can i ask you about trust (*ﾟｰﾟ)ゞ
- Co-founder, CTO (former CEO) of Fauna
- ~2017 Global transactional database tech based on Twitter experience
- serverless before there even was 'serverless'
- in 2016 people wanted servers not APIs
- ~2018 found early adopters building GraphQL interfaces for Fauna in the Jamstack
- pivot to developer-led db as a service
- & the rest is history
- him & co-founder (Chief Architect) Matt Freels ex-Twitter anarchist hippies
- Twitter: home of the weird
- off the shelf solutions like Cassandra & MongoDB wouldn't work for what they needed
- considering the journey of other small teams and how to help them
- "fundamentally motivated by anger and rage" fave quote of the show & why Fauna came around to help
- Moore's Law pun
- there are only so many large companies w specialized dbs
- for the rest of us they wanted to make off-the-shelf dbs that would grow with company
- LAMP era analogy & steak dinners & web 1.0
- data replication and inability to modernize
- 100s of millions of dollars a year on Oracle
- from the developer out
- provisioning microservices and GraphQL
- the permission chain of architectural change
- "you don't know what the future is going to be you just know you need to iterate"
- the bigger the company the harder it is for them to trust third parties
- is the foundation stable & secure
- making distributed strictly serializable
- Calvin algorithm
- giving people more information & transparency
- region groups
- knowing just enough to be dangerous
- some Phil puns
- no one migrates their database
- you can port if you want to...but maybe don't
- decoupled architectures besides the Jamstack
- mixing and matching
- sad rags
- more Pheels about Phil philtting and philing shirts
- not getting rid of old things...on purpose
- computer treasures, we want that data
- ummmmbilical cords & Beautician and the Beast
What is Remotely Interesting?
Join us to chat about Jamstack, coding the web, the people who code the web, and sometimes, lollies. With love from Netlify 💙.
[00:00:00] Previously interesting.
Hello, and welcome to remotely. Interesting. This is remotely interesting. Well, that seems a little presumptuous. That's the name of the show.
**Tara Manicsic:** Hey everyone. Welcome to another episode of Remotely Interesting. We're very excited to have you here along with a very special guests to tell us about his Jamstack journey. Today, we have Evan Weaver, CTO of FaunaDB. Um, and I, I don't want to sound biased, but we are fans of FaunaDB. We do use it and we like it. Um, but, uh, I want to just go ahead and jump right in to Evan telling us who are you and how did you find the Jamstack, or did it find you?
**Evan Weaver:** I am Evan Weaver, co-founder, [00:01:00] CTO, former CEO of Fauna. So we're, our Jamstack journey started, I guess it was late 2017, 2018. So far has been around for a while. We're a deep tech startup. We started. And working on global transactional database technology based on our experience at Twitter, which I can talk about in a minute.
**Evan Weaver:** And we were the first serverless database, but we were too early. Serverless didn't have a name. Like people thought what we were doing was strange and weird. They wanted to run traditional databases on hardware. They could kick if it wasn't working. Right. We, we, we, we, we initially launched the alpha, the service in 2016.
**Evan Weaver:** And people were like, this is crazy. Like, ha you need servers. So like I'm not going to use an API for my data. Like I would rather be able to delete my own bin log whenever I please and cause my own downtime. And we went on a little bit of an Odyssey in our go to market in pursuing enterprise [00:02:00] options for Fauna.
**Evan Weaver:** But while we were doing that, we kept the cloud service. We kept the database as an API, right. You know the whole time. And we had a small community of users and we, we found and of 2017, early 2018, and there were other vendors in what is now the Jamstack space, like, like Netlify and Vercel who are interested in Fauna.
**Evan Weaver:** We had a bunch of our early adopter, SAS, database users, all building GraphQL interfaces for Fauna so they could use with Jamstack components like Vue and React and Next.js and that kind of thing. And I was like, what is this? Like, all these people are doing the same thing. You know, we feel like we're beating our head against the wall a little bit to get enterprise doctors to switch from SQL ,but meanwhile, we have a bunch of people, like we're not even spending time with who are excited about the database and doing something useful for it. And we should probably invest in that. So. We surveyed those users. We talked [00:03:00] to them a lot. We ended up pivoting the company, you know, back to the developer led, you know, adopt adoption motion.
**Evan Weaver:** The developer led database as a service. Sort of product like growth kind of, kind of angle we had initially started with. And, you know, the rest is history from, from our perspective, you know, it was very popular in the GraphQL and Jamstack community, and we've continued to grow since, and now we're focused on maturing the product and making it easier to use for, for anyone who wants to adopt it, whether you're building a greenfield Jamstack application or augmenting a legacy app.
**Tara Manicsic:** What made you what made you, is it two of you that founded it?
**Evan Weaver:** Yeah, my co-founder who's now our chief architect is Matt Freels. He was on my team at Twitter.
**Tara Manicsic:** What made you and Matt. Okay. With a weird, like everyone else that, that serverless databases were weird, but you're like, no, it's cool. Like why?
