The Changelog: Software Development, Open Source - Blockchains and Databases at OSCON (Interview)

Episode Date: December 14, 2017

We went back into the archives to conversations we had around blockchains and databases at OSCON 2017. We talked with Monty Widenius, creator of MariaDB the open source forever fork MySQL, Brian Behlendorf, Executive Director of Hyperledger, the open source collaborative effort hosted by The Linux Foundation to advance blockchain technologies, and Tague Griffith, Head of Developer Advocacy at Redis Labs, the home of open source Redis and commercial provider of Redis Enterprise.

Bandwidth for Changelog is provided by Fastly. Learn more at Fastly.com. And we're hosted on Linode servers. Head to linode.com slash changelog. This episode is brought to you by Auth0. Auth0 makes authentication easy. We love building things that are fun and let's face it, authentication is not always fun or easy to build.

It can be a pain, it can take hours to implement, and sometimes even days. And even after you have it all in place, you have to keep it up to date, keep it secure, and Auth0 makes it super easy and fast to implement real-world authentication and authorization architectures into your apps and APIs. You can allow users to log in however you want, regular username and password, Facebook, Twitter, enterprise ID providers like AD or Office 365, or let them log in without passwords,

just like we do on changelog.com. To get started, grab the SDK for your platform of choice. Add a few lines of code to your project. This can be a mobile app, a website, or even an API. They all need authentication. Sign up for Auth0 and get the free plan, Thank you. changelog. No credit card required. Once again, ozero.io slash the changelog. You're listening to the Changelog, a podcast featuring the hackers, leaders, and innovators of open source. I'm Adam Stachowiak, Editor-in-Chief of

Changelog. On today's show, we're going back into the archives, the conversations we had around blockchain and databases at OSCON. We talked with Monty Vadanius, the creator of MariaDB, the Opens for Forever fork of MySQL, Brian Bellendorf, Executive Director of Hyperledger, the open source collaborative effort hosted by the Linux Foundation to advance blockchain technologies,

and also Tad Griffith, Head of Developer Advocacy at Redis Labs, the home of open source Redis of developer advocacy at Redis Labs, the home of open source Redis and commercial provider of Redis Enterprise. So we're joined here by Monty Badanius, one of the founders of MySQL and CTO of MariaDB. MariaDB. Maria.

Oh, see, Adam and I were arguing how you pronounce it just a few minutes ago, weren't we, Adam? And I said Maria. And I said, this guy's with MariaDB. And he goes, isn't it Maria? And I said, I don't know. At least you got MySQL right and not... It's not MySQL. Oh.

We also have interviewed Richard Hipp of SQLite. And we learned there that he likes to call it SQLite, like a meteorite. So fun times arguing over how to pronounce things. But, Monty, first of all, thanks for joining us. Secondly, thanks for MySQL. And tell us about MariahDB. Give us just a brief history of that. I want to get into your talk about how to make money with open source.

That's what he's talking about today. But before that, give us just the brief rundown of the history of the MySQL, the MariaDB, and where you're at today. I mean, I was the original creator of MySQL. In other words, I did most of the development and eventually we sold to Sun and then Oracle took over Sun and me and other group of people was afraid that MySQL would not survive as an open source project so we went out from Sun and created a fork of MySQL called MariaDB which is guaranteed open

source and almost all of the original core people left MySQL, developed and joined MariaDB. So I'm working with the same team I have been working with since 95. So is that a point now where if you're still using MySQL, there's just, you should be on MariaDB now? There's no reason to use MySQL anymore. All distributions, Linux distributions, except Ubuntu has switched to MariaDB as default.

Ubuntu is also including MariaDB. So if you want to use what everybody else is using, you should use MariaDB. And was it a straight code fork in terms of... Originally, yes. But now we have some 30 man years on top of MySQL. A lot more features, better application, and much more secure. Very cool. So I had heard the story of Maria, I never heard anybody say MariaDB out loud, obviously, but I've read that that was happening, but I didn't know the state of the new thing,

which is great to hear that it's production ready, it's out there. For five, six years already. Five, six years, so. The MariaDB company, we're 100 people, almost half are developers working on MariaDB and things around MariaDB. So we're doing really good.

Really cool. So you're talking about how to make money with open source. Yes. MySQL was open source. You sold a company around it to Sun, then to Oracle. Now MariaDB is open source, but yet MariaDB Corporation.

So tell us about how you're going about running a 100 person company on an open source thing like this. Yeah, so with MySQL it was easy because we did own all the code, so we come up with how to do dual licensing on GPL, and we were the first project to ever did that. You were.

And the thing with being a product company where you create a product, you can't just survive by license revenue. And we did know that from the start. So that was also why we started with dual licensing because I wanted to work full time on MySQL and I couldn't do that just by doing support and consulting. So we were the second project who did dual licensing at all

and first with GPL and that allowed us to grow and then eventually got sold to Sun for one billion dollars. Oh wow. And the license revenue was the one that made it the difference compared to other open source companies. And nowadays I'm helping a lot and advising a lot of companies how to be a success in a similar way that we did with MySQL.

With MariaDB we don't know the code because we are a fork of MySQL so we are bound by the GPL. So what we do is then is support and subscription around the core product, but then we have some other products on the side that has a licensed revenue that allows us to grow. Because if you want to compete with the big guys, you need to have a similar revenue model as they have.

