 Hi, this is Shemal Dyer here with another episode of Supervisor TV. This is our first episode of 2016, and so we thought it's best to kind of take this moment to reflect on the past year, as well as discuss what might be in store this year. With me, I have Boris Rensky from Morantis and Randy Baez from EMC. Boris and Randy, would you please introduce yourselves? Yes. My name is Boris Rensky, and I am a Chief Marketing Officer at Morantis. I'm Randy Baez. I was formerly the CEO of Cloud Scaling, and now I'm Vice President of Technology at EMC, where I advocate around open source, particularly open stack. Great. Well, thank you to both of you for making time to join us today. So let's get this thing going. So what I'd like to ask you first is, let's recap a bit. What were the most significant developments in open stack in the past 12 months? And what do you think drove those changes? Some of the things that just kind of come to my mind as important developments from last year, I think all of them are necessarily exactly the last 12 months in that ballpark. So the first thing is, I think, kind of a change in the open stack landscape that has happened. So we have seen quite a bit of consolidation. Some of this consolidation were successful exits, like, for instance, Randy's startup Cloud Scaling. Some of them were a little bit less graceful, like Nebula just posting on their website that they are no longer in business. But in my view, it's probably the most important thing that happened, not simply because there's companies going away and new ones coming in, but because it marks a new era in my view in open stack, where we're moving from kind of the open stack 1.0 period where it was a lot of it was driven by kind of hype and excitement and a lot of guys screaming, a lot of things that weren't necessarily relevant or on point to now the new era, where the landscape is kind of much cleaner, the people that are seriously considering using open stack are kind of more serious in terms of their intentions and the players that remain on the market are better funded, bigger and tougher and more serious. So all in all, that's kind of a good thing, in my opinion. Another thing that jumps to mind is, I think, fairly credible set of motions in the open stack community to embrace containers and developers. I think that we're kind of a long way from being where we need to be. But if you were to just look at open stack two years ago, we didn't think about that at all. Last year, Magnum got quite a bit of attention. Kola is getting off the ground and there's a lot of conversation at the open stack summit about the shift in power from the CIS admins to developers. So that's definitely an important and positive development in my view. And finally, another thing that's kind of like a byproduct of the first thing I would say is that I'm seeing analysts kind of starting to shift their tune about open stack, such as for instance Gartner, from all its piece of crap that will never work and it doesn't work and it has all these problems, to kind of being a little bit more neutral and acknowledging the open stack wins in potential in a little bit more of an optimistic fashion, so to speak. So these are some of the things that kind of just kind of jump into my mind. Randy, what do you think? I mean, I saw a lot of the same kinds of things, but I guess my viewpoint and interpretation of them is probably a little bit more customistic. You know, the consolidation is good. I do think we move into a new era that things are more mature in terms of the vendors that are around table in many ways. But I also think that, you know, from my perspective, you know, a lot of the core components of open stack have not gained velocity. They've just continued to slow down. Things like development around Nova and a lot of the other kind of core components really have gotten very mired in the process for like making changes gotten very mired. So that's concerning. And you know, I think that, you know, there was this window that open stack can be successful in and a part of me wonders whether it's starting to close and whether kind of for third platform type applications, whether people might potentially go in Kubernetes and Mesa and sort of simpler type systems because if you look at, you know, why, you know, Docker, like you look at the last user survey from open stack and, you know, it's very telling that people compare Docker along with Puppet and Chef and Ansible and, you know, it's got this huge uptake. People are trying to use containers really for a configuration management or application management type tool. And, you know, it's a little stunning, right? Because it's not the same type of tool, right? It's not an infrastructure as a code type system. It's a little bit different. And the reason is it's just more developer friendly. And open stack itself is not very developer friendly. API has changed a lot. You know, there's so many configuration options that a lot of the different open stack deployments wind up being snowflakes. And that sort of kind of leads me to the last thing there, which is, you know, I've got some pretty great concerns about open stack's ability to be adopted by the mainstream enterprise. Yes, Walmart can like, you know, hire a lot of people and throw it at them, throw it at the problem, ATT, so on. A lot of big announcements from people around, you know, kind of their open stack. But these are fairly large businesses with deep pockets