WEBVTT 00:00:00.246 --> 00:00:05.532 >> Ladies and gentlemen, please welcome Steve Guggenheimer and John Shewchuk. 00:00:07.814 --> 00:00:12.019 >> All right. Good morning, everybody. 00:00:12.492 --> 00:00:14.781 Welcome to day two of Build. My name is Steve Guggenheimer. 00:00:15.008 --> 00:00:18.025 My colleague is John Shewchuk. They often call me Guggs and Shew. 00:00:18.278 --> 00:00:20.277 Bad joke there. We'll let that one go. 00:00:20.851 --> 00:00:24.072 We hope you guys had an outstanding day one, and our goal is just to keep the 00:00:24.201 --> 00:00:26.822 momentum going on day two. The first thing I want to do is thank our 00:00:27.014 --> 00:00:29.495 deejay, Alan Smithson. I thought he did a great job. 00:00:33.839 --> 00:00:37.449 The Emulator board that he had here, there's only 25 in existence. 00:00:37.555 --> 00:00:39.977 He helped co-create that. He'll be at the party tonight, so if you 00:00:40.086 --> 00:00:43.315 want to make your own mixes and use the board, you're welcome to come out and see him. 00:00:43.686 --> 00:00:45.604 Put it to the test and have a little bit of fun with that. 00:00:45.958 --> 00:00:50.588 Now, what John and I wanted to do is sort of take some of the conversations we have 00:00:50.718 --> 00:00:52.406 on a regular basis and bring those to life. 00:00:52.753 --> 00:00:56.306 As I hope you got a sense from the video, we basically spend all of our time talking 00:00:56.559 --> 00:01:01.918 to developers, from students to startups to large and small ISVs, to folks who work 00:01:02.160 --> 00:01:06.081 in large and small commercial enterprises, people who build software for a living. 00:01:06.323 --> 00:01:09.716 And the trick here is to take those many conversations and to bring them to life. 00:01:09.951 --> 00:01:14.508 They range from types of apps to technology areas to business models to 00:01:14.643 --> 00:01:16.954 tool chains and more. John, what do you think about that as 00:01:17.050 --> 00:01:19.620 a starting point? >> Yes, when Guggs and I are out there 00:01:19.756 --> 00:01:23.472 talking to developers, one of the things that's extremely exciting for folks 00:01:23.590 --> 00:01:26.475 is just the vast array of capabilities that are out there. 00:01:26.602 --> 00:01:29.871 Different kinds of devices you can target, different cloud platforms with 00:01:30.273 --> 00:01:36.273 PaaS, with IaaS, and then these new emerging platforms, software as a service, 00:01:36.551 --> 00:01:38.592 like Salesforce or what we're doing with Office. 00:01:39.001 --> 00:01:42.430 So when they come, and we have those conversations, it's almost never about 00:01:42.669 --> 00:01:45.955 this Microsoft technology or this particular product. 00:01:46.195 --> 00:01:49.460 It's about how do we put all this stuff together to build great applications 00:01:49.807 --> 00:01:52.811 and make applications that's help us make money and be successful? 00:01:53.036 --> 00:01:56.710 So we thought we'd share a little bit about how those conversations are going, 00:01:56.970 --> 00:02:00.687 how we're putting these pieces together to build world-class next-generation apps. 00:02:01.561 --> 00:02:04.413 >> So that will be our framework for today. You see behind me a set of the 00:02:04.674 --> 00:02:06.907 conversations we're going to bring forward. We're going to have some of our 00:02:07.033 --> 00:02:08.912 colleagues come out and help us throughout the morning. 00:02:09.159 --> 00:02:12.013 We're going to try to bring these to life through two things -- as much as possible 00:02:12.232 --> 00:02:15.371 use code that partners or others have built, ISVs that we've worked with, 00:02:15.625 --> 00:02:18.806 second, do as many demos as possible. We always try and get as many in as short 00:02:19.012 --> 00:02:21.254 timeframe as possible. And third, show code. 00:02:21.625 --> 00:02:24.428 I saw a little meme yesterday that said there was a lot of code, maybe too much. 00:02:24.923 --> 00:02:27.084 I don't know how other folks run their dev conferences, but if you didn't like 00:02:27.236 --> 00:02:30.114 the coding yesterday, you're not going to like today, because we'll probably show more. 00:02:30.610 --> 00:02:31.911 And with that, I think we ought to dive in. 00:02:36.879 --> 00:02:39.563 I wanted to start out with fun in the media and music space. 00:02:40.210 --> 00:02:44.357 One of the cool things that's going on is the mixture of IoT, predictive analytics, 00:02:44.761 --> 00:02:49.666 Bluetooth, and it's enabling some crazy ideas. This is one our friends at Muzik showed us. 00:02:49.917 --> 00:02:51.610 When I got a chance to work with it, 00:02:52.807 --> 00:02:57.222 you see the drum kit on the Surface, you see the virtual drums? 00:02:57.753 --> 00:03:02.019 It's like things possible I would have never thought of. 00:03:02.200 --> 00:03:04.731 The cool thing here is obviously I'm not a drummer, but the folks at Muzik, 00:03:05.000 --> 00:03:08.113 they're working with folks who do this for a living. Let me show you a little bit about what 00:03:08.253 --> 00:03:09.064 they're talking about. 00:03:10.757 --> 00:03:13.351 >> Thanks, Guggs. This is Questlove of The Roots and 00:03:13.661 --> 00:03:14.971 a member of the Muzik team. 00:03:15.144 --> 00:03:19.090 I just wanted to say, I'm really excited we're working with Microsoft on our motion 00:03:19.350 --> 00:03:20.661 instruments and our accessories. 00:03:21.035 --> 00:03:26.295 Personally, I can't wait to see all the things that you guys create using Xbox 00:03:26.557 --> 00:03:29.697 and your mobile devices with our game-changing technology. 00:03:30.438 --> 00:03:33.901 Let's work together on the future. The future of Muzik, the future of 00:03:34.188 --> 00:03:37.193 gaming, and the future of technology. Let's go. 00:04:04.595 --> 00:04:07.463 >> So pretty cool stuff. Jason Hardy and the guys at Muzik, 00:04:07.810 --> 00:04:09.183 a little thumbs up for Questlove. 00:04:09.792 --> 00:04:12.379 This is something we couldn't have contemplated a few years ago. 00:04:12.653 --> 00:04:14.489 So there must be some work going on between what they're doing and what we're 00:04:14.605 --> 00:04:17.858 doing to make this possible. >> Think about this particular application. 00:04:18.214 --> 00:04:21.893 You've got professional musicians, you're using these devices with accelerometers, 00:04:22.302 --> 00:04:27.053 they're doing predictive analytics and a whole bunch of other pretty complicated 00:04:27.339 --> 00:04:30.652 stuff, but at the end of the process, what they've got to do is they've got to play 00:04:30.795 --> 00:04:32.460 a sample, and they've got to do it quickly. 00:04:32.855 --> 00:04:37.477 Now, what I've got on the screen here is some code that we would have had to write 00:04:37.597 --> 00:04:41.978 back in Windows 8. This is some pretty complicated C++ code, 00:04:42.393 --> 00:04:48.214 page upon page of it, all of this to go take that event and play that sample. 00:04:48.689 --> 00:04:51.481 And even when you were done with this, you would have had a problem. 00:04:51.734 --> 00:04:54.731 The latency would have been a little high. So the engineers in Windows have been 00:04:54.857 --> 00:04:55.612 hard at work. 00:04:56.419 --> 00:05:01.208 They've been optimizing the video, display channels, the audio channels, and they've 00:05:01.543 --> 00:05:05.760 actually made it incredibly easy to program. So here are some new APIs that we're introducing. 00:05:06.059 --> 00:05:09.521 Here's the new Audio Graph API. This will do audio processing. 00:05:09.692 --> 00:05:12.650 Notice we can create a device output. 00:05:12.859 --> 00:05:15.610 Here's where some of the magic happens right here. 00:05:15.955 --> 00:05:19.480 Notice we set the quantum size selection to the lowest latency. 00:05:19.603 --> 00:05:22.754 This is what lets the thing play. We connect it to an output device, 00:05:23.243 --> 00:05:26.128 and then we go load the samples here and build the file nodes. 00:05:26.269 --> 00:05:31.048 And that's the core of one of these applications. Obviously, more stuff going on with the 00:05:31.159 --> 00:05:34.043 Bluetooth stack and so on. Let me actually just show you this thing. 00:05:34.388 --> 00:05:36.243 Here's a quick example. 00:05:36.654 --> 00:05:41.900 I'm just going to use touch, and I can now play on this. 00:05:42.183 --> 00:05:46.186 Hey, Steve, you appeared on the screen. Now, remember, last year when we were 00:05:46.298 --> 00:05:48.725 doing this, I remember you wanted to switch gears a lot. 00:05:48.929 --> 00:05:49.645 >> Let's switch gears. 00:05:49.828 --> 00:05:53.728 Let's, let's, let's switch gears. >> Super-helpful, John. 00:05:53.906 --> 00:05:55.730 This is why we don't let him play with the music side. 00:05:56.101 --> 00:05:57.978 Maybe we'll give that to Alan later on for tonight. 00:05:58.330 --> 00:06:01.928 So let's switch gears. Sorry, it just was there. 00:06:02.774 --> 00:06:06.374 The client side is one area of innovation. The other is obviously on the cloud side 00:06:06.532 --> 00:06:08.012 and with streaming media overall. 00:06:08.377 --> 00:06:13.167 Yesterday, we streamed out over Channel 9 the video from this presentation and all 00:06:13.308 --> 00:06:16.163 the rest of them. We hit about 300,000-plus users and about 00:06:16.290 --> 00:06:17.404 75,000 concurrently. 00:06:17.780 --> 00:06:20.260 For us, that's big scale. We're pretty excited about that. 00:06:20.500 --> 00:06:23.125 This year, working with NBC, we did the Super Bowl, and it was over 1.2 billion 00:06:23.415 --> 00:06:25.727 concurrent users. When you think about over the top 00:06:25.994 --> 00:06:30.197 transmission, when you think about streaming media, music, video, all of it, 00:06:30.603 --> 00:06:33.181 it's coming together now, and a lot of work is going on the cloud side for that. 00:06:33.529 --> 00:06:37.985 >> Now, what you can see here is once you've created great video, great audio, 00:06:38.147 --> 00:06:42.210 the ability to get it out across all the different platforms is very important, 00:06:42.457 --> 00:06:46.601 and we're looking at the actual streaming that's coming through the Azure Media 00:06:46.736 --> 00:06:51.101 Services, the Channel 9 discussion. So there it is, doing cloud encoding 00:06:51.440 --> 00:06:53.637 and a whole bunch of other stuff. What I really wanted to do, though, 00:06:53.974 --> 00:06:58.039 was just spend one second. Here's a very simple player where we can 00:06:58.189 --> 00:06:59.313 see what's happening. 00:06:59.685 --> 00:07:01.313 I just want to show you the code behind this. 00:07:01.612 --> 00:07:06.467 This is what it takes to go build one of these cloud-distributed, 00:07:06.586 --> 00:07:11.657 high-scale DVR-based systems. All of this code we're making available, 00:07:11.991 --> 00:07:16.221 it's incredibly easy to use and again, at the core of some of these huge projects 00:07:16.331 --> 00:07:17.052 like the Olympics. 00:07:17.773 --> 00:07:21.130 >> And the nice thing with all these pieces is, we're starting to see a set of 00:07:21.422 --> 00:07:23.928 ISVs building them and giving us feedback and helping us along. 00:07:24.323 --> 00:07:26.783 Let me run a video from a partner or an ISV that's running on top of this 00:07:27.111 --> 00:07:28.095 and helping us learn more. 00:07:29.355 --> 00:07:30.740 [Video Plays]. 00:07:36.300 --> 00:07:40.449 >> Propellerhead is a leading developer of software for music making, both on desktop 00:07:40.785 --> 00:07:44.107 and mobile, allowing musicians of any skill level anywhere in the world to make 00:07:44.303 --> 00:07:47.125 music together. In a cloud-first and mobile-first world, 00:07:47.623 --> 00:07:49.727 our customers expect us to be cross platform. 00:07:49.957 --> 00:07:55.164 Microsoft allows us to develop on one device and rapidly deploy across other devices. 00:07:55.784 --> 00:07:57.970 We started Propellerhead 20 years ago, 00:07:58.383 --> 00:08:02.531 and our aim was to democratize music making, to make it available to anybody, 00:08:02.799 --> 00:08:07.165 regardless of their income or skill level. In a way, deploying on mobile brings us 00:08:07.316 --> 00:08:11.680 back to those roots, and we're able to deliver for free incredibly powerful music 00:08:11.995 --> 00:08:17.962 making tools. As music brings people together, so does technology. 00:08:18.773 --> 00:08:23.072 With these technologies, we'll be able to help people make their musical dreams come 00:08:23.187 --> 00:08:25.868 true anywhere in the world, on any device, at any time. 00:08:26.336 --> 00:08:29.891 We're really excited about what all this technology can do for people going forward. 00:08:30.533 --> 00:08:32.190 [End Video]. 00:08:40.504 --> 00:08:44.397 >> So my thanks to the folks from Propellerhead. Look, we're getting feedback from networks. 00:08:44.676 --> 00:08:50.122 You see TSN, CTV up here, Rdio and Tunein and iHeartRadio are giving us help on the 00:08:50.280 --> 00:08:52.314 client side, so we're getting a lot of feedback on how to make all of the 00:08:52.432 --> 00:08:56.014 underlying and underpinnings for what we're doing better and better and better. 00:08:56.586 --> 00:09:00.286 So I'm going to move off of media now, and given our audience, we know a lot of 00:09:00.386 --> 00:09:03.009 you work in the commercial space for a living, we just thought before we get into 00:09:03.230 --> 00:09:06.306 some of the deeper details, we'd do a couple commercial pieces. 00:09:06.539 --> 00:09:09.484 So first off, let me just say, hey, we've got great partners here who are also 00:09:09.617 --> 00:09:11.673 giving us feedback. We heard about some of them yesterday 00:09:11.787 --> 00:09:15.676 with SAP, and you saw DocuSign onstage. GE has been doing some great work with us. 00:09:15.834 --> 00:09:18.066 Salesforce. Salesforce, yesterday, we saw the 00:09:18.238 --> 00:09:20.795 integration with Office. They're also building a Windows 10 00:09:20.931 --> 00:09:24.933 application, working with us not just on the application but on the SDK. 00:09:25.182 --> 00:09:26.495 There's also infrastructure. 00:09:26.718 --> 00:09:29.746 We saw Docker yesterday. We're doing some work with Mesosphere. 00:09:30.103 --> 00:09:31.464 I wanted to talk a little bit about 00:09:31.723 --> 00:09:35.205 Autodesk at this point, because one conversation that is bridging, I think, 00:09:35.468 --> 00:09:38.536 from consumer to commercial, is the area of 3D printing. 00:09:38.635 --> 00:09:41.381 This is an area where we have a lot of discussions, and this is one of those 00:09:41.496 --> 00:09:46.036 areas where today we do a lot with 3D printing for prototyping or modeling, 00:09:46.291 --> 00:09:49.782 and folks want to move to the area of manufacturing. So really make that step into 00:09:50.077 --> 00:09:52.763 a commercial space. Now, to do this, we have to get all the 00:09:53.002 --> 00:09:55.114 systems to work together. You have to get the software to work with 00:09:55.358 --> 00:09:58.370 the hardware, the printers themselves, as well as the materials. 00:09:58.725 --> 00:10:00.722 And today, it's not quite as smooth as we'd all like. 00:10:00.968 --> 00:10:04.100 So we've done some work working with Autodesk. Autodesk has been building an engine they 00:10:04.221 --> 00:10:07.883 call Spark, and Spark allows us to have a platform across the software, the hardware 00:10:07.989 --> 00:10:11.565 and the materials, with a set of APIs in an open way that we can all work on 00:10:12.030 --> 00:10:15.096 and use and what I'm happy to announce today is we're going to integrate that Spark 00:10:15.267 --> 00:10:18.010 engine into Windows 10. So now we have a common platform we can 00:10:18.137 --> 00:10:21.568 all work from and build on as we want to go out and do 3D modeling. 00:10:21.788 --> 00:10:24.361 And the cool thing here is they're starting to build sort of a line of 3D 00:10:24.493 --> 00:10:28.174 printers that take advantage of the Spark engine. This is the first time anyone's seen one 00:10:28.308 --> 00:10:30.977 of these new Ember printers. I love the fact that on the Autodesk 00:10:31.237 --> 00:10:33.964 side, they do nice design. And when you think about manufacturing, 00:10:34.344 --> 00:10:37.061 you can think about lots of scenarios. Up here, you see an earbud. 00:10:37.306 --> 00:10:41.138 The professional musicians, the racecar drivers, they have these custom-made ear 00:10:41.295 --> 00:10:45.063 inserts, and the notion of moving from doing this as sort of a project through 00:10:45.294 --> 00:10:48.870 multiple phases to basically having a manufacturing ability to do this is the 00:10:49.159 --> 00:10:51.657 direction we're headed in. So I'm pleased with the work we're doing 00:10:51.782 --> 00:10:54.039 with Autodesk, and I wanted to highlight that today, because that's an area where I 00:10:54.185 --> 00:10:56.977 get lots of questions, we have lots of fun, but we're going to move from modeling 00:10:57.379 --> 00:10:58.564 I think to manufacturing in the near future. 