So the end user is still in control. 13:10.000 --> 13:14.000 It's not like some random robots taking control of your document. 13:14.000 --> 13:17.000 Here's a more advanced demo. 13:17.000 --> 13:19.000 Some of this is still to come. 13:19.000 --> 13:22.000 But most of it is available already. 13:22.000 --> 13:24.000 So here's a template document. 13:25.000 --> 13:30.000 And now I demonstrate that it's also possible to connect certain tools. 13:30.000 --> 13:33.000 So look up some information about the European Union. 13:33.000 --> 13:37.000 It does a web search on the back end with an AI tool. 13:37.000 --> 13:40.000 And now you see it making those suggestions in the document. 13:40.000 --> 13:41.000 I can review them. 13:41.000 --> 13:46.000 And as a user I can say, OK, accept these changes and then they become part of the document. 13:46.000 --> 13:51.000 So connecting a web search might be one thing that you like. 13:51.000 --> 14:00.000 But maybe you want to connect like a rag system or your own database or have it look up something about your colleagues or about other documents that you've written earlier. 14:00.000 --> 14:06.000 You can make all these integrations yourself. 14:06.000 --> 14:07.000 Yes. 14:07.000 --> 14:09.000 Suggestions and versioning. 14:09.000 --> 14:11.000 I'm really looking forward to this. 14:11.000 --> 14:13.000 This is something that we expect. 14:13.000 --> 14:20.000 We've done a prototype on this last November and December in close collaboration with Kevin Yans. 14:20.000 --> 14:24.000 From the YGS project. 14:24.000 --> 14:28.000 This was sponsored by Deanum and Zendis. 14:28.000 --> 14:32.000 And so this is about what we call asynchronous collaboration. 14:32.000 --> 14:35.000 So making different versions. 14:35.000 --> 14:40.000 I think I'm in front of the screen. 14:40.000 --> 14:47.000 See being able to see and they've changes between versions. 14:48.000 --> 14:50.000 And also to go into suggested mode. 14:50.000 --> 14:53.000 So what you might know as track changes suggestions. 14:53.000 --> 14:58.000 You can maybe have a user that doesn't have permissions for the base documents. 14:58.000 --> 15:07.000 But once you suggest some changes and then the owner of the document can accept or reject those changes. 15:07.000 --> 15:10.000 And we want to as part of that. 15:10.000 --> 15:14.000 We also want to make a possible so that you can store metadata. 15:14.000 --> 15:16.000 Keep in the version history. 15:16.000 --> 15:20.000 Keep a complete overview of who they want and when. 15:20.000 --> 15:23.000 So we call that attributions. 15:23.000 --> 15:26.000 That's going to be. 15:26.000 --> 15:28.000 Let me quickly see. 15:28.000 --> 15:32.000 We're hosting a local first dev room tomorrow. 15:32.000 --> 15:35.000 And it's going to be a presentation about the technical stuff behind that. 15:35.000 --> 15:39.000 And a little bit of sneak peek of what's upcoming and also what's coming in the new version of 15:39.000 --> 15:49.000 BIGS to enable all this. 15:49.000 --> 15:50.000 Yeah. 15:50.000 --> 15:53.000 Some other things that are ahead of us. 15:53.000 --> 15:55.000 So I mentioned a sync collaboration. 15:55.000 --> 16:01.000 Of course there are some features that we have to remind that we would love to work on. 16:01.000 --> 16:06.000 We also want to keep working on some of the foundations like we have done last year. 16:06.000 --> 16:08.000 We've always things to improve there. 16:08.000 --> 16:12.000 Well, I think this was illustrated very clearly by us in the last talk. 16:12.000 --> 16:15.000 So we don't want to lose a side of that. 16:15.000 --> 16:20.000 And I think also as our project might mature as we have more people working on it. 16:20.000 --> 16:22.000 As we have more people depending on it. 16:22.000 --> 16:24.000 I think actually there's also. 16:24.000 --> 16:29.000 We want to spend more time looking at the overall ecosystem around blocknotes. 16:29.000 --> 16:33.000 So that means okay, how can we be transparent in terms of our roadmap? 16:33.000 --> 16:35.000 How do we govern the project? 16:35.000 --> 16:39.000 But also thinking more about things like extension systems. 16:39.000 --> 16:41.000 Can we make that even easier? 16:41.000 --> 16:44.000 We've done some investments in the extension API last year. 16:44.000 --> 16:46.000 But can we make this even easier? 16:46.000 --> 16:52.000 Maybe like extension gallery where the community can create and share extensions, etc. 16:52.000 --> 17:02.000 But also think about other interoperability related features like standardization of these kind of documents. 17:03.000 --> 17:13.000 Yeah, and really look beyond just the editor or just the new features, but really at blocknotes as a project. 17:13.000 --> 17:14.000 Thanks. 17:20.000 --> 17:23.000 There's seven minutes for questions. 