 Well, I'm John Furrier, the founder of SiliconANGLE Media and co-host of theCUBE. We're here in Palo Alto at our studio here. I'm joined with Joseph Jackson, the founder and general partner of OSS Capital, open source software capital OSS stands for. He's also the founder of KubeCon, which now is part of the CNCF. And it's a huge conference around Kubernetes. He's a cloud guy, knows open source, very well respected in the industry and also a great guest friend of theCUBE, a Kube alumni. Joseph, great to see you. Also known as JJ. JJ, good to see you. Thank you for having me on again, John. Hey, great to have you come on. I know we've talked many times on theCUBE, but you've got some exciting news. You've got a new firm, OSS Capital, open source software, not operational support like a telco, but this is an investment opportunity where you're making investments. Congratulations. Thank you. So I know you can't talk about some of the specifics on this fund size, but you are actually going to go out, talk to entrepreneurs, make some equity investments. Yep. Around open source software, what's the thesis? How did you get here? Why did you do it? What's motivating you? What's the thesis? A lot of questions in there. I mean, this is a really profoundly huge year for open source software on a bunch of different levels. I think the biggest kind of thing everyone anchors towards is GitHub being acquired by Microsoft, just a couple of weeks ago, we had the two huge Hadoop vendors joined forces that I think surprised a lot of people. MuleSoft, which is a big open source, middleware company getting acquired by Salesforce, just a year after going public, which is a huge outcome. I think one observation, just to sort of like summarize the year, 2018 is actually starting in January, almost on sort of like a monthly basis. We've observed a major sort of open source software company outcome. And sort of kicking off the year, we had CoreOS getting acquired by Red Hat, Brandon and Alex, the founders over there, built a really interesting company in the Kubernetes ecosystem. And I think in February, Al Fresco, which is an open source content portal, taking privatization sort of outcome from a private equity firm, I believe in March we had Magento getting acquired by Adobe, which is an open source based CMS, PHP CMS. So just a lot of activity for significant outcomes, a multi-billion dollar outcomes of commercial open source companies. So an open source software is something like 20 years old, 20 years in the making. And this year in particular, we've just seen just a huge amount of large scale outcomes that have been many years in the making from companies that have taken lots of venture funding. And in a lot of cases, sort of partially focused funding from different investors that have an affinity for open source software and sort of understand the uniqueness of the open source model when it's applied to business, when it's applied to company building, but more sort of opportunistic and sort of affinity oriented as opposed to a pure focus. And so that's kind of been part of the motivation. I'd say the more authentically kind of compelling motivation for doing this is that it just needs to exist. This is sort of a model that is happening by necessity. We're seeing more and more software companies be open source software companies. So open source first, they're built in a distributed way. They're leveraging engineers and talent around the world, which is part of this open source kind of philosophy. And they are fundamentally kind of commercial open source software companies. And so we felt that if you had a firm, basically designed in a way to exclusively focus on those kinds of companies and where the firm was actually backed and supported by the founders of the largest commercial open source companies in the world before sort of the last decade, that could actually deliver a lot of value. So we've been sort of blogging a little bit about this. You wrote a great post I read about open source monetization. But I think one of the things I'm seeing as well that supports your thesis and I'd like to get your reaction to it because I think this is something that's not really talked about, but open source is still young. I mean, you go back, I remember the days when we used to have to hide in the shadows to get licenses and pirate stuff and do all those crazy stuff. But now it's only a couple of decades away. The leaders that we're investing were usually entrepreneurs that have been successful. The Rob Bearden's, the Amur Awadallah's, the guys who did Spring, all these different open source Linux, obviously a great success story. But there hasn't been any institutional, yeah, you got Benchmark, other things, done some investments. A discipline around open source where open source is now table stakes in all software development. Cloud is scaling, scaling out globally. There's no real, there hasn't been a firm that's been focused on just open source from a commercial while maintaining the purity and ethos of open source. I mean, is that, you agree? 100%. Yeah, that's been the big part of creating the firm is aligning and solving for a pure focused structure. And I think what I'll say abstractly is the sort of venture, venture capital venture style approach to funding enterprise, technology companies, software companies in general has been to kind of find great entrepreneurs in an abstract way that can build great technology companies, can bring them to market, can sell them and can scale them and so on and either create categories or dominate existing categories and disrupt incumbents and so on. And I think while that has worked for quite awhile in the venture industry overall and in the 50, 60 years of the venture industry lots of successful firms, I think what we're starting to see is a necessary shift toward accounting for the fundamental differences of open source software as it relates to new technology getting created and new software companies kind of coming into