 Okay, we're back here live in Santa Clara, California in Silicon Valley in the heart of big data innovation. This is O'Reilly Stratoconference, this is Silicon Angles theCUBE, our flagship program. We go out to the events and extract the signal from the noise and bring that independent perspective to the event and obviously this is a three day event, our three day, we call blanket coverage, wall-to-wall coverage, interviews all day. Nine to six, yesterday we went to 6.30, just so much content, so many stories to tell. Great data to share and we want to share that with you. My next guest is Simon St. Laurent. He is the chair of the Fluent Conference, which is developers, HTML5, JavaScript, all the programming that is needed in the future and very relevant around big data, so welcome to theCUBE. Thank you. Great to have you, looking forward to covering your shows, but really the important thing about big data that we've been hearing today besides security, which we covered earlier, is end user experience and design, which you guys at O'Reilly have really, really highlighted and are doing a lot of work around, but it's the developers out there that are really wanting to extract the data, use visualization, present the data. Well, we had Todd Papiano on from Continuity who's building a fabric for developers and it's not just IT, like we talk about IT with all the EMC players kind of here, but it's just not IT anymore, although that's a massive growth market. There's a ton of developers building mobile apps, web presence, and just in general purpose, web or mobile programming, and they need to act on big data and use data as code, as we say. So, one, what's your take? You're here mingling around, you're probably sharpening your pencil for your show. What's your take on big data and how do you connect that to your event? Well, the big data story for us is less about the data itself. It's more about how you present that data to an audience, how you let that audience interact with the data. Some of its visualization, I think we're just at the beginning of user interface design that can really work effectively with large data sets efficiently. It's not just a massive dose of statistics and computation. So, on the front end, we're bringing these pieces, we're making them accessible to developers and combining a variety of programming and user interface technologies. What are some of the hot, for the folks out there that are in the big data space who want to understand the tools and techniques in the UI and user experience side of the development, what kind of tools and tricks to the trade are out there that you're seeing that are either expanding in relevance or emerging onto the scene that you see as heightened focus? Well, over the last five or six years, we've seen a real renaissance in JavaScript that's gone from being a simple object manipulation language to being a whole much better respected language in which you can build frameworks. And for the data side of things, these frameworks typically combine a traditional web framework with more specific tools for visualization or for interacting with data. So, to some extent, the folks coming to Fluent are figuring out how to apply these frameworks. Some of them are there to figure out how to build their own frameworks and do it efficiently. It's all about building smooth layers of abstraction on top of whatever is on the other side of that connection. So, one of the things that was impressive is, and that we've covered in the past couple of years, but once it's our fourth season of theCUBE, we've got the events. And one event that we did that we were excited about was the Node.js Summit. Yes. And that was in 2012, 2011, I should say, at the end of the year. And it was really motivating to me because what you saw was, you saw developers who were in JavaScript who were doing some really cutting-edge work, especially around the dynamic edge of the interface. And what was needed was more headroom. They wanted to do more, so you saw them going server-side scripting. That opened the door to a massive amount of opportunity. How does that play into your event? And those developers that are growing up, you know, they're expanding, they're growing up, and they need more headroom. But the server side was not obvious that that was going to happen. No, I mean, I remember way back in 98, I was working with Netscape's server-side JavaScript. And if you told me that I would be doing server-side JavaScript, again, I would have been surprised. It just didn't seem likely. Node takes a whole variety of paradigms and expectations and reapplies them very differently. A lot of the things that front-end engineers have learned to do in the browser actually turn out to be useful in the server. Restructuring the way we look at server-side applications has been maybe more exciting and adventure than a lot of people thought Node would be originally. And it's a key part of our story. I mean, most affluent is about the front-end. Most of the pieces are definitely pieces that came from the front-end. But Node gives you the possibility of doing things differently, applying the same sets of tools to very, very different problems and making it work in a server context. So one of the things that we announced this past year that was really groundbreaking research from our Wikibon research team was this notion of software-led infrastructure. So obviously, we cover the enterprise, convergence space and big data, but convergence was kind of an old definition. And I kind of hated it. I didn't really like it. I liked it when it was five years ago when you had servers, storage and networking and it was elegant way to describe the confluence of those three things. But it never really was updated. You have flash memory, you've got all these new web apps going on where IO was really the critical bottleneck. And so we were seeing that software-defined networking and software-defined compute. It's not defined yet. And the use cases are different. So Node.js was like, to me, the really interesting DevOps kind of environment where the guys doing Node were really DevOps guys who had to do everything. But then Hadoop kind of sucked the oxygen out of the room around DevOps. Cloud came in around the data center. So DevOps has kind of been in transition in terms of a market. Yet the people are changing. So in a way, the DevOps side is turning into design ops. Design, where it's not so much cloud monkeying around with cloud code. That's cloud ops. You're seeing a lot of design ops. I guess, I mean, I don't know if that's a term, but if it is, we'll just go with it. But user experience. I mean, just IO, fast IO for whatever application. That's a design objective. It's a user experience requirement. What's your perspective on this kind of design ops concept? Well, I see the software development cycle and the operation cycle meshing ever tighter and tighter and tighter. One of the things that's been interesting watching people do in Node is that you can shift where you make decisions from the server to the client