who can order to invest into, you know, sort of like open stack plans. I'm a little concerned that while the maturity of the vendors has increased, I'm not sure that the maturity and maturity of the market, I'm not sure the maturity of open stack has increased at the level I would like. And the focus isn't where I'd like it to be, which is just on the core infrastructure as service components, plus maybe containers, but all these other pieces seem highly distracting. And so, you know, it's a lot of the same stuff that Boris is saying, but I maybe have kind of a more pessimistic interpretation of it. I would like to see, you know, some nice, clean, like, I'd like to love to see Maranthus just take off and have, you know, Maranthus open stack kind of get like a lot of market dominance and have most of the Maranthus open stacks be like pretty templatized, and from one Maranthus open stack point to another, it looks the same. Like, that would make me really happy. And I think it would show me that open stack has gotten to where, you know, mainstream enterprises can deploy it and manage it easily, and, you know, there's some hope for actually getting to the hybrid cloud future. Got it. That makes sense. So, you know, both of you mentioned adoption and changes, if you will, from sysadmin to developer and developer focus. When you're out there talking to customers in your day jobs, are you seeing any differences in how open stack is being applied to business use cases? Are we seeing a shift in use cases, the perception, the approach, if you will, when you went out and about and talked to people in 2015 that are seriously considering the platform? No, not really. So, I don't think that there is any major change in use cases. I don't think that the reasons for adopting open stack has changed last year or the years before. We can talk about the reasons separately later. But what has changed is, I think, that the set of expectations that people come into open stack with and the type of people that are actually, you know, pursuing doing open stack. So, some of that is related to what Randy is saying. So, in my view, what we're seeing now is that, you know, after five years of kind of, you know, extreme hype and, you know, all of the early movers that are, you know, very easily moved by hype having kind of, you know, tried and some places succeeded and some places failed. We're now observing the second wave of more conservative folks coming into open stack with their expectations already kind of being tuned by the first five years of successes and failures. And I think that a lot of the guys that we have seen start working with open stack in the latter half of last year, they already kind of understood that, you know, either you go with, you know, highly prescriptive, highly opinionated approach that's been tested and tried and go that way, or if you want to build a snowflake, then you need to arm yourself with proper operations expertise in-house. And, you know, in both cases, the outcome of open stack implementation being successful is greatly increased. And that's kind of, you know, the wave of open stack 2.0 that we're observing. There's no change in the reasons why people are doing it, but there is a change to the types of organizations and the types of expectations that they're coming into open stack with. Yeah, and I think part of that also is that, like, I'm having less of the conversation of, you know, open stack is for a third-platform or cloud-native applications and DevOps. But, like, that's happening a lot less frequently, at least in North America, where I don't have to do the dog and pony show and explain why open stack isn't a cheaper VMware. And so I think that does go to, of course, with saying just now about kind of expectations. I think those are starting to get dialed into what open stack actually can do. And so that's good. And now what it really comes down to is sort of, you know, is it going to meet the needs of these folks who understand this for a third-platform? And can that broaden out into all the mainstream enterprises? Are they all going to get that they need to adopt DevOps and start working on mobile and big data and analytics and, you know, focusing on the agility of their business? And then do they see open stack is foundational for that? And can their open stack vendor of choice, you know, make that path to getting open stack up and running? It's one that's very, very straightforward. It's okay if it's not simple, but it needs to be something where they're not spending much time on the infrastructure because that's not where a lot of value is. They need to be, you know, changing the culture internally and moving to DevOps, learning how to iterate much, much faster and open stack should really be out of the way and just something that, you know, just works. Got it. And from a workflow perspective, are you seeing generally all-purpose cloud-native or does something stick out to, you know, be it NFV, big data, e-commerce, et cetera? Well, I sort of see, you know, I see NFV as a big bucket. The carriers obviously are glomming on the open stack as a way to solve those problems. I think hybrid cloud to Amazon and to Google and maybe Azure is now starting to get the attention that it's always deserved that I've kind of, you know, talked about for a long time because HP's exited, you know, public cloud really leaves only rack space and I think people are starting to understand that, you know, there's