00:10:59.013 --> 00:11:03.764 Now, one other one that when I think about commercial, then, is true line-of-business applications. 00:11:04.137 --> 00:11:07.165 And here, you can take a vertical scenario by industry or horizontal. 00:11:07.539 --> 00:11:11.032 You can think ERP, CRM, supply chain management, SCM, etc. 00:11:11.265 --> 00:11:14.157 And so I thought we'd take a horizontal solution, in this case, ERP. 00:11:14.411 --> 00:11:18.708 Our friends, ISVs in the Northwest called Acumatica, who built a cloud-based ERP -- 00:11:19.114 --> 00:11:23.060 why don't you show us what we have here, John? >> So many of these vertical systems at 00:11:23.306 --> 00:11:28.295 the core have reporting, and here what we see is a typical sales order summary, 00:11:28.714 --> 00:11:32.750 and it's got a lot of information, but -- >> As usual, it's close, John, but I need 00:11:32.860 --> 00:11:36.509 a couple of changes. >> And normally, as a developer, often, 00:11:36.635 --> 00:11:40.708 those changes are things that I would have to go in and do relatively manually, 00:11:41.115 --> 00:11:44.761 code it up, add it into the list of reports that these folks can do, and that means 00:11:45.026 --> 00:11:48.858 that I'm not necessarily working on the broader aspects of the project. 00:11:49.413 --> 00:11:52.781 Reporting can take a lot of effort. So we were having that conversation with Acumatica. 00:11:53.210 --> 00:11:57.564 How could we leverage what's happening in the cloud, this new distributed world, 00:11:58.030 --> 00:12:00.665 to take what they've already got and make it much better? 00:12:00.909 --> 00:12:02.162 And so we did this pretty cool thing, 00:12:02.709 --> 00:12:05.241 which is we added to their query system 00:12:05.441 --> 00:12:10.778 the ability to expose all their data out to the world through the Internet, 00:12:11.163 --> 00:12:12.464 through REST interfaces. 00:12:13.014 --> 00:12:16.712 It's called OData, and the way we did that is actually pretty straightforward. 00:12:18.084 --> 00:12:22.662 What you're seeing here is the actual code inside of Acumatica. 00:12:22.911 --> 00:12:26.759 They have an object-relational model. They hooked up a simple model that 00:12:27.066 --> 00:12:31.513 exposed that ORM through OData, and let me show you how cool that is. 00:12:31.910 --> 00:12:37.029 Here's the Power BI Builder, which is one of the tools that can consume OData. 00:12:37.514 --> 00:12:39.912 Lots of other things, Excel and other stuff can do that. 00:12:40.210 --> 00:12:43.183 All I need to do is I just click on this, 00:12:43.433 --> 00:12:50.366 and I go add an OData feed, and I connect to it, type that URL into the Acumatica 00:12:50.576 --> 00:12:54.786 data, and off we go with now the new Power BI tools. 00:12:55.042 --> 00:12:57.207 In fact, I think we've got it running over here, so you can -- 00:12:57.365 --> 00:12:59.666 >> So we might as well have some fun, since we have all this great new hardware 00:12:59.800 --> 00:13:00.628 to play with. 00:13:00.908 --> 00:13:03.984 Now, as the end user, I really don't care what John's doing. 00:13:04.265 --> 00:13:06.510 I just want to be able to get access to the information, and so I've got Surface 00:13:06.675 --> 00:13:07.434 Hub here. 00:13:07.715 --> 00:13:09.711 Sorry, let me reboot it. 00:13:10.915 --> 00:13:14.113 So I've got my Surface Hub here. I have smooth inking on this, so I could 00:13:14.280 --> 00:13:17.069 choose to type in, but you know what? I'm just going to go up here. 00:13:17.558 --> 00:13:19.863 I'm actually just going to ask for sales 00:13:21.688 --> 00:13:26.061 by city, let that come up. And you know what? 00:13:26.210 --> 00:13:29.161 I'm curious. I'd like to see that as a map, so I'm 00:13:29.308 --> 00:13:33.385 going to say as map, and you'll notice, it just automatically goes. 00:13:33.482 --> 00:13:35.961 So for me as the end user, I can sort of pick the reports I want. 00:13:36.311 --> 00:13:39.097 I can zoom in if I want. It's a pretty cool way to take what 00:13:39.366 --> 00:13:42.512 John's working on and not have to go back and ask him over and over again for reports. 00:13:42.912 --> 00:13:46.063 Now, John, this is pretty cool. What I saw yesterday in Satya's section 00:13:46.314 --> 00:13:50.034 is he also hooked into the Office Graph, so what I'd really like to do is see this by user. 00:13:50.358 --> 00:13:51.768 In fact, I'd like to see this by your team. 00:13:52.113 --> 00:13:57.070 >> So really what we're talking about is doing dynamic joins across different data 00:13:57.247 --> 00:14:00.066 sets using the OData protocol underneath the hood. 00:14:00.435 --> 00:14:04.872 What I've got here on the screen is a very simple, single-page web app. 00:14:05.510 --> 00:14:09.368 This is the kind of thing you would do if you were talking to a Facebook API, 00:14:09.587 --> 00:14:11.907 anything, and it's pretty straightforward. 00:14:12.114 --> 00:14:14.185 You do a redirection to a login page. 00:14:14.413 --> 00:14:17.700 You get back a token. You use that token to access information, 00:14:18.163 --> 00:14:19.905 but here's where the real work occurs. 00:14:20.324 --> 00:14:23.460 We're going out to that new Office Graph. 00:14:23.714 --> 00:14:27.837 We're going to graph.windows.net, we're taking our me token, and notice I'm 00:14:28.161 --> 00:14:32.765 looking at the direct reports. So let's actually just run that, and what 00:14:32.887 --> 00:14:36.765 we can see here is we want to go retrieve these things. 00:14:37.233 --> 00:14:40.261 It reaches out, it gets the data. We've printed out the data, and as you 00:14:40.467 --> 00:14:44.033 can see, here's the JSON that comes back. This is pretty clean, 00:14:44.159 --> 00:14:47.239 pretty easy-to-parse stuff. Anybody in almost any language can take 00:14:47.472 --> 00:14:52.913 advantage of this, but again, this has been put out in a way that's consumable by 00:14:53.283 --> 00:14:55.759 a lot of products. It's got this OData characteristic. 00:14:56.234 --> 00:15:00.785 In fact, we've also linked that one up to the same Power BI, so now we can do those queries. 00:15:01.276 --> 00:15:04.012 >> Yes, so you see the Microsoft Graph API. We've hooked it up, as well. 00:15:04.240 --> 00:15:11.784 So here I can now go orders by manager. 00:15:12.162 --> 00:15:16.785 And you'll notice, this is the true magic of software. 00:15:17.165 --> 00:15:20.314 John is selling more than anyone else, and all of us know in this room that John 00:15:20.455 --> 00:15:22.581 is not selling anything to do with software right now. 00:15:22.731 --> 00:15:25.516 He's completely doing the coding. So I like the fact the magic of software 00:15:25.644 --> 00:15:28.674 allows him to sell more, and really, that works out as a nice connection between the 00:15:28.788 --> 00:15:31.784 end user and John on the back end of this piece. 00:15:32.213 --> 00:15:35.185 So this is a good spot. I think we've done a little bit to get 00:15:35.331 --> 00:15:38.184 going on the consumer, commercial, two conversations we have on a regular basis. 00:15:38.670 --> 00:15:41.165 One thing we heard a lot of feedback on yesterday that people want to talk about 00:15:41.298 --> 00:15:44.771 is Windows and building for Windows 10, in particular, taking advantage of the 00:15:44.918 --> 00:15:48.111 Universal Windows Platform, as well as some of the bridges, so with that, 00:15:48.237 --> 00:15:51.233 I'm going to introduce our good friend David Treadwell to come on out. 00:15:51.510 --> 00:15:54.563 David's coming from that way, and help us out, take us on a journey through Windows 10. 00:15:54.818 --> 00:15:57.966 David, it's nice to see you. You're up, man. 00:16:00.519 --> 00:16:02.078 >> Good morning, Build. 00:16:04.765 --> 00:16:09.182 So at Build, we've been talking a lot about our conversations with developers. 00:16:10.411 --> 00:16:14.579 I'd like to tell you about a conversation that Terry and Kevin Gallo and I had last 00:16:14.712 --> 00:16:16.516 summer with one well-known developer, 00:16:17.814 --> 00:16:19.593 a guy named Bill Gates. 00:16:20.327 --> 00:16:25.226 We were reviewing with Bill our plan for making Windows 10 great for developers 00:16:25.325 --> 00:16:28.269 like you. As you might imagine, review with Bill 00:16:28.427 --> 00:16:32.215 is always a big deal, so we were well prepared, but we were also nervous. 00:16:33.637 --> 00:16:37.116 One of the points that I made is that we've been getting lots of feedback from 00:16:37.368 --> 00:16:41.560 people like yourselves that they didn't like one thing about the modern platform 00:16:41.830 --> 00:16:46.670 in Windows 8, and that's that it really only supported full-screen applications, 00:16:47.165 --> 00:16:48.586 no windowing. 00:16:49.281 --> 00:16:53.718 Full screen is great on a smaller screen like a tablet, but many people like large 00:16:53.835 --> 00:16:55.973 screens, multi-monitor setups. 00:16:56.212 --> 00:16:57.988 I love the three-monitor setup that I have. 00:16:59.567 --> 00:17:04.515 So I told Bill one of the things that we're doing in Windows 10 is we're 00:17:04.767 --> 00:17:07.384 bringing back this fine feature from 1986, 00:17:08.958 --> 00:17:11.514 resizable, overlapping Windows for apps. 00:17:18.724 --> 00:17:23.277 And Bill paused, and he looked at me kind of intently for a second, and then he 00:17:23.509 --> 00:17:27.919 said, wow, that's a really great idea. 00:17:28.847 --> 00:17:30.280 You know what we should do? 00:17:30.845 --> 00:17:32.913 We should name the product after that feature. 00:17:36.092 --> 00:17:38.404 So I told him that the joke in the engineering team is that we really should 00:17:38.587 --> 00:17:40.522 have called it Window 8. 00:17:48.920 --> 00:17:50.965 True story. Happened just like that. 00:17:51.210 --> 00:17:52.933 And that story does show something important. 00:17:53.890 --> 00:17:56.277 While it is valuable to have experiences 00:17:56.921 --> 00:18:02.570 that span multiple devices, we as developers must put the user at the center 00:18:02.798 --> 00:18:04.466 of these experiences. 00:18:05.964 --> 00:18:10.113 What works great on a tablet may not work great on a huge modern desktop monitor. 00:18:11.561 --> 00:18:15.115 This important idea from our conversation with Bill carries through both of the 00:18:15.441 --> 00:18:19.364 developer conversations that Kevin and I are excited to talk with you about today. 00:18:20.560 --> 00:18:26.317 First, how can Windows 10 enable you to provide a continuum of experiences across 00:18:26.491 --> 00:18:29.387 device types that put the user at the center? 00:18:30.962 --> 00:18:35.873 And second, how can Windows 10 help you take the code that you have today 00:18:36.361 --> 00:18:38.789 and reach over 1 billion Windows users with 00:18:39.089 --> 00:18:41.517 more personal computing experiences only 00:18:41.697 --> 00:18:43.513 possible on Windows? 00:18:45.291 --> 00:18:47.112 Let's start by discussing the first one. 00:18:48.215 --> 00:18:52.518 With Windows 10, the Universal Windows Platform now lets you reach the entire 00:18:52.704 --> 00:18:59.315 family of Windows devices, from laptops and desktops to tablets, phones, HoloLens, 00:18:59.664 --> 00:19:02.388 Surface Hub, Xbox, and even Raspberry Pi, 00:19:03.364 --> 00:19:05.769 all with one amazingly flexible platform. 00:19:07.148 --> 00:19:10.763 Whether a device has a large screen or a small screen or even no screen at all, 00:19:11.120 --> 00:19:15.316 a mouse, keyboard, touch, a game controller, pen, camera, sensors, you name it, 00:19:16.566 --> 00:19:21.595 the Universal Windows platform gives you a way to exploit the abilities of the device 00:19:22.219 --> 00:19:26.169 and create more personal computing experiences for your customers. 00:19:28.568 --> 00:19:31.104 We've made improvements across the board for developers. 00:19:31.519 --> 00:19:34.040 It's now easier to design, debug, 00:19:34.214 --> 00:19:36.873 develop, and distribute your app. 00:19:38.327 --> 00:19:42.910 Last year here at Build, I told you about how Universal Windows apps could adapt one 00:19:43.272 --> 00:19:49.170 app across all devices or tailor the design to each device. 00:19:50.361 --> 00:19:53.721 Universal Windows Platform on Windows 10 makes this even easier. 00:19:54.727 --> 00:19:58.364 There's now one design language and one control set across the platform. 00:20:00.046 --> 00:20:04.190 This gives you a predictable, consistent toolkit for creating your user experiences 00:20:04.667 --> 00:20:06.270 across all Windows devices. 00:20:07.301 --> 00:20:09.195 For example, the pivot control, 00:20:09.416 --> 00:20:14.480 which originated on phone, is now available across devices in a way that works great 00:20:14.712 --> 00:20:15.853 on each one. 00:20:17.321 --> 00:20:21.874 Universal Windows Platform on Windows 10 includes brand-new controls too, like 00:20:22.164 --> 00:20:24.162 a fully featured calendar control. 00:20:25.020 --> 00:20:26.917 Controls work across screen sizes, 00:20:27.249 --> 00:20:32.479 with mouse, touch, keyboard, pen, and now even game controller right out of the box. 00:20:33.566 --> 00:20:37.314 And all the controls are accessible to all users. 00:20:39.927 --> 00:20:44.768 The one design language and one control set make it easy to adopt your app for 00:20:44.993 --> 00:20:45.999 common scenarios. 00:20:46.417 --> 00:20:50.172 Things that every app designer needs to handle, like window size and input type 00:20:50.424 --> 00:20:54.094 are often handled automatically by the platform for you. 00:20:54.305 --> 00:20:56.153 For example, menus appear with 00:20:56.398 --> 00:21:00.433 mouse-optimized or touch-optimized presentation, depending how they're 00:21:00.689 --> 00:21:04.591 invoked by the user, but you can write the code just once. 00:21:05.995 --> 00:21:09.852 Universal Windows Platform on Windows 10 gives you more power to tailor each 00:21:10.177 --> 00:21:12.639 individual device as well. 00:21:13.594 --> 00:21:18.178 The new adaptive triggers feature lets you respond to changes such as window size 00:21:18.544 --> 00:21:21.948 and automatically adjust your layout, all in markup. 00:21:23.792 --> 00:21:28.739 Beyond just size, you can tailor based on input, orientation, availability of 00:21:28.971 --> 00:21:33.420 sensors, app services, or pretty much any other change you can detect. 00:21:35.295 --> 00:21:40.572 You can even create a totally custom XAML view for each device, all sharing common code. 00:21:42.903 --> 00:21:45.922 And last year, I also told you about how you could share common code for your 00:21:46.293 --> 00:21:48.569 Universal Windows app across phone and PC. 00:21:49.890 --> 00:21:53.691 With the Universal Windows Platform on Windows 10, you can now literally write to 00:21:54.144 --> 00:21:56.640 one universal API and produce one package 00:21:56.894 --> 00:21:58.345 to reach every device. 00:21:59.238 --> 00:22:02.189 It's incredibly straightforward as well as flexible. 00:22:03.544 --> 00:22:07.888 Using feature detection and API contracts, you can take advantage of any 00:22:08.190 --> 00:22:12.464 device-specific APIs while still writing the overall app to the one universal API. 00:22:14.445 --> 00:22:18.593 Universal Windows Platform is really just v10 of the same modern platform you've 00:22:18.746 --> 00:22:20.456 been using on phones since Phone 8.1 00:22:20.799 --> 00:22:22.946 and on PCs since Windows 8. 00:22:23.513 --> 00:22:24.992 Your code still works. 00:22:25.498 --> 00:22:27.626 Your concepts still hold true. 00:22:29.239 --> 00:22:32.594 You can continue writing apps to the Universal Windows Platform in your 00:22:32.737 --> 00:22:38.595 language of choice, whether that's C, C++, C#, Visual Basic, or JavaScript. 00:22:39.714 --> 00:22:45.170 The Universal Windows Platform brings together the best of Windows, Phone, and Xbox. 00:22:46.246 --> 00:22:49.374 Now you can access Cortana and Action Center on the PC. 00:22:50.147 --> 00:22:51.496 You can print on the phone. 00:22:52.041 --> 00:22:55.990 You can use Xbox Live services on every Windows device. 00:22:57.287 --> 00:23:00.541 And there are tons of new things in the Universal Windows Platform as well. 00:23:01.037 --> 00:23:05.438 DirectX 12, Holographic, Active Pen, Windows Hello. 00:23:06.285 --> 00:23:12.347 Over 2,500 platform features, over 8,000 new functions unlock new capabilities for 00:23:12.636 --> 00:23:14.461 your apps. 00:23:15.632 --> 00:23:19.229 Now, I'd like to ask Kevin Gallo to show you how to use the Universal Windows 00:23:19.347 --> 00:23:23.343 Platform to build an app that reaches every Windows device family. 