Even with open source, the nice thing that I loved with MySQL was that only one in a thousand paid. And that mean that we was able to really help a lot of people to get a good database, and at the same time having a few who made it possible for us to do that. Which is awesome, absolutely. Yeah, and that was the best thing

with open source licensing, that you can actually do a good thing and at the same time earn enough money so you can continue doing that. Yeah, at the same time earn enough money so you can continue doing that. Yeah, so I'm just, I got stuck a little bit on your billion dollar sale. I didn't realize that was that big. Personally, Monty, why didn't at that point you say, well, I'm going to go get a mojito

and hang out in Cancun or St. Lucia for the rest of my life? Retire. Yeah, why didn't you retire at that point? You wanted to keep going. So first I wanted to ensure that Sun would get the most benefit of MySQL. That's why I joined Sun originally, to see that they would be able to steer the project, MySQL, in the right direction. And I stayed there for one year.

And then there was friction between people who were used to working at a small company and didn't like the atmosphere of Sun. Nothing wrong with Sun but a big company there's lots of bureaucracy and everything else and people started to leave. So I talked with them, CTO, and said that the best way to ensure that you don't lose the minds of these bright people is that I go out, I create a new company, I hire these people and then I can take care of them as they would be in a small company, but you would have access to them as they would have worked for you. Santo, that was a great idea, so I left Sun, and one month later, then Oracle announced

that they will buy Sun. So instead of helping Sun, I refocused to start the work of working on the project. Did that completely blindside you, the Oracle acquisition? Yeah. Yeah. And the thing was that I didn't ever plan to do a MariaDB fork, and I probably would have been working or being in Hawaii with Mojito, very likely.

But the thing is that when you have put your whole life into an open source project and feel that you have made a difference and helped a lot of people, just letting everything collapse just because you want to have your mojito, I would have felt bad about it. money that I got from the Sunsail I put into ensuring that MariaDB would survive so that both the project I have worked on since 1981 and the people that I worked with for tens of years would have a workplace for life. That's so great. I mean, a sense of responsibility and a sense of purpose is better than a mojito. And you can take vacations, but- You got to think about legacy.

You got to think about future. You got to think about the future. And all the people using the software. Yeah, yeah. I mean, I felt really bad for all the people that I got to believe in open source. It's here to stay if my scroll just disappeared. I would basically have let down all of them.

And that would not allow me to sleep. And I do value my sleep. So it was easier to work more allow me to sleep. And I do value my sleep, so it was easier to work more and get good sleep. So, give us some examples in Maria's model where you have the core product, you don't have licensing on that because of the reason that you said.

But then you have these side add-ons or? So we have, what we have now, we have a max scale, which is a really good proxy. We also be able to do failover and other things, but also use you more control of using MariaDB, especially in larger environments. And that we have a license under something we call the Business House license.

This was a license that me and David Axmark, the other founder of MySQL, come up with some 10 years ago. And now we have started to make it more popular. Because I see a lot of companies struggling with that. They would love to do open source, they see the benefit of it, but they are afraid that if they would become open source,

they would lose all the revenue, the customer would stop paying them. And that causes a problem with conflict. You want to do the right thing, but you can't. And the business source license was invented to solve exactly this thing. To allow more people to do eventually open source, but, and also give the users of the software

many of the freedoms that you expect with the open source software. So it's not an open source license, but it's an eventually open source license. It's not an open source license, but it's an eventually open source license. It become within four years guaranteed that the code that you have is open source. But you have all the code from the start. So you get the benefits that if there's bugs,

you can fix it. If you want to add something, you can do that. You still have to pay for it like with any closed source software, you can do that. You still have to pay for it like with any closed source software, or commercial software actually. But you still have the freedom that you are not held by the one company who can do any changes or just go out to business or stop producing the software. So this is an entirely different thing.

You said it's business source license. Business source license. It's a closed source license, but with lots of open source ideas around it. And we have released MaxScale on this one, and going from having MaxScale as a free component in our subscription offering, now we are selling licenses and deal sizes,

not anymore 10,000, they are in the 100,000 or even in the million dollar range. subscription offering, now we are selling licenses and deal sizes not more, any more 10,000 they are in the 100,000 or even in the million dollar range. So it totally transformed the business. Other option you have as a company is to go with OpenCore. I hate OpenCore with passion because what I expect from software I use, I want to be in charge of my own destiny. With OpenCore, I lose all the freedoms. I can't do bug fixing anymore. I'm totally dependent on the company's whims and so on.

Business Source is there to be a bridge between closed source and open source. And in the end, produce more open source companies who couldn't afford to do it before. Are there other people taking this business source license that you're using at MariaDB and using it for their companies? Is it a part of Steam?

I've been holding talks about this now for two years and there have been lots of interests. Still, they needed somebody to show the way and that's why we released, was it nine months ago, MaxScale as a business source and make lots of noise about it. And today also talking about that today. And we need to spread the word because personally, I feel it really bad as an open source advocate

to see that on the phones, almost all apps are closed source. And the only reason for that is that these companies want to earn money and the open source model doesn't work. With business source, they could earn as much money as they do now, but still getting the benefit of getting other people helping them develop the software. Very cool. We're definitely going to check that out. Adam, I had not heard of the business source license. I've got a pull up right here. It's BSL 1.1 right now it seems. And it seems like if they contribute back through the business source license it will eventually be open source as well. So is it? That's the idea. The idea is basically that with the business

source initiative is that i want to make business source or bsl as easily understood as gpl gpl is a complex license when i started with my sql um when my sql year 2000 become gpl i spent so much time explaining gpl to companies and the effects of that. Right. And no people know what it is. To get BSL popular, you need to do the same thing. And that's why we have done it trivial for anybody to adopt it. So you just basically change your header files and add that this is now in the BSL and then you copy one file from our website

and change three lines that describes your product name, the restrictions, when you have to pay, and when your source code becomes open source. And that's the only thing you have to do. So it's very trivial to do that. So last question. look forward five years, three to five years, MariaDB both on a technical sense and as your business, where do you see it going?

