 and welcome back for our second panel of the day with Johanna and Alex. Okay, so Alex, I'm going to ask, I'm going to ask you the same questions we asked to other speakers. Is there anything around you, your surroundings that is worth sharing or if not, well, you can share anything that like it is controversial about yourself. Well, I would, I would have skipping rope on my desk. Okay, boring meeting. I just hop outside and do, you know, a few of my numbers just to keep the COVID five away. So you're doing that during the meeting? I put on my headphones and then Okay, that's a nice suggestion for our next boring meeting. Okay, John, if you want. Yes, the question, same question for Johanna. Is there anything interesting about your surroundings or anything you'd like to share that's quirky or strange or normal? Apart from my camera flickering. That's pretty cool effect. Which is something I've planned and especially for you guys. No, I'm in the juxtapose, which is pretty cool. I've got Alex and Malcolm just out of shot keeping an eye on me because I'm moving house today. I've moved house yesterday. So I'm, I guess, interesting. Nice. I bet that's fun moving. Your video reminds me a bit of Max Hedron, the jerky cartoon character, the graphic character they made that basically just overlaid some graphics over somebody and he was kind of like jerking around just to make him like seem more artificial. I can turn it off if it's distracting, but well, the whole video, but otherwise I'll stay. Oh, it's fine. It's fine because the voice is coming through because the clear. So that's important. Okay, so let me grab the first question for Alex that came through this code. This is Mark Jensen. Thanks for the talk, Alex. And could you please pause the links that you add at the end of the talk somewhere about modeling with data log and so on, if that's possible. Maybe. Yes. In this good. Yes. So for Joanna, we have quite a few questions. The first question is how does grab differ from La senior? So grab was is more modern than listening and listening has been around for a while. The grab was written after the, I forget the name of it. SDL schema definition language was standardized. But it is a bit more stripped back than listening. Yeah. I think there might be some information on the grab read me on that. Nice. Thank you very much. And another question for Alex. Well, I'm unable to pronounce the name, but it's coming from Zoom. One of the reasons I like Firebase for MVPs, minimum viable products, is to eliminate the need to spin up an application server. So my browser JavaScript and or closure script can talk directly to Firestore, including authentication and so on permissions and using the JavaScript APIs of Firebase. So is there any plans to integrate fire comic with closure script? In a way, so I just put it still above my pay grade. But the folks at data high code building a conserve interface in CLJs. So once that lands, do the five base bits, and then we can connect. And then it'll be similar to data script that you can just run in your browser and connect straight into Firebase. I think the data like server will come first and then then that piece will come with later. So it's kind of. All right, thank you. Joanna, how does site deal with resolvers that need to fetch data from other data sources? So site is intended to be built on one day to base, which is actually, but you can put custom resolvers into site. And you can also load closure namespaces into site. So if you really need to, you can do it yourself. So would you do that? So you do the insight, not. Yeah, so you can have like a GraphQL engine be a view of multiple databases. So is that what site is doing? Okay. So the resolver would give you the data that the GraphQL is asking for. Okay, great. Thanks. Here comes another question for Alex. So see you are a Postgres QL user. And I was wondering, are there particular reasons why you would favor that data hike? So data log over Postgres QL. And why, what, and what might be the use case for that to hike data log over Postgres QL? Okay. So I'm not super familiar with them. Well, I do use Postgres, but not a lot. What I generally found, I mean, it may have changed now, but when I started writing closure or that you often had to write the SQL queries in, you know, a DSL or in text and then imported into your code. And that for me always seems a bit touch and go, especially because you can make mistakes there that you don't pick up because, you know, there's no syntax highlight highlighting. It's not really compiled. So the benefit for me there is that basically writing closure all the way through in your queries. And also with, with, I find with, you know, with data log databases, you can, you can, as you go, change your schema and, you know, without being complete chaos, where they go working in that Postgres, you know, you have to do migration, which can be dangerous. You dump all your data and then you write to your schema and you import them and you're good to go. So that's kind of my preference for using data log, especially because I do a lot of prototyping first. So as a result, you know, I always know I'm going to throw away a lot of stuff. So this log helps me throw away and restart without losing the data. Yeah. Yeah. Thank you, Alex. And another for Johan. So have the jokes team been able to implement streaming queries or subscriptions in GraphQL using Grub or Andor site? No, but because site is still in an experimental stage, it's something that we're definitely