00:23:25.511 --> 00:23:29.321 Go get 'em. >> Thanks, David. 00:23:30.085 --> 00:23:33.422 I'm excited, because today I'm going to be able to give you a lap around the 00:23:33.544 --> 00:23:34.744 Universal Windows Platform. 00:23:35.394 --> 00:23:39.074 What I've done is written a single application that runs across a plethora of 00:23:39.254 --> 00:23:40.584 devices that I'm going to show you today. 00:23:40.836 --> 00:23:44.091 I decided to focus on a home remodeling application. 00:23:44.342 --> 00:23:47.795 I was inspired because my wife and I started a project in our home, and as we 00:23:48.068 --> 00:23:51.469 were working with a firm to go through the process, I kept thinking about how I could 00:23:51.788 --> 00:23:56.479 streamline and improve the experience that they had between the firm and with us, 00:23:56.710 --> 00:24:00.254 the customer, and really also with the device family that we have with Windows. 00:24:00.522 --> 00:24:03.143 There's a bunch of things they could do that would be unique, and they could 00:24:03.262 --> 00:24:04.467 differentiate from their competitors. 00:24:05.114 --> 00:24:08.321 So let me just dive right in here and talk about this application. 00:24:08.517 --> 00:24:10.586 So it has five phases. 00:24:10.836 --> 00:24:16.690 It has a design phase, a planning phase and a capturing phase, then an 00:24:16.821 --> 00:24:19.862 organizational phase, and then kind of this last phase, which is they present the 00:24:20.056 --> 00:24:23.256 final product. We'll ultimately try to enclose that. 00:24:23.693 --> 00:24:27.266 For this design phase, I think the best device that I could think of is kind of 00:24:27.385 --> 00:24:30.264 a creative sketching device, which is a Surface Pro 3. 00:24:30.636 --> 00:24:31.831 So here I have that running. 00:24:32.340 --> 00:24:35.346 You can see on the left hand side, I have a split view UI here at the top. 00:24:35.744 --> 00:24:39.693 I have the five phases that I talked about. On the right hand side is a canvas for me 00:24:39.797 --> 00:24:42.697 to draw on. So I'll just start drawing, and you'll 00:24:42.837 --> 00:24:47.468 notice the beautiful ink support that we have with the universal platform, and I 00:24:47.602 --> 00:24:48.945 can go draw my room here. 00:24:49.592 --> 00:24:53.346 I'm actually not very good at sketching, but it all is beautiful with this ink, 00:24:53.635 --> 00:24:57.836 low latency and you can see it's very fluid. This is all built on the new Direct Ink 00:24:57.976 --> 00:25:01.690 APIs that we support on the Universal Platform and this is all hosted inside of 00:25:01.882 --> 00:25:05.703 an ink canvas. But I wouldn't be the one who really was 00:25:05.837 --> 00:25:09.262 doing the initial design. Instead, I would have gotten that from a designer. 00:25:09.749 --> 00:25:15.183 So we can also save and load ink. When it loads the ink, this is real ink. 00:25:15.659 --> 00:25:19.060 This is not some image that can't be modified because we want to collaborate on it. 00:25:19.233 --> 00:25:24.239 So here, I'm going to move my pen into selection mode, and I'm going to draw 00:25:24.345 --> 00:25:27.093 a lasso around that, and then I'm going to use my finger. 00:25:27.206 --> 00:25:30.438 This app is aware and knows that, when I'm using my finger, I'm in editing, 00:25:30.893 --> 00:25:33.122 I'm not drawing. I can also draw with my finger, but here 00:25:33.386 --> 00:25:36.301 I'm going to differentiate between pen and touch., and then I can go and say, 00:25:36.444 --> 00:25:38.462 you know what? I really want my entryway to be a little 00:25:38.661 --> 00:25:41.782 bit bigger. And so I can add my creative -- whatever 00:25:41.968 --> 00:25:44.645 creative I have to this application for us. 00:25:44.915 --> 00:25:48.586 So let me go show you the code that we wrote to create that lasso. 00:25:49.718 --> 00:25:53.587 So I'm going to hop one over here to Visual Studio, and what you'll see here 00:25:53.947 --> 00:25:59.788 is this is my XAML for my design pane, and so all I had to do is add this ink canvas 00:26:00.066 --> 00:26:02.496 control that's new in the Universal Platform. 00:26:02.910 --> 00:26:06.615 So once I have this in here, then I can wire up some code to create that lasso. 00:26:06.937 --> 00:26:11.342 So let's go look at how to write that code. There are really three parts that I have, 00:26:11.584 --> 00:26:14.348 three event handlers that I need to do. My first press, when I want the 00:26:14.639 --> 00:26:18.238 selection, and then as I move the pen, I get another set of events, and then when I 00:26:18.342 --> 00:26:20.586 release it, I need to go perform the operation. 00:26:20.940 --> 00:26:24.101 So let's start with the code for creating the lasso on my press. 00:26:24.486 --> 00:26:26.946 So when I press, I'm going to create a poly line. 00:26:27.063 --> 00:26:28.153 This is the line that I'm drawing for the lasso. 00:26:28.393 --> 00:26:30.440 I go set some attributes on the line itself. 00:26:30.636 --> 00:26:35.142 I add the first point, and then I go and I say, for the canvas, I want this to be 00:26:35.547 --> 00:26:38.891 a selection versus the actual part of the canvas drawing. 00:26:39.388 --> 00:26:41.902 The ink canvas is smart enough to know the difference between selection 00:26:42.187 --> 00:26:43.441 and drawing on the canvas itself. 00:26:43.996 --> 00:26:48.444 Then I'm going to go here to where I move it, and all I do is add another point 00:26:48.782 --> 00:26:52.381 every time I move, and that's nice and simple, and then when I release it, I need 00:26:52.638 --> 00:26:55.586 to process it. And so of course, never forget to add the 00:26:55.832 --> 00:26:59.371 last point so it looks right, and then create a bounding rec here. 00:26:59.498 --> 00:27:04.299 What I do is I get the poly line that I drew around it, and I ask the ink canvas 00:27:04.698 --> 00:27:08.984 to give me all of the points that are within that poly line, and that creates 00:27:09.150 --> 00:27:13.639 this bounding rec, and then right around it, I draw that bounding rec, and that's 00:27:13.816 --> 00:27:16.951 all I had to go do. So very straightforward, a little bit of 00:27:17.180 --> 00:27:20.839 code here, and I have a great selection and a great experience on this device. 00:27:21.314 --> 00:27:24.694 Now, this is not the only device that supports pen. 00:27:24.995 --> 00:27:29.350 And we know that Universal applications run across many devices, and any one that 00:27:29.590 --> 00:27:33.047 supports pen will support it. So here, we have the largest of our 00:27:33.262 --> 00:27:37.907 pen-enabled devices. This is an 84-inch 4K Surface Hub. 00:27:38.380 --> 00:27:41.297 So let me go load the same application. 00:27:42.397 --> 00:27:45.790 This literally is the same binary running here on the Surface Hub. 00:27:46.265 --> 00:27:50.268 And I can go and draw, beautiful fluid ink. 00:27:54.093 --> 00:27:59.501 Build -- I shouldn't have done that, oh, well. Made an error there. 00:27:59.917 --> 00:28:03.245 You know, can't spell, but I can load the same exact thing here. 00:28:03.469 --> 00:28:06.268 All very fluid, beautiful ink, all with the same application. 00:28:06.884 --> 00:28:11.237 But this is not just for drawing. It really is a phenomenal collaboration 00:28:11.471 --> 00:28:14.989 device, and having this in the showroom would be a great place to interact with 00:28:15.231 --> 00:28:19.758 their customers and kind of not just draw the sketch, but also begin to plan the trip. 00:28:20.120 --> 00:28:23.745 The next phase is going in and out, going to different stores and looking at 00:28:23.986 --> 00:28:28.466 different materials and carpeting and flooring and countertops and deciding what 00:28:28.709 --> 00:28:31.590 you want to do. Here in this phase, we're looking at a map. 00:28:32.203 --> 00:28:35.066 And this is a built-in map control. If you remember from yesterday, 00:28:35.353 --> 00:28:40.068 Joe showed you our Map app, and our Map app is using the same exact map control that 00:28:40.238 --> 00:28:44.080 is available to every developer using the Universal Platform. 00:28:44.558 --> 00:28:47.550 Now, this is a very rich control. I can go move it around. 00:28:47.714 --> 00:28:51.238 I can stretch it. I can actually do a tilt view. 00:28:52.010 --> 00:28:53.564 All this works offline. 00:28:54.320 --> 00:28:57.307 I also have a 3D aerial view built in. 00:28:57.559 --> 00:29:01.313 All you have to do is change the mode and say you want an aerial view, but I'm going 00:29:01.440 --> 00:29:04.239 to move back here. So the map's not very useful if you can't 00:29:04.356 --> 00:29:09.264 put anything on the map, so we also have the ability to pin points of interest, 00:29:09.555 --> 00:29:10.935 or in this case, stores. 00:29:11.158 --> 00:29:15.457 So I've set up a little filter here where I can go choose what type of stores I want. 00:29:15.709 --> 00:29:19.914 They have a list that they work with and partner with, and I typically like more 00:29:20.039 --> 00:29:22.353 modern stores. My wife likes traditional, but since 00:29:22.519 --> 00:29:25.287 she's not here, I win this time. I'm going to go look at modern stores. 00:29:26.157 --> 00:29:30.108 Your mileage may vary. And you'll see that I also have a list 00:29:30.359 --> 00:29:32.755 here, because I want to actually see what they are, and then the points of interest 00:29:32.987 --> 00:29:35.702 just show up on the map. All I had to do was data bind that list 00:29:36.189 --> 00:29:39.185 to both of these, and they show up. So now I'm going to hope over here 00:29:39.440 --> 00:29:43.041 and show you the code that I wrote to create those points on the map. 00:29:44.340 --> 00:29:46.491 So here I am, back in my Visual Studio, 00:29:47.025 --> 00:29:50.636 and I'm going to go to my plan trip XAML right here. 00:29:50.890 --> 00:29:52.970 So all I had to do was add a map control, 00:29:53.583 --> 00:29:58.961 set some basic zoom levels on it and then create a list box, data binding the stores 00:29:59.214 --> 00:30:03.933 that I was going to have here in this list box, and then that's all I need to do in 00:30:04.182 --> 00:30:06.132 my XAML. Then, every single time I do the 00:30:06.299 --> 00:30:10.781 dropdown, and I optimize which ones I want, whether I select all of them, I get 00:30:10.893 --> 00:30:15.168 an event handler here on my click, and let me just pop in the code here. 00:30:15.596 --> 00:30:18.593 So I do some simple processing right here, go through. 00:30:18.718 --> 00:30:22.758 This is pretty simple, just see which things are checkboxed, create some queries 00:30:22.878 --> 00:30:26.115 for which of the stores we wanted. And then I just go and I update my 00:30:26.312 --> 00:30:28.519 observable collection. That's data bound. 00:30:28.913 --> 00:30:31.920 And then I update that and it will automatically will go show the points of 00:30:32.065 --> 00:30:32.998 interest and new list. 00:30:34.036 --> 00:30:37.078 Very straightforward. Easy to add, easy to optimize. 00:30:37.554 --> 00:30:42.501 But now once I've gone and I sort of planned my trip, I want to go take photos 00:30:42.917 --> 00:30:48.153 and see the different furniture and carpets in a nice, optimized fashion. 00:30:48.389 --> 00:30:51.175 You just kind of walk around and go, and it's kind of heavy to go lug one of these 00:30:51.279 --> 00:30:54.282 big things around. So let's go from our largest device to our 00:30:54.620 --> 00:30:57.870 smallest device, which is our phone, and it's the most portable, so I can bring it 00:30:57.986 --> 00:30:58.823 with me. 00:30:59.067 --> 00:31:03.115 So here, I'm going to launch the same exact application. 00:31:03.395 --> 00:31:07.624 And what you'll notice is, it has automatically adapted its user experience 00:31:07.995 --> 00:31:11.520 to run on the smaller form factor and look appropriate on this device. 00:31:12.118 --> 00:31:15.872 And basically, the map is now at the top and the list at the bottom. 00:31:16.196 --> 00:31:18.020 The menu is now in overlay mode. 00:31:18.446 --> 00:31:21.605 All of that very straightforward, so let's go see here. 00:31:22.103 --> 00:31:27.604 Go into capture, and you'll see here we have a new camera API that allows you to 00:31:27.822 --> 00:31:31.870 go into a preview mode and take a picture. And this conveniently located end table 00:31:32.105 --> 00:31:35.735 over here is a perfect one that I think I'll put maybe in my living room. 00:31:36.002 --> 00:31:37.499 So I'll just go, I'll take a picture of that. 00:31:38.347 --> 00:31:42.330 And I'm able to go say, hey, tell my wife, I think this would be great for maybe 00:31:42.468 --> 00:31:44.003 a bedroom or a living room or something like that. 00:31:44.546 --> 00:31:49.406 So let me now show you the code that we wrote to build that adaptive user 00:31:49.621 --> 00:31:52.397 experience that we've been talking so much about at this conference. 00:31:53.768 --> 00:31:58.647 So I'm going to pop back here to the XAML that I have in that planning page. 00:31:59.132 --> 00:32:04.683 So what I didn't show you above this is right here, right above the map and the 00:32:04.914 --> 00:32:07.653 list, they're inside of a relative panel. 00:32:07.932 --> 00:32:12.045 Relative panel is a new layout container that we've created that really provides 00:32:12.322 --> 00:32:17.722 a lot of the magic behind the adaptive user experience. It allows you as developers to either 00:32:18.197 --> 00:32:23.409 dock items -- items inside of it to each other or to the side, to the top, bottom, 00:32:23.650 --> 00:32:26.522 and give you a lot of flexibility in setting attributes. 00:32:26.972 --> 00:32:31.858 Above this, I have set up a few visual state groups with adaptive triggers that 00:32:32.117 --> 00:32:34.452 David talked about. And so I have two states. 00:32:34.717 --> 00:32:37.423 One, when it's between 0 and 599 pixels, 00:32:37.649 --> 00:32:42.253 and then when it's 600 pixels and above. So my narrow state is that smaller one, 00:32:42.602 --> 00:32:46.384 and so basically, my adaptive trigger says, hey, when it's zero, then we'll go 00:32:46.596 --> 00:32:50.599 less than the other one. It will go and do a bunch of setters. 00:32:50.970 --> 00:32:55.652 What these setters do is it makes the map go above and the list below and then take 00:32:55.807 --> 00:33:01.653 the full width of the screen. When it's 600 pixels, this adaptive 00:33:01.757 --> 00:33:06.714 trigger will fire and apply these setters. And these setters do what you see on the 00:33:06.854 --> 00:33:11.730 screen over here, the map on the left-hand side and the list on the right-hand side, 00:33:11.904 --> 00:33:13.727 and then they take up the full height. 00:33:13.911 --> 00:33:18.193 So here you can set up a bunch of adaptive triggers based on the size or any other 00:33:18.325 --> 00:33:21.572 attributes and adapt the user experience very straightforwardly. 00:33:21.924 --> 00:33:25.374 Here's where it provides all the magic, relative panel plus adaptive triggers, 00:33:25.661 --> 00:33:27.194 plus the Visual State Manager. 00:33:27.413 --> 00:33:30.456 And that gives you all that richness that you've seen in all these different user 00:33:30.620 --> 00:33:31.995 experiences over the last few days. 00:33:32.622 --> 00:33:37.760 So now that we've taken the photos, and we have a bunch of those that we've 00:33:38.002 --> 00:33:41.963 collected, it's really about arranging those photos and deciding you want to 00:33:42.080 --> 00:33:43.738 create storyboards for different rooms. 00:33:43.982 --> 00:33:47.811 Now, when I think about going and creating a storyboard and kind of 00:33:48.181 --> 00:33:52.390 manipulating lots of different pictures and photos, I think a desktop is a much 00:33:52.688 --> 00:33:57.286 better device for doing that, use a mouse, my keyboard, maybe to type things. 00:33:57.449 --> 00:33:58.858 So I'm going to move over to my desktop 00:34:00.710 --> 00:34:03.972 and show you the same exact application running here. 00:34:04.812 --> 00:34:08.666 And what you'll notice in this organized phase, first of all, is we're using a pivot. 00:34:09.111 --> 00:34:11.998 This is, as David said, we've added pivot now to the desktop. 00:34:12.108 --> 00:34:15.832 This is a great user control to be able to view the rooms and then be able to 00:34:16.078 --> 00:34:18.228 assign different things to different rooms, so I can pivot between each of the 00:34:18.365 --> 00:34:21.710 rooms, and I start to create storyboards for each of these, and it works very 00:34:21.825 --> 00:34:22.988 nicely with touch, as well. 00:34:23.292 --> 00:34:27.410 And then I think the right way to do this is really to use a drag-and-drop metaphor, 00:34:27.850 --> 00:34:29.764 where I'll go to my living room, and I'll say, you know what? 00:34:30.056 --> 00:34:32.564 I don't really like this, I'm going to remove that, so I use right-click. 