Where would you love for it to go in five years? So the interesting thing with open source software is that you are not the leader of where you are going. You are working with the community and if you have the right community, we basically are the one who is predicting and steering the future, and you are working with them, you're basically part of it.

And that makes open source so strong. And that's why I want to have more companies being involved in that, to be a part of that. The only thing I know is that that I see the trends just now is that much more distributed environments and setups where you're using lots of server and exchange data behind them. Like some of the NoSQL solutions but still with a strong asset compliance so that you know that you never lose something and we have already have a product for

that in them all up areas in other words and in analytics but we also want to do the same thing in in in a transaction environment so that that's we see at least one trend yeah another trend that we are working on is make it easy to transist from closed source commercial databases to open source databases. So we're adding in 10.3, we're adding a PLSQL layer on top of MariaDB. Oh really? That's interesting. So where do people go if they're interested in the BSL, MariaDB, and want to learn more, considering using BSL for their software, where do you

go? Where do you send them? You just search with your favorite search tool, search for the business license, and you will find the thing on top. It's easy. Same thing with MariaDB. Okay. Or you just go to your favorite website that's probably already using MariaDB. Gotcha.

There you go. Very good. Well, Monty, thanks so much for joining us. This episode is brought to you by DigitalOcean, who just launched Spaces, a beautifully simple object storage service designed for developers who want a simple way to store and serve a vast amount of data, such as hosting web assets, storing user generated content such as images and large media files, archiving backups in the cloud, and storing logs. Just like you use S3, Spaces has an ecosystem of S3 compatible tools and libraries that can be used to manage your space and it's available independent of DigitalOcean servers. You don't need to use anything else but just Spaces if you want and to make it easy to

try for both new and existing digital ocean customers you can get started today with a free two-month trial of spaces by going to do.co slash changelog and for new customers only you'll also receive a $10 credit to use for digital ocean droplets or other services. Once again, do.co.changelog. And by TopTal, the best place to work as a freelancer or hire the top 3% of freelance talent for developers, designers, and finance experts. In this segment, I talk with Jeff Mazur. My name is Jeff Mazur, and I'm a TopTal finance expert. And Jeff has a pretty awesome background for a freelancer. I have a background both in finance and law, and I actually have a CFA and a JD.

I've worked both in finance and as a practicing lawyer for major law firms in New York, Washington, and Chicago. So I asked Jeff to share what differentiates TopTal as a global talent network and the process he had to go through to ensure he could be trusted as a finance expert for TopTal. The differentiator I see between TopTal and some of the organizations that are comparable or try to offer a similar type of service is that the people who are part of TopTal have really gone through pretty extensive screens. So in my case, for example, I probably spent 20 hours of preparation and conversation and interviewing to make sure that I was the right fit for TopTal.

So what I offer and what other TopTal finance experts offer is we offer just really deep expertise in the areas that we talk about. So for example, if you look at, you know, ICOs, if you just did a Google search or you went into another platform and you look for ICO experts, ICO coordinators, ICO finance experts, you'd come across thousands of people. I mean, given the frothiness of the market and the level of interest in the market, you know, everyone's a finance expert right now in ICOs. But in the case of TopTal, what they've really done is that they winnowed that huge group out to come up with people who really are experts in the field. So if you're looking to freelance or you're looking to gain access to a network of top

industry experts in development, design, or finance, head to TopTal.com and tell them Adam from the Change Law sent you. That's T-O-P-T-A-L.com. And for those out there wanting a more personal introduction, email me, Adam at ChangeL Law sent you. That's T-O-P-T-A-L.com. And for those out there wanting a more personal introduction, email me, adamatchangelaw.com. So I love interviewing people who have Wikipedia pages. Oh. Because it makes my job really easy. Like I don't have to do any research. I just pull up Wikipedia.

I assume everything's 100% accurate. It's absolutely accurate. I watch it every day. You watch it every day. You've got, like, a program that checks it. Jimmy won't let me edit my own Wikipedia page. That's what I hear. I was speaking with the fellow that started MySQL

on MariaDB. Yeah. And I was asking him about his... Yeah, Monty, thank you. About his Wikipedia page. And you can't edit your own page. You're not allowed to. Oh.

I think that's actually a pretty good policy. It's near miraculous, then, that mine is actually updated at all. I'm just inaccurate. So thank you out there to anyone who's editing my Wikipedia page. Yeah, so Brian Baelendorf. Been around a long time,

doing lots of things, Apache Web Server, part of the Mozilla Foundation, EFF on the board there. First of all, thanks so much for sitting down with us. Thank you, and it's great. You of all, thanks so much for sitting down with us. Thank you. And it's great you mentioned all my hobbies. And not your job. All the things that I do have done for fun.

So what do you do for not fun? What do you do for business? Okay. So my day job, my 150% time kind of gig, is full-time executive director of Hyperledger, which is a collaborative project hosted at the Linux Foundation. Okay. So Hyperledger, tell us more. DAN GALPIN- Sure.

It's a open source project. It's actually a community of projects that is focused on building technologies, kind of building blocks for doing distributed ledgers and smart contracts. MARK MANDELBAUM- OK. So I look at somebody with a Wikipedia page, a history in software.