market. And so we actually fundamentally believe that commercial open source software companies are fundamentally different functionally in almost every way as compared to proprietary closed source software companies of the last 30 years. And the way we've sort of designed our firm and we'll be about 10 people pretty soon, we're just about a month in. So we're growing the team quickly but sort of a small focused team is- A 10 is not focused small. I mean, I know venture firms two billion in management that don't have like more than 20 people. Well, we have portfolio partners that are focused in different functional areas where commercial open source software companies have really fundamental differences. If you were to sort of stack rank by function where commercial open source software companies are really fundamentally different sort of top to bottom, legal would be probably the very top of the list. Right in terms of license, compliance, management, structuring all of the sort of protections and provisions around how intellectual property is actually shipped to and sold to customers. The legal licensing aspects of commercial software licensing. This is quite a polarizing hot topic these days. The second big functional area where we have a portfolio partner focused on this is finance. Finance is another area where commercial open source software companies have to sort of behaviorally orient and apply that function very, very differently as compared to proprietary software companies. And so we're crazy honored and excited to have world experts and very respected leaders in those different areas sort of helping to provide sort of different like sort of pillars of wisdom to our portfolio companies, our portfolio founders in those different functional areas and we provide a really focused kind of structure. Well, I want to ask you that kind of the question that kind of bridges the old way and new way because I'd definitely see you guys definitely being new and different, which is good. Or as Andy Jackson would say, you can be misunderstood for a while but as you become successful, people will start understanding what you do. And that's a great example of Amazon. So but the pattern with success is traditionally the same. You kind of encapsulate the difference between open source old and new and that is you have something of value and you're disrupting the market and collecting rents from it, okay? Or revenue or profit. So that's commercial, that's how businesses run. How are you guys going to disrupt with open source software the next generation value creation? We know how values created certainly in software that open source has shown a path on how to create value in writing software if code is value and functionality is value but to commercialize and create revenue which is people paying some for something. That's a little bit different kind of value extraction from value, the value creation. So open source software can create value in functionality and value product. Now you bring it to the market, you get paid for it. That's, you have to disrupt somebody you have to create something. How are you looking at that? What's the vision of the creation, the extraction of value, who's disrupted? Is it green field, new opportunities? How do you think, what's your vision? A lot of nuance and complexity in that question. What I would say is Well, open source is creating products. Well, open source is the basis for creating products in a different kind of way. I'll go back to your question around let's just sort of maybe simplify it as the value creation and the value capture dynamics. We've sort of written a few posts about this and it's subtle but it's easy to understand if you look at it from a fundamental kind of perspective. We actually believe and we'll be publishing research on this and maybe even some sort of more principal scientific perhaps even ways of looking at it and then blog posts and research. But we believe that open source software will always generate or create orders of magnitude more value than any constituent can capture. And that's a fundamental way of looking at it. So if you see how cloud providers are capturing value that open source creates, whether it's elastic search or Postgres or MySQL or Hadoop and then commercial open source software companies that capture value that open source software creates whether it's companies like Confluent around Kafka or Cloudera around Hadoop or Databricks around Apache Spark or whether it's the creators of those projects the creators of Spark and Hadoop and elastic search. Sometimes many of them are the founders of those companies I mentioned some of them sometimes they're not. We just believe regardless of how that sort of value is captured by the cloud providers the commercial vendors or the creators the value created relative to the value captured will always be orders and orders of magnitude greater. And this is expressed in another way which is maybe easier to understand so sort of reinforcing this kind of assertion that there's orders of magnitude value created far greater than what can be captured. If you were to do a survey which we're currently in the process of doing and I'm happy to sort of say that publicly for the first time here of all the commercial open source software companies that have projects with large significant adoption whether say for example it's Docker with millions of users or Apache Hadoop how many Hadoop deployments there are how many customers companies are there running Hadoop deployments or it may be even MySQL how many MySQL installations are there. And then you were to sort of survey those companies and see how many end users are there relative to how many customers are paying for the usage of the projects. It would probably be something like if there were a million users of a given project the company behind that project or the cloud provider or say the end user the developer behind the project is unlikely to capture more than say 1% or a couple of