00:34:32.805 --> 00:34:35.476 That works natural. Then I'll just drag in another one 00:34:35.722 --> 00:34:38.368 and say, you know, I want this instead. So here are the floors that I think would 00:34:38.557 --> 00:34:43.477 be great for it. I'll go to my master bedroom, say I think fireplace. 00:34:43.727 --> 00:34:46.370 Don't you think I should have a fireplace in my master bedroom? 00:34:46.734 --> 00:34:49.166 I think it would be great to have a fireplace in my master bedroom. 00:34:49.327 --> 00:34:50.879 So I drag that in there, as well. 00:34:51.133 --> 00:34:56.192 So now I have -- able to basically add a few things here, but on the desktop, 00:34:56.523 --> 00:34:58.361 it's not just about the single application. 00:34:58.521 --> 00:35:02.728 It's about using multiple applications and really multiple windows, like David said. 00:35:03.175 --> 00:35:08.078 So I'm also able to support drag and drop between applications, so here we have the 00:35:08.193 --> 00:35:12.288 photo application that's built in, but you could also use the File Explorer. 00:35:12.559 --> 00:35:17.680 So if you had something on an SD card, or if you had a WPF or Win32 application, 00:35:17.958 --> 00:35:22.071 all of those would work as well. So we can now drag in, and this will go 00:35:22.367 --> 00:35:26.482 take this bitmap that I had, or this JPEG I think it is, and it will go and import 00:35:26.482 --> 00:35:28.594 it in here and add it to the filmstrip at the bottom. 00:35:28.850 --> 00:35:31.761 It allows me to do some of the palate that I'm choosing from for my rooms. 00:35:32.098 --> 00:35:35.456 So now I'm going to go back to Visual Studio and show you how I implemented all 00:35:35.878 --> 00:35:37.338 of that drag-and-drop functionality. 00:35:41.456 --> 00:35:45.158 So I'm going to move over to my organize page, and this is the code behind that 00:35:45.405 --> 00:35:49.399 I've got, so I have a handler called grid drop, so whenever I drop it onto that 00:35:49.536 --> 00:35:51.210 grid, this fires. 00:35:51.450 --> 00:35:54.206 It's the same exact function that fires whether or not I've dragged it from with 00:35:54.532 --> 00:35:57.202 inside the app or from outside, from a different application. 00:35:57.685 --> 00:36:00.206 But I get different content. When I dragged it from outside the 00:36:00.538 --> 00:36:05.046 application, I got a bitmap or I got a JPEG or some sort of image format. 00:36:05.374 --> 00:36:09.011 But when I dragged it from inside, I had already a furniture that I had defined in 00:36:09.256 --> 00:36:10.532 my collection there. 00:36:10.937 --> 00:36:12.689 So I detect those in two different ways. 00:36:12.854 --> 00:36:16.463 So when it comes in here, it's one of the -- it's a storage item. 00:36:16.618 --> 00:36:20.969 Then I'm going to know it came from outside. So let me drop in the code here and show 00:36:21.123 --> 00:36:25.859 you how I add it to the list of furnitures. So I go and find if there are any items, 00:36:26.252 --> 00:36:29.283 and then I go and I get the bitmap. I import it. 00:36:29.411 --> 00:36:32.010 So whatever format. We have a bunch of decoders built into 00:36:32.290 --> 00:36:35.053 the platform, and they will automatically decode it. 00:36:35.285 --> 00:36:38.965 Then I have three different attributes on every furniture item, which is a name, 00:36:39.254 --> 00:36:41.323 a description, and then I have this ID that 00:36:41.576 --> 00:36:44.528 I have, and it gives it a unique identifier for every piece of furniture, 00:36:44.734 --> 00:36:48.100 and I actually use that later on so that I don't have more than one copy, 00:36:48.452 --> 00:36:50.260 because nobody wants more than one copy of everything. 00:36:50.503 --> 00:36:53.858 And then I set the image, and then all I do is add it to the list of captured images. 00:36:54.624 --> 00:36:58.888 If I actually am dragging it from within the application -- I have some code here 00:36:59.050 --> 00:37:03.091 I've written. Here, I pass -- the way I do that is I 00:37:03.323 --> 00:37:07.456 reference just the unique identifier of the furniture, and then I want to assign 00:37:07.655 --> 00:37:09.593 it to a specific room in the pivot. 00:37:09.976 --> 00:37:13.377 So I get that ID. It was passed in during the drop operation. 00:37:13.724 --> 00:37:18.927 Then I go, ask which of the pivot tabs -- and those are the three rooms that I was 00:37:19.057 --> 00:37:22.409 having remodeled -- it was in. And then I just add it to the list of 00:37:22.505 --> 00:37:24.811 captured images for that room. 00:37:25.055 --> 00:37:28.578 So very straightforward. I was able to use drag and drop between 00:37:28.689 --> 00:37:32.326 applications or within my applications as a great way to build an experience that 00:37:32.686 --> 00:37:34.530 felt natural for someone using a mouse. 00:37:35.656 --> 00:37:40.474 So now, once we kind of get through this phase of we've kind of assigned painted 00:37:40.605 --> 00:37:45.253 storyboards for each room, what we also like to do is kind of do a slideshow. 00:37:45.519 --> 00:37:48.812 How do I get a feeling for what that room's going to feel like and the 00:37:48.915 --> 00:37:52.259 characteristics of that room? So here, I'm going to go back and get 00:37:52.445 --> 00:37:54.606 this side. We have a slideshow mode, where if I hit 00:37:54.868 --> 00:37:56.975 this, it will cycle through all the photos 00:37:57.262 --> 00:38:01.606 of the captured images for that room, so I can get a feel for the aesthetic and say, 00:38:01.759 --> 00:38:03.256 yes, do they all match or do they not match? 00:38:03.578 --> 00:38:08.492 While this is great to do for my room, I think it's great also for this customer. 00:38:08.643 --> 00:38:11.126 I was thinking, well, if they could build some kiosks and have some of these 00:38:11.405 --> 00:38:15.654 pre-done, some area of their showroom that says, okay, here's the kitchens, 00:38:15.932 --> 00:38:19.276 over here, here's the living room, or here's a different type of aesthetic, 00:38:19.400 --> 00:38:22.490 like traditional, they can have these dedicated kiosks. 00:38:22.781 --> 00:38:25.977 That would kind of advertise some of the ways that their aesthetics works 00:38:26.254 --> 00:38:29.811 and attract more of their customers and excite them and give them maybe some ideas. 00:38:30.182 --> 00:38:33.606 And I think the best way to do that is with some dedicated devices. 00:38:34.090 --> 00:38:37.656 So I suggested we use a Raspberry Pi 2 00:38:38.141 --> 00:38:42.534 with Windows IoT Core running on it, because I can run the exact same application. 00:38:42.828 --> 00:38:48.366 What you see here is a Raspberry Pi, and I have a proximity sensor hooked up to it, 00:38:48.794 --> 00:38:51.049 so it can tell if someone walks over to the room. 00:38:51.201 --> 00:38:53.462 I can give them a custom experience as they do that. 00:38:53.601 --> 00:38:56.249 So let me show you here running, because I have one running here. 00:38:56.379 --> 00:39:00.394 So when I walk in front of it, you'll see that it will detect me. 00:39:00.569 --> 00:39:02.629 I'm here. And it will start the slideshow, and it 00:39:02.852 --> 00:39:06.524 will start rotating through several of the images. And then as I move closer, it'll say, 00:39:06.682 --> 00:39:09.404 well, I know they're closer, so I'll get denser, and even closer, it will get even 00:39:09.580 --> 00:39:12.825 more dense. When I move away -- good, I'm glad it worked. 00:39:13.778 --> 00:39:19.653 Those things are -- it's always nice when they do. 00:39:20.141 --> 00:39:22.302 And it will just turn off when I'm not in front of it anymore. 00:39:22.533 --> 00:39:24.647 So that would be great. And the nice thing about it is, 00:39:24.901 --> 00:39:28.225 since it's the Universal Windows Platform, it runs on those devices, as well. 00:39:28.604 --> 00:39:32.337 So let me hop over and show you the code that I wrote to talk to that sensor. 00:39:34.673 --> 00:39:37.978 So let me move over to my sensor code here. 00:39:38.453 --> 00:39:42.597 And so that's a little bit of special code there that, in the past, if you were 00:39:43.025 --> 00:39:46.344 running on anything other than Windows 10, you would have had to go write a driver 00:39:46.697 --> 00:39:49.702 and you'd have to talk to this special bus and add a bunch of support. 00:39:49.933 --> 00:39:54.021 It was a fairly complex amount of code. But with Windows 10, we now directly 00:39:54.299 --> 00:39:56.727 support talking to these various different buses. 00:39:57.098 --> 00:40:02.401 And so here, I'm talking to an I2C bus. And so what I need to do is in my 00:40:02.626 --> 00:40:05.226 function here, the first thing I'm going to do is enumerate the devices that are on 00:40:05.454 --> 00:40:07.896 the bus. And then I will establish a connection, 00:40:08.131 --> 00:40:10.052 and a lot of this information comes from the device. 00:40:10.344 --> 00:40:11.297 Got to read the manual. 00:40:11.666 --> 00:40:14.535 You can try and guess the parameters here, but typically, read the manual and get the 00:40:14.654 --> 00:40:17.905 SDK for it. And then set up a timer that will go 00:40:18.137 --> 00:40:20.853 and sample that periodically. Now, once I have all of those samples, 00:40:21.104 --> 00:40:25.103 then I can go hook up how close I was and how fast it changed. 00:40:25.337 --> 00:40:29.500 And then I can recreate that behavior. Very straightforward, easy code now that 00:40:29.670 --> 00:40:31.453 we have support for all these buses in the 00:40:31.648 --> 00:40:35.249 Universal Windows Platform. 00:40:36.271 --> 00:40:40.646 Now, once we've kind of set up all these things in the showroom, the most important 00:40:40.901 --> 00:40:44.500 thing is the presentation. They will go off, they'll finalize it, 00:40:44.755 --> 00:40:48.103 they'll do some sketches, and they'll come back with a design, and they've got to 00:40:48.216 --> 00:40:50.455 close the deal. They want the best presentation that they 00:40:50.794 --> 00:40:54.569 can possibly set up. So we think the best way to present that 00:40:54.753 --> 00:40:59.850 is on a large-screen TV, because we all watch our home remodeling shows, don't we, right? 00:40:59.968 --> 00:41:02.206 Everybody watch those? And they're always on our TV, so that's 00:41:02.344 --> 00:41:03.910 kind of the best way to go present it. 00:41:04.096 --> 00:41:09.847 So here, I've chosen to show this on an Xbox One running on the TV here that you 00:41:10.102 --> 00:41:12.406 can show on a TV. And so I'm going to run the application, 00:41:12.653 --> 00:41:16.704 and again, because it's part of the Universal Windows family, it can run any 00:41:17.055 --> 00:41:18.995 app that was built to Universal Windows Platform. 00:41:19.353 --> 00:41:22.698 And so I'm going to go into presentation mode here, and here we've got a sketch of 00:41:22.830 --> 00:41:25.345 a 3D model. That's the room, and so you can go present 00:41:25.452 --> 00:41:29.550 that, and they can look around and see the different rooms and whether they like them 00:41:29.736 --> 00:41:32.951 or not, and sure it's a little void right now, because I'm really not very good at 00:41:33.204 --> 00:41:35.819 designing, but a nice, great experience. 00:41:36.071 --> 00:41:39.650 But I was thinking, if you want to be the best one out there, you want the best 00:41:39.812 --> 00:41:43.401 experience for your customers, there's another device in the Windows family that 00:41:43.747 --> 00:41:47.807 will provide an even better experience? Can you join me in welcoming Felicia from the 00:41:48.226 --> 00:41:53.938 HoloLens team out here to show you this same exact app running on the HoloLens? 00:41:59.011 --> 00:42:04.503 And so what you'll see here is the same app that we wrote runs on the HoloLens, 00:42:04.911 --> 00:42:07.602 because Windows Holographic is Windows 10. 00:42:07.753 --> 00:42:11.990 And so here we have the same exact model, and she is able to look at it, walk around 00:42:12.227 --> 00:42:16.626 it, look at the flooring. 00:42:17.356 --> 00:42:21.631 See, it looks a lot nicer than it did on the Xbox, and kind of just look around 00:42:21.863 --> 00:42:25.414 this and get just a great experience using the same exact application. 00:42:25.918 --> 00:42:29.394 Now, of course, we probably would have gone and built a more customized 00:42:29.641 --> 00:42:33.301 experience, something that she could be immersed in and walk through it, and that 00:42:33.549 --> 00:42:36.462 would be all great, but I wanted to show this exact same app running to show you 00:42:36.742 --> 00:42:40.841 that a very simple app can be moved over to the HoloLens just as easy as any app 00:42:41.027 --> 00:42:43.306 can move among any of the device families. Thank you, Felicia. 00:42:43.502 --> 00:42:44.277 I appreciate it. 00:42:47.838 --> 00:42:50.815 So what I'm going to show you is the actual code that we wrote to make that 00:42:51.036 --> 00:42:54.633 work on the HoloLens versus what we had on the Xbox. 00:42:54.865 --> 00:42:57.862 So before I jump into that, I want to explain a little bit about how the camera works. 00:42:58.358 --> 00:43:03.550 In HoloLens, the camera moves all the time. Whenever the person's head moves or tilts 00:43:04.037 --> 00:43:07.959 or they move around, you must keep up the camera, or the object will not look correct. 00:43:08.356 --> 00:43:11.431 In the past, the camera has been controlled by the developer or the 00:43:11.558 --> 00:43:14.499 programmer in most cases. But here, it's controlled by the end 00:43:14.636 --> 00:43:19.503 user, and so it's also a stereoscopic camera. You have left and right eye, whereas on 00:43:19.757 --> 00:43:23.104 the Xbox, the flat 2D. So the camera on the Xbox, the 3D model 00:43:23.356 --> 00:43:27.558 is the same -- it's the same exact 3D model. But the camera was a fixed 2D camera that 00:43:27.956 --> 00:43:32.143 would take it, create an image from the three-dimensional object and then show 00:43:32.261 --> 00:43:35.073 that on the screen. On here, we're going to have to keep 00:43:35.352 --> 00:43:39.353 updating it with the left and right eye and update the parameters. 00:43:39.587 --> 00:43:43.182 So simply on each update -- this update gets called whenever she moved or looked 00:43:43.324 --> 00:43:45.727 around or tilted. And I need to go update the left and the 00:43:45.849 --> 00:43:49.731 right eye, so I'll read those from the HoloLens, and I'll update those parameters 00:43:50.041 --> 00:43:52.589 in my rendering loop. And then I need to go update the rendering 00:43:52.716 --> 00:43:56.583 resources and the shaders based on those new camera positions. 00:43:56.733 --> 00:43:59.598 So this goes and now creates an interactive experience where she can walk 00:43:59.868 --> 00:44:03.132 around and see and interact with that model. So fairly straightforward for a very 00:44:03.265 --> 00:44:05.835 simple application. Obviously, for some of the more advanced 00:44:05.943 --> 00:44:08.993 stuff we saw yesterday, it's a little bit more of an investment in code, but then 00:44:09.145 --> 00:44:10.585 you get a lot more benefit for the customer. 00:44:10.931 --> 00:44:14.961 So I've kind of gone through a lot of devices and shown you the same exact app 00:44:15.194 --> 00:44:19.960 running on all of them, but the goal really being that it was centered around 00:44:20.094 --> 00:44:22.566 the user. They were able to use that app with the 00:44:22.712 --> 00:44:26.129 right device that made sense for the task and the part of the operation that they 00:44:26.237 --> 00:44:28.591 were doing. And they can provide -- this company can 00:44:28.882 --> 00:44:30.332 provide -- a great experience for their 00:44:30.642 --> 00:44:34.731 customers and really could differentiate versus the competition. 00:44:35.083 --> 00:44:36.033 Thanks. Back to you, Dave. 00:44:43.095 --> 00:44:46.343 >> Great job, Kevin. One app across all the devices. 00:44:46.586 --> 00:44:47.656 It's really pretty remarkable. 00:44:49.291 --> 00:44:54.588 Visual Studio 2015 makes it easier not just to develop your apps but to debug 00:44:54.791 --> 00:44:55.647 them, as well. 00:44:56.582 --> 00:45:02.482 The improved XAML Designer makes it easy to review and adapt your UI for each device. 