I don't want to structure you too much, but you could probably do lots of things at this point in your career. Why is this something that you're excited about? Why are you working at Hyperledger? I've kind of had career ADD. I've done a couple of startups. I worked for the World Economic Forum. I worked at the White House. I was a venture capitalist at one point. Never really worked for a giant company, but kind of had been swimming around them a lot. And I had this unnerving kind of feeling over the last 10 years, having been on the internet since fairly early days, like early 90s, right? That the last 10 years has been this move towards centralization rather than decentralization,

right? That as we've gotten more digital, we started to depend upon central companies more and more and been forced to trust them rather than choosing to trust them. And that's because it's always easier to set up a big central database than it is to build a federated or decentralized network, decentralized system. But the reason why the internet created so much value and was so amazing was because of its decentralization. Yes, absolutely. And I'm still one of those freaks who runs my own mail server, which I'm sure is going to get DDoSed upon somebody hearing this.

So you go into your spam assassin and update your Bayesian filters? Right, exactly. There's a lot of attacks there. So, no, it's not easy, but I actually enjoy it because I'm weird like that. So when I started hearing about Bitcoin, I was kind of skeptical because I'm not a currency speculator at heart. And I don't necessarily agree with a lot of the political kind of side of that. But I was really attracted to the idea that, hey, here's a bunch of people now thinking about decentralization, right? And autonomy, sovereign presence on the internet, all that good stuff.

So I followed that a little bit from the periphery, followed the Ethereum community from the periphery. But I was still passive. And it was as an investor at my gig just before this, where I started to talk to Bitcoin companies and blockchain companies, and finally heard the use case that made me go, whoa, which was land titles in emerging markets. Land titles.

Land titles. Real estate. Real estate, which might sound completely bizarre, but anybody who's ever bought property or a house or sold property or a house knows how much BS there is in dealing with paperwork, but also how much we depend upon, you know, registering paperwork with the right government officials and signing the documents.

Well, we even have companies that do these things like closing fees and entitlement companies, right? Right. And it might seem like, why bother thinking about this as a problem? Well, in emerging markets where they had started to digitize land records, it actually made it, in order to give people not only a sense of ownership over the property they perhaps lived on for generations, the idea was also allow them to be able to take out loans to start businesses, right? There's an economist named Hernando de Soto who wrote a book

called The Wealth at the Bottom of the Pyramid, which was all, let's help India build their economy and South America and others by allowing people to build businesses on top of the ownership in the land that they have. So there's a big movement to digitize these systems, but it was discovered that this also made it easier for corrupt bureaucrats and those who had root- Which is already pretty easy, but this is even easier now, right? And those who had root on that database, to be able to change entries in that database without any accountability, without any sort of paper trail,

and made it easier. Which is the problem with centralized systems is whoever holds the keys holds all the power. All right. So there was actually an Austin-based company called Factum, which has been working on land titles in addition to a bunch of other related projects

and using a distributed ledger to track changes to a public database of who owns what property where. Okay. And in doing that, making it much harder for people's land to be taken away from them because then it becomes obvious when there's an entry added that violates principles, right? That violates this person's consent, this landowner's consent, et cetera. So that made me go, aha, and I started digging more at it. And my next question was, could you do this kind of thing without a cryptocurrency?

Right. Because, you know, I run a mail server. I don't charge people to send me mail. If I did, I'd be a very wealthy man. Perhaps that's my mistake. Or you'd get way less mail, right? But I also think the internet, like I was on Prodigy before I was on the internet, right? Prodigy was this BBS ran by IBM, right?

Where it cost you 25 cents to send an email. Oh, wow. Yeah. That would stop some spammers. Exactly, right? You never got any spam, but it was also 25 cents to send an email. You also never got an email. You had mailing lists, in fact.

Well, no, I won't go there. So what was great about the internet is you didn't have to charge people to set up mail servers or send people mail. You didn't have, you know, and websites, the same thing. You could set up mail servers or send people mail. And websites, the same thing. You could set up a web. So I really wanted to figure out, were these two highly dependent or could you tease them apart? And it turns out that cryptocurrencies are an application of distributed

ledgers and smart contracts. And so once I saw that architectural approach, I started going, oh, who's working on that kind of thing? And then I saw Hyperledger get announced by the Linux Foundation, right? And to some degree, it kind of looked like a 45-year-old. I'm 44, by the way. A 45-year-old showing up at a skate park with a skateboard going, you know,

hey, you guys make it look easy. Hey, dude, let's do this. Hold my beer. You don't want to be that guy. No, no. So learning more about it and spending time with Jim Zemlin, who's the head of the Linux Foundation,

spending time with folks who are part of pulling that together from IBM, from a startup company called Digital Asset, from a bunch of others, learned more about kind of the rationale and said, okay, this is important to work on. It's not against the cryptocurrency side of the world, but it's like, in fact, there's ways that we need them

and ways to work together. But maybe there's some real opportunity here to build something that has some lasting impact. A standalone thing that's all about the ledger, it's all about that public record. It's about getting this idea out, getting it in the form of Legos and building blocks that people can piece together,

and making sure that all the good ideas out there about how to do it, all the controversial kind of architectures, and different ways of doing smart contracts, give them a place where they can be built independent of a single vendor. Because this is another thing that had been on my mind is a lot of the open source community had kind of forgotten about the role of neutral kind of homes for the governance of the projects, right?

So Apache has done that very well. Right, foundations. Free Software Foundation, other, Software So Apache has done that very well. Right, Foundations. Free Software Foundation, other software conservancy have done that very well. But much of the open source world had gone to a company kind of releasing it and then setting up a.org

and saying the community edition. Which makes it harder and harder for us using open source or even just following open sources. And something that we do at the Change Dog all the time is help us tease apart the business from the open source thing

and see clearly through the mud. And I really felt it was important. If this was how we re-decentralize the net, if this is how we rebuild the building blocks, that that plumbing be as publicly owned as possible, right? And be as multi-stakeholder, if you will. Like this multi-stakeholder concept is something that I've believed in for a long time and at the World Economic Forum is a big part of how they operate too. This idea that the best policies and the best projects in the world come from people with wildly different agendas coming together and figuring out what's the greatest common denominator of what you can work on, of what you can build, of what you all independently want and then

can build together. That's why the Hyperledger approach appealed to me. Tell us the approach specifically. Help us understand it in your context. Hyperledger approach kind of appealed to me. So tell us the approach specifically. Help us understand it in your context. Hyperledger, the open source thing. This is your 150% day job, right? Yeah. So the business.