percent of those end users to companies to paying companies to paying customers. And many times that's high. Many times one to 2% is very high. Often what we've seen actually anecdotally and we're doing principled research around this and we'll have data here across a large number of companies many times it's a fraction of 1% which is sort of maybe sometimes 10% of 1% or even a smaller. So the practitioners will be making more money than the actual vendors. Absolutely right. And users and practitioners always stand to benefit far greater because of the fundamental nature of open source. It's permissionless, it's disaggregated the value creation dynamics are untethered and it is fundamentally freely available to use freely available to contribute to with different constraints based on the license. However, all those things are sort of like disaggregating the creation of technology into sort of like an unbounded network. And that's really a really incredible. So first of all, I agree after your premise 100%. We've seen it with theCUBE or videos are free. And that's a good thing. All those things are good. And Dave Vellante says this all the time on theCUBE and we actually pointed this out and called this in the Hadoopika system in 2012. In fact, we actually said that on theCUBE and turned out to be true because look at Hortonworks and CloudDare had to merge because again the market changed very quickly because value was created around them and you had cloud, et cetera. So the question is that changes the valuation mechanisms. So if this is true, which we believe it is to say it is, then the traditional net present value cash flow metric of the value of the firm, not your firm, but like if I'm an open source firm if I'm only one's portion of the extraction I'm a supplier and I'm an enabler the valuation on cash flow might not be as great as the real impact. So the question I have for you have you thought about the valuation because now you're thinking about a bigger construct, community, network effects. These are new dynamics. I don't think anyone's actually crunched the valuation model around this. So if someone knew that say for example an open source project created all this value and they weren't necessarily harvesting it from a cash flow perspective there might be other ways to monetize it. Have you thought about that and what's your reaction to that concept of okay, because capitalism would kind of shake out the system because why would someone be motivated to participate if they're not capturing any value? So if the value shifts are they still going to be able to participate? You follow the logic I'm trying to give you? I definitely do. I think what I would say to that is we expect and we encourage and we will absolutely heavily invest in more business model innovation in the area of open source. So what I mean by that is and it's important to sort of like qualify a few things there. There's a huge amount of polarization and lack of consensus, lack of industry consensus on what it actually means to have or implement an open source based business model. In fact, there's a lot of people who just sort of point blankedly assert that an open source business model does not exist. We believe that many business models for monetizing and commercializing open source exist. We've blogged and written about a few of them. There's services and training and support. There's open core, which is very effective in sort of a spectrum of ways to implement open core. You can have a, you know, around the core, you can have a thin crust or a thick crust. There's SaaS. There are hardware based distribution models, things like source fire and cumulus networks. And there are also network based approaches. So for example, project called storage or storage A being developed and run now by Ben Golub, who's the former CEO of Docker. Cube alumni. Ben's really great, open source veteran. You know, this is a network, kind of decentralized network based approach of, you know, sort of right sizing the production and the consumption of the resource of a storage based open source project in a decentralized network. So those are sort of four, I guess four or five ways of commercializing value. However, however, four or five ways of commercializing value. However, what we believe is that there will be more business model innovation. There will be more developments around how you can better capture more or in different ways the value that open source creates. However, what I'll say though is it is unrealistic to expect two things. It's unrealistic and in fact unfair to expect that any of those constituents will contribute back to open source, proportional to the value that they receive from it or the benefit. And I'm actually paraphrasing Doug Cutting there who tweeted this a couple of years ago, very profoundly deep wise tweet which I very strongly agree with. And it is also unrealistic to expect a second thing which is that any of those constituents can capture a material portion of the value that open source creates which I would assert is many trillions of dollars perhaps tens of trillions of dollars. It's really hard to quantify that. And it's not just dollars in economics sense, it's dollars in productivity, time saved, new markets, new areas and so on. Yeah, I think this is interesting and I think that will be an open book. But I will say that what I've observed in looking through all these CUBE interviews, I think business model innovation absolutely is something that is an IP. We need it. Well, it's now an actual property. The business model isn't, hey, aren't the business school learn this at Babson or Harvard. I learned this business model. We're going to do SaaS, freemium. Okay, I get that. There's going to be very interesting new innovations coming and I think that's the new IP. Because open source, if it's community-based, there's going to be formulas. So that's going to be a wrench. So now let's get back to the actual funding itself. You've got to do an early