00:45:03.330 --> 00:45:08.846 You can jump between phone, desktop and other views of your same XAML layout in line. 00:45:10.409 --> 00:45:13.443 The new live visual tree lets you select, 00:45:13.813 --> 00:45:17.987 inspect and edit elements in real time in your running app. 00:45:18.782 --> 00:45:22.990 It's like browser F12 tools for app development. It's so cool. 00:45:24.090 --> 00:45:27.730 Visual Studio pushes the state of the art at diagnosing performance issues by 00:45:28.038 --> 00:45:31.790 allowing you to identify performance issues without even leaving the debugger. 00:45:32.891 --> 00:45:36.134 You can see how long code takes to execute through perf tips. 00:45:36.526 --> 00:45:40.254 You can visualize your app's CPU and memory utilization through the CPU 00:45:40.408 --> 00:45:41.553 and memory graphs. 00:45:42.039 --> 00:45:45.641 All this without having to switch and run a separate profiling tool. 00:45:46.614 --> 00:45:51.482 Windows 10 makes it simple to distribute a Universal Windows app and offers more 00:45:51.590 --> 00:45:56.837 powerful insights into your app's usage. There's one submission process for all devices. 00:45:57.302 --> 00:46:00.392 There's one set of certification policies to go through. 00:46:00.872 --> 00:46:03.327 No other platform makes it this easy to 00:46:03.574 --> 00:46:07.462 distribute your app to such a broad range of devices. 00:46:09.032 --> 00:46:13.274 New features from the store let you analyze your usage and monetize your apps. 00:46:14.236 --> 00:46:18.785 App insights are integrated right into your apps and dev account, enabling you to 00:46:19.030 --> 00:46:23.117 answer questions like, how many users are getting stuck in a level of a game? 00:46:23.974 --> 00:46:26.553 How is a specific feature actually being used? 00:46:27.890 --> 00:46:32.587 Campaign attribution lets you compare the success of your app marketing strategies 00:46:32.833 --> 00:46:37.442 and get your data within a few hours, so you can adjust your strategy in real time. 00:46:40.122 --> 00:46:43.983 So that's how the Universal Windows Platform on Windows 10 helps you design, 00:46:44.466 --> 00:46:47.524 debug, develop and distribute your app. 00:46:48.773 --> 00:46:52.274 Those of you that have already bet on the Universal Windows Platform can benefit 00:46:52.545 --> 00:46:54.232 from all of that in Windows 10. 00:46:55.572 --> 00:46:59.317 For those of you who don't yet have Universal Windows apps, let's talk about 00:46:59.548 --> 00:47:01.231 how we're going to help you get your code 00:47:01.602 --> 00:47:03.273 to the Universal Windows Platform. 00:47:04.885 --> 00:47:08.783 We all know the best apps are the ones developers have poured their blood, 00:47:08.899 --> 00:47:10.126 sweat and code into. 00:47:11.433 --> 00:47:15.730 The whiteboard sessions working through your app's design, the late nights, 00:47:15.986 --> 00:47:18.435 hunting down and fixing that gnarly bug, 00:47:19.573 --> 00:47:23.828 the excitement and anticipation of pushing out each flight -- it's not just your 00:47:24.033 --> 00:47:29.350 code, but all that has gone into your code, that ultimately makes your app what 00:47:29.582 --> 00:47:30.525 it is. 00:47:31.773 --> 00:47:35.667 At Microsoft, we live and breathe developers. 00:47:36.177 --> 00:47:37.956 We recognize this fundamental truth -- 00:47:38.326 --> 00:47:41.526 your code is the soul of your app, 00:47:42.631 --> 00:47:44.100 and that's why we're so passionate about 00:47:44.486 --> 00:47:49.559 building Universal Windows Platform bridges that let you start with your code, 00:47:50.285 --> 00:47:54.960 no matter where it was born, and bring it to the Universal Windows Platform and store. 00:47:57.004 --> 00:48:00.179 The first bridge that we're going to talk about today is how you can bring your web 00:48:00.433 --> 00:48:01.635 code to Windows 10. 00:48:02.284 --> 00:48:04.129 There's a new web rendering engine, 00:48:04.368 --> 00:48:07.992 Microsoft EdgeHTML, that's built for the modern web. 00:48:08.838 --> 00:48:13.242 It powers both the new Microsoft Edge Browser that Joe showed yesterday, as well 00:48:13.552 --> 00:48:15.526 as web content in Universal Windows apps. 00:48:16.977 --> 00:48:21.671 Microsoft Edge leaves behind all of the Internet Explorer behaviors -- all of the 00:48:22.120 --> 00:48:25.391 behaviors specific to Internet Explorer that have been built up in the last 20 years. 00:48:26.726 --> 00:48:29.123 And it has over 4,200 interoperability 00:48:29.376 --> 00:48:34.126 improvements to ensure that the web just works. 00:48:35.320 --> 00:48:37.247 For developers, this means less time 00:48:37.638 --> 00:48:40.267 fixing up your web apps to work across browsers. 00:48:43.203 --> 00:48:49.242 Microsoft EdgeHTML also features over 45 new web standards, including HTTP2, 00:48:49.695 --> 00:48:52.354 Web Audio and a bunch more that you can see here. 00:48:53.135 --> 00:48:54.613 This is just a snapshot. 00:48:54.937 --> 00:48:58.361 We've been doing our Microsoft EdgeHTML planning in the open, with new features 00:48:58.617 --> 00:49:01.284 coming into Windows Insider builds each month. 00:49:03.096 --> 00:49:04.527 While improving interoperability 00:49:04.788 --> 00:49:10.331 and adding new features to EdgeHTML, we didn't forget to pay attention to performance. 00:49:11.688 --> 00:49:15.439 On the JetStream benchmark created by Apple, which simulates complex web 00:49:15.684 --> 00:49:20.443 applications, Edge HTML is over 1.5 times faster than IE 11. 00:49:21.695 --> 00:49:24.929 Beyond that, it's also ahead of the latest 64-bit browsers. 00:49:27.013 --> 00:49:31.844 Next, on Google's Octane benchmark, created by Google, Microsoft Edge is over 00:49:31.962 --> 00:49:33.759 twice as fast as IE 11. 00:49:34.695 --> 00:49:39.814 And again, Microsoft Edge leapfrogs into the lead ahead of the other 64-bit browsers. 00:49:41.466 --> 00:49:45.564 Rest assured that great web performance will continue to be a priority for us. 00:49:47.378 --> 00:49:51.726 Beyond making the web run great on the browser itself, Windows 10 provides 00:49:51.865 --> 00:49:55.040 a powerful new path for sites to become apps. 00:49:56.292 --> 00:50:00.880 With Windows 10, you can reuse your existing website code and create an app 00:50:01.297 --> 00:50:03.816 that points directly to your URL. 00:50:04.863 --> 00:50:08.262 You don't need to rewrite your server code into packaged client code. 00:50:08.736 --> 00:50:12.815 You can access Universal Windows Platform APIs directly from the web code that came 00:50:13.082 --> 00:50:16.460 down from your server, and you can distribute the app you've made in the 00:50:16.689 --> 00:50:17.866 Windows Store. 00:50:18.630 --> 00:50:22.954 This new path is fully integrated into your existing web workflow. 00:50:23.480 --> 00:50:28.176 You can continue to use the same tools, libraries and services that you use today. 00:50:29.275 --> 00:50:33.376 You can continue to deploy to your servers to make updates. 00:50:33.873 --> 00:50:36.747 Kevin, show us how to make a website into an app. 00:50:37.023 --> 00:50:38.463 >> Great. Thanks, Dave. 00:50:38.867 --> 00:50:41.028 I have here a website that I've built, 00:50:41.317 --> 00:50:45.035 and I'm going to show you how I've made it into that hosted web app here. 00:50:45.497 --> 00:50:47.945 First, I'm going to show you the site running in Microsoft Edge. 00:50:48.350 --> 00:50:50.951 It's a Flight Arcade app, flight simulation, kind of cool to play. 00:50:51.521 --> 00:50:53.427 And so here I'll just go start, 00:50:53.915 --> 00:50:59.022 and you'll notice here that it's a beautiful-looking app, great graphics. 00:50:59.467 --> 00:51:02.237 It's also got great audio, because audio's got to be there in every game to 00:51:02.498 --> 00:51:06.262 be cool, but this audio here is actually using that Web Audio, and as you hear me 00:51:06.522 --> 00:51:10.424 throttle up and down, it's actually a single audio file that we have that we're 00:51:10.532 --> 00:51:12.097 using Web Audio to modulate. 00:51:12.280 --> 00:51:16.500 But I'm really not very good. My team told me not to fly, but I've got 00:51:17.144 --> 00:51:21.232 a game controller here, and we've hooked this up to the new game pad support that 00:51:21.470 --> 00:51:28.508 we have in Microsoft EdgeHTML, and so I'm able here to try -- I like to fly fast and loose. 00:51:28.741 --> 00:51:32.751 And not very good. So we'll kill that before I go crashing 00:51:32.867 --> 00:51:33.832 into the Earth. 00:51:34.547 --> 00:51:37.866 Now, there's a couple other features that I wanted to add that really weren't 00:51:38.070 --> 00:51:41.900 available on the web platform. One is integration with Xbox Live, and so 00:51:42.065 --> 00:51:47.498 here, I'm going to launch the same app, the same site, but hosted inside of an application. 00:51:47.748 --> 00:51:51.985 So I've already deployed the application here, and when I run it, it logged me into 00:51:52.143 --> 00:51:54.552 Xbox Live. I had my cached credentials from the 00:51:54.668 --> 00:51:57.152 machine, and it automatically just logged me right in. 00:51:57.374 --> 00:52:02.227 The second feature that I wanted to go add is Cortana integration, the ability to 00:52:02.470 --> 00:52:04.144 launch this application using Cortana. 00:52:04.431 --> 00:52:10.308 So I use Sublime to edit, and I also use Git as my repository. 00:52:10.598 --> 00:52:14.548 So what I'm going to do here is edit the code. I'm going to submit it to Git. 00:52:14.826 --> 00:52:18.752 It will publish it back up to Git, and now I have it automatically deployed to my 00:52:18.888 --> 00:52:21.662 web server. I'm not going to reinstall the application. 00:52:21.807 --> 00:52:24.832 I'm just going to rerurn it, and automatically, it's got all that 00:52:24.950 --> 00:52:28.185 functionality with no update, no ingestion process or anything like that. 00:52:28.675 --> 00:52:34.099 So here, let me uncomment the code. Save. 00:52:41.915 --> 00:52:44.341 Adding Cortana support. 00:52:52.442 --> 00:52:53.989 Wait for that to go work. 00:52:54.881 --> 00:52:59.911 There we go. And now, I can go to Cortana and say, 00:53:00.082 --> 00:53:03.881 Flight Arcade, go to waypoint alpha, bravo. 00:53:08.732 --> 00:53:11.438 >> Okay, going to alpha, bravo in Flight Arcade. 00:53:12.530 --> 00:53:17.869 >> And so you can see here that I was able to launch the application using Cortana, 00:53:18.192 --> 00:53:22.076 go to a certain point that it deep linked inside. I had set that up automatically, and it 00:53:22.191 --> 00:53:26.105 did also that integration with Xbox Live, and of course, my team set it up so that 00:53:26.336 --> 00:53:28.393 by doing nothing, I can get an achievement. 00:53:29.994 --> 00:53:32.733 Everybody want that? Everybody gets an achievement doing nothing. 00:53:34.090 --> 00:53:37.841 So there you go, very simple, able to go do that and continue in the developer 00:53:37.985 --> 00:53:40.347 workflow that web developers are typically used to. 00:53:40.444 --> 00:53:42.112 Back to you, Dave. >> Thanks, Kevin. 00:53:43.787 --> 00:53:45.632 Let's shift gears, to use a Guggs-ism. 00:53:46.592 --> 00:53:50.138 Was everyone here at PDC in 2003 when we introduced WPF? 00:53:52.191 --> 00:53:55.384 How about 2000, when we introduced Windows Forms? 00:53:56.185 --> 00:54:00.504 Still a few hands. Let's go way back, 1992, Win32? 00:54:01.525 --> 00:54:02.729 All right. 00:54:03.370 --> 00:54:08.540 Later this year, Windows 10 will provide a bridge for you to the Universal Windows 00:54:08.680 --> 00:54:09.734 Platform in store. 00:54:10.637 --> 00:54:14.771 Apps written to these classic platform technologies will be able to be packaged 00:54:15.189 --> 00:54:17.456 and deployed with APPX. 00:54:17.835 --> 00:54:22.142 You'll get the same fast, safe, trusted deployment as apps written to the 00:54:22.283 --> 00:54:23.495 Universal Windows Platform. 00:54:24.391 --> 00:54:25.941 You'll be able to distribute these apps 00:54:26.241 --> 00:54:27.789 via the Windows Store. 00:54:28.331 --> 00:54:29.878 You'll be able to call even more of the 00:54:30.134 --> 00:54:35.718 Universal Windows Platform APIs directly from your existing code -- Live Tiles, 00:54:36.073 --> 00:54:38.480 notifications, Cortana, and more. 00:54:39.679 --> 00:54:42.521 Now, I'm going to turn it back over to Kevin to show you how to use the Universal 00:54:42.745 --> 00:54:46.945 Windows Platform APIs from your existing code. 00:54:47.676 --> 00:54:49.942 >> Thanks, Dave. So, of course, since I've written nearly 00:54:50.194 --> 00:54:53.227 all these platforms -- I've shown you web, I've shown you Universal Platform, now, 00:54:53.354 --> 00:54:54.540 I'm going to show you a WPF. 00:54:55.033 --> 00:54:59.138 So let me just run it here, just to give you kind of a look at it. 00:54:59.517 --> 00:55:02.792 It's a very simple application that I've written, but basically, it's a personal 00:55:03.032 --> 00:55:06.984 trainer application that allows me to track information about things that I eat 00:55:07.342 --> 00:55:10.945 and my workouts, my health information, a bunch of different things. 00:55:11.557 --> 00:55:14.338 And what I want to do here is add two specific features. 00:55:14.623 --> 00:55:19.078 One is a very simple feature, the ability to just pop a toast, and I want to show 00:55:19.238 --> 00:55:22.122 you just how simple it is, to show it's not a lot of complex work, just to call 00:55:22.345 --> 00:55:27.996 one of the new Universal APIs from a WPF application. The second is one of our more complex ones. 00:55:28.369 --> 00:55:31.830 We've added a feature called App Services. This is the ability for Universal 00:55:31.988 --> 00:55:35.881 applications to provide services to other applications so they can call them, 00:55:36.346 --> 00:55:40.374 and they can basically work together on different data and different services. 00:55:41.128 --> 00:55:42.546 So the first thing here is when I click 00:55:43.193 --> 00:55:45.595 the launch button, I just want to pop a toast. 00:55:45.877 --> 00:55:48.082 So it's very simple code. It's three lines here. 00:55:48.315 --> 00:55:50.743 First, you'll notice that I had added some namespaces up here. 00:55:50.881 --> 00:55:53.520 Otherwise, it wouldn't compile. These are just the Universal API 00:55:53.738 --> 00:55:58.032 namespaces that I use. Then, for the toast, I go, I get the XML 00:55:58.032 --> 00:56:00.138 for it. I create a new toast, and then I just 00:56:00.275 --> 00:56:02.521 say, show a toast. So let me run this code. 00:56:06.877 --> 00:56:09.230 And so when I click launch, this will pop 00:56:09.400 --> 00:56:11.148 what we call an interactive toast. 00:56:11.327 --> 00:56:15.924 This toast actually allows me to open an application from it, so what I did is I 00:56:16.028 --> 00:56:19.155 popped this here, and the XML said open the Microsoft Health application. 00:56:19.457 --> 00:56:22.123 So I'm going to click on that, and it actually opens this app here. 00:56:22.854 --> 00:56:28.036 This is a sample app that allows you to gather information from the Microsoft 00:56:28.277 --> 00:56:31.516 Band, and I've been wearing that for a while, so it's tracked my heart rate 00:56:31.878 --> 00:56:38.017 and different things that it tracks, and so it has also an app service that we've implemented. 00:56:38.573 --> 00:56:41.397 And so what I want to do is be able to call that app service to get all that 00:56:41.502 --> 00:56:43.078 information from the last week. 00:56:43.277 --> 00:56:47.647 The nice thing about App Services is I don't need this app running to be able to talk to it. 00:56:47.885 --> 00:56:51.573 So I'm going to close the application, and then when this calls it, it will just 00:56:52.060 --> 00:56:53.742 directly connect to the application in the background. 00:56:54.227 --> 00:56:58.179 So here is I'm going to get my click handler for that start button that I had 00:56:58.422 --> 00:57:02.576 in there, and I'm going to pop the service here, talk to the code here, talk to that service. 00:57:02.949 --> 00:57:06.474 So the first thing I do is I establish a connection. I do a strong connection here. 00:57:06.746 --> 00:57:09.555 I want to know that I'm talking specifically to that application. 00:57:09.878 --> 00:57:13.666 So I have a GUID here and an App Service name, and then I just establish the connection. 00:57:14.277 --> 00:57:18.458 The next thing I want to do is send it a message, so I want to send it a message to 00:57:18.578 --> 00:57:23.550 get the past week's data, so I send it this message here, I do an await here, 00:57:23.779 --> 00:57:24.975 I send it asynchronously. 00:57:25.148 --> 00:57:29.770 And then when it's done, it comes back, I update my user interface with the data 00:57:30.026 --> 00:57:34.437 that I got from it, so again, very straightforward and not a huge amount of code. 00:57:36.269 --> 00:57:40.777 So here, I'm going to hit start, it will take a few seconds, not even. 00:57:41.028 --> 00:57:44.032 And then it goes, it talks to that app service in the background, and it's pulled 00:57:44.144 --> 00:57:47.868 that data and integrated it in directly into my WPF application. 