no chance of like this alternative open stack ecosystem that challenges Amazon and Google. And I think that's good and we need to play with those guys because it creates a bigger pie for everybody if we do that. So I see that kind of hybrid cloud use case as actually being something that's important in Robin. And then I'm really excited about PaaS. I think PaaS is really starting to take off tools like Cloud Foundry, you know, are really starting to get a fair amount of traction. It's number one PaaS for Robin's second. I know the Marantz guys have had a lot of success with Pivotal Cloud Foundry as well. And what I see is that, you know, the average enterprise, the bulk of enterprises are never going to be cognitive experts, just like they were never e-commerce or web experts either, right? The average enterprise developer needs kind of like, you know, sort of like, you know, gutter rails, like a bully now, right? They need some way to make sure that they can kind of stay in the middle of the pocket in tools like Cloud Foundry to allow you to do that. And so I think that open stack is an underpinning for that is a really great use case for both of those. It allows Cloud Foundry to operate on a low cost Amazon-style cloud that can be easily grown and scaled, but it abstracts away a lot of sort of the messiness and is much more developer-friendly interface. So I think NNV, big data, you know, PaaS in the form of tools like Cloud Foundry and that hybrid cloud use case. I'm excited about all four of those. So, yeah, from my viewpoint, there's, you know, the mainstream use case that we've seen the most is the DevOps and you can kind of, you know, put the Cloud Foundry, Agile, IT, Agile enterprise into that bucket. But the bottom line is that it's a building a platform on top of which the ever-growing developer community inside the larger enterprises can build new applications and experiment with. That's by far has been the biggest use case for open stack and it's a use case incident of which Amazon has started as well. And the other two that we've been seeing quite a bit is the NFV that Randy's mentioned and the IoT. So on NFV, frankly, my opinion is that, you know, we are just kind of entering this hype cycle right now. So NFV stories with open stack is where open stack itself was five years ago. Isn't NFV more happy than IoT, dude? No, it's the same. It's the same. I haven't touched on the IoT just yet. But the whole idea is that, you know, everybody's very excited about this notion that NFV is great, this, you know, doing everything using open fabric and throwing out all of this, you know, expensive bulky networking gear that we've been paying millions of dollars to establish providers for. So a lot of this is happening and it will happen there eventually, but I think that we are still, you know, about to hit a lot of disappointments. So right now everybody's talking about it. Very few people actually starting to do some things around NFV with open stack in my opinion. So we've, you know, everybody wants it. Everybody has aspirations, but very few actually have, you know, production network functions of any sort running on open stack. And we're going to hit some tear amount disappointments, I think, in the next couple of years as more and more of those start coming up. And I mean, it's natural. It doesn't mean that necessarily open stack is not a relevant player in the space. This is just kind of, you know, something to be expected and we're kind of bracing for it at Mirantis. And that's just my personal opinion. I think we're a little bit further ahead on NFV than you think. I mean, AT&T was our early adopter. I guess it depends on how you define network function, but, you know, you look at a lot of services, AT&T is running on silver lining. And, you know, there are some folks that are out ahead, but I agree. I mean, it's early days for NFV for sure. But IOT, man, you got to tell me about IOT because I don't even know how to define it. I feel like it's like the word cloud in 2008. No, I mean, it's, so, you know, we'll talk a little bit about some of the specific use cases later in a year. But, you know, the umbrella story there is that there is a lot of devices that nowadays need connectivity to cloud like purpose-built devices. So all of the guys that make these devices, and typically those are big guys, right? They need some sort of backhand. And, you know, they've been talking about building the backhand for a while. And now some of these initiatives are starting to kick off. And, you know, simply because of the scale and aspirations of a lot of these things, they want to kind of own and control the infrastructure behind it, which is why OpenStack is a good choice for a lot of them. You're saying that the use cases are more like somebody like GE wants to have all of their sensors or whatever, everything that runs in the house, be able to report back to a central system. And they're going to run SaaS services effectively on a substrate like OpenStack for each of those different sensors and probably aggregate data into some kind of big data system, right? So it's more like, it's like, it's more like multiple use cases. We've seen the past, but all kind of put together for the purposes of servicing vendors who've got IoT devices. Exactly correct, yeah. Okay. All right. Awesome. Thank you. So now let's focus on this new year that we're in, 2016. And tell me a little bit about where you think, you know, OpenStack will be