So I'm the pointy-haired boss. I'm the diplomat. I'm the one who try to cajoles the companies into joining the project and then working nice with each other, right? Right. Because it is too easy in open source projects for one vendor to run away with the title, run away with the brand, run away and say, we're the only Linux company or we're the only Apache company. Sure. Lots of concerns about that in the early days of any of these projects.

Absolutely. And so part of it is keeping that at bay. And that means doing boring things like thinking about trademark and thinking about events to be at and how to get the right messages out. And then there's the other side, which is having the substance. It doesn't matter what you talk about if you don't have substance. So how do you get the communities together? How do you bring the right kinds of additional projects in, the right kind of additional thinking in, so that it's not just one person's view of what should be built, but a collective

view and where you can be challenged on that, where there's a healthy community of people who feel empowered to say, maybe there's a different approach, maybe you're wrong. It's the multi-stakeholder thing, right? Let's get to rough consensus and running code as quickly as we can. Really emphasize running code, right? It's something that doesn't matter which ideas are big, you can't boil it down to running code. So get these communities together, ship it,

get them using the best practices of the open source community, whether that's how Apache works or on a security perspective, the CII badging process. Like, how do you run security conscious, security sensitive kind of projects out there? And then finally, a lot of the companies useful, and not have to divulge details about your client or your private business. A lot of people are very skittish and still unsure how to do that. And so 20 years later, it feels like a lot of the same issues. But these are ever-growing needs. Everything old is new again.

Give us the purview of the technology, where it stands, what it is, the meat of the matter. So there's eight different projects at Hyperledger, all of them except one in incubation status still, because we're still growing the community cultures around these projects. The one that's graduated is called Fabric. Okay. And it's one of four, which you might think of as frameworks at Hyperledger.

It is a tool for being able to stand up a distributed ledger amongst a network of participants. So you and I and eight other people would independently run a Fabric node, right? And we'd point at each other, right? We'd be permissioned, actually, amongst each other.

And then between us eight, between us ten, whatever, there would be this common system of record, right? And you could write entries, I could write entries, somebody else could write entries. All verifiable, irreversible. Well, yeah, exactly. The immutability, because it's a Merkle tree, which is a string of signatures with data,

so that you'd have your copy, I'd have my copy, but the integrity of that would be verifiable. Right. that you'd have your copy, I'd have my copy, but the integrity of that would be verifiable. And then Fabric, on top of that, has a format for being able to write scripts that get deployed across the network that are written in Go, that run inside of a Docker VM. And so what this means is that we can encode, like you and I could encode a promise

that says, I promise to transfer to you some tokens in this network if the price of rice in China is this on this certain day or something like that, right? An options contract or an insurance contract. And the cool thing is because it's a contract that now is shared on the network, then it's verifiable. Everyone can see, oh, yeah, Brian made a promise that you would do this thing. And I can't pull it back.

So your risk that I wouldn't make good on that promise or pay up is gone. So this is called counterparty risk. And that's eliminated in a setting like this. And normally that's a big issue. Counterparty risk is why we have financial crises. Exactly. So this is a tool to try to fight that kind of thing. Cool.

So you got excited because of the land title situation. Obviously with contracts, I think about land titles all the time or property. But is Hyperledger specific to that? Or you can use it for any kind of contract, any sort of thing? Oh, it's very generalized. You could use it for all sorts of digital assets. So you could use it as a cryptocurrency system if your currency are things like frequent flyer miles, or loyalty points, or carbon emission credits, or other types of currency-like things.

What's distinctive about Ethereum and Bitcoin is that they are cryptocurrencies whose total volume is limited by the hash rate. It grows a little bit. Eventually, all of them will be mined. Right. It grows fast right away, and then slower, slower, slower, slower. Exactly. But that's just one way to do it. Ripple is a different approach where Ripple has

a fixed set of XRPs and I think they reserve the right to mint new ones. I'm not exactly sure, but that's a different way of just managing a currency. But you could create currencies using Fabric that are either completely independent, not tied to anything, or you could say it's tied to a promise to have a share

of a bar of gold sitting in a safe somewhere, right? Now, there's still the risk that that company doesn't make good on that promise, right? Or it could be your title to your home or a title to a big diamond. That latter example is something I think we're likely to see one project move in, which is using Hyperledger Fabric,

which is towards, it's all about the supply chain and diamond industry and trying to keep blood diamonds from entering circulation and the provenance of where these diamonds come from, right? So a titling system for knowing who owns what diamonds and who has the right to sell it. So in the future, if you have like a giant rock that you want to sell me, I can look up whether you actually own that in the ledger.

And if you don't, I probably shouldn't buy it from you because you probably got it from a warlord somewhere. Yeah, exactly. So that's very cool. You've got eight projects, seven in incubation, so it's still early days. You've got fabric, which is ready to roll. Exactly. So there's Sawtooth, which is a different approach to building distributed ledgers.