17:24.000 --> 17:36.000 We use YGS CRDT for powering the collaborative features. 17:36.000 --> 17:42.000 Basically we think like YGS is kind of like a industry standard for this. 17:42.000 --> 17:49.000 We also directly work with Kevin for this kind of more advanced features. 17:49.000 --> 17:51.000 We always keep an eye on different things. 17:51.000 --> 17:56.000 So tomorrow on the local first demo, I think there might be a session related to Loro as well. 17:56.000 --> 18:01.000 But really it is like okay we're pretty into YGS at this moment. 18:01.000 --> 18:04.000 So I've done haven't done like an advanced comparison. 18:04.000 --> 18:06.000 I'd make might be more into this topic. 18:06.000 --> 18:09.000 But maybe that's more on the side. 18:09.000 --> 18:10.000 Yeah. 18:14.000 --> 18:15.000 Yes? 18:16.000 --> 18:21.000 I worked in the company, which was using the solution. 18:21.000 --> 18:24.000 For me, because how do you read about them? 18:24.000 --> 18:28.000 And I don't really speak about 80 things in the community. 18:28.000 --> 18:31.000 And now I'm going to move to another company. 18:31.000 --> 18:36.000 I don't know what they have, but I have to suggest them. 18:36.000 --> 18:39.000 Would my life be there? 18:40.000 --> 18:41.000 Nice. Thanks for your questions. 18:41.000 --> 18:45.000 The question is, was difficult for you. 18:45.000 --> 18:50.000 I expect for accessibility reasons to work in a notion that the previous company that you worked, 18:50.000 --> 18:55.000 and would your life be better working in block-not. 18:55.000 --> 19:00.000 It's a little bit challenging to say because the notion is like an end user application, 19:00.000 --> 19:02.000 and we are just like an editing service. 19:02.000 --> 19:05.000 So also very much depends on the editor around it. 19:05.000 --> 19:13.000 We've done some sprints on accessibility, and we expect also a collaboration with some of the sponsors 19:13.000 --> 19:18.000 to do larger sprints on top of accessibility this year. 19:18.000 --> 19:22.000 So I can't guarantee you that as of this moment, 19:22.000 --> 19:29.000 but I will be really happy if the end of this year, 19:29.000 --> 19:32.000 I can say with 100% confidence. 19:32.000 --> 19:35.000 So let us give a try to accomplish it. 19:37.000 --> 19:39.000 Will I? 19:39.000 --> 19:46.000 Yes. So I already said that we are using block-not-not-not-appal project 19:46.000 --> 19:51.000 and of course we extended this schema to have links to work packages, 19:51.000 --> 19:52.000 which are my entities in our project. 19:52.000 --> 19:55.000 But I don't know the good ideas that I have in a year. 19:55.000 --> 19:59.000 Like, how my schema should actually should've looked at it like in the beginning. 19:59.000 --> 20:06.000 in the beginning. So my question is, in your example, you had an alert with, I think, like, 20:06.000 --> 20:18.000 for a very used, let's say we replace Aera by Adal. Adal. And we want to migrate, let's say, 20:18.000 --> 20:24.000 100,000 documents, change that. Is there any concepts how we can do this? Like, oh, yeah, 20:24.000 --> 20:31.000 we can prove that, make this structure. So the question is, okay, how to deal with 20:31.000 --> 20:38.000 migrations of the schema. So let's say you support the specific kinds of blocks. 20:38.000 --> 20:45.000 And let's say you change the severity that make them out from Aera to Fatal. I'll skip 20:45.000 --> 20:51.000 the Spanish part. How do you go about that? I think that's a good question. 20:51.000 --> 20:58.000 There is, as in most migrations, this is, I would say this is similar to most migrations. 20:58.000 --> 21:07.000 So at this moment, you have multiple options. Like, either make your new application 21:07.000 --> 21:13.000 compatible with both formats. Then on the clients, you can choose to either upgrade it 21:13.000 --> 21:26.000 to show the same interface when it's Aera or Fatal. Or you find a way to migrate your data. 21:26.000 --> 21:33.000 I know there has been quite some research in systems that kind of like keep migrations 21:33.000 --> 21:40.000 in the clients and automatically try to upgrade things. I know that are currently 21:40.000 --> 21:46.000 react. So most of the UI components are based on react. So we keep the option open for the future, 21:46.000 --> 21:48.000 but I don't see it happen this year. 21:48.000 --> 21:53.000 I have a follow-up on this in Crystal, the project from Twiki, the audio.js. 21:53.000 --> 21:59.000 And so we have plenty of modules that we don't specifically to allow the integration of blocks. 21:59.000 --> 22:07.000 So we have plenty of modules that we also submit to the modules of blocks. 22:07.000 --> 22:13.000 And I have a follow-up to the map also. In Tenzu, we use block-plat and we are in Angular. 22:13.000 --> 22:18.000 And so, if you want, for example, of block-plat working in Angular, and I should focus on this