headed and what will change the most significantly in OpenStack this coming year, in your opinion. Okay. Arend, do you want to do the honors or should I tackle this first again? I'll go for it. And so, I mean, I'll just talk from my perspective. You know, it's sort of, I feel like we're kind of moving into a crux point where we've gotten a certain amount of adoption and we kind of have to push on through the other side or we won't get there. Like I remember talking to some of the investors for Cloud Era in the early days and they told me Cloud Era got out there and then they like hit like 60 customers and they just like lat line for a while and it took a little bit for them to kind of get to the other side. And I feel like that's where we're at. I look at the user survey data. I look at some of the other stuff and it's, I feel like we've kind of stalled just a little bit, but the interest is higher than ever, right? Like I don't talk to any enterprise where there's not a conversation by OpenStack. Like it's literally off the charts, the level of interest. So I feel like we just need the final crystallization. Like, you know, the winners need to show up, whether it's Marantys or Red Hat or whomever doesn't really matter and they need to really start to, you know, get enough traction in enough places and there needs to be a maturation of those OpenStack-powered products along with use cases, which we're all talking about in such a way that, you know, we really kind of clinch it, right? Because if we don't, like I said, I feel like the windows shutting. And I think there's like some, like EMC for example, we have sort of a turnkey appliance that's coming out in another few months and not really a lot of talk about it. It's not launched, called Neutrino. And we're hoping that, you know, kind of like the post-nebula era is ready, like mainstream enterprises like that are not super functional, can't build their own operation scenes if we give them sort of like a, you know, turnkey OpenStack solution that that'll help them, you know, actually get there, right? Because the most interesting things that, you know, as Boris was saying is in the DevOps and kind of like the application pieces above it. But, you know, I don't know that it's there. I don't know if we're going to get there. But I'm very hopeful, continue to be hopeful. And other than that, I guess the big things that I want to see are the hybrid cloud use case. I want to see focus on that. I want to see us look at Amazon and Google as partners. And let's see what else there. And I think that there's been some really interesting turnover on the board. And I'm hoping that maybe we'll see some actual action on some of the things that haven't been there for a long time, like around interoperability. So from my end, I think that in the coming year, the narrative is kind of changing from OpenStack being all about open because that's been kind of the key narrative for many years before. It's all open. It's, you know, Apache 2. You can change it however you want. You can customize it, whatever, to more about, you know, making it work. So move kind of from it's all open to it's about, you know, it's about making it work. And a lot of it I completely agree with Randy is about kind of a move towards a more prescriptive type of set of reference architectures, different consumption models. So appliance that you guys are launching definitely is the right move. As you know, we're doing the same stuff on our end. We've unlocked appliances and working with a variety of vendors on pushing very prescriptive hardware architecture underneath OpenStack to make it much more turnkey. Part of the change that narrative though, Boris, is something you're responsible for, which is, you know, pushing the idea that, like, this stuff, you know, the no vendor lock-in kind of narrative, which has been a little frustrating to me because open source doesn't mean you don't have any vendor lock-in. Like, I don't know, like, if that was the case then the financial services guys wouldn't be so, like, wired into Red Hat. Like, I would love to see the conversation not only changed to being about making work, but I'd love to see a conversation to be about switching costs. Like, if I can move between a proprietary and an open source solution and the tooling's all there to do that so that my switching costs are low and I can change the vendor relationship so that I'm in charge and not the vendor, that's what it seems like people want. Like, whether it's open or not seems irrelevant to, like, reducing, you know, how much that the vendor has power over me, which seems to be the key issue. Yeah, I mean, I think that a lot of the vendor lock-in narrative is not necessarily about the specific vendor and switching the vendor costs, but it's about the ability to influence the roadmap of the product that you are using. Yeah, but that's not lock-in. Well, no, it is lock-in, in my opinion. Vendor lock-in has pretty specific denotation, which is I am locked into that vendor. It doesn't have anything to do with influencing road. Yeah, that's a great question. Sure, but I think that the underlying issue to vendor lock-in is exactly about roadmap influence. So if I can, for example, have confidence that if I need a certain set of features or I need a particular set of hypervisors or particular hardware certified around my cloud fabric and if a vendor does not do it, I can just go ahead and do it myself because open