looking to playing with and trying out, but we haven't looked at that yet. And are you taking poor requests for that feature? Are we taking poor requests? Yeah, we definitely are taking poor requests. There we go. Excellent. Another one for Alex from me. And I, I think I heard you mentioned in your bio that this is nighttime work for you or like site projects. So I guess you're not using currently any of the fire tall make or even closure in your production environment or current role. So if that is correct, is any plan for you to like introduce that in like in your role somehow, or change your role? Okay. Yeah. So I do use it in production. Production just gets attended to after six p.m. So I work in a design team. So UX, UI and process engineering. So you don't actually do much good. So I think I actually like the code and do close and open source to keep the, the creator in me alive in my day job. I mean, I'm a manager of managers. So, yeah, I mean, I'm actually quite happy that through that way, because then I can explore fully. You know, whereas when you're introducing new technology into a bank, you know, it's not just like closure. You know, it's like three years and then closure. So, so yeah, so no plans to change that at the moment. But I still have a mean, especially now with all the lockdown, they have a lot of time to write code. So no complaints on that. Okay. So in a sense is a positive that you can play with it, but it's a negative you cannot spread the word. So I hope one day you can, you can do that. Thank you. And the next question for Johanna is, can grab actors the endpoint to federated GraphQL services? I'll actually divert that question to Malcolm. I think he's responding on this. Oh, because that's more a grab related thing. I want to do it justice. Oh, there he is. So you just really super small libraries any four namespaces and it's really just a sort of a library to use. If you were building a bigger GraphQL system and you just wanted a parser and something to execute the GraphQL according to the rules of the spec. So it's really just a faithful implementation of the current well, you know, the previous, the, the, the June 2018 version of GraphQL, which is quite modern and includes native for the schema definition language, as, as Johanna said. And so far as a federated system, and it's not, you could, you could use grab if you wanted to create your own a little proxy, but the use of grab insight is really to give X2DB a kind of GraphQL schema, you know, entry, entry point into the database. So you can, you know, define shapes. It's really the schema, a schema layer on top of X2DB, because X2DB is a schema less database. And so it doesn't have an indoor schema. So you kind of have to bring one along. And if you like GraphQL schema, then, then, then site provides that. But you could, you could, if you were going to create a big federated system of GraphQL, you might want to use site as a sort of microservice or a node or a link node in that federated system, if you wanted to provide a GraphQL interface to a bunch of data that you wanted to stand up quickly and then stitch up with other GraphQL services to create a bigger graph. Excellent. Thank you very much. Welcome. Enjoy. So we currently don't have any more questions for Alex. So if I have some more, yeah, I have some questions for not for, not for, yeah, not for Alex, but for Johanna again. So I'm quite curious. GraphQL is kind of relatively new for a lot of developers, even though it's becoming very popular quite quickly. Was there a lot of learning up front? And did you kind of get a sense of how different designing a GraphQL API would be compared to something like REST? I think there's barely any learning up front. Once you get your head around the schema language, which isn't that difficult to pick up the very basics to build a simple API using GraphQL compared to REST is very straightforward. There's obviously an increase that you need to know if you want to start doing clever things. But GraphQL also has really good tutorials and really good resources for learning. So it makes it that much easier. Anything you want to do, you can just look up. Yeah, and it looks like there's quite nice tooling around there as well. I've been using the Apollo Studio, which seems to be quite a popular user interface, and you can do it on the web as well. And it kind of helps you experiment with that. I guess did you find the actual design, deciding what goes into the GraphQL was kind of an interesting challenge? As with everything, I think it is. I think your first draft is always something that you can go back and think while that won't actually work in reality. But no more difficult than REST. Yeah, sure. So you've got lots of options. I just need to decide which options you actually want to get. Cool. So while we're waiting to see if there are any more questions for Alex as well, Alex, is there anything else you wanted to add that you couldn't get into the talk? Not particularly. I think I said what I wanted to say. But I guess just to give a shout out to the folks at Data Hype, they really helped me understand a lot of the stuff and work me through the papers and all that. So that was just pretty neat. Okay. I have another question for Johanna. Is that how you pronounce your name, Johanna? I'll take anything, but I say Johanna. Johanna, okay. What was I going to ask? Oh, yeah. Sites are relatively new projects. So you might not have answers that. But what kind of projects have you built with or like looking to