00:57:48.124 --> 00:57:50.906 So I've showed that you're able to do very simple things with a few lines of 00:57:51.027 --> 00:57:54.086 code, or you can do one of the more complex things, integrating with these App 00:57:54.288 --> 00:57:57.881 Services a brand-new feature with the platform coming in Windows 10. 00:57:58.149 --> 00:57:59.759 Back to you, Dave. >> Thanks, Kevin. 00:58:05.914 --> 00:58:09.863 Now, let's talk about the bridge we're building to help you take your Android 00:58:10.108 --> 00:58:11.906 code and get it to Windows. 00:58:12.555 --> 00:58:16.530 You can reuse existing Java or C++ code you've written for your Android app. 00:58:16.864 --> 00:58:20.756 You can continue to work in your existing tools, even using Android Studio to 00:58:21.031 --> 00:58:22.262 produce the app for Windows. 00:58:24.551 --> 00:58:28.150 When you make the app for Windows, function calls and even services route to 00:58:28.322 --> 00:58:30.611 deliver the correct Windows experience. 00:58:31.432 --> 00:58:36.301 You can update your code to take advantage of key features of the Universal Windows 00:58:36.557 --> 00:58:37.983 Platform, like Live Tiles. 00:58:38.236 --> 00:58:42.184 You can distribute your app through the Windows Store to phones running Windows 10. 00:58:42.438 --> 00:58:43.615 Let's take a look. 00:58:46.284 --> 00:58:51.544 Yesterday, you saw how you can start with code you wrote for Android to create an 00:58:51.672 --> 00:58:55.945 app for Windows that integrates with Windows shell features like keyboard, 00:58:56.176 --> 00:58:57.779 touch, and navigation. 00:58:58.401 --> 00:59:01.557 Today, I'd like to show you how starting with your Android code you can integrate 00:59:01.830 --> 00:59:04.025 with some more advanced Windows capabilities. 00:59:05.277 --> 00:59:11.373 This app, Lose It, empowers people to live healthier lives and achieve their 00:59:11.670 --> 00:59:14.630 goals through rich experiences and insights about their lifestyle. 00:59:15.077 --> 00:59:18.635 This version of Lose It running on a Windows 10 Phone was built starting with 00:59:18.866 --> 00:59:20.430 the original Android code base. 00:59:21.482 --> 00:59:24.349 All this talk up on stage, it's like exercise for me. 00:59:24.470 --> 00:59:26.223 I'm sweating a lot. It's kind of fun. 00:59:26.582 --> 00:59:27.668 Let me log this exercise. 00:59:33.477 --> 00:59:36.125 You can see I have stage talk as a regular exercise in addition to running 00:59:36.427 --> 00:59:39.179 and bicycling. We're supposed to be up here about 45 minutes. 00:59:39.424 --> 00:59:41.380 We're going to be about that by the time we're done. 00:59:42.824 --> 00:59:46.477 We'll go ahead and save. Now we've got stage talk logged as an exercise. 00:59:47.322 --> 00:59:49.953 Terry and I, sometimes we're in competition with our exercise. 00:59:50.118 --> 00:59:54.554 So I'm going to share this stage talk workout that I did with Terry. 00:59:54.924 --> 00:59:58.615 This app is using the Windows share contract instead of the default Android experience. 01:00:00.227 --> 01:00:02.726 If I were to click into messaging -- I'll not actually send it to him, but you can 01:00:02.980 --> 01:00:05.262 see how it brings up the Windows experience for sharing. 01:00:05.883 --> 01:00:10.598 This is just one example of how Windows code to reuse -- Windows apps that reuse 01:00:10.766 --> 01:00:14.326 Android code and can interact with other Windows apps. 01:00:15.982 --> 01:00:20.642 These apps can also use key Universal Platform APIs like Live Tiles. 01:00:21.778 --> 01:00:24.905 If you look in the upper right on the screen, you can see my start screen is 01:00:25.183 --> 01:00:30.004 a Live Tile for the sample app, Timber. That's Timber with a B as in bravo. 01:00:30.984 --> 01:00:32.660 It occasionally changes state. 01:00:34.782 --> 01:00:38.679 Timber was another Windows app that was also built starting with an Android code base. 01:00:39.208 --> 01:00:42.554 With just a few lines of code, it was able to integrate a Live Tile. 01:00:46.276 --> 01:00:50.145 Apps can also integrate with Microsoft Cloud Services with minimal effort. 01:00:51.129 --> 01:00:54.446 Here, Timber is showing me a map of my favorite national park. 01:00:55.996 --> 01:01:00.672 Maps are one example of cloud services that modern apps want to use. 01:01:01.033 --> 01:01:04.464 It was trivial for the developer to invoke Bing Maps instead of Google Maps when 01:01:04.728 --> 01:01:05.799 running their code on Windows. 01:01:07.651 --> 01:01:11.402 Part of the bridge we're building for Android code will include a rich SDK for 01:01:11.680 --> 01:01:15.883 Microsoft services in Java to make it easier for developers reusing Android code 01:01:16.081 --> 01:01:17.754 to build a successful Windows app. 01:01:18.858 --> 01:01:21.283 And as far as tooling, I'm not going to show you this right now, but there's 01:01:21.431 --> 01:01:25.346 a session later on today where you can learn to reuse your current IDE and workflow to 01:01:25.577 --> 01:01:28.282 build great Windows apps from your Android code. 01:01:36.806 --> 01:01:39.326 That was a big step for us, in case that wasn't obvious. 01:01:40.694 --> 01:01:45.039 Now, let me walk you through the Universal Platform bridge for your iOS 01:01:45.186 --> 01:01:46.534 code as well. 01:01:47.126 --> 01:01:49.483 It lets you import an Xcode project into 01:01:49.657 --> 01:01:53.955 a Visual Studio solution while sharing the exact same source code files across the 01:01:54.197 --> 01:01:55.437 two projects. 01:01:56.400 --> 01:02:00.930 You can work in the familiar, powerful Visual Studio experience to edit, debug, 01:02:01.152 --> 01:02:03.122 profile and ultimately build your Windows app. 01:02:03.784 --> 01:02:07.399 You have full access to the API set the platform provides, including things like 01:02:07.508 --> 01:02:11.929 Cortana, and you can integrate C and C++ code in with the same project. 01:02:13.132 --> 01:02:14.932 You're producing a Universal Windows app 01:02:15.199 --> 01:02:18.331 that you can deploy to the Windows Store. Let's check it out. 01:02:22.779 --> 01:02:26.609 Here we have a directory with the source code files for the Math Dream project that 01:02:26.845 --> 01:02:29.598 Terry showed yesterday. Same exact Objective C files. 01:02:30.410 --> 01:02:34.953 In the command-line interface, we're going to run a little batch file, and what 01:02:35.104 --> 01:02:38.608 this batch file does, it runs the code that's going to do the conversion of the 01:02:38.778 --> 01:02:43.952 Xcode project to the Visual Studio Solution. If you watch in the Explorer window over 01:02:44.156 --> 01:02:47.546 here, a bunch of new files showed up, including the solution file. 01:02:48.512 --> 01:02:52.135 Double-click on it. We'll bring up Visual Studio, and we've 01:02:52.368 --> 01:02:55.139 created this Math Dream project from the Xcode project. 01:02:57.717 --> 01:03:02.890 It still has the original Objective C code from the iOS application, and you can 01:03:02.991 --> 01:03:04.568 see it in the Visual Studio view. 01:03:06.136 --> 01:03:07.683 Simple F5. 01:03:08.421 --> 01:03:09.969 We'll build and run it. 01:03:17.222 --> 01:03:19.860 And for a simple app like that, that's all it takes. 01:03:20.139 --> 01:03:22.017 Here's the app running, the same one you saw yesterday. 01:03:25.314 --> 01:03:28.810 We can add the apples, simple arithmetic application. 01:03:29.424 --> 01:03:32.409 But now, let's take a step and bring this one step further. 01:03:33.988 --> 01:03:38.252 We know that great Windows apps also want to take advantage of all the compelling, 01:03:38.667 --> 01:03:41.720 unique Windows features that we've been talking about in the Universal Windows 01:03:41.943 --> 01:03:43.463 Platform, so let's actually do that. 01:03:46.028 --> 01:03:53.139 First, we're going to add a couple of files that we've already written into the project. 01:03:58.264 --> 01:03:59.696 These two files. 01:04:06.995 --> 01:04:09.817 And then what we're going to do is we're going to make a couple of small 01:04:09.943 --> 01:04:13.411 modifications to this source code. You're not going to enjoy watching me 01:04:13.556 --> 01:04:19.008 type source code, so I'm just going to cut and paste this from the popular Notepad application. 01:04:35.222 --> 01:04:37.488 All right. Now, when we rebuild this project, 01:04:39.819 --> 01:04:42.086 we'll rebuild it with the changes for the new 01:04:42.247 --> 01:04:45.043 interactive notifications. It's the same concept that Microsoft -- 01:04:45.370 --> 01:04:48.637 that Kevin showed with Band, but now we're going to have an interactive toast added 01:04:48.747 --> 01:04:49.840 to this particular application. 01:04:53.971 --> 01:04:55.397 F5 to run it. 01:04:58.084 --> 01:05:02.637 What we've done is, if it you do a math operation that results in 10 -- we like 01:05:02.763 --> 01:05:10.002 the number 10 -- it brings you this interactive toast down at the bottom right. 01:05:10.762 --> 01:05:14.396 And what we can do is the compelling feature of unlocking division inside the 01:05:14.641 --> 01:05:18.683 application, so that we can have all four basic arithmetic operations in the app. 01:05:23.487 --> 01:05:25.171 Six divided by 3 is, in fact, 2. 01:05:27.215 --> 01:05:33.946 So that's how you can add unique Universal Windows Platform capabilities to an 01:05:34.285 --> 01:05:38.890 existing iOS Objective C application. All right, thanks for your attention. 01:05:39.449 --> 01:05:50.841 That was all a look at how you can take your Universal Windows Platform bridges to 01:05:51.142 --> 01:05:55.287 take the code you have today, no matter where it was born, bring it to Windows. 01:05:56.412 --> 01:06:00.036 Then you can build experiences only possible on Windows, across all our family 01:06:00.244 --> 01:06:05.401 of devices, by taking advantage of unique Windows features like Xbox Live Services, 01:06:05.771 --> 01:06:10.393 Cortana, interactive notifications, Live Tiles, adapting and tailoring your UI, 01:06:10.690 --> 01:06:13.794 inking, sensors, holograms and much, much more. 01:06:15.046 --> 01:06:19.491 I'm psyched to see what you all do to change the world using the Universal 01:06:19.715 --> 01:06:22.714 Windows Platform. Thank you very much. 01:06:25.971 --> 01:06:27.294 >> Come this way, David. 01:06:27.537 --> 01:06:29.236 >> There you guys are. >> There you go. 01:06:29.429 --> 01:06:31.241 Great job. That was really good. 01:06:31.719 --> 01:06:34.599 Big hand, those guys cranked it out. That's exactly what we hope you were 01:06:34.770 --> 01:06:37.639 looking for, right stuff? Everybody's awake now? 01:06:38.568 --> 01:06:41.832 All right. The nice thing for us is we get to take 01:06:42.168 --> 01:06:45.595 the work that Kevin showed you and David showed you, and we bring it to life. 01:06:46.091 --> 01:06:50.401 We've already started working with developers building solutions on the technology. 01:06:50.889 --> 01:06:53.888 So you see a slide here of Shazam. Shazam is taking a web experience, 01:06:54.417 --> 01:06:58.238 turned it into an application. It's a nice mix because it uses C++ code 01:06:58.493 --> 01:07:01.400 for doing the music recognition, but then it's all web-based experience. 01:07:01.584 --> 01:07:02.932 We put it into the store. 01:07:03.184 --> 01:07:07.029 They can deep link, so from within the app itself, you can go in and download music. 01:07:07.400 --> 01:07:10.804 It's got links back into the platform itself. From there, there's actually an 01:07:10.965 --> 01:07:13.971 opportunity for monetization through a referral program. 01:07:14.453 --> 01:07:17.966 The Rdio one we saw earlier is also a web-based experience. 01:07:18.279 --> 01:07:22.228 We're starting to bring these things to life. I've got a couple here just to take from 01:07:22.352 --> 01:07:24.875 where Kevin was. First one is one of my favorites. 01:07:25.182 --> 01:07:29.032 I'm a long time Fitbit user, about 2,500 miles a year. 01:07:29.284 --> 01:07:31.567 Here, two things that come out of what Kevin was showing you. 01:07:32.000 --> 01:07:36.402 One, today, if you use Fitbit to connect to the PC, you need to use the Bluetooth 01:07:36.634 --> 01:07:39.203 dongle, so you plug it into the side of the PC, and then when you're working 01:07:39.478 --> 01:07:42.626 and you're moving around, that's how you do the Bluetooth connection. 01:07:42.880 --> 01:07:45.078 Don't need the dongle anymore. Connects automatically. 01:07:45.320 --> 01:07:48.119 Don't have to use that for the PC. Works just like the phone. 01:07:48.603 --> 01:07:51.325 The second thing is building from the ground up, taking advantage of the 01:07:51.644 --> 01:07:52.783 responsive design. 01:07:53.083 --> 01:07:55.773 In this particular case, you'll notice, here's my Control Panel. 01:07:56.000 --> 01:07:59.312 This is my dashboard. If I start to slide down, it gets very 01:07:59.420 --> 01:08:02.806 clean, as I come down, and here sort of tablet form. 01:08:03.144 --> 01:08:07.464 If I go a little smaller here, this is phone UI, and John's got it up and running 01:08:07.802 --> 01:08:10.937 on the phone next to me, and you'll notice -- let's see if they can project that. 01:08:11.307 --> 01:08:15.001 The phone UI looks exactly the same now as what we've done on the desktop. 01:08:15.348 --> 01:08:18.749 So you have this really great ability to take the app and scale it across all of 01:08:18.865 --> 01:08:20.602 the devices without having to do extra work. 01:08:20.850 --> 01:08:24.195 Now, the other side of this particular conversation would be up to the Xbox. 01:08:24.715 --> 01:08:27.990 So if we switch over to the Xbox, I can take that same dashboard. 01:08:28.154 --> 01:08:32.524 >> And here I'm going right into the Xbox, and notice, I'm using a pretty radically 01:08:32.807 --> 01:08:34.112 different way to navigate around this, 01:08:34.397 --> 01:08:37.627 but I get all the features through that Universal app. 01:08:37.828 --> 01:08:39.916 >> I think there's a magic of code here, because a couple of your guys have more 01:08:40.029 --> 01:08:42.666 steps now than I do, which there's just something wrong on that one, as well. 01:08:42.911 --> 01:08:44.467 I've been pacing backstage for hours. 01:08:45.277 --> 01:08:49.322 So here's a good example on the consumer side. Fitbit, we want to thank them for letting 01:08:49.423 --> 01:08:51.851 us show it. Now, on the other side of the coin, 01:08:52.001 --> 01:08:55.344 on the commercial side, this is an application from Siemens. 01:08:55.599 --> 01:08:58.628 This is called JT2Go. And if you think about design, this 01:08:58.774 --> 01:09:00.222 is the designer for physicists. 01:09:00.474 --> 01:09:04.188 So if you want to build bridges, if you want to build ships or, in this case, 01:09:04.411 --> 01:09:08.198 if you want to build rocket ships, this is literally the toolset that they use, 01:09:08.426 --> 01:09:09.428 very high end. 01:09:09.716 --> 01:09:12.006 It has all the physics applied, along with the math. 01:09:12.261 --> 01:09:14.547 Now, the cool thing is I have it here on the Surface Pro. 01:09:14.917 --> 01:09:18.830 I can use it with touch and interaction. But same application, if we go back over 01:09:19.065 --> 01:09:24.661 to the Surface Hub, now we see JT2Go up on the big screen, and here, we can 01:09:24.915 --> 01:09:28.945 collaborate, work on it together. John can use the smooth inking. 01:09:29.431 --> 01:09:30.870 >> So this is the actual rocket from 01:09:31.370 --> 01:09:35.340 orbit, and you notice that you can see pretty amazing stuff like this connection here. 01:09:35.596 --> 01:09:37.664 Maybe that's a little too close, something like that. 01:09:37.953 --> 01:09:39.022 >> Yes, and we could work on it together. 01:09:40.111 --> 01:09:43.316 This pen has decided to give up on me. 01:09:43.724 --> 01:09:46.347 Well, we could work on it together, but we're going to use your pen for now. 01:09:48.912 --> 01:09:50.863 It's all real code. That's the challenge. 01:09:51.594 --> 01:09:54.267 Go ahead. >> So one of the great things that we can 01:09:54.519 --> 01:09:59.663 do in addition to what's happening here is we can bring in, for example, experts. 