 Okay, we're back live here in San Francisco, day two of the Q Silicon Angles broadcast of the latest events. And this is the O'Reilly media conference, the Fluent Conference. This is where the action is for developers, front-end developers, JavaScript, jQuery, Node.js, a slew of other technologies. That was once just a web design framework now exploding full stack, full programmability. And really the developer story is all about the web, the web, it's not just mobile and web, it's all the one thing. These developers making it happen. I'm John Furrier, the founder of Silicon Angle. My co-host is Jeff Frick from Silicon Angle. And our special guest is Roger McGoldless. Welcome back, research director of O'Reilly media. Thank you for coming again. Cube alumni, multiple times, repeat guest. I like Cube alumni. Roger, we love having you on theCUBE because, you know, O'Reilly's a very special place. We've talked about it before on other cubes about the culture. But really you have to be on the cutting edge of the marketplace to spot the trends, spot the people, to spot all the action, making up in this time more than ever a transformation tech business. Maker Faire was great this year. We had Brady Forrest on earlier. We're talking about Internet of Things. Nest has got a thermostat that is programmable. It's more like an iPod as a user interface. It's very Star Trek-like, Google Glass. We had our first ever Cube guest on theCUBE wearing Google Glass. And really riffing on how elegant it is. And we talked about go Google search Apple One and see what it looked like, the original Apple, and then Apple Two, and then the rest is history. We know what Mac's doing. So Google Glass isn't about what it looks like today, it's about what it could become. So Internet of Things, obviously on the big data side, you're doing a lot of data science stuff in general. It's your job to be putting your fingers in a lot of different pies in the marketplace and some more than others where you get excited. So what are you working on right now? What are you thinking about and what's your current view of the landscape? Sure. So as far as what we're working on, one of the things that I'm interested in and I've been doing some work on is how health and data interact. So maybe not the most JavaScript-y topic, but health data has been traditionally kind of siloed where you did a study on whether this drug did this thing and you had 100 people doing it. Well now we've got sensors and electronic medical records and genetics and all this stuff that you might be able to see that there's a hundred different things that you can actually look at instead of just one. You know, there's an example we often use is that no one understands the reason why, but people who floss have less congestive heart disease. What? Right? Flossing immediately tonight. Right, and so it's all like there's a whole correlation-causation thing. Is it just that people who floss take better care of themselves? But they're obviously, they think there's some physiology thing around the microbes, things around the microbiome, environment things. People in noisy places as an example, seeing the different health outcomes of people in quiet places. So we're trying to create this notion of a platform that helps bring lots of data sources together, apply the strata data science stuff to changes in healthcare. And there's things going on in health because of the Affordable Health Care Act that make bringing data in something that can help. Well, healthcare, we had the strata editor on earlier and she's, you know, Julie handles mostly, most of Julie Steele handles most of the view of the strata RX conference. But you know, healthcare, it's always been compute intensive. You know, you go back to all the healthcare organizations from hospitals to clinicals to everything else, it's data driven. But it's old infrastructure and regulations like HIPAA have caused all kinds of problems with real time, right? So I mean, come on, have you gone up against that wall? What's your take? Come on, give us your view on that all the trash. Yeah, so it is tough and you want to do the right thing, right? You don't want someone that have their life ruined or bad things happen because of data leaking out. But if you're familiar with Meaningful Use, which was part of the high tech act and how all those EMRs were paid for, there's a thing called Stage Two, Meaningful Use Stage Two, and that's people are entitled to their EMR data. They can do whatever they want with it, including donate it to disease groups. And that's what we're trying to set up, is that a new cultural norm is that donating your data, we would de-identify it and so forth, becomes a new way to put things in and there's no HIPAA because it's not a culture. It's open source, open source strikes again, right? It's open source sharing for the greater good of trying to solve these hard problems. And we think that, you know, given that we can provide the right protections and we're working on all that kind of thing, that people will actually like having the better outcomes and finding out more about different things in their lives that might have an effect. I think most people know that if they weigh too much, there's probably