**Evan Weaver:** Well, we were, we were early engineers at Twitter, so it's a [00:04:00] strange question to ask Twitter was the Twitter was the home of the weird in terms of, you know, we can't like so, so real talk.
**Evan Weaver:** Twitter was founded by anarchists and hippies, and that doesn't exclude me or my co-founder. I joined Twitter in 2008. So one of the first, what you would call experienced engineers on the team. About four other engineers at the time, a lot of the early team had been let go because the site was having so many problems.
**Evan Weaver:** I came from CNET networks had seen it. I did shout out, come and urban. Baby baby was a threaded real-time web chat for moms. So if you delete the for moms, it was basically Twitter. You know, I had some, I had some applicable skills in the Ruby on rails and my SQL space, and I wanted Twitter, Twitter, the product to serve.
**Evan Weaver:** And I, I ended up, I was employee 15. I ended up running what we call the software infrastructure team. I hired Matt onto one of the database teams for that. [00:05:00] And our challenge at Twitter was we were building a soft, real time global product. And there weren't any databases that were even marketed to solve that class of problems.
**Evan Weaver:** And, you know, that was the social media era. It was the nurse equal era. Like we looked at Mongo, we invested a lot in Cassandra and that kind of thing. At least for our purposes, none of those off the shelf solutions ever really arrived. Like we couldn't use Cassandra for operational data. It was, it was too, too inconsistent and availability problems.
**Evan Weaver:** It was difficult to scale Postgres or any of the off the shelf already BMS is at the time, you know, it was even like the pre-cloud era. We were in a co-located data center and had to rack HP hardware and it made it very difficult to scale. It made it very difficult to iterate on the product. We ended up building point solutions for all the core data objects, but we knew, like we knew that this was a general problem that could be solved generally.
**Evan Weaver:** And we didn't want other small teams like [00:06:00] ourselves. To have to go through that same journey. Everyone in the database industry is fundamentally motivated by anger and rage at the previous database.
**Tara Manicsic:** Hahaha I have no idea what you're talking.
**Evan Weaver:** No, no one, no one works on databases because they love them so much and find them pleasant experiences. They work on them because they're frustrated and they want them to be better. They know that. Yeah, something closer to perfection as possible, but it's just not happened yet. And that was the same, same reason we ultimately started following it.
**Evan Weaver:** And initially we were a consulting company for awhile just exploring the data space. But, you know, as the industry continued to evolve, we saw that no one was really working on this process. From the ground up, you know, there are people building middleware for existing databases. There are people trying to extend eventually consistent in a SQL systems and that kind of thing, but no one was really considering what does it take to build a modern consumer SaaS application like Twitter [00:07:00] from first principles?
**Evan Weaver:** What kind of data platform would you actually want for that use case and those same needs and goals and vision lined up essentially. Perfect.
**Tara Manicsic:** That's awesome. And it's like, we constantly are using more and more data. That's a little Moore's law on. Um, but like, sorry. But, uh, so it, it's funny to think just the fact of like, yeah, we're all gonna run into this problem. We're all gonna deal with this. And for, for you to just be like, yeah, let's, let's help with that. I just sit there and complain.
**Phil Hawksworth:** And there's such a, like a broad class of problems that you might be trying to solve by, you know, providing this service that you do. And because, because there are types of sites that don't need any kind of data storage or any kind of data, data layer, but the bump into that quite quickly, and it might be something very small and trivial in bite.
**Phil Hawksworth:** I need somewhere to put data. I need some way to get data [00:08:00] from. Right. The way through to I'm architecting something at a huge scale. I mean, you, you kind of talk about the Twitter scale and I, I imagine that you haven't thought, well, let's build a service that if you wanted Twitter could just use us as their data API, but there must be somewhere in that spectrum that you're thinking that it goes from this end of the spectrum to that end of the spectrum. So. How, how did you decide how broad to make it or where to kind of focus first if, if that's, uh, if that question makes sense.
**Evan Weaver:** Yeah. Yeah, it definitely does. One of the things we learned from our consulting experience and we mostly worked with social media and video game companies. Um, One of the things we learned from that was that, you know, there's only so many Twitters, Googles, Netflixes, have you in the world.
**Evan Weaver:** And if you're already at that size, you're going to essentially do custom databases. Like we did a Twitter and th that's fine. You know, you've already gone on this maturity, this growth and scalability journey. And you're kind of at [00:09:00] the end where specialization to your particular product makes, makes economics sense.
**Evan Weaver:** But for everybody else, you know, the 99.9% of the rest of the market, you know, there was nothing that you could get off the shelf that would really let you grow from a small MVP, a prototype, you know, a couple of people working in a garage or on now, you know, on a zoom trying to get something out the door.
**Evan Weaver:** I grow through your journey to maturity, both product maturity and complexity, operational maturity, availability, scale, global distribution, that kind of thing. So that was really our target, you know, be the, be the general purpose database that lets you start small, get to market as fast as possible. And grow.
**Evan Weaver:** And I think the best analogy for, for what we're doing and where we are in this space is comparing to the LAMP era databases in particular have gone through. Sort of these, these generations [00:10:00] and they follow this stack. The operational database is always the last thing to change because it's the riskiest thing to change.