I can go into that if you're interested. Another one called Iroha, which came from some of our Japanese members, which is tightly coded C++ designed for small systems, mobile clients, that sort of thing. Actually, an interesting project there with the Bank of Cambodia, of all places. And then the fourth one is called Burrow. And this is actually an implementation of the Ethereum virtual machine, which is really exciting because there's a lot of good academic research as well as a lot of market validation going on around the Ethereum VM and a smart contract language called Solidity. Yeah.

So now being able to grasp that on top of the other DLTs is suddenly much more possible than it was before. We did a show on Ethereum maybe nine months ago and had a lot of excitement in our community around it. Smart contracts, of course, that was right. It was kind of interesting. It was right through the crisis. In fact, we scheduled a show before the crisis, then we had to reschedule, and then everything. It was one of these things, you know, technology moves fast.

Well, cryptocurrencies are the fastest moving things. Everything had changed by the time we actually did this show. So it was very interesting that way. But, yeah, smart contracts, very interesting. Making the actual scriptability or the writing of those contracts, bringing that down to where it's more, not acceptable, but what's the word? Accessible. Very close to acceptable.

More accessible to other developers is very cool. It sounds like these different projects, it's not like we have these eight and they mesh together and perform these sound like different things so those four are very different things they solve at 100,000 foot view the same problem it's kind of like comparing MySQL to Cassandra to CouchDB

exactly, now the other four just real quick, one is kind of a GUI tool for building these kinds of blockchain, what they're calling business networks, which a mix of all these things, but it's kind of like Eclipse, right? What Eclipse was for Java, and it's called Composer.

Another project is called Explorer. It's simply a way to navigate around your DLT and see who's published what, and it's like a debugging tool, but also a way to understand what's going on. Sometimes all you see are these hashes pointing to other hashes. It can be kind of confusing, right?

There's another tool called Indie, Hyperledger Indie, which is all about trying to build an identity layer that spans multiple distributed ledgers. It's trying to answer the question of how do we get to a world where we have identities on these different systems without creating a master database of one organization, government or private,

that knows everything about us, right? But how do we get back to sovereign identity control? And this is a recurring theme that has been a part of our industry for 15, 20 years. No one's ever really solved well. I think we have a key part of a solution to this. I'm really excited about that. So I wanted to ask you.

To the distributed identity problem. Yeah, you're actually keying into what I was going to ask for my final segment here, which is about the future and what would it look like, assuming that everything that you're up to with Hyperledger is successful and you build the things that you're trying to build and the community thrives and people accept and they all build around it. Assuming the success, what does the industry, digital assets, and all these things that we're involved in look like five years from now? I mean, is it utopia? Is everything perfect

now? It's not magic pixie dust. Some of us idealists have certainly seen our technologies used to create some really horrible things. So you have to be very cognizant of that. No, I think for businesses, it means that a lot of the processes that today are and rules about how markets work that are enforced by lawyers and enforced by people having to watch what goes on and can't automate what they do. There'll be a lot more automation. And because of more automation, there'll be more auditability. And it'll just be either impossible or much harder to commit certain kinds of fraud. Hopefully it means more stability in financial markets because I think the mortgage crisis had that industry used distributed ledgers and smart contracts to define these structures. We would have had much less of a run-up and also

a bit more of an orderly unwinding. I don't know if you watched, saw the movie, The Big Short. The chaos in there potentially could have been a lot less. Now that's an idealist speaking, right? And so we have to be careful about that. For consumers, it might not look that much different. This is still largely a backend story. I mean, but where I think it plays out is, so the idea of Bitcoin wallets, right? Really made real for people,

like the fact that this is actually a currency. Imagine that concept now applied to your health records or to your land title and car title and other important documents that today you print out, or if you probably don't even have a digital form, it might just be pieces of paper you keep in a filing cabinet, which you hope you remember to grab with you when a forest fire is coming for your house. Well, I was going to say, just thinking me personally, we have our title in a safe, just in case the house burns down, It's like fire-resistant safe.

Right. So that should be digital. There should be a way to take that with you. And for things like medical records, people don't even feel connected to their medical records, right? No. So my hope is that in all these important parts of life, as they get more digital, they actually get more sovereign. They get more about things that you know are portable, that you don't have to depend upon a particular company or another to hold for you and look out for you. And by that, it gives you more power as a consumer, more control, more flexibility.

Mad Fientist Awesome. Brian, final thoughts, anything you want to say? Brian Fitzpatrick No, it's really fun to be back at an open source conference. I've been coming to this for 20 years, 22 years, I think we realized the other day, 21 years. Wow. And to be amongst my peeps, I guess, and to see the interest and energy here. People still want to understand what it's about, and so that's why I'm here. Yeah. But more and more people are going, yeah, we really want to get our hands dirty with this.

Cool. Well, I'll let you get back to your peeps. Thanks so much for sitting down and having a talk with me. Thanks. This episode is brought to you by GoCD. GoCD is an open source continuous delivery server built by ThoughtWorks. It provides continuous delivery out of the box with its built-in pipelines, advanced traceability, and value stream visualization. With GoCD, you can easily model, orchestrate, and visualize complex workflows from end to end. It supports modern infrastructure with Elastic On-Demand Agents and cloud deployments, and their plugin ecosystem ensures GoCD

will work well in your unique environment. To learn more about GoCD, visit gocd.org slash changelog. It's open source and free to use, and there's also professional support and enterprise add-ons available from ThoughtWorks. Once again, go cd.org slash changelog. So I'm joined by Tag Griffith, who is the head of developer advocacy for Redis Labs. So cool that there's a Redis Lab, like there's businesses around Redis, right?

Yeah. Or maybe just one. I don't know if there's more than one, but just the fact that Redis is a thing that still exists in the world so much and is so vibrant that there's like a cottage industry around it. To me, it's just cool. Yeah, well, I would say it's even more than just a cottage industry. I mean, I got this job. I come from a Redis developer background.