an issue, but they're finding all these weird things like taking too many antibiotics and how that might help promote obesity. Well, again, another weird connection that we might be able to find stuff out if we have enough data on what's going on. Well, I think it's relevant to this conference because, you know, you're talking about data visualizations. It's been a programmatic issue. I'm obviously back in things. You're talking about JavaScript in mainstay. You're not going anywhere. JavaScript is now, you know, spreading and use the disease metaphor, but, you know, virally spreading into servers, the databases. So you're seeing server sites with Node.js. So you're seeing that ability to program. So it is kind of like, there is a small tie in data with Fluent. It's the developers going to be building these apps. Yeah, and they'll be just bringing up the whole JavaScript thing, because about three years ago in O'Reilly, I came up with this notion that JavaScript is the most important language to know. And a lot of it had to do with Node when it made the jump to the server side. Because I know as a programmer myself, I like to stay in context. Like, I don't mind, you know, I've know a lot of different languages, but switching can be kind of hard. But when you just need to get stuff done and you're trying to be fast, like the easiest thing is to do what you know. Yeah, and we also had a guest on who wrote the Bootstrap book that you guys just put out on O'Reilly Media. And we teased out the question, creativity versus efficiency of code. I've had many, many conversations with folks saying, hey, I wanted to be efficient, but I really was in a creative mood. I didn't want to really care about the code. And some say, no, my code's great, but it's not very creative. It's a balance, right? You're getting something done. Yeah, and the one thing with JavaScript, so I'm a Python guy mostly, but JavaScript is the fastest interpretive language. When Google did V8, that was an important part of the puzzle. It runs fast. JSON is now the most used. And WebSockets is coming around the corner, through heavily as the browser connector. And because Node helps with a scaling issue that's a pretty common scaling problem, it becomes just more of a go-to place to be. And even what's the problem with JavaScript? It might be a little syntax-y. Well, there's coffee script. All the scripting language for a reason. That's right. So you can use coffee script to get kind of a higher level view. Yeah, but I think what's important about JavaScript is, first of all, a good call on pointing out the obvious to you and not to many other people, but you're right. It's easy to use, right? And you can get stuff done fast. It's kind of scripty, so the purists might want to say it, but we had a guy here earlier who runs a powerful service for Node developers. He said, quote, JavaScript is the assembler for the browser. That's kind of a metaphor, but what he was trying to get to is that native browser OS is like Firefox announcing here at Fluent, and then as we know it's going on with Chrome, you can really get some efficiencies on the stack if you go direct, which is why this WebSockets concept is really interesting to us, because that's not really being talked about that much here, but WebSockets is not a new concept. However, the game changes a bit when you're at your Java script server side. What's your take on all that? Yeah, so when I totally agree. The quote is the assembler language, would you like that quote? You know, the only thing that's wrong with that is that the assembler implies a certain like esoteric knowledge of hardware that I don't think you'd need esoteric knowledge. It's still software. I mean, JavaScript. It's no poor dumps in JavaScript. Right, and you also aren't moving things from you don't need to have any registers you have and stuff like that. So it's still a higher level-ish language, but I do think like getting to the bones of things is why real apps are being done in JavaScript. I mean, I use Google Docs. It's a JavaScript app. I use Google Spreadsheet, whatever it's kind of known as called. But I think it's called Google Spreadsheet. Yeah. Well, I'm using this. Gsheet. These are apps, right? That are doing a lot of work for me. And I'm happy that it's there. I don't say, gee, it's going to be slow because I'm there because it's not slow. You know, so I think on the server side, you've got this kind of notions of high performance. You've got the kind of connections that makes you be able to build that kind of modern, in fact, I heard that term used modern JavaScript, like this whole architecture, which you're describing as the modern JavaScript architecture. And this baggage with JavaScript has some legacy, but the future is a lot more headroom. There's node shows, there's some headroom on the server side. You're seeing some tooling come out around. That's right. In fact, I think this is kind of a funny story. When I was saying how JavaScript was really important and I did an email to Tim O'Reilly about node and I was suggesting that there'll be more NBC stuff in JavaScript. And Tim sent it to a couple of people and they said, that guy is full of it. That's crazy. Well now? The mainframe because there's