source, open development process that OpenStack has adopted enables me to do that, that alleviates the issue. That's the interest in open, not the interest in no vendor lock-in. We can just agree to disagree, but after five plus years and dozens of conversations with customers, I've never heard one tell me that that was their interpretation of a vendor lock-in me, not one. Well, okay. My experience has been different, but okay, let's agree to disagree. Yes, sir. Your other predictions or other things that need to happen this year? Yeah, so the other prediction, I don't know if it's a prediction, but I think that it would be a good thing for OpenStack. Another good thing for us is I think that from the messaging product delivery standpoint, I think that in the early days, the notion of OpenStack as an abstraction layer for infrastructure versus an OpenStack as a substitute for VMware has been forced towards the latter by the vendors that have vested interest oftentimes in selling a completely integrated, consolidated stack. So when you come to a lot of the customers, for instance, that we engage with, the first impressions they have is like, okay, well, Marentus is also a Linux vendor and a hypervisor vendor and an SDN vendor and everything. And this entire thing is OpenStack, whereas OpenStack originally was envisioned as kind of an abstraction layer, as kind of an orchestrator of various API endpoints. So I would like, and I think that some of the newer vendors, actually newer startups that are coming to market, like Platform9 or ZeroStack, they have done a pretty good job kind of lobbying this notion of decoupling OpenStack as a control plane, as an abstraction layer or infrastructure underneath and OpenStack as this completely integrated thing. But I think a lot more of that needs to happen and I see a lot of the folks that we compete with actually lobbying the different direction, so it's like one thing, it's like VMware and we're competing with VMware of OpenStack. I would like to see this change. I think that it would benefit not just Marentus, but it would benefit OpenStack community overall because it's also much more in line with the architectural principles and what OpenStack was envisioned to begin with than what it ended up being interpreted as. So you're saying you'd like to see more integrated stacks with an opinion? No, no, no, no. I want for people to understand that OpenStack is not a hypervisor and OpenStack is not a Linux distro. That OpenStack has been originally envisioned as an orchestrator of various, you know, API endpoints over which in many cases it can be multiple hypervisors. And OpenStack talks to the hypervisor and spins VMs and stuff like that and there is some interdependencies, but OpenStack is not the same as a hypervisor and you shouldn't be choosing OpenStack as your hypervisor because hypervisors are different things. And then it pulls a whole ball of wax with it which is like the certification of guest VMs and a whole bunch of things that become very messy if you treat OpenStack as this just, you know, complete universal integrated infrastructure fabric. Got it. Well, thank you. From the customer perspective, you know, you talked about more prescriptive reference architectures, et cetera. What are some of the external factors or influences around OpenStack, you know, be it the economy of operating OpenStack or the technology behind OpenStack and the underlying, you know, abstraction, if you will. What are some of the most influential drivers that are driving these changes towards needing more, you know, opinionated reference architectures? The desire to make OpenStack work well, I think, is probably the primary driver. And that's it. I mean, I don't know, Randy, do you have an opinion here? Yeah, I mean, we kind of have to circle back up to the conversation that we just had, right? One of the fundamental challenges in OpenStack from the very beginning, both organizationally and from a technology perspective is, you know, just kind of attempt to sort of like be a big tense to try to accommodate as many people's needs possible to get as many people around the table as possible while at the same time trying to deliver something that was interoperable and customers could turn on to look the same from deployment to deployment to deployment, something that we've struggled with. And those two things are just inherently at odds, right? I mean, they're just fundamentally at odds and I don't know that we've gotten a very good resolution on them. So, you know, Platform 9 and whomever else from ZeroStack, I guess you mentioned, you know, they may be coming to the table with an opinionated stack, but I'm pretty sure they weren't the first. In fact, I can think of this other company that was run by somebody. Oh, yeah, cloud scaling had this opinionated stack for OpenStack. And, you know, so what I think is that, like, you know, I really wish that we could reconcile this, right? And we just haven't. And I have lobbied to reconcile it a few different ways and I feel like you either have to kind of, like, start saying, hey, we're going to limit options. We're not going to support all different hypervisors and all different storage and all different networking. We're going to have a default set of, you know, choices. But then you wind up being exclusive and you get people who are going to, you know, maybe want to fork it or go elsewhere, play in a different pool, a different pond. But, you know, the other