**Evan Weaver:** Like in the eighties, early nineties, we had Oracle and mainframes and closet, you know, smart clients or rich clients, whatever kind of client you would call it. It was a desktop PC on the line. You know, running, uh, running the app. And we had that client server architecture for a long time. And then we got the web, people were dialing up, but 14.4 kilobits per second, and everything was great.
**Evan Weaver:** Um, but you, you needed, you know, dynamic web apps. Systems like Apache and lamp develop to which lowered the barrier to entry gave, gave hobbyists in particular, the ability to build an app and get it out the door without having to have a steak dinner with Oracle sales person and that kind of thing. And, um, you know, that really lowered the barrier to entry brought about the web 1.0 era.
**Evan Weaver:** And my sequel in particular was for its [00:11:00] time, you know, great at the core transactional workloads and adequate and everything else you needed to get an app out the door. It was general purpose. It didn't force you into like architecture just to build a basic web app. You didn't have to cobble together four or five different data services.
**Evan Weaver:** You didn't have to spend all your time on integration and duplication. And we, we found that the industry had kind of returned. To that failure pattern and people needed to launch at relatively large scale, you know, apps get distributed to a global audience out of the gate. Most of the time anyone can load a progressive web app or install a mobile app on Android or iOS.
**Evan Weaver:** And there was, there's kind of this like probably glad persistence movement at the time that said, you know, plan upfront, use the best tool for each individual job. Duplicate your data into all these different systems, you know, make sure the systems are replicated for availability. And we found that, you know, your typical, your typical [00:12:00] customer client at the time was replicating the same data, 15 times over spending all their time on integration, running these data systems that like single digit percentile utilization and just like having a terrible time.
**Evan Weaver:** Getting something out the door,
**Tara Manicsic:** This doesn't sound familiar at all. Well, so we basically did the same thing. Um, at the company I was working for everything from like, uh, cause we would build based on usage. We were hosting, we were hosting sites, uh, mostly like node applications and we were using Mongo, uh, mongoose, Cassandra, Cassandra, Cassandra, I don't know. I guess it depends on what dialect do you.
**Tara Manicsic:** Speak words and, um, was about six years when I think we settled on Cassandra and, um, it was like, we constantly like, uh, had so many replications of the data, so many ways to try and, uh, make it so [00:13:00] all fall backs because infrastructure was constant. Like we couldn't rely on infrastructure whatsoever to.
**Tara Manicsic:** Keep us, you know, make sure we have the data for our users and this is how we were getting money. Right. So based on their usage, we were billing them. And if we didn't have those usage numbers, we couldn't properly bill people and we couldn't get money. So it's not, it wasn't anything we were willing to be Willy nilly about.
**Tara Manicsic:** Um, so instead we just overdid it. We over-engineered it. And, uh, it was. Such such a pain. Um, and then like, cause like let alone trying to access things in a very efficient way and make sure that everything was, you know, persistent across every copy is just such a headache. Oh.
**Cassidy Williams:** I definitely have been on one of those teams where, because we wanted to modernize, but we couldn't fully modernize different parts of our data stuff. We had combos of [00:14:00] reticent, Cassandra and Mongo. Throughout the code base. And then we would end up having to duplicate code in different parts to make sure that the data was actually being saved properly. Cause we didn't know which was the source of truth at any given time. It's not an ideal situation.
**Phil Hawksworth:** And even just the, the on-ramp to getting to use these things is it's evolved so much as well. I mean, haven't you, can you kind of joked about having to buy an Oracle salesman, a steak dinner, um, but that.
**Phil Hawksworth:** Very close to the truth. Isn't it? I mean, that's not really a, that's not really a joke at all. I mean, I, I think it's long enough ago now that I'm probably safe to talk about it, but like a decade ago I worked at, um, uh, British telecon, um, and BT had had this, you know, they've got 150,000 employees. They've got like huge infrastructure.
**Phil Hawksworth:** And of course that means that they're a prime customer for a large database kind of vendor. And they. And they were, you know, we're contracted to, I [00:15:00] think, hundreds of millions of dollars a year from Oracle licensing. So if you were building a project, no matter how small, no matter how trivial, if you needed a database by crikey, it was going to be Oracle because we spent all of this money.
**Phil Hawksworth:** So there was a policy. Yeah. Oracle first and Oracle only. And you know, I, I joined the team at the point that, you know, we were building open source products and trying to evangelize about open source and kind of modern web development techniques, you know, 10 years ago as they were. And we, we wanted to use SQL Lite for something, or, you know, just anything else that would give us an just anything anything else apart from anything else, we wanted to be able to start writing code today, please. So. So, you know, start making progress or prototype something and having a very many months cycle to either acquire licenses or more likely just to get something provisioned from the infrastructure that you're allowed already just has such a huge barrier to entry.