been a lot of mini computers, he doesn't know what you're talking about. So now there's a lot of them, you know, people are doing, you know, NBC. I think that's a logical progression that JavaScript has become a multi-purpose tool. So before when I said it was the most important language to know, that didn't mean it was the best or the only one you're going to use. It was important that the web is a really important progression. That's right. Is that if you don't know it, like I'm a data guy, well we do D3. And that's how we visualize. It's a great visualization tool. Is it the best? Maybe, maybe not, but it does a lot of the work we do and so we're in it and we have to use it. So take us your take on what's coming around the corner. Obviously we're proud to announce that, you know, we're now doing theCUBE with O'Reilly and all your events. We're excited to bring the kind of that mojo game day, CNN kind of conversation. But you know, we're on the people. We're going to do a velocity and strata. Strata we know, like the back of our hands, we've done all the stratas. But talk about those two conferences, velocity and strata, because you're really talking about you guys have got this concept with your events recently with strata where it's just not one thing, it's multiple vectors coming in. Strata, you had data science, you had platforms, you had business users, you had geeks. Velocity is you got DevOps, you got cloud, you got visualization, you got users, you got usability. You know, so you have multiple vectors on all of these different events. So breakdown velocity and strata at a high level, what are those shows? And what are they targeting for content? Sir, so velocity is my favorite word to describe is DevOps, it's got lots of different names. But this is where you are operationally developing an infrastructure. It's mostly around the web, like 95% around the web. And what it means is that you need things to perform in a certain way, reliably. You need to be able to do rolling upgrades. You need to be able to keep up with huge, huge workloads. Lots of users. Like hyper scale. That's right. And the companies that are leading that were lucky enough that some of the people from them are willing to share how they get that done. And we're also fortunate of velocity is here in the Bay Area that those, the biggest scaling companies are right here. So when you think about between Google and Twitter and Facebook and LinkedIn and even Zynga, these are... It's a cloud show in a way. That's right. But not cloud, like just pure, it's not cloud. So it's the confluence of the future. Cloud infrastructure is really the app. It's the architecture of a back end and a front end, a web front end. At the end of it. Right, whether that cloud is a private or public cloud, it's still work is being done in another tier of the architecture. And the faster that's done, the more connections you can serve, the more you can take value out of the data you have. So people who use the graph, the social graph. So there's the tag lines from a marketing standpoint are speed and strength. Be strong, be fast. So okay, that means it's under the hood. So virtualization would fit into that, right? So self-defined networking is a topic, maybe? Yeah, more and more. But the big thing is that like, how do you get a performance in a JavaScript world? Like how do you do tracing and how do you do profiling? How do you cut out that last little bit of what's going on? And that includes server side and client side because the client is relying on the server providing some of those functions. And again, it's not one shoe fits everybody for these conferences. There's a lot of vectors coming in. That's right. And one of the things I think is that those companies have focused a lot of mental, intellectual effort into making that work. Every company needs to know this is beyond a pretty small size because potentially they're going to be handling a lot of people and a lot of work. So I think that- So that's velocity, right? So velocity is essentially infrastructure of the future which is not the old infrastructure. It's not like cloud, you know, infrastructure. It's still a bit different. Strata, obviously strata is now global. You have healthcare, RX, strata RX. What are the vectors in strata? Anything new or? How do you put that in a box? Yeah, strata, what we think is going on is a maturing towards how do you make data an operational, useful part of an organization. So there's tools, there's techniques, the statistics, the machine learning and all that. That's starting to happen. How do you take what's there and make that work in the company? It's almost like a cultural problem. How do you ask the right question? How do you do a hypothesis? How do you integrate qualitative work with quantitative work? And I think that that's pretty important that qualitative piece matters. When I talk to people, I talk to a lot of people, particularly in the data space because of the work I do, there's a thing you always hear. I didn't know, I needed a taxonomy to make sense of this data and I didn't know how to do that. So something as simple and really in a way not as like data, like statistical, but just being able to put things into the kind of buckets that makes sense to manage. Construction of the model. It's hard work. It's hard work. Data modeling. I think data modeling is super important. In fact, we're seeing this kind of emerging architecture where there's an ingestion engine because you're trying to pull data in and then there's a mid-tier that does, organizes the data usually in some kind of variant of a dimensional model so that people can hit it. Once that's there, you're kind of like doing Pareto. 