way to go, which is the way that I really wish we had gone and lobbied at the board many, many times and just never got anywhere on, which is to start to try to look more like Apache Software Foundation, where instead of sort of promising that OpenStack is a thing and promising that it'll ever have interoperability, we instead say, you know, it is a framework that can be molded into a variety of different use cases, which is really what Linux is, right? I mean, Linux on an Android handset looks nothing like Linux on your server in the data center. Those are two different Linux things, even though it's in code base. And, you know, but I, we never really got there. We still have this kind of centralized technical committee, you know, we have this process, even though we've got the big tent process, it's still like you have to go asking if you can be part of it, instead of us just having kind of like a, you know, more of a free crawl where it made the best, you know, projects win. And then we could have competitors like in networking, you could have direct competitors to Neutron, you could have competitors Cinder and so on. And people like Marantis could come in and say, look, here is, you know, Marantis OpenStack for NFV, which has these best of breed choices that we've worked with carriers to choose. Here's OpenStack for AWS compatibility, which is a different flavor of OpenStack that you can deploy with Marantis that, you know, has these choices made for you and so on. And that's the way I'd always hope that it would break, but I've unfortunately failed to get enough mindshare with the Movers and Shakers to try to make that really coalesce. I think that you're completely spot on about the change that needs to happen. My kind of thoughts on this is that, you know, this happens naturally with adoption. And it's not like, I mean, like open source and upstream development is always going to be what it is. I don't think that a board or even technical committee can mold OpenStack into this very prescriptive set of things and reference architecture. No, no, that's not what I've lobbied for. I'm just so clear before I get too far. What I've lobbied for is that to get out of the way of that occurring naturally, as you're saying, there are certain unspoken rules and sort of conventions that aren't necessarily written down, like TC basically will block a competitor or nobody. You can't come in and bring it to something that competes with nobody, like they just won't have it. So, you know, things like that actually potentially get in the way, right? Because if neutrons aren't going to work for NB, then why can't OpenJLIP be a legitimate component, for example? There's not really any reason, except it's not OpenStack. I don't know, but, you know. So, but with BigTent, you can do that now, right? No, you can't. You can't replace a component like Neutron and call it OpenStack. Well, they have the Acanda project now, which is basically kind of like a competitor to Neutron. It is the crack in the door that suggests that we might get there, but it does not compete completely with Acanda. It competes with a subset of Neutron that most people agreed was failing. And that's what the TC has been clear about in the board meetings, is that they've said that they don't want gratuitous competition. I think that, yeah, well, I mean, there was, you're right, that I think that in the early days of OpenStack, as a community was kind of hashing itself out, there was this notion that on the one hand, OpenStack is going to be this kind of very prescriptive product and there's not going to be any competing projects on site and we're going to have a very solid definition of what projects responsible for what. Yet we want to be an ever-pluggable and ever-moldable type of thing. And that was the thing that's at odds. I think that now from the community standpoint though, like as far as things that are happening upstream, there's been a fair amount of credible motions such as BigTent, for instance, that actually enable you to get healthy competition. So like the community stuff that's happening upstream is like an innovation bucket, right? It's kind of like a constant development bucket. It's an environment. It's a bunch of ideas kind of clashing and emerging and then people can take those ideas and turn them into something that is more prescriptive and workable for different use cases. And from my standpoint, I think that the process is kind of like a natural process of evolution. It's kind of hard to force it. And the revolution is ultimately a subject to people adopting and using OpenStack, right? So in the first day when OpenStack just came out, people wanted to use it as a framework. People wanted to build snowflakes. And the people were just handful of large guys like PayPal and Walmart and stuff like that. Now that's moving a little bit more to mainstream, we see more and more our customers coming to us and saying, well, what's the Miranda's opinion on this, right? Like, how does your OpenStack work? I don't care. I'm bringing you because you're the OpenStack guys. But before, you know, we'd go into PayPal and they'd be like, we are the shit. We know everything about infrastructure. You're nothing. So just, you know, we'll tell you what to do with OpenStack and you make it happen for us. Now it's a different conversation. And that's pretty typical of I think a lot of the open source kind of, or new things, right? So, and I think that's happening