stage. Can you take us through the approach? We're very focused on early stage. Investing and backing teams that are just sort of welcoming the idea of a commercial entity around their open source project or building a business fundamentally dependent on an open source project or maybe even more than one. And the reason for that is this is really where there's a lot of structural inefficiency in supporting and backing those types of founders. And I think one of the things with Coral West, I know is with that acquisition, they were pure on the open source side, doing a great job. Didn't want to push the business model too hard because the open source, let's face it, you've got people that want to get caught on the business side and get revenue, perverse incentives might come up or fear of incentives that might be different or not aligned. It was a great value for Red Hat. So Red Hat got a steal on that one. But as you go forward, there's going to be certainly a lot more stuff. You're seeing a lot of it now in the CNCF, for instance. I want to get your thoughts on this because being the co-founder of KubeCon and donating it to the CNCF, Kubernetes is the hottest thing on the planet as we talked about many years ago. What's your take on that now? It's the exciting things happening. What is the impact of Kubernetes in your opinion to the world and where do you see that evolving rapidly and where is the focus here that people should be paying attention to? I think that Kubernetes replaces EC2. Kubernetes is a disaggregated API for distributed computing anywhere. And it happens to be portable and able to run on any kind of compute infrastructure, which sort of makes it like a liquid disaggregated EC2 like API, which a lot of people have been sort of chasing and trying to implement for many years with things like OpenStack or Eucalyptus. But interestingly, Kubernetes is sort of the right abstraction for distributed computing because it meets people where they are architecturally. It's sort of aligned with this current movement around distributed systems, first designs, microservices, packaging things in small compartmentalized components. Good for integrating existing stuff. Absolutely, and it's very composable, unopinionated architecturally. So you can sort of take an application and structure it in any given way. And as long as it has the sort of isolation boundary of a container, you can run it on Kubernetes without needing to sort of retrofit the architecture, which is really awesome. I think Kubernetes is a foundational part of the next kind of computing paradigm in the same way that Linux was foundational to the computing paradigm that gave rise to the internet. You know, we had commodity hardware, meeting open source-based sort of cost reduction and efficiency, which really Linux enabled and the movement toward scale out data center infrastructure that supported the internet's sort of maturity and infrastructure. I think we're starting to see the same type of repeat effect thanks to Kubernetes basically being really well received by engineers, by the cloud providers. It's now the universal sort of standard for running container-based applications on the different cloud providers. And I think having the non-technical opinion posture, as you said, or architecture posture, allows it to be compatible with a new kind of heterogeneous. Heterogeneity is critical. Heterogeneity is key because not just within the environment, it's also within each vendor or customer has more heterogeneity. So okay, now that's key. So multi-cloud, I want to get your thoughts on multi-cloud because now this goes into some of the things that might build on top of if Kubernetes continues to go down the road you say it does. Then the next question is stateful applications, service meshes. A lot of buzzwords, a lot of buzzwords in there. I mean, I think like, so stateful applications real because at a certain point in time, you have a maturity curve with critical infrastructure that starts to become appealing for stateful mission-critical storage systems, which is typically where you have all the crown jewels of a given company's infrastructure, whether it's a transactional system or reading and writing core customer, financial service information, whatever it is. So Kubernetes starting to hit this maturity curve where people are migrating really serious mission-critical storage workloads onto that platform. And obviously, we're going to start to see even more critical workloads. We're starting to see edge workloads because Kubernetes is a pretty low footprint system that you can run on edge devices, even run on microcontrollers. We're sort of past the experimental, fun and games with Raspberry Pi sort of towers and people actually legitimately doing real-world edge kind of deployments with Kubernetes. We're absolutely starting to see multi-geo, multi-replication, multi-cloud sort of style architectures becoming real as well because Kubernetes is this API that the industry is agreeing upon sufficiently. We actually have agreement around this sort of surface area for distributed system and style computing that if cloud providers can actually standardize on it in a way that lets application-specific vendors or new types of application deployment models innovate further, then we can really unlock this sort of tight coupling of proprietary services inside cloud providers and disaggregate it, which is really exciting. I forget the Netscape, Jim Barksdale, bundling, unbundling. We're starting to see the unbundling of proprietary cloud computing service APIs, things like Kinesis and ALB and ELB and proprietary storage services and these other sticky services get unbundled because of two big things. Open source, obviously, we have open source alternative data paths and then we have Kubernetes, which allows us to sort of disaggregate things out pretty easily. I want to get your thoughts, one final concept before we break because I was having a private conversation with three people