80% of the people can go against that. But because you've got this ingestion engine that's flexible and schema-less, you can then quickly build up sandboxes to answer any question that might be slow to build it. But once it's built, you can iterate fast through it. If that is an effective analysis, that becomes part of the formal structure. So we love metaphors and analogies in theCUBE. We use sports analogies. Maureen doesn't like Maureen Jennings. She's a fashion, you know, for the foodies out there, we use restaurant metaphors. But this one, we use a car. The car, the engine is velocity and the tires and the body. And the dashboard and the instruments is strata. Yeah. Would that fly? Would that be a way to... Yeah, and the power and the strength is the engine and the mechanics. Yeah, and the control, particularly as cars become more automated, right? The driverless car, we're moving closer to that, that you have a lot of machine learning. You've got a lot of, there's really a lot of math that's going on, right? So a car becomes a lot of math. And the math is controlling the devices. You mentioned the internet of things. I mean, that's something I just was, in fact, on the call today about that stuff. Increasingly, we're bringing software into the hardware world. And there's control, but I think the data part is going to be an increasing part of that. General Electric just did a big investment in Pivotal and they called the Industrial Internet, which is a great name, which is they're trying to rebrand internet of things, which is MIT really kicked off. But really, that's now relevant. So you're seeing that just starting to hit the conversation cycle of this notion of wearable computers, or whatever you want to call it. But it's sensors, it's everything else. So final question for you. Just what's your current take of the tech business right now? Honestly, I just saw a post on Facebook from a friend saying Washington Post just laid off all their photographers. I mean, Washington Post, the Chicago Tribune just laid off all their photographers, the media businesses upside down and morphing. You've got the financial markets that are relying on big data. All the big incumbent players in the tech business are retooling and transforming enterprises or modernizing their infrastructure. I mean, we're living in a very special time right now. As a researcher and a geek programmer, you look at that, what's your take on that? And share your personal opinion on that. Sure. So one, I think it remains like really exciting. There's so much new going on and that's kind of the vibe here. That there's all this kind of new stuff. And yes, new means disruption and that disruption can be destructive. So we're in publishing, a major publisher system. Now it's big layoffs. So things are happening around it. But if you look at things, everyone quotes Clayton Christensen. I don't know that everyone really even reads it. Yeah, it's a great book. It's the all-time best business book, I think. But that whole thing that if you're not like trying to keep up with things and you're trying to protect your old business. You're done. Not that you're done, but you're going to have a slow decline. And someone is going to come in who sees the opportunity for what it is and is going to bounce on it. And I think that that's what keeps things exciting and there's still enough of the VC community. There's still enough successes. Things can go on. And what I think is interesting is that there's almost like the ladder has gotten more sophisticated on what people are doing. So before it was just like, oh, I need to connect two people together to make a network. Oh, I'm going to have a graphical interface. And now it's like, I'm going to create virtual worlds for you. I'm going to show you your social network. I'm going to figure things out about you and help you out and do things. I might drive your car. It's like this increasing intellectual input is going on. Will there be issues? Are there some things that are not really that worth it, of course, but we'll filter through that and go ahead. So I'm pretty sanguine about the future of that. There's a lot going on. It's pretty exciting and we'll figure it out. So the bad things will, they'll be bad and we'll know them and they'll be gone. Yeah, well, we're here to extract that signal to help figure things out. We broadcast live at the O'Reilly events. Look for theCUBE. We're going to be talking to the smartest people you can find, whether it's the research guys at O'Reilly like Roger or entrepreneurs, developers. We want to, we talk to everybody and want to share that with you. This is theCUBE. We'll be right back with our next guest here. Day two coverage of Fluent Conference right after the short break. We'll be right back.