**Phil Hawksworth:** [00:16:00] So tools like Fauna DB that allow you to get started quickly, look at an API document and say, okay, I can start putting data here and getting data back and start playing with it and understanding it quickly just feels like a huge shift to me. And so I think, I think that experience just as unlocked a lot of different expectations from developers around the world, and I assume just kind of infrastructure teams and technical architects and everyone, but I suspect it kind of percolates gradually from the developers.
**Phil Hawksworth:** Out. I mean, you mentioned right at the top of the, the kind of podcast kind of repositioning foreigner to think about being a bit more developer centric. So I'm assuming that this was kind of part of the motivation for that to kind of think, well, these people are going to be convinced quickly and get it. Is that, is that fair to say, or am I putting words.
**Evan Weaver:** Yeah, absolutely. And you know, like our, our DNA is developer DNA. I'm, I'm just the least technical co-founder. So I ended up as [00:17:00] on the business side of the house for a long time, but I'm, I'm an engineer by training and, um, yeah, I think the, the provision in point you brought up is a good one, which we see today.
**Evan Weaver:** Like, you know, one of the reasons fun as popular as the GraphQL, the native GraphQL interface, and in particular that fills a need where people today who are building modern applications, that one GraphQL access to the backend are exactly where you were talking about, where you have to like, you know, file your bug Zilla ticket with it, to get a VM installed somewhere.
**Evan Weaver:** And like 10 months later, you get the wrong class of VM. And you're like, I can't use this. And then you're like, your access doesn't work. Like, you know, they're doing the same thing, but it. microservices and Graph QL now. And you know, if you're working on a front end application team, you want to store some additional data.
**Evan Weaver:** Well, you can go find the backend teams, convince them to add a table to Postgres somewhere, make sure it's the right Postgres to make sure it's the [00:18:00] right team. Then get someone to implement. You know, some glue code and a go microservice that runs in Kubernetes somewhere. And it's the Kubernetes monitor.
**Evan Weaver:** No one knows what's monitored. Now make sure like we have telemetry and then we'll connect it to Apollo. And Apollo was great, but I didn't need Apollo. I just wanted to just store my data and you can do all that and like build a traditional GraphQL backend, but like that's, that's fine for GraphqL is original purpose.
**Evan Weaver:** It's fine for integrating legacy and traditional data storage. Yeah, more usable way for the front end, but it's not fine for greenfield or augmentation stuff. You just want to store data. You want to GraphQL API. You can access ubiquitously immediately without having to ask permission. The other big, one of the biggest advantages of MySQL or the innovations is just the fact that it was open source and could run, run on a workstation.
**Evan Weaver:** You didn't have to, you didn't have to have. To get it, you know, you can just use it. And if it worked, if you built something valuable, then you could go [00:19:00] figure out how to operationalize it. And you were in the same place again, as Jamstack you know, kind of follows that trajectory from static applications and sites, just like web 1.0, you know, static to dynamic rich, full featured SAS and consumer experiences.
**Evan Weaver:** When you have the same needs for an accessible data back in this general purpose, that's flexible that doesn't make you architect everything upfront because you don't know. You don't know what the future's going to be. You just know you need to iterate.
**Phil Hawksworth:** Can I, can I ask you about trust as well? God, I've just realized phrasing it. Like that sounds like I've caught you reading my diary or eating by sandwich from the fridge. I need to teach enough. Trust me. Sure. But that's a, that's a real. Stumbling block. I may be bringing some kind of baggage from my days, working at an agency, um, where we had to think about, well, what architecture we're going to use for this?
**Phil Hawksworth:** What tools are we going to use for it? And particularly if there's a big client, you know, who had certain expectations, um, [00:20:00] the idea of. Well, yeah, we're not going to have servers that we can kick, as you mentioned earlier on, but we're actually going to use some third-party to like, provide all of our kind of data storage and all of the, all of these critical parts of the infrastructure.
**Phil Hawksworth:** Um, I think the larger the enterprise, the slower they are to trust a third party to do that, or maybe the, you know, it's just, they've got different compliance hoops to jump through and all the rest of it, but it seems to me like we've, we've reached. I think a bit of a tipping point where. Those battles are kind of being won more often than they're being lost in terms of finding that trust.
**Phil Hawksworth:** Did have you witnessed that? I mean, does that something that, that played into your decision to found foreigner in the first place thinking, okay, we've got over that threshold or is that still a kind of a fight that's going on and kind of a regular battle to try and try and win that trust so that you can, you can operate for large enterprises.
**Evan Weaver:** Yeah, that's absolutely correct. And we have a whole fun and trust page [00:21:00] specifically for that, for that reason. Now I think the two, the two hurdles we had to overcome in, in our journey like databases, operational databases are the, and basically the hardest part of enterprise software because they're the riskiest.
**Evan Weaver:** And that also makes them, you know, the riskiest for the customer to adopt. Plenty of people were burned by Mongo and Cassandra back in the early days. You know, people, people want to know that the foundation of their house that they're intending to build is going to be stable and secure and there for them.
**Evan Weaver:** Long-term and, you know, I think that the first hurdle we had to overcome was just, it is what fun is doing possible. Cause if you remember back in the day, like the no seagull vendors, like DataStax and Mongo were aggressively preaching that distributed transaction. Couldn't work, you know, then it was a law of nature that if you wanted to replicate to more than one node, you would have to give up consistency and well, you want to scale don't you.