build with or the kind of examples of what people have done with site already? Yeah. Yeah. So we're using site in one of our client projects. We're also using site for our internal HR network, just the whole internal juxtap. And we're hoping that all of the juxtap developers and even less technical people will eventually be able to easily put in a schema, create their own applications and grow from that. We really like the idea of developers being in control of the systems we use. Excellent stuff. And are there tutorials for that already? Or is that still a bit early days yet? There's obviously all of Malcolm's building site series, but that's more building site. And then there'll be offshoots posted regularly to the same YouTube channel of short tutorials. We're planning to do as we build them ourselves on how to actually build on top of site. So there's the building site and then there's also how to build one site. There will be. Cool. Excellent stuff. Is there any question that the speakers wants to do to themselves? That's a difficult one. Not really. You need to be prepared for that. That is putting people on the spot a little bit. I do have a quick one for Joanna. Do you know what the name of the Emacs theme you're using? I'm sure some people are curious. I do know the name of it and it's called Joanna has decided over the course of however many years that this is background color to use and this is the font color to use. And maybe for reclosure, I'll make this one a bit bolder so it's clearer so that it's just evolved over. I think I started with the closure for the brave and true sets up and then from there I've just incremented it. Nice. Excellent. It could be in control of things sometimes. My first boss tried to convince me to use Emacs. He's still trying. There are other editors. Okay, still waiting for questions for Alex, but if there are no more, is there anything else more for Joanna? This code here. Are there any features in the atomic missing from data hike that you miss? Same question with the on-prem and cloud versions of the atomic. So it's confession time. I once tried to run the atomic like five years ago and it asked me for a license so then I closed that and moved on with my life. So all my atomic knowledge is really from reading all the docs and watching all the videos, not actually using it. So yeah. Thanks for like asking the question and also answering the question. Any more for Johanna? I have one question, but it might be a bit too open-ended. I was just curious what it's like working at trucks. Sorry. No, it's great. It's really good. I have to say that. No, it's great. It's a great environment to work at. I can't complain. I couldn't if I wanted to, but I can't complain. Excellent. Thank you. You've been talking about XTDB in your presentation today. I just wanted to remind people that there was an XTDB workshop yesterday by Jeremy. Very detailed. Went into a lot of the theory, but also a lot of the practical use of XTDB. And I think, yeah, there are some workshops around using that specifically as well out there. So I think that's a good point. Thank you. From juxt. Yeah, but yeah, there's some tutorials online and workshops all available on XTDB.com. Com. Yeah. Excellent. Thank you. And also remembering our audience tomorrow. Paula. I hope I pronounced her last name correctly. It's going to talk about that a log. I think it's going to be in a, in a more. Like a brand free flavor. So going to the root of that a log and understand what does it mean to talk about that the log without talking about data make necessary or XTDB necessarily. So I think there are a few connections for you. Other talks in the conference regarding. Data log and yeah. Databases in general. I think that's a good question in the discord. To both. Oh, it's saying this might be an unfair question, but how to choose between dead hike and XTDB. Oh, I'd be interested to know what Alex thinks, because he's obviously done a lot of research into these databases. I, I think XTDB all the time, every time. But if Alex has looked at XTDB and has any answers, then I'd like to hear. So I mean, I haven't used XTB. I mean, I've read about it and I followed it since it was cracks. And if you want to look, there's a bunch of links, but I'll post a link that shows the differences between the two that all the three and the atomic. But for me, what I like about data hike is the fact that it's open source and easy for me to understand. Look, for me, a big selection criteria and things that pick it is that easy for me to understand. And this is easy for me to teach. And at the moment, data hike is the top of that list. So I would choose data hike every time. We'll have to put out some more XT tutorials then. And I'll also get to writing. Excellent. Thank you both. Okay. So let's just check if we are out of question. And I think we are. So if we are out of question, then we are going to just thanks once again, Johanna and Alex for the X-Men talks and be here at the panel and their effort and putting together these talks. Very appreciated and contributing to the conference. And we are going to take our break. We're supposed to be like in 12 or so minutes, but there's no problem is lunchtime in UTC. So I'm going to probably use that in this case. And we are going to see you all, let's say five minutes to two, just to be sure that you connect correctly and everything is working fine. So we're going to start at 2 p.m. with the next talk by Artem. All right. Thank you very much. Enjoy some visual art.