with OpenStack. So I'm kind of optimistic. I think you're right about where it needs to go. But I also think that it is kind of going there, maybe not fast enough, but I think it's going to get there. That's the key. It's not going fast enough. I've been talking about this window closing and that. And so I, you know, I'd like to push this along. In fact, I'm going to announce some stuff, you know, in the next month or two, you know, about some work that we've been doing to overload RustStack to sort of test different flavors of OpenStack for particular purposes for hosting PAS, Cloud Foundry, or BNA, DFBS compatible and stuff like that. So, you know, from my perspective, I just feel like the work around DevCore and trying to have like one interoperability standard is a place where we haven't made any movement because we insist that there will be one interoperability standard. And I think we'll get to that next level. But you're talking about, as soon as we just admit that we'll never get to a single unified interoperability standard across all OpenStack, and then you start moving to like, okay, what are the workloads in the use cases? You know, what do we need to be able to test for interoperability for those? How do I make sure that OpenStack can, a particular OpenStack deployment can host Cloud Foundry? How do I make sure that a particular OpenStack deployment or products, you know, like Cater to the NFV workloads that carriers are working on and so on? And I very much like to get there. So I, myself and my team, we're in a bunch of code that we're going to be unveiling here soon that will hopefully take that to the next level. But I'm still worried that the community, even though it's got the big tent stuff, there's still this mentality that we're going to build this one monolithic OpenStack that, you know, is interoperable and everybody has and it's all the same. But oh, by the way, you can also pick any hypervisor, any storage, any network, and use it. Like I just, like I still see that kind of, you know, you know, wishful thinking going on. But I just don't, I'd like to see that get shut down. Fair enough. Well, thank you. This is actually time flew by. This was a really enlightening conversation. So first of all, thank you for coming on. And as we wrap up, I have one final question considered almost a closing statement, if you will. Just what other predictions do you have for OpenStack's future? This is not time down anymore. There's just overall predictions on the future of the platform and community and everything. Well, you know, again, I think that I remain convinced that OpenStack has established itself in terms of mind share as the open alternative to the movement that AWS has started effectively. And it might not be maybe moving fast enough or we can argue until the cows come home, you know, with respect to whether or not it's prescriptive or not, stuff like that. But my kind of general premise is that AWS has shown the world how to do infrastructure in a innovative way. And after that happened, everybody kind of scrambled because they understood that there is going to be some sort of open alternative. There was, you know, the period of many years where it was like battles between OpenStack and Eucalyptus and CloudStack and many versions of OpenStack and all that stuff. So now all of that is kind of gone to a large extent. And OpenStack has solidified itself in the mind share of, you know, ultimate consumer in the mind of the CIO as that open standard, as that open alternative. And provided this is the case, I think that it's just a matter of time before it starts taking massive advantage of the disruption that AWS has started. And it's inevitable that OpenStack will succeed, in my opinion. And, you know, some people might think that maybe it's kind of an extreme view, but then again, you know, I do put my money where my mouth is and, you know, I'm fully committed to Mirantis and Mirantis being an OpenStack company and making this decision happen. For me, it's pretty simple. The OpenStack needs to be, it needs to be able to have the same level of focus on the end user application developer that Amazon handles. Amazon's customer that they're laser focused on are the aging co-cross of this world. The people who are going to build cloud native applications and stick them on Amazon in order to get leverage. And, you know, to date, OpenStack has not reached out to those people. It has not gathered a lot of information from them. They don't sit on the TC and the board. And, you know, we just don't really have that. I've heard murmurs that, you know, one of the larger companies in Germany might try to go become a Gold member. You know, that would be fantastic if there was an actual customer on the board. I'd like to see just a lot more laser focus on the customer, the end user, and particularly the application developer that put applications on top of OpenStack. I think if we could do that, that we could make that window that I'm talking about. And if we can't do that, some of these possibly alternative ecosystems that do focus on those folks may actually pass us by. And I wouldn't like to see that happen. But, you know, I think that we're just... We haven't made as much progress in really talking to and understanding that user and driving the requirements back into OpenStack as we could have in the past five years. Well, thanks again for coming on Subriger TV. That concludes our episode. Have a great evening.