**Evan Weaver:** So like bite the bullet and, and do it, you know, try to fix [00:22:00] things in your app. You can never fix everything. But the basic experience is probably going to be okay. Well, that's not true, but it took, it took a lot of marketing and PR in particular by Google and by us. And technical marketing and technical education to get people to understand that there are new classes of algorithms that can actually make distributed strictly serializable acid transactions possible, or Jepson report was a big part of that.
**Evan Weaver:** We use an algorithm called Calvin, which is relatively unique, gives us better latency properties, and multi-region scenarios. You know, others in this space like Google spanner is, is. And algorithm based on atomic clocks and lacking, but it's now widely accepted that with the correct implementation, you can actually do transactions at scale.
**Evan Weaver:** So those sort of like, do you trust the architecture? Do you believe that what we're building can actually work kind of like the quantum computing kind of attitude now, our challenges are exactly what you said. You know, no [00:23:00] one wants to actually operate. Software hardware. Like everyone just wants things to work all the time, or if there's a problem, they want it to be someone else's problem.
**Evan Weaver:** But then, then that doesn't mean. That they want opacity. They still need operational transparency into how the system is running, because that lets them qualify. If it's working correctly for them, it gives them confidence and comfort and understanding about the system they're building. If you don't know if a bug is, you know, the vendor's fault or your fault, if you can't fix it, you need to understand exactly what's happening behind the scenes.
**Evan Weaver:** So a big area of investment for us is now exactly the operational transparency, even in a serverless context. You're giving people more access to telemetry about their data patterns, performance, and cost data log access, the ability to manage your own backups. And instead of relying on our internal backups and that kind of thing, and compliance also goes along with that, you know, if you can't have transparency, if you [00:24:00] can't see into your data services, you can't prove that they're compliant.
**Evan Weaver:** So our region groups feature, which allows you to choose exactly where your data is replicated and that kind of thing are all designed. Really to meet these kind of mid-market maturity requirements. Like if you build an app as a hobbyist, if you build a new app in a larger organization, You're not so worried initially about these kinds of concerns, but as soon as you start using existing user data, as soon as you start to grow, soon as you start to become a successful, it becomes paramount to get to the next stage of here, your maturity and your evolution.
**Tara Manicsic:** I feel like that's the thin line too, of like, I want transparency in a product, but I also want it to just work without me having to know anything. And so it's just like, I want, I want it to just work. And then when it, and not know anything until it. Then when it breaks, I need to dig in and have the information to be able to dig in.
**Tara Manicsic:** And it's like somebody that's making a product. That's like just such a, such a [00:25:00] balance beam there.
**Evan Weaver:** Uh, our goal is for it never to break. You know, obviously we, we do have occasional issues, but you know, we want you to be able to trust that, you know, finances, secure foundation. The product and the application and the, the, the business you're you're trying to build.
**Evan Weaver:** So our goal is to, you know, expose that foundation in a way that lets you understand it, that lets you trust it. This speaks to the security controls too. One of the reasons we didn't do a SQL interface. So the SQL doesn't give you the kind of security control you need to enforce access and identity and.
**Evan Weaver:** Um, you know, safety on the web. So we have a very robust attribute based security model, which lets you encode these properties similar to the way you can encode the geographical replication properties in the API, in the database itself and ensure that your data is only being accessed by the users who really should be.
**Tara Manicsic:** So [00:26:00] that, uh, well, one of the points that I really like, um, uh, on the founder's website is the line thinking data, not database operations, because that seems like such a like cognitive leap to me, because for all the years that I've been coding, I've been working with data, but every time I try to get or put data, it's always in a database operation that I'm thinking of it.
**Tara Manicsic:** Like, it's just such a. Deep down in concept to me, like, has that been, have you like hit walls against people? Like with that constant.
**Evan Weaver:** Th th the answer is yes. I mean, databases are conservative area of the market and customers are often very conservative as well. For good reason. Like we talked about the risks of choosing the wrong database, choosing something that's unstable, that kind of thing.
**Evan Weaver:** When, when people use it, they understand it. They get the idea of a programmable data API. [00:27:00] They get the idea that, you know, if the consistency and performance property. Of the backend. Don't vary based on where you're calling it from. Like, you don't have to worry about how it's provisioned. You don't have to worry about the location and size of the nodes and that kind of thing.
**Evan Weaver:** You know, there's nothing, there's nothing that mandates that every database be processed, isolated in a VM with, you know, excess memory just in case. I think it's something you have to experience and you, you may not experience it in databases. Initially, you may experience it with something like a graph, QL, a graph, QL API for an existing SAS service like GitHub or something, um, or Stripe that gives you more access to data you already have in, uh, a web native, easily accessible, secure way.