Okay. So I was doing large, like 25 billion operation a day systems using Redis. Wow. And so I'd been talking a lot about the work that I had been doing with Redis, and it was really kind of a natural transition. And there really is a huge ecosystem around Redis. And then we have Redis Labs who provide,

you know, the standard enterprise version of an open source product. And we sponsor the open source product. But I wouldn't even call it a cottage industry. Like I said, I mean... It's bigger than cottage. Yeah. We were just at a Docker con a couple of weeks ago right here in Austin.

And we're actually the number one downloaded database container on the Docker hub now. Really? Is Redis. Is Redis. Yeah. So, so many uses, so many, it fits into so many use cases. You know, if you have a, if your favorite relational database of choice, everybody can just throw Redis into the many use cases, if you have your favorite relational database of

choice, everybody can just throw Redis into the mix and have some immediate win. So it makes sense that some people are pro this, pro that, but everybody loves Redis. I think one of the things is it's very easy for developers to approach Redis. You've got all these data structures

that are very similar to what you work with in Python or Java, whatever language you're working with. And it's very easy to do incremental projects with Redis. So you start maybe just as a hobbyist, and then you drop it in as a cache and your first actual real production system. Then you start finding other data structures and uses of Redis where it can really either accelerate performance, store data that doesn't fit into SQL databases really well. And then even you start looking at more of the commercial products to hit on the high availability, the scaling, and some of that. And you've got this whole range of systems and uses you can apply this one product to.

Refresh me on the history of Redis a little bit. So you had, you know, and I always call them anti-res because I just say it in my head. Maybe it's Antiriz. I don't know. Salvatore Sanfilippo started this years ago. He actually became sponsored by, I think he got hired by VMware or something for a while.

I don't know. It was kind of like one of these, the frontier of like paying people to do an open source thing and how that was going to work out. And I think we've even had shows with him throughout that part. But that was years ago. I haven't kept up with where Redis is, except for as a user. Tell us a little bit of the history, maybe the recent history of the Redis project. So I think it was around 2009 when Salvatore first released Redis.

And I always pronounce it anti-Rez. It actually comes from he and a friend were looking for handles, and it comes from Trent Reznor. Okay. I always kind of wondered where that came from. Yeah, me too. And it comes from Trent Reznor. Okay.

And so his friend was. He doesn't like Trent Reznor. No. He's anti-Rez. I don't know. It actually is slightly different. Okay. I think he likes Trent.

I'm trying to start a mythos here. His friend was going by the handle Reznor or Rez. And so he just sort of was the opposite. Gotcha. Okay, very good. That's how I've heard it. Yeah, so he started it. It kept growing and growing.

I sort of came to Redis. I used to work at Flickr and Yahoo and were very early adopters of Redis. Then all this community started building around it, so you had sort of the first Redis conferences. Then you had Salvatore VMware started sponsoring it. Then Pivotal was providing some support. And then eventually Redis Lab showed up, we started as a cloud provider of Redis

and then thought it would be strategic and useful for us to sponsor the open source project. And now we sponsor open source Redis. And Salvatore works with us very closely. We contribute back to the product. As well as we have evangelists like myself and Itamar and our employees who go out and support both the purely open source Redis and our commercialized version. Cool. So is he still able to work on it full time now through your guys' sponsorship? He does. And he spends most of his time coding. In fact, he resists really anything that kind of gets in the way of coding. Good for him.

I'm very jealous. I come from a developer background, so I love writing code. But he's definitely, you know, he works with us, but he ultimately owns the agenda. So it's very much, it is his open source project. We support him. We support the project. Yeah. So how does that feel from a business perspective and kind of being in the hands of somebody else? I mean, I guess that's

how you got to where you are now because he's had good hands. He's proven he's had good hands. But is there any like lack of control can sometimes scare businesses? But do you guys have any of that? Sure. I don't think it's a problem. I think it works really well. I think kind of the bigger things are trying to be sensitive around community is that we need to do the right thing by the open source community as well as our users. So balancing in that out, being fair to the open source. I always feel like it's very important for us that we support both purely the open source as well as our commercial users. I think it involves different ways of working. If you look at a lot of corporations, they like to have plans. And really what we have to do is we have to take his technical direction,

and he sets his direction. We work with him to contribute. And that means sometimes he may not like our contributions. Other times he does. And so we work in the same framework as every other contributor to the open source Redis. Excuse me if I'm a little distracted. There's a lot of different people that come to OSCON, but there's an actual pigeon that's walking around behind you. So I'm trying to pay attention, but I'm also like, you know, when you see a squirrel and you're like, a squirrel.

But there's like a literal pigeon that just keeps flying around. Yeah, the pigeon dropped by our booth earlier to learn about Redis. So even the pigeons love Redis. It's like everybody's coming to OzCon. Yeah. So you're here, you're at an event like this, you guys have a booth. What are the kind of things, you know, you're talking about Redis evangelists. What are the kind of things that you guys say that would get somebody who's like,

what's Redis or why would I use Redis? What's the most compelling bit about it? Sure. I think one of the big things is the flexibility. It's very easy to approach. It has a lot of data structures that you can use to solve various engineering problems. Like we talked about earlier, the sort of incrementalness. So it fits into kind of these agile models.

And also you can sort of start using Redis alongside other systems to improve your performance, to store data in a way that doesn't work well in SQL relational databases. And then you start really iteratively building on your success with earlier projects to build more and more Redis. That's how I got started when I was a developer.