and back again. You can do all kinds of preprocessing. You can adapt to different circumstances. Bandwidth is our perpetual key constraint. We've gotten JavaScript to the point where it performs fine and we can do things. But if you can't get the data down, how do you respond? How do you change? And this is a whole new world for operations because you can flip things around on the go. You can set a set of rules and expect that your software will adapt based on the profile that you set for it in ways that you never would have imagined even, I don't know, three years ago, two years ago. It's become much more flexible and I think for operations folks, much more intricate. We're going to get to HTML5 in a second but I want to ask you about, I'll see. One of the things we've been watching pretty heavily is the spring source acquisition that VM were acquired years ago. And then I'll see most of the top spring guides bailed or bailing and then ultimately VMware is taking those guys out again with Pivotal, these other, away from VMware. And spring represented that environment where those frameworks, the framework concept is very interesting. People like frameworks, they want toolkits. They want to have that framework. So the two-part question is, talk about the front-end frameworks where there's some programmatic interfaces and then compare and then try to connect that to all the bumping around on HTML5, right? Mobile's big, native HTML5, do I go with some sort of third-party app? Just as a dynamic's going on right now around front-end frameworks and then do I build my own native HTML5? And then there's new startups coming and say, no, no, no, use ours, we'll port you across all environments. You guys talk about that, is that something that's? Yes, yes, we definitely do. The framework's question is getting interesting because we've had an explosion of frameworks in the last five years. Ed Dumble, who's the chair of the Strata Conference, wrote a piece about the future of programming in which he worried we had wobbly tea cups, the wobbly tea cups sitting on top of wobbly tea cups as we just stack these frameworks further and further up. The situation is kind of an explosion of possibilities right now. I sometimes say it's the Cambrian explosion and we aren't sure what's going to happen, you know, what's going to survive out of this. There are too many frameworks for people to easily keep track of and cover at this point and a lot of what developers are doing and creating new frameworks is trying to abstract away the layers of complication in the layers below. The explosion is continuing, I don't know how long we'll keep seeing this burst. One thing that we're also seeing is stuff that takes capabilities that were just on the server before and really moves them to the client. So I started doing Model View Controller seriously in Rails, but now I see JavaScript frameworks that have a complete MVC stack just on the client. It's definitely getting more ambitious. Yeah, I mean it's a blurring of an area right now so like in here the conversation tends to be about Hive, right, and real-time querying engines. And it's awkward and not elegant if you're a purist on the front end, it's awkward, right, and not elegant. But yet works great if you're a data warehousing guy, right, or in that world. How's that bridge connect? A lot of the times it misses. So we have a lot of people who were used to building applications with server-side technologies. You would structure your entire application around, say, PHP or around Rails routing expectations and that was how you built the app. But now that we have people building apps that are really much more front-end dominated, you start with the UI, you don't start with the file structure that you're gonna have to build it in, we're seeing some cranky conversations. These two sides have to figure out how to meet in the middle. They need to find ways to wrap the expectations of the interface on the client, on the front end, with the API expectations on the back end. And we're really only beginning to have that conversation for real. You know, Simon, we'll get you to the last word here. We're getting a little bit over time, but I want to ask you this important kind of perspective. And there may or may not be a right answer. I don't know, because I'm not deepening the weeds on this like you are, but from a high level, kind of understand the general issues around what's going on mobile and web and cloud, et cetera. But really, at the end of the day, when markets are growing, there's always kind of jockeying, people are elbowing each other, elbows get sharp, and do I have the right position, the platform or all that stuff, et cetera, et cetera, happens, but at the end of the day, the user experience of the developer is about simplicity and ease of use. So again, that's kind of the core. Whoever does that wins, right? So what's your perspective on that? And we'll end on that question of what is out there and what needs to be done better to get to that goal of simplicity, easy, reduce steps to get a job done, and I don't want to say push button programming, but making the program's job easier and simpler. Right, we're actually kind of going away from push button programming right now. I feel like we reached a peak of power and ease in the JavaScript world with JQuery's sort of dominance about two years ago, when it was like 85% of the market, and you could count on it to do enough for you, you could count on the rest of the community speaking the same rough language. With this burst, this additional burst of frameworks, with people saying things like JQuery isn't even a framework because it doesn't provide this set of parts, we've jumped up to another layer of abstraction, and eventually that will ease some of our pain, but in the meantime, as we sort out what that layer should look like, it is very noisy, elbows out, we're not really sure where we're going. The good news, I think, is that it took a few years to get the experience that led to that kind of JQuery comm, and we really are testing these things out, we're trying them out and we're seeing what happens, and so I'm guessing in the next year or two, I don't know that we'll ever see something as dominant as JQuery again, but I think it'll be a much simpler conversation with a much saner set of choices for programmers. Okay, so for the folks interested, the Fluent Conference, fluentconf.com is O'Reilly's conference, this is the JavaScript and beyond is kind of the title, really what it's teasing out is the future of front end and programmatic scale for great user experiences. Simon, thanks for coming on theCUBE, we really appreciate it, and we hope to see you in San Francisco on the 28th and 30th with theCUBE. We'd be great if we could bring theCUBE there and we'll continue to have those conversations. This is theCUBE, Silicon Angles, exclusive coverage of O'Reilly media's Stratoconference. We want to extract the signal from the noise, a space for big ideas to grow. We'll be right back with our next guest after this short break.