**Evan Weaver:** And then you start to think, you know, can I, can I, can I get this experience elsewhere? Like the rest of my application is running in JAMstack tools. There's no server side [00:28:00] component. Do I really want to spin up all this legacy apparatus just to take the next step into essentially dynamic behavior you don't, and then you, you see weird stuff, you see people's shoe horn, arbitrary data into other API APIs and start using it as if it was a database.
**Evan Weaver:** I'm sure you've all seen that before. Yeah. So I think like, like people, especially people who are using newer, newer technologies, aren't completely beholden to the. The incumbent web stacks. You absolutely understand this interaction model and want it. They just may not have a name for it.
**Phil Hawksworth:** Yeah. It amazes me how, um, how I've managed to convince clients in the past that I know enough about database operations to be able to be, yeah, this team, we can do that stuff as well.
**Phil Hawksworth:** Whereas actually they just, we needed a service to do it for us, but people would [00:29:00] sometimes be happier. Oh yeah. Well, You want to build this for us, right. And you'll manage the infrastructure and you do the database stuff. I didn't know what I was doing. I couldn't write, I couldn't write some select statements.
**Phil Hawksworth:** I can write some SQL but I'm just, that's dangerous. Just enough to be dangerous. And that, that bar isn't very high. It turns out. But, um, I think there's a, there's a, there's a big, I'm part of a big cohort. I think who. Yeah, why couldn't write to a database and I can get data and I can do a few bits and bobs, but that does not make me a DBA.
**Phil Hawksworth:** And that does not make me an infrastructure specialist. And so I, I rushed at the chance to be able to say, oh, there's an entire team that I can get, who are experts at that? And the way I get access to them as I use their API and I sign up for this, thank you very much. I'll take it. But you know, it it's, it, it feels like it's taken a while for.
**Phil Hawksworth:** The broader user base customers, uh, [00:30:00] environment to kind of be comfortable with that and say, oh, okay. Yeah, that's that, that works. You, it turns out it's better if, if you don't have access to that server, Philip, um, is this too personal now in my turn, am I sharing too much of my,
**Tara Manicsic:** you already asked him about trust. I didn't know if we were going to need a segment called fillings so that you can let all of those out, but now
**Cassidy Williams:** to fill up the time.
**Tara Manicsic:** Well, I mean, like you do bring up a good point where it's, you know, we talk, we're talking about this, but, uh, this new approach, but so many people have legacy approaches, especially to data, let alone, you know, legacy code that you had. Switching from the way that you handle data, uh, especially like you're talking about enterprises earlier.
**Tara Manicsic:** I feel like that would be the hardest switch because trying to get decisions made quickly for changes of legacy code in an enterprise company is takes [00:31:00] forever. Yeah.
**Evan Weaver:** I mean, I think, I think like one of the dirty secrets of operational databases of which there are many, um, is that no one really migrates their database.
**Evan Weaver:** You build an app is like it's been running on DB two version one since 1983, like is fine. Like no one is going to bet their career on changing that thing. Like, all you can do is lose, but that doesn't mean that doesn't mean you want to continue investing in that architecture. You're not going to like go learn COBOL.
**Evan Weaver:** Cold fusion and, you know, various stored procedure dialects and all kinds of stuff, because like you think it's great. It's just what you have and, you know, maintain it. And so, you know, most products that operational database lasts the lifetime of the product. And when you build a new product or you augment an existing product, Existing, you know, a [00:32:00] set of solutions, then you're at a decision point when you, where you can actually adopt something new and evaluate what will be most efficient and productive for your dev team.
**Evan Weaver:** Um, you know, that's what we're targeting to, you know, we're not looking for people to take their existing. No data sets that have been running fine on, on my seagull for 20 years and port them over like you can if you want, but you're, you're also undertaking, you know, fair amount of application rewrite in the process, but we do want to make sure that.
**Evan Weaver:** Everything new you build or everything you augment, you know, you no longer have to invest in that, that legacy stack, you no longer have to pay the maintenance costs. You no longer have to have a DBA to make sure your, your query is properly optimized. Like your devs can do it with modern technology and the way they understand and the way they want to build going forward.
**Phil Hawksworth:** We talk, obviously we talk a lot about JAMstack architectures on this, on this podcast and just in this team in [00:33:00] general, but we wouldn't, we wouldn't want to suggest that that's the only way that you can ever architect anything, no matter what you're building on the web, because you know, that's, that's not how the world works.
**Phil Hawksworth:** And I, we, we do talk a lot about. Decoupled architectures in general and this kind of move to using more services as APIs. And I, I kind of imagined that yes, do using a tool like foreigner is prime for someone who's reaching for a JAMstack architecture, but I kind of assume as well that. That's not the only place that you live.
**Phil Hawksworth:** Right. I mean, I'd like to think that other architectures are also embracing this kind of decoupled nature so that you've got an equally thriving kind of market there as well, which is more kind of quote unquote traditional stack, but still moving to decouple it would I, would that be a fair summary?
**Evan Weaver:** No, that's, that's absolutely true.
**Evan Weaver:** You know, not like we, you know, we, we talk about these abstract, like stack paradigms and stuff that most applications are not that PR, you know, they've been [00:34:00] developed for awhile, that people coming in to build them, have a mix of skills or the new that they want to apply. And we see lots of people building, you know, sort of the, the pure JAMstack architectures, but we also see lots of people, you know, augmenting.