I started with the basic key value paradigm using Redis as a key value source, and then started moving into storing other pre-computations, persisting data, all of that. And it's very natural. You can just build and build and build on that experience. I mean, Redis is fast. Everything is in memory. Just in any sort of well-structured Redis is fast everything is in memory just in any sort of well-structured Redis database you're almost always gonna have all your your operations served in under a millisecond which is extremely fast you generally there's very few systems that can offer you that level of low latency so there's I've never used Redis as a primary store or as a persistent store where this is like the one place my data lives.

It's always like a cache or something where I can run some stuff that's ephemeral. I can recalculate it if I have to, right, in the case of data loss. Surely there are people that are using Redis as their one and only or, you know, in persistence. Tell us that story. How does it work in that sense? Sure. I actually think that's a very common way people come to Redis. Yeah.

Everyone knows Redis as sort of the key value store. A flexible cache that you can use in unique ways. Everyone loves Redis as a cache. That's actually almost kind of one of the biggest challenges of my job is that everyone knows Redis as the cache. They put it in the corner. And I'm kind of like, hey, there's all these other things you can do with Redis. And it takes a lot of people a while to get to the notion or the idea

that you can use Redis persistence. There's multiple modes of persistence. There's sort of the snapshot version of persistence. There's also the changelog version of persistence. So you can tune, and this is one of the nice things with Redis, is that you can tune those trade-offs between performance, persistence, durability guarantees, your sort of window of loss. So you can find kind of the sweet spot that works the best for you as a developer. Nice. But I do, I know a lot of

people who are like, oh yeah, Redis is great. It's this cache. It's this great cache. And it's a big kind of leap of faith to go from cache to, oh, but I can use this as a persistent store. It feels, because it's so easy and lightweight and flexible, it feels disposable. It feels ephemeral. Like you always think it's like an add on. Yeah.

But reality is very capable in other ways. Very, very capable. In fact, I gave a birds of a feather at OSCON to really talk to people about getting beyond that notion of just as a key value. And I think so many people come to Redis as a key value, as a cash. Everyone knows it as that.

They often never get beyond that and look at all the other powerful things you can do with Redis yeah one of the things that Salvatore added probably a couple years ago now which is really cool is the scripting stuff through Lua uh-huh lots of you know improvements to Redis over the years what what's the future look like you're obviously involved and where it's headed you know push us out a year or two like what's the new stuff that's gonna to be coming down the pipeline? So I'll make a little bit of a plug.

There's going to be some new stuff announced at RedisConf, which we run in San Francisco. Okay, when's that? It'll be at the end of the month. So we'll be revealing more of the, well, really, Salvatore will be revealing more of the roadmap and some cool changes that are coming in. I think the big one for this show, which was announced last year,

or this conference that was announced last year, is modules. That's really the new, really, really powerful thing we're working on. And so this has been a collaboration between Salvatore and Redis Labs. We have several of our engineers contributing both modules and support for the modules code. And what modules are going to give Redis developers is now you have a way of extending and modifying and changing Redis without having to convince Salvatore that this is the thing that Redis needs.

I mean, it's one of the great parts of open source is that you can, it's now a modular extension to the Redis needs. Right. I mean, it's one of the great parts of open source is that you can, it's now a modular extension to the Redis without needing to go through the change, you know, change your class. Right, right. And generalization. It's actually a great proving ground, too. Yeah.

For things that could go into core. I mean, once you have, like, a plugin or module model, you don't have to convince them. But if you build something that's, you know, super useful and all of a sudden everybody loves it and it becomes huge and it's like, actually, you know, that makes sense to be part of Redis. And so it's cool that way that you can have it proving that. And it's actually one of the maybe interesting things is it's also going backwards from that. So Salvatore did a fork to provide basically a message broker kind of system based on Redis. And now he's realized that he can take that and turn it into a module.

And instead of, so his fork is essentially collapsing. Was that D message or was that? Disque. Disque, okay. Yep, I remember that. And so he's realized that he could take that and provide it as a module. So he's hoping to find time to work on that. I see.

But it's sort of interesting. It's almost the de-evolution of a fork into something else. It's like he forked it, but now he modified the main thing so that he wouldn't have had to fork it. He could provide that as a module now. Very cool. And, you know, for Redis Labs, I mean, there's a lot of advantages is that we can provide things for our customers

in a way that doesn't necessarily impact open source in a little bit of a faster way, faster turnaround. We're trying to do a lot of our modules open source. We're also trying to put together a module registry so that you have a good, authoritative place to go look for new modules. We have on our GitHub, you can go there and you can see a lot of the modules

that our engineers have written. Right now we've written one to make JSON a native data type within Redis, re-JSON. There's also a secondary indexing and search module. And then also some things around machine learning. So you can use Redis to sort of parameterize machine learning models. Well, thanks for the fun. Thanks so much for talking to us. Great. Thanks for having me. All right. Thank you for tuning in to this episode of The Change Log.

If you enjoyed the show, share it with a friend. Rate us on Apple Podcasts. Go on Overcast and favorite it. Share it with a friend. If you enjoyed the show, that's what you got to do. Share it with a friend. Thank you to our sponsors, Auth0, Digital Ocean, TopTout, and GoCD. Also thanks to Fastly, our bandwidth partner.

Head to Fastly.com to learn more. And we host everything we do on Linode cloud servers. Head to Linode.com slash changelog. Check them out, support the show. And the changelog is hosted by myself, Adam Stachowiak, and Jared Santo. Editing is by Jonathan Youngblood. And the awesome music you've been hearing

is produced by Breakmaster Cylinder. You can find more episodes just like this at changelog.com or by subscribing wherever you get your podcasts. Thanks for listening. Thank you. Bye.