01:10:00.030 --> 01:10:04.111 So if Guggs and I, who happen to not be rocket scientists, wanted to understand 01:10:04.354 --> 01:10:06.863 whether we've got a problem here, just bring up Skype, have the conversation, 01:10:06.992 --> 01:10:09.580 and off we get. >> And Skype's built in, so take advantage 01:10:09.803 --> 01:10:11.658 of all the different services with the 01:10:12.167 --> 01:10:15.430 applications, and so good set of conversations. There's a whole set of partners we're 01:10:15.571 --> 01:10:19.605 working with this and ISVs. I wanted to build off of where David 01:10:19.717 --> 01:10:22.861 and Kevin was, and link into one of the conversations we do have on a pretty 01:10:23.033 --> 01:10:25.195 regular basis. We often get questions -- they showed us 01:10:25.462 --> 01:10:28.750 how to build a Universal Windows app from scratch, how to bridge in. 01:10:29.279 --> 01:10:32.743 Sometimes, we'll get the question, if I'm starting from scratch and I want to build 01:10:33.022 --> 01:10:37.038 an application that runs across all the devices, what's a good way to do that? 01:10:37.280 --> 01:10:40.266 And so there are multiple plans or ways to do that. 01:10:40.517 --> 01:10:44.070 Last year, no, two years ago, we showed Unity for the gaming platform. 01:10:44.447 --> 01:10:47.997 Last year, we took what Kevin had built the day before, on day one, and used -- it 01:10:48.115 --> 01:10:50.831 was a .NET application. We used Xamarin to bring it over to iOS 01:10:50.988 --> 01:10:54.361 and Android, so this year, we thought we'd pick up on the web theme, and if we wanted 01:10:54.464 --> 01:10:57.517 to build a web experience or a web application that runs across all devices, 01:10:57.634 --> 01:11:01.060 how could we do that? >> So as you saw earlier today with 01:11:01.360 --> 01:11:07.402 Kevin's demo, with the flight game, what Shazam and others are doing, we're able to 01:11:07.568 --> 01:11:11.227 take a web application. Here's a web application, very simple, 01:11:11.672 --> 01:11:15.398 using modern frameworks, and what we'd like to be able to do is very quickly 01:11:15.515 --> 01:11:16.627 generate a Windows app. 01:11:17.123 --> 01:11:22.811 The key to this is a new format from the W3C called a manifest, and I've got the 01:11:22.917 --> 01:11:25.850 manifest for this particular app sitting over here. 01:11:26.064 --> 01:11:29.922 And really, it's the simplest thing. It's just a JSON description of where the 01:11:30.036 --> 01:11:34.472 app's located out on the web, whether you want this thing to be full screen, icons, 01:11:34.673 --> 01:11:36.725 various things like that -- again, all standardized. 01:11:37.323 --> 01:11:41.923 We load that into the Windows Store by giving it the URL to the app and it goes 01:11:42.213 --> 01:11:46.207 and produces the Windows application, and we saw great examples of Kevin showing 01:11:46.475 --> 01:11:49.643 that up and running. Now, wouldn't it be cool if you could take 01:11:49.876 --> 01:11:54.762 this same manifest, that same approach, but also create that same kind of 01:11:54.890 --> 01:11:58.839 experience but on, say, iOS or Android or other platforms? 01:11:59.013 --> 01:12:02.479 So we're pretty excited to announce a new capability today. 01:12:03.262 --> 01:12:08.764 If you're a web developer, chances are, you know JavaScript, and chances are, 01:12:08.879 --> 01:12:12.641 you're probably using Node on the back end. So we've added, and we're announcing 01:12:12.877 --> 01:12:16.879 a new thing -- you're probably using the Node Package Manager, and if you typed npm 01:12:17.045 --> 01:12:25.439 install -g manifoldjs, that will go download this toolset that I'm about to show. 01:12:25.690 --> 01:12:28.322 Now, I've already done that, so let me just actually have this work. 01:12:28.489 --> 01:12:34.083 Here's manifoljs. There's the URL to my app. 01:12:34.526 --> 01:12:36.478 I'm just going to turn on logging with 01:12:36.981 --> 01:12:40.240 debug level so you can see what happens. So there it is. 01:12:40.482 --> 01:12:44.925 It's reached out across the web, it's picked up that manifest, and it's using it 01:12:45.209 --> 01:12:48.205 to go generate a bunch of applications. In fact, let's go see what it's done. 01:12:48.436 --> 01:12:52.812 So we're going to down into the Shifter application, and I'm just going to launch 01:12:53.128 --> 01:12:57.330 Explorer here, so you can kind of see what's going on, and notice, it's now 01:12:57.562 --> 01:13:03.275 automatically generated apps for Android, for Chrome, for Cordova, Firefox, iOS, 01:13:03.530 --> 01:13:04.809 web, you name it. 01:13:05.130 --> 01:13:10.525 These are now ready to go as apps. So let's dig into the Android example. 01:13:10.810 --> 01:13:14.759 I'm going to launch this. That's going to bring up Android Studio, 01:13:15.612 --> 01:13:20.521 and it's going to use a very powerful piece of open-source technology, Cordova, 01:13:21.059 --> 01:13:26.192 so that it can use that set of web technologies as an Android application, 01:13:26.437 --> 01:13:27.880 and we can see that coming up now. 01:13:28.383 --> 01:13:30.641 So here's our Android application up and running. 01:13:31.130 --> 01:13:34.240 We could go through this, look at all the Java and so on. 01:13:34.482 --> 01:13:36.608 Now, I'm just going to hit run, and that 01:13:37.094 --> 01:13:40.935 would go off and compile it, but Steve, you have the apps up here and running. 01:13:41.190 --> 01:13:42.809 >> Let me bring that back up. 01:13:43.211 --> 01:13:45.758 So I have two devices here. I have an Android device and I have an 01:13:45.884 --> 01:13:49.886 iOS device, and we have that same Shifter app basically built and running on both of 01:13:50.156 --> 01:13:53.569 the platforms, straight from what John did, real easy to work with, and of 01:13:53.815 --> 01:13:56.566 course, we could have some fun picking our favorite frameworks here if we'd like. 01:13:57.087 --> 01:14:01.268 >> Now, this might almost seem like magic, but it's amazing how many partners we're 01:14:01.426 --> 01:14:03.935 working with that have actually started this way. 01:14:04.214 --> 01:14:06.131 >> And in fact, the conversations -- we talked about Salesforce before. 01:14:06.409 --> 01:14:08.732 That's exactly how they operate. They basically do their dynamic content 01:14:09.010 --> 01:14:12.610 up on the web with HTML5 and JavaScript, then do native code locally. 01:14:12.980 --> 01:14:16.164 They use Cordova, and that's how we started the project together in terms of 01:14:16.453 --> 01:14:17.537 building the Windows 10 application. 01:14:17.894 --> 01:14:21.858 >> So what I'm doing here is I'm closing down everything, and I wanted to talk for 01:14:22.129 --> 01:14:25.367 a second about one of the challenges when you start building apps this way. 01:14:25.708 --> 01:14:29.528 It's fantastic to be able to generate these apps and put them out there, but we 01:14:30.015 --> 01:14:34.214 lose one of the most powerful features of the web, which is the fantastic debugging 01:14:34.483 --> 01:14:36.521 capabilities that are built into the browser. 01:14:36.932 --> 01:14:41.640 So we've got a cool additional project that we wanted to announce today called Vorlon. 01:14:41.814 --> 01:14:46.883 Now, what Vorlon does is it's a package that you add to your web app, and it 01:14:46.981 --> 01:14:51.271 reaches out across the web to a Vorlon server running in Node.js, and as you can 01:14:51.467 --> 01:14:54.245 see, this is connected to our Android and iPhone app. 01:14:54.476 --> 01:14:57.959 I'm going to go start up the Windows version of the app, and as you can see, 01:14:58.330 --> 01:15:00.886 it's now popped that one up. I can select it. 01:15:01.129 --> 01:15:06.364 Not only that, I can reach out anywhere. This app could be running in Australia. 01:15:07.092 --> 01:15:08.528 We would see that thing pop up. 01:15:08.780 --> 01:15:12.578 I can identify it, just to make sure that I'm thinking about the right one, so we 01:15:12.959 --> 01:15:14.286 have the ability to interact with it. 01:15:14.657 --> 01:15:18.011 I can look around at the DOM. I have a lot of capabilities here. 01:15:18.413 --> 01:15:23.284 In fact, I can go and do things like write to the console, all that stuff. 01:15:23.698 --> 01:15:27.209 I'm just going to do the simplest thing. I'm just going to call a function that 01:15:27.564 --> 01:15:31.164 would be equivalent to if I clicked that I liked one of these frameworks, and I'm 01:15:31.410 --> 01:15:32.480 just running that remotely. 01:15:32.811 --> 01:15:34.414 Notice, I reached out across the web, 01:15:34.659 --> 01:15:39.684 injected that into a running Windows application, running iPhone application, 01:15:40.102 --> 01:15:44.212 all nice, new ways to develop using the web technologies. 01:15:44.815 --> 01:15:46.117 >> Slick stuff, John. 01:15:46.392 --> 01:15:49.285 Then the other one then is obviously the Asm -- let's give John a hand for that 01:15:49.438 --> 01:15:50.285 and the team. 01:15:50.936 --> 01:15:56.650 >> So the whole web space is just the level of innovation going on there 01:15:56.765 --> 01:16:01.130 is incredible, and we're seeing stuff like the Babylon work that we announced last time. 01:16:01.335 --> 01:16:05.728 Another project that we wanted to announce today is the Asm.js project, 01:16:06.133 --> 01:16:10.485 which gives us a way to write down in JavaScript a subset of that that can be 01:16:10.684 --> 01:16:12.681 easily compiled and provide high performance. 01:16:12.991 --> 01:16:17.815 So, again, we're very supportive of both this cross-platform way to do things but 01:16:18.109 --> 01:16:21.068 also bringing in open source into the whole approach. 01:16:21.583 --> 01:16:24.152 >> Thank you, John. And I think in terms of open sourcing 01:16:24.407 --> 01:16:26.738 and sharing code, we've done a lot of work with GitHub. 01:16:26.887 --> 01:16:29.174 John says all the projects here will be up on GitHub. 01:16:29.413 --> 01:16:31.485 One of the things they've done is they've worked really well with us on students, 01:16:31.616 --> 01:16:33.776 a bunch of areas. I'm happy today to announce we now have 01:16:33.882 --> 01:16:37.062 GitHub Enterprise for Azure up and running. That means if you're behind the firewall 01:16:37.216 --> 01:16:40.022 at a commercial entity, you can use your Active Directory for sharing code. 01:16:40.393 --> 01:16:43.484 We can scale to all of the developers out there and work together on Azure. 01:16:43.623 --> 01:16:46.132 So again, just trying to contribute into the community for a bunch of different 01:16:46.374 --> 01:16:49.025 areas, make sure we can share code and projects and work together. 01:16:49.786 --> 01:16:53.051 Now, I'm not going to use switching gears, because that's been overused. 01:16:53.288 --> 01:16:56.083 Why don't we go to Python-esque with a now and for something completely different? 01:16:56.336 --> 01:16:57.171 We'll go that way. 01:16:57.671 --> 01:17:00.654 One of the conversations that I thought Scott did a good job on yesterday was 01:17:00.933 --> 01:17:04.847 around data, and scaling out data, security for data, Data Lake. 01:17:05.136 --> 01:17:07.287 There was just a whole bunch of parts to that, and we end up with a lot of 01:17:07.619 --> 01:17:09.826 conversations where people want to talk about data as a platform. 01:17:10.210 --> 01:17:12.927 Data as a platform is incredibly interesting, but you have to be able to do 01:17:13.066 --> 01:17:15.561 something with it. The ability to understand and act on data 01:17:15.898 --> 01:17:19.220 and bring it to life, that's where the magic is in a lot of the cases. 01:17:19.383 --> 01:17:20.973 There's nobody better to talk about that 01:17:21.229 --> 01:17:25.965 than our own Joseph Sirosh, who leads the Machine Learning Group, an expert in the area. 01:17:26.279 --> 01:17:28.970 So I thought we'd have him come out and sort of give us a walk around the park, 01:17:29.224 --> 01:17:32.934 in this case, building on top of the data with machine learning, real-time analytics. 01:17:33.298 --> 01:17:34.730 Let's see where Joseph's coming out from. 01:17:35.621 --> 01:17:37.528 >> All right. >> There he is. 01:17:38.129 --> 01:17:39.674 >> Joseph, welcome to the stage. 01:17:40.135 --> 01:17:42.279 How are you? >> Thank you, Steve. 01:17:42.416 --> 01:17:43.853 >> Have some fun. We'll be back in a little while. 01:17:44.033 --> 01:17:44.832 >> Yes, thanks. 01:17:46.939 --> 01:17:48.997 Software is eating the world. 01:17:49.822 --> 01:17:56.358 Marc Andreessen said that about four years ago, but today, the cloud is eating software. 01:17:58.098 --> 01:18:03.372 Software is being delivered as services in the cloud, and most pieces of software 01:18:03.732 --> 01:18:04.687 are connected to the cloud. 01:18:05.485 --> 01:18:10.028 Now, the cloud is not only eating software, but it is eating data. 01:18:11.052 --> 01:18:13.267 And when the cloud digests software 01:18:13.709 --> 01:18:16.716 and digests data, it can produce intelligence. 01:18:17.680 --> 01:18:19.240 Let me tell you what I mean. 01:18:20.460 --> 01:18:23.033 For human history, most of human history, 01:18:23.517 --> 01:18:24.952 the world was analog. 01:18:25.607 --> 01:18:27.632 Once upon a time, in the 1980s, 01:18:28.520 --> 01:18:31.650 analog data dominated. Digital data was just being born. 01:18:32.760 --> 01:18:38.004 And come the '90s, you have the Internet. You have client-server computing. 01:18:38.681 --> 01:18:40.000 You have digital music. 01:18:40.366 --> 01:18:41.807 You have DVDs. 01:18:42.058 --> 01:18:43.513 And the digital data exploded. 01:18:44.114 --> 01:18:46.918 But with Internet, one thing very interesting happened. 01:18:47.366 --> 01:18:49.712 Data started to have an IP address. 01:18:50.699 --> 01:18:52.117 It became connected data. 01:18:52.764 --> 01:18:54.867 It lived on datacenters on the cloud, 01:18:55.116 --> 01:18:59.570 PCs and devices, but it was starting to have an IP address. 01:19:00.288 --> 01:19:04.466 And fast forward to today, digital data dominates. 01:19:05.117 --> 01:19:10.717 Analog data is going down, and about half of all digital data is connected. 01:19:11.408 --> 01:19:14.760 It lives in the cloud. It's flowing in from mobile devices that 01:19:15.007 --> 01:19:18.671 are connected, the Internet of things is just starting to emerge, and there are 01:19:18.782 --> 01:19:20.844 10 zettabytes of it. 01:19:21.085 --> 01:19:23.244 That's 10 million petabytes. 01:19:24.210 --> 01:19:28.125 Look at the future. Look at 2020. 01:19:28.587 --> 01:19:30.921 In 2020, we're going to have 50 01:19:31.407 --> 01:19:35.439 zettabytes, 50 million petabytes, of data. 01:19:36.214 --> 01:19:41.785 The vast majority of that is in the cloud and flowing in it from connected devices. 01:19:41.972 --> 01:19:44.031 Analog has gone down almost to zero. 01:19:44.435 --> 01:19:46.121 The cloud is where the data lives. 01:19:46.641 --> 01:19:51.356 That's where the planet's data, intelligence and information resides. 01:19:52.095 --> 01:19:54.365 And that has profound implications for developers. 01:19:55.688 --> 01:20:01.650 So the best way -- so when I talk to developers about this, we tend to have 01:20:02.045 --> 01:20:03.373 about four conversations. 01:20:04.233 --> 01:20:09.127 So developers want to look at historical data and look at it retrospectively. 01:20:09.325 --> 01:20:11.640 It's big data. You want to analyze it. 01:20:11.919 --> 01:20:12.870 You want to report on it. 01:20:14.098 --> 01:20:15.983 And then there's real-time data, 01:20:16.245 --> 01:20:22.188 data flowing in over the wire, and developers want to analyze that asset as flowing 01:20:22.446 --> 01:20:24.599 through the wire and get insights from it. 01:20:25.643 --> 01:20:30.181 Developers also want to look at historical data, learn patterns from it, and predict 01:20:30.445 --> 01:20:31.271 the future. 01:20:31.524 --> 01:20:36.241 And they want to bring all of this together into intelligent SaaS apps. 01:20:37.574 --> 01:20:41.477 To bring this to life, let me tell you a story. 01:20:45.326 --> 01:20:47.468 My story is about the connected cow. 01:20:48.488 --> 01:20:50.522 Yes, there is such a thing as connected cows. 01:20:51.527 --> 01:20:53.190 It has all the buzzwords of today, 01:20:53.922 --> 01:20:57.126 the Internet of things, analytics, the cloud. 01:20:58.125 --> 01:20:59.963 It's perhaps the sexiest story in the cloud. 01:21:01.200 --> 01:21:02.876 But it's also about human ingenuity 01:21:03.127 --> 01:21:08.562 and how this new technology is revolutionizing even some of the world's oldest 01:21:08.672 --> 01:21:10.270 industries, such as dairy farming. 01:21:11.979 --> 01:21:15.365 So these are the connected cows that I'm talking about. 01:21:15.691 --> 01:21:19.824 Look at those pedometers on their legs. They're connected to Wi-Fi. 01:21:20.441 --> 01:21:23.552 I have one of those in my hand, the cow pedometer. 01:21:24.548 --> 01:21:32.250 It counts the steps of every cow on the dairy farm and sends that data to Microsoft Azure. 