**Evan Weaver:** Agile applications that have server side application code people, running things in serverless operational environments, but not necessarily using like the newest serverless frameworks and that kind of thing, and really anything that it needs to access data. In, uh, you know, a performance and distributed way is a, is a candidate for fun.
**Evan Weaver:** And we see a lot of use from, you know, edge clouds, people who want their data to be lower latency and don't want to. So like, you know, in a back haul to USC is to wine from everywhere in the world. You want a database which can distribute things closer to you. In a serverless way, give you a better performance experience, that kind of thing.
**Evan Weaver:** And there's so much mixing and matching to, you know, the old [00:35:00] database is left in place, but no one wants to alter a table anymore. It's too risky. So new data starts going into fun and that kind of thing. And we also see people use. Really it's that integration point, which is kind of a paradigm that went out of style for a while integrating through your database.
**Evan Weaver:** But with funny, you can do it cause you get the compliance and security controls. You get the unlimited scalability, you get the, the low latency, regardless of where you query the data from. And that lets you, you know, enforce consistency and constraints. In the data model itself in a way which makes having multiple applications access it and collaborate through it very easy.
**Tara Manicsic:** So we've gotten a lot of data into our brains about this. And, um, I think, I think it's time to put, uh, some, some even more special data. In our heads with our very serious section, we call it tidbits and thought things. Um, so this is where we, uh, ask a question that [00:36:00] is related or not so related whatsoever, uh, about each other's experiences and today's tidbits and thought things is what is something that you own that is old, but you have a hard time getting rid of.
**Cassidy Williams:** It's the one that comes to mind for me first is like really comfy. T-shirts like when you wear those comfy, t-shirts just to death. Those, those t-shirts, they don't look good anymore. They're pilly. They're not great, but they're comfy. I'll cling to them until they just kind of fade away into a rag.
**Tara Manicsic:** I think in one. And you can then use it as a rag now.
**Cassidy Williams:** Yeah, I do. But then I'm sad about it cause I can't wear the shirt anymore.
**Tara Manicsic:** So cleaning and you're sad that you're cleaning and sad that you're cleaning with the rag that you
**Cassidy Williams:** Overall. It's just a really sad experience.
**Phil Hawksworth:** I was going to go there as well. Cause I relate to that very much and I [00:37:00] do have one particular t-shirt.
**Phil Hawksworth:** I can't throw away cause it's for sentimental reasons, but I don't know how long I can keep wearing it. Cause it's starting to fall apart. And also it actually has the date on the front it's it's uh, it's a t-shirt from the basketball club. I was a part of when I was at university and it says Aston university basketball.
**Phil Hawksworth:** I beg your pardon. I was like question again. Do I still face it?
**Phil Hawksworth:** That's right? Yes. I'm, I'm much more buff now than I used to be. So you don't fit in.
**Phil Hawksworth:** It's just a bit, it's a bit tight around his shoulders. Loads of rooster. Rumor knows what it is. It is, it has the date on it as well. So it's 95 to 96, which also means that the style at the time was those t-shirts were. Which means that yes, I do. Still fit it. Thank you for that one, but oh yeah. I don't know what else I would put into this.
**Phil Hawksworth:** I don't want, I've [00:38:00] got to come up with something other than a t-shirt, but I need more thinking time.
**Tara Manicsic:** Do you have one Evan while he thinks? Cause mine is weird. So I should probably wait until last.
**Evan Weaver:** I mean, I, I, I really like old things with the exception of old databases. I put all my energy for the new and to, and to find out and that's basically it, you know, I, I always try to repair everything.
**Evan Weaver:** Retro game console. Is that on any TV and that kind of thing? You know, my recent, my recent upgrade to the, to the gaming system was, uh, a new converter for the RGB skirt. And although again, a massive improvement. So I'm very happy with it.
**Tara Manicsic:** That's why I like that answer. I like, I like not getting rid of old things. I definitely have things that I've patched and patched and okay. Just like old code.
**Phil Hawksworth:** That sounds nice and deliberate as well. And like, it's not that you just couldn't get rid of it. It's also, you still use it and upgrade it and maintain now all of the rest of it that goes along with it. I mean, [00:39:00] I, I, my other answer, I think until we're building up to Tara's weird one, I suspect also it doesn't fit her anymore.
**Phil Hawksworth:** Yeah. Okay. Would, would be computers. I do have a couple of computers that I just can't get rid of and it's not because I, like, I still need them or I text them. It's just that I haven't necessarily got a hundred percent confidence that I've got every last photo or thing. That's, I've been through it, but I don't, I never finished that job.
**Phil Hawksworth:** So they're in the attic and I don't know, I need a house. In order to fall, but they've traveled with me through our house move.
**Cassidy Williams:** So I have a laptop that's legitimately like five pounds that I need to get rid of it. It's so heavy. It's all old, but there might be something on it that I might want to save.