besides myself, a big time CIO of a company that if I said the name, everyone would go, oh my God, that guy's huge. He's seen it all going back many, many ways. Currently, he's done a lot of innovation. A hardcore network chip guy who knows networking, old school infrastructure and then a cloud native application founder who knows a lot about software development and is state of the art cloud native. So cloud native, all experience, old school, kind of about my age. A cloud native app developer, it's big time CIO and a chip networking kind of infrastructure guy. And we're talking, and one thing that came out, I want to get your thoughts on this. He says, so what's going on with DevOps? How do you see this service mesh as a state for web, top of the stack, no stacks, horizontally scalable and the comment that came out was, storage and networking have had this relationship with everything since day one. Network moves a packet from point A to point B and nothing happens in between, maybe some inspection and storage goes from here, now to then, as you store it. He goes, that premise moves up the stack, so then the cloud native guy goes, well that's what's happening up at the top. There's a lot of moving things around workloads and or services, provisioning services and then from here, from now to then, state, real time. And what dawned on the next conversation the CIO goes, well this is exactly our challenge. We have under the hood, infrastructure being programmable. My phone's calling me, programmable phone. So you got the programmable at the top of the stack too. So the CIO said, that's exactly the problem we're trying to solve, we're trying to solve some of these network storage concepts now at an application level, your thoughts to that. Well I think if I could tease apart everything you just said, which is a profound synthesis of a lot of different things, I think we've started to see application logic leak out of application code itself into dedicated layers that are really good at doing one specific thing. And so traditionally we had some crud style kind of behavioral semantics implemented around business logic. And then inside of that you also had libraries for doing connectivity and lookups and service discovery and locking and key management and encryption and coordination with other types of applications and all that stuff was sort of like shoved into the single big application binary. And now we're starting to see all of those like language, runtime specific parts of application code sort of crack or leak out into these dedicated, highly scalable, UNIX philosophy oriented sort of like layers. So things like Envoy are really just built for the sort of nervous system layer of kind of application communication fabric up and down the layer two through layer seven sort of protocol transport stack which is really profound. We're seeing things like Vault from Hashicorp handle secure key storage persistence of application authentication, authorization, metadata and information to sort of access different systems and endpoints. And that's a dedicated sort of stateful layer that you can sort of fragment out and delegate sort of application specific functionality to which is really great for scalability reasons and on and on and on. And so we've started to see that and I think one way of looking at that is it's a cycle. It's the sort of bundling and unbundling aspect. The sort of micro services. Or the brand new level of services are getting really low level. Yeah, it's a sort of like bundling, unbundling and so we've got all this unbundling happening out of application code to these dedicated layers. The bundling back may happen. I've actually seen a few sort of Bay Area companies go like we're going back to the monolith because it actually gives us lots of efficiencies and things that we thought were trade-offs before. We're actually comfortable with a big monorepo and one or two core languages and we're going to like build everything into these big binaries and everyone's going to sort of live in the same source code repository and break things out through folders or whatever. I mean there's a lot of really interesting things. I don't want to say we're sort of clear on where this bundling, unbundling is happening but I do think that there's a lot of unbundling happening right now. And there's a lot of opportunity there. And the open source obviously driving it. So final question for you. How many deals have you done? Can you talk a little bit about the firm? And exciting things and plans that you have going forward. Yeah, we're going to be making a lot of announcements over the next few months. And we're, I guess, extremely thrilled. I don't want to say overwhelmed because we're able to handle all of the volume and inquiries and inbound interest. We're really, really honored and thrilled by the reception over the last couple of weeks from announcing the firm on the 1st of October sort of before the Hortonworks Clutter Merger the JFrog funding announcement that week, the Elastic IPO. Just a lot of really awesome things happened that week. It was obviously before Microsoft open sourced all their patents. We'll be announcing more investments that we've made. We announced our first one on the 1st of October as well with the announcement of the firm. We've made a good number of investments. We're not able to talk too much about our first sort of initiative but you'll hear more about that in the near future. Well, we're excited. I think it's the timing's perfect. I know you've been working on this kind of vision for a while and I think it's a really great timing. Congratulations, JJ. Thank you so much. Thanks so much for having me on. Joseph Jackson, also known as JJ, founder and general partner of OSS Capital, open source software capital. Co-founder of KubeCon, which is now part of the CNCF. Real great player in the community and the ecosystem. Great to have him on theCUBE. Thanks for coming in. I'm John Furrier. Thanks for watching. Thanks, John.