**Phil Hawksworth:** Go to be a treasurer.
**Tara Manicsic:** Speaking of with data storage, I have an external hard drive that no drivers work on it anymore. If anyone has a [00:40:00] really good, like YouTube tutorial to open it up so that I can access what's on there. I'm home in this, it's in this drawer next to my desk. I'm like, I want it. I want that data so bad.
**Tara Manicsic:** I want it. There's gotta be something good. I'm going to open it up. It's just going to be like really. Really stupid things that I, you know, brought down up of Napster and just getting a snack. That's going to be a old game of selfies or something. Um, yeah. So my weird one is, um, whatever is, this is pretty gross.
**Tara Manicsic:** You're so I have two kids and I have the umbilical cord. Belly, but a lot of moms do that. And it's just like, why? Like my mom had had my teeth for like my baby teeth and I was like, that's disgusting. Get rid of that. But now I get it. I have. Weird shriveled up pieces of things that don't fit me anymore.
**Tara Manicsic:** Fill your right. Um,[00:41:00]
**Tara Manicsic:** oh, I want to take back. I want to take that. We need to unsay something and then it just like, like my port, my daughter is like, I mean, they're not one is like almost three one's seven months. My daughter's is like in a drawer and I'm like, I should do something with that.
**Tara Manicsic:** It's the Castro.
**Phil Hawksworth:** I remember once upon a time, I remember you talking about chaos draw. It's not in the cave drawer. He's just losing the chaos draw. It's a baby. It's like weird, like a baby things like chaos. Yeah. I'll say as a K a straw. It's like, I don't know. There's like a valve adapter for pumping up.
**Tara Manicsic:** Yeah there. Are there things that you can use to clone humans in it? Yes.
**Phil Hawksworth:** Um, is there, is there a plan for how you're going to keep an present? The, I mean, are they going to go in a front, on the wall [00:42:00] in their bedroom to give them nightmares or?
**Tara Manicsic:** I don't know. Cause I have one that I have almost a three-year old one and it's seven month old one. And I'm just like, what am I supposed to do with. I don't know, but it just feels so weird to throw it away. I don't know, in the yard.
**Tara Manicsic:** That sounds, that sounds better than I think,
**Evan Weaver:** I think we plan it. We planted a lavender Bush over them.
**Phil Hawksworth:** That's nice. That's probiotic. It's lovely. Throwing it in a drawer is not that.
**Cassidy Williams:** Have any of you seen there's an old friend Drescher movie called beautician and the beast. And you, have you seen or heard of this movie?
**Cassidy Williams:** First of all, it's campy and excellent. If you just want to watch an old friend Drescher movie, cause you can kind of guess the general genre of the movie based on the title, but there's this line in the movie. Her mom is trying to guilt her into stuff. And she's [00:43:00] just like, don't travel away. I love you. And she's just like, ma you love me too much. You got to let me go. And she's like going through her mom's purse and she finds a little film canister and her umbilical cord is in it. And meanwhile, her mom is trying to convince her that she's not clingy. And it's just the perfect, comedic timing in that scene.
**Cassidy Williams:** So anyway, great. Highly recommend. But there yet. Yeah. Okay. All right. Unless you want to take after Fran Drescher smile beautician in the beast.
**Phil Hawksworth:** Th the show notes for this episode are going to be quite a mixture on there. Incredible. All right. I think, I think we must be closing in on time. I th I th I don't know how much more of Evan's time we can, we can hold him hostage. So, Evan, thank you so much for, for joining us today. We really appreciate it. It's great to hear some of these perspectives about fauna. Now, before we let you go. Where do people go to sign up and get started using for? And I was what's the scoop,
**Evan Weaver:** the Fauna Fauna is [00:44:00] completely free to get started with no credit card, just an email or a get hub account or a Netlify counter or for sale counter, any account. And really, um, so just go to fauna.com, F a U N a and click sign up, go through the tutorials. We've you know, much improved docs that we just launched and yeah, let us, let us know how you like it. Trying to build something. Join the Slack community Yeah. Have fun with, with modern databases.
**Phil Hawksworth:** Love it. That sounds great. Awesome. Well, thanks. Thanks to Evan. And so, uh, so that's, that'll do it, I think for another episode of Remotely interesting. Um, we do have another one just over the horizon, so don't, don't, don't fail to come back for that, for that next time. Uh, we've been looking forward to this, actually, we're going to have a debate in switch, which is a named.
**Phil Hawksworth:** I think to celebrate on its own. I think so our debate, the switch, we're going to be talking about monopolizing technology who knows what could unfold there. [00:45:00] Uh, so, uh, so until next time on Remotely Interesting, uh, I've been Phil there's a fly in my soup Hawksworth
**Evan Weaver:** This is Evan I can't do this again Weaver
**Phil Hawksworth:** noted. I think that's the earliest. We've got a review on one of our podcasts,
**Tara Manicsic:** two stars. I can't do this again.
**Cassidy Williams:** I'm Cassidy because it's a seagull, not a bagel Williams.
**Tara Manicsic:** I'm Tara interrupting cow. Manicsic. Thank you. That's a real