01:21:32.757 --> 01:21:38.222 It has Fujitsu's Gyuho, which means cow step, service in the cloud. 01:21:40.302 --> 01:21:43.327 So do cows need to take 10,000 steps a day too? 01:21:45.624 --> 01:21:51.451 Before I give you the answer, let me tell you about the modern dairy farm. 01:21:51.780 --> 01:21:56.497 These days, every company is a data company, even the ones you least expect. 01:21:56.977 --> 01:22:00.181 A dairy farm has all the constraints of a modern business. 01:22:00.669 --> 01:22:05.529 It has a fixed herd. It has a pasture and labor, which often 01:22:05.643 --> 01:22:07.580 is very expensive, and their output 01:22:07.869 --> 01:22:12.421 is milk and beef, and they have to optimize that output under all of these constraints. 01:22:13.428 --> 01:22:15.218 So what can a farmer do? 01:22:15.832 --> 01:22:17.501 It turns out he has two controls. 01:22:18.745 --> 01:22:22.308 He can detect health issues in cattle early and prevent loss. 01:22:23.150 --> 01:22:27.370 He can also improve cattle production by accurate detection of estrus. 01:22:28.670 --> 01:22:33.362 If you remember high school biology, estrus is when the animal is ready to mate 01:22:33.638 --> 01:22:37.691 and goes into heat, when the time is right, when the magic is ready to happen, 01:22:37.844 --> 01:22:38.813 so to speak. 01:22:41.613 --> 01:22:44.169 But how could you detect estrus in 01:22:44.817 --> 01:22:46.365 hundreds of cows in a dairy farm? 01:22:47.220 --> 01:22:50.962 Before I tell you, let me tell you about the importance of doing this. 01:22:51.342 --> 01:22:54.168 These days, with artificial insemination, 01:22:54.418 --> 01:22:57.768 the probability of conception is pretty good. It's about 70%. 01:22:58.263 --> 01:23:01.969 But detection of estrus has been done by age-old methods. 01:23:02.338 --> 01:23:03.435 It's just really close observation. 01:23:04.091 --> 01:23:07.366 In the best of cases, you get about a 55% accuracy. 01:23:08.897 --> 01:23:11.327 And then the pregnancy rate is about 39 percent. 01:23:11.883 --> 01:23:13.766 But if you could move that detection rate 01:23:14.019 --> 01:23:16.449 up to 95 percent, look what happens. 01:23:16.989 --> 01:23:20.706 You get a pregnancy rate of about 67 percent, which is a big improvement. 01:23:21.484 --> 01:23:26.840 It's very material for the dairy farmer. But this is hard. 01:23:27.301 --> 01:23:31.338 Estrus lasts only 12 to 18 hours every 21 days, and it's variable. 01:23:31.748 --> 01:23:33.675 And it occurs, like many of these things, 01:23:33.883 --> 01:23:35.837 mostly between 10:00 p.m. and 8:00 a.m., 01:23:36.099 --> 01:23:39.341 when the farmer is taking his hard-earned rest. 01:23:41.350 --> 01:23:44.708 So how can farmers tell when the time is right for hundreds of cows? 01:23:45.586 --> 01:23:47.132 Could technology help? 01:23:47.386 --> 01:23:50.834 Could you develop a heat map of the dairy, so to speak? 01:23:55.188 --> 01:23:59.708 Well, that's a question that a farmer in Japan asked Fujitsu, our partner. 01:24:00.991 --> 01:24:05.162 Fujitsu engineers consulted some researchers and came up with a very 01:24:05.276 --> 01:24:10.062 ingenious system, something that was indeed 95% accurate for the detection of estrus. 01:24:10.885 --> 01:24:13.703 It's literally the hottest system for detecting heat. 01:24:14.690 --> 01:24:16.464 Now, this is that service. 01:24:17.508 --> 01:24:22.637 Data from the pedometers that I showed you from the cows are sent over the wire 01:24:23.559 --> 01:24:29.022 to a service on the cloud on Microsoft Azure, where it analyzes all that data. 01:24:29.556 --> 01:24:35.232 And from the footsteps of cows detects when the animal goes into heat and sends 01:24:35.535 --> 01:24:37.369 alerts on the farmer's phone. 01:24:37.691 --> 01:24:41.165 So the farmer knows which cow went into heat and exactly when. 01:24:43.218 --> 01:24:45.763 Well, it turns out there's a simple secret 01:24:46.157 --> 01:24:47.934 to detecting when an animal goes into estrus. 01:24:48.689 --> 01:24:51.036 Let me explain that to you with a graph. 01:24:51.485 --> 01:24:53.659 The X-axis is time of night on this graph. 01:24:54.215 --> 01:24:57.013 Y-axis is the number of steps a cow is taking. 01:24:57.951 --> 01:25:03.109 So this is a cow on a normal night. Sleeping sometimes. 01:25:03.688 --> 01:25:07.369 Let's see what happens when the animal goes into heat. 01:25:08.086 --> 01:25:10.507 Yes. The number of steps she takes goes up. 01:25:10.960 --> 01:25:13.963 An animal goes into heat, she starts walking around furiously. 01:25:15.284 --> 01:25:19.544 And that turns out to be about 95% accurate for the detection of estrus. 01:25:20.383 --> 01:25:25.215 And the optimum time for artificial insemination for maximum conception rate 01:25:25.560 --> 01:25:26.839 is 16 hours from that. 01:25:27.908 --> 01:25:30.636 And that's when AI meets AI, 01:25:30.819 --> 01:25:32.240 artificial insemination, that is. 01:25:36.415 --> 01:25:41.738 Now, Fujitsu researchers found something else that was very amazing. 01:25:42.765 --> 01:25:45.373 There was a window around that optimum. 01:25:46.390 --> 01:25:51.570 If in the first four hours of that window, if you performed AI, you would get 01:25:51.835 --> 01:25:56.619 a female, and the second next four hours you would get a male. 01:25:57.090 --> 01:26:00.940 So the farmer now had an amazing control at his disposal. 01:26:01.483 --> 01:26:04.141 He could choose the gender of the calf 01:26:04.418 --> 01:26:06.106 based on the needs of his daily farm. 01:26:06.683 --> 01:26:11.363 Not only that, Fujitsu found that by analyzing the step patterns of cows, 01:26:12.021 --> 01:26:14.348 they could detect about eight different diseases. 01:26:15.414 --> 01:26:20.661 And remember that 5% false positive rate, 95% accurate, 5% false positive? 01:26:21.592 --> 01:26:25.311 Even some of that turns out to be very important because some of it is when the 01:26:25.658 --> 01:26:27.910 cow jumps over the fence and skips the farm. 01:26:30.485 --> 01:26:32.818 Well, it's an amazing application. 01:26:35.092 --> 01:26:40.737 But it's an application that today every one of you could build in a few days with 01:26:41.468 --> 01:26:43.712 the platform services that's available in Azure. 01:26:44.417 --> 01:26:46.240 And you can scale that out globally, 01:26:46.391 --> 01:26:50.297 distribute it, and this would never have been possible before. 01:26:52.091 --> 01:26:56.968 So that's the power of intelligent apps that live on the cloud. 01:26:57.304 --> 01:26:59.906 And now let's look at the power of real-time analytics. 01:27:00.786 --> 01:27:06.944 Data is always moving these days. It's constantly in motion. 01:27:07.473 --> 01:27:09.937 Connected data coming from sensors, 01:27:10.191 --> 01:27:14.462 from websites, from apps in the cloud, from social apps, it's streaming over the wire. 01:27:15.832 --> 01:27:20.573 Now, developers want to easily process this data and get insights real time. 01:27:22.052 --> 01:27:25.466 So now let's take a look at an example of a solution that does that on Azure. 01:27:27.651 --> 01:27:32.263 So with me is Corum, who is going to help me do a demo. 01:27:32.674 --> 01:27:35.982 We're going to create an app that's going 01:27:36.775 --> 01:27:39.675 to be a fun, engaging demo, and you are 01:27:40.048 --> 01:27:43.717 going to actually help us create data in motion, okay? 01:27:44.202 --> 01:27:47.514 When we started this, we wanted to build a fun and intelligent application, 01:27:48.380 --> 01:27:51.706 something that would capture the attention of people worldwide. 01:27:52.875 --> 01:27:58.622 So we went to the Machine Learning Gallery, and today we have an exciting 01:27:58.831 --> 01:28:02.467 collection of intelligent APIs on that Machine Learning Gallery. 01:28:02.828 --> 01:28:05.731 It's gallery.azureml.net. All of you can go there. 01:28:06.180 --> 01:28:11.057 And yesterday afternoon, Harry Shum, the head of MSR, announced Project Oxford to 01:28:11.467 --> 01:28:16.147 bring some of the amazing, intelligent APIs that have been created by Microsoft 01:28:16.432 --> 01:28:21.065 Research and Bing to developers like yourselves. There are face APIs. 01:28:21.401 --> 01:28:24.002 There are image analysis APIs. There are speech APIs. 01:28:24.677 --> 01:28:26.111 So we took the face API. 01:28:26.715 --> 01:28:30.477 It's a really fun API that allows you to submit a picture. 01:28:30.724 --> 01:28:36.880 It detects the faces in that and it tells you an estimate of their age and gender. 01:28:38.583 --> 01:28:42.326 So the URL for this demo is howold.net. 01:28:42.882 --> 01:28:44.625 Now, I would like everyone here to 01:28:44.776 --> 01:28:48.599 actually take a minute and go to this URL, especially everyone watching online. 01:28:49.383 --> 01:28:52.192 Please navigate to the site and take a look. 01:28:54.979 --> 01:28:56.534 So let's take a couple of pictures. 01:28:56.976 --> 01:28:58.379 Let's look at that happy family. 01:28:59.061 --> 01:29:01.355 Let's submit that, and you will see, 01:29:01.723 --> 01:29:03.942 it identified the faces in the image, 01:29:04.681 --> 01:29:07.738 the grandparents, the parents, the children. 01:29:08.133 --> 01:29:11.026 It got that -- it looks like it got pretty accurate. 01:29:11.383 --> 01:29:12.840 So let's try another picture. 01:29:13.401 --> 01:29:16.530 Let's take, say, Da Vinci's Mona Lisa, a painting. 01:29:20.253 --> 01:29:24.796 And so we're using the Bing search API to search Bing, get the images from online, 01:29:25.292 --> 01:29:29.474 and hopefully we can bring -- yes, Mona Lisa's image up. 01:29:30.087 --> 01:29:34.038 So let's submit that photo, and she is 23. 01:29:34.404 --> 01:29:39.239 Surprisingly, when that model posed for Leonardo, she was actually about 23. 01:29:40.283 --> 01:29:42.430 So now let's all try your pictures. 01:29:43.323 --> 01:29:45.242 Take a picture from your mobile phone, 01:29:45.493 --> 01:29:47.635 from a PC, especially if you're watching online. 01:29:48.254 --> 01:29:50.938 Feel free to take a selfie and send it, as well. 01:29:51.333 --> 01:29:53.178 Let's see how it performs. 01:29:53.797 --> 01:29:58.434 So, Corum, maybe you can take one of our photos from the desktop. 01:29:58.675 --> 01:30:01.436 So that's my family picture. Just uploaded it. 01:30:09.680 --> 01:30:11.236 So it's analyzing that. 01:30:14.717 --> 01:30:17.274 So we can actually see how it's doing on 01:30:17.583 --> 01:30:19.356 our dashboard, as well. 01:30:25.780 --> 01:30:28.952 Yes, the upload is taking a little while, Corum, so let's go look at the dashboard 01:30:29.177 --> 01:30:30.440 and see what people are doing online. 01:30:30.692 --> 01:30:34.001 It's amazing. This is the Power BI dashboard. 01:30:34.198 --> 01:30:36.005 John actually showed some of it earlier. 01:30:36.982 --> 01:30:42.753 People now from all of the world seem to be coming to howold.net, and that map 01:30:43.000 --> 01:30:46.683 shows the distribution of people. So what's different about this dashboard, 01:30:46.995 --> 01:30:48.353 by the way, it's pretty amazing that 01:30:48.850 --> 01:30:54.203 you're seeing real-time flowing data being analyzed, and you see that graph that just 01:30:54.429 --> 01:30:57.932 spiked up. That's all of you bringing data in motion. 01:30:59.133 --> 01:31:03.337 Now, let's peek behind the curtain a little bit and see how we built this 01:31:03.688 --> 01:31:06.329 streaming analytics and the dashboard. 01:31:06.799 --> 01:31:09.797 So the data from the howold.net website 01:31:10.597 --> 01:31:12.152 is captured as a JSON file. 01:31:12.780 --> 01:31:18.145 There is -- it looks at the face data, the age data, looks at the browser. 01:31:18.959 --> 01:31:22.929 It looks at the location from which the IP address is coming, a collection of things. 01:31:23.334 --> 01:31:28.731 The JSON is sent to Azure Event Hubs. Azure Event Hubs is a fully managed 01:31:28.842 --> 01:31:32.204 service on the cloud to ingest data at 01:31:32.448 --> 01:31:34.238 millions of events per second. 01:31:34.733 --> 01:31:36.882 And then, we use Azure Streaming Analytics. 01:31:37.602 --> 01:31:43.038 Azure Streaming Analytics is a fully managed service for complex event processing. 01:31:43.201 --> 01:31:49.040 It lets you analyze streaming data, put complex queries in it and get aggregates 01:31:49.183 --> 01:31:52.002 from that. And then we use those aggregates 01:31:52.350 --> 01:31:53.440 and display it in a dashboard. 01:31:53.694 --> 01:31:55.208 So now let's take an example. 01:31:55.400 --> 01:32:01.761 Let's take that dashboard that is there, that trending graph, so that trending 01:32:01.855 --> 01:32:03.760 graph shows two lines. 01:32:04.460 --> 01:32:11.414 You have the blue line showing the count of females in the faces submitted, the pictures. 01:32:11.783 --> 01:32:13.816 The magenta is the number of males. 01:32:14.579 --> 01:32:18.485 And all developers here, if you have tried to construct a trending graph like 01:32:18.728 --> 01:32:22.255 this over streaming data, you know how hard it is. 01:32:22.477 --> 01:32:23.920 It takes hundreds of lines of code, 01:32:24.332 --> 01:32:27.436 if not more, to set something like this up. 01:32:28.325 --> 01:32:33.017 Let's now go to Streaming Analytics, the Azure Stream Analytics query, and show you 01:32:33.174 --> 01:32:36.813 how simple it is. So Azure Stream Analytics makes it really 01:32:36.956 --> 01:32:42.282 easy to configure such analytics using just about eight lines of SQLite code. 01:32:42.734 --> 01:32:46.038 It's a select statement. It's selecting the system's timestamp as 01:32:46.161 --> 01:32:49.528 out time, the face.gender as gender 01:32:50.060 --> 01:32:54.683 and the count of gender from the streaming input and then aggregating it over 01:32:54.927 --> 01:32:58.253 a window over a 10-second interval, right? 01:32:58.410 --> 01:33:02.590 And that's all there is to it, and that's the data that is now being displayed in 01:33:02.907 --> 01:33:04.357 the Power BI dashboard. 01:33:05.308 --> 01:33:09.130 So this is the magic of fully managed services, like Azure Stream Analytics in 01:33:09.663 --> 01:33:11.684 the cloud and Power BI. 01:33:11.939 --> 01:33:16.733 It lets you analyze real-time flowing data and embed intelligence around that 01:33:17.211 --> 01:33:18.293 into your applications. 01:33:21.180 --> 01:33:25.013 So now let's go from understanding the present to predicting the future. 01:33:27.659 --> 01:33:37.734 So if we switch to slides, so in Feb. of this year, we launched a powerful 01:33:38.014 --> 01:33:40.177 service called Azure Machine Learning. 01:33:41.137 --> 01:33:45.215 It allows you to learn patterns from historical data and predict the future 01:33:46.175 --> 01:33:51.604 and create APIs in the cloud that can be easily hooked up into any application. 01:33:52.258 --> 01:33:58.531 For example, Ziosk is using Azure Machine Learning to produce recommendations on 01:33:59.008 --> 01:34:00.982 Ziosk tablets on Chili's restaurants. 01:34:01.709 --> 01:34:08.086 We have eSmart Systems of Norway using Azure Machine Learning to actually predict 01:34:08.827 --> 01:34:12.495 the energy usage in the future and control energy demand. 01:34:13.143 --> 01:34:16.664 The Microsoft Band uses Azure Machine Learning for recommendations. 01:34:16.941 --> 01:34:21.668 We have companies, more traditional companies such as Gaffey Healthcare using 01:34:21.891 --> 01:34:27.590 Azure Machine Learning to estimate when an insurance claim will be paid, so they know 01:34:27.918 --> 01:34:31.594 how to act on that particular claim. It's a very powerful service. 01:34:32.583 --> 01:34:38.213 So the best way for me to bring this to life is to tell you another story. 01:34:40.383 --> 01:34:44.869 For those of you who are watching this from outside the US, there is an annual 01:34:45.132 --> 01:34:47.886 sports event here called March Madness. 01:34:48.860 --> 01:34:54.331 It refers to the NCAA Men's Division 1 Basketball Tournament, so it's played each 01:34:54.689 --> 01:34:58.289 spring with 68 collegiate teams to determine the national championship. 01:34:59.633 --> 01:35:02.531 And for Americans, we find it as a time to gamble. 01:35:03.336 --> 01:35:07.141 We try to predict who will make it to the Final Four teams and ultimately who will 01:35:07.610 --> 01:35:08.330 win it all. 01:35:09.628 --> 01:35:12.291 It seems that everyone has a prediction for the winners. 01:35:12.634 --> 01:35:19.183 Our President Obama, our CEO Satya Nadella, and Satya's predictions got a lot 01:35:19.416 --> 01:35:21.811 of press, because it seemed to be extremely accurate. 01:35:22.515 --> 01:35:26.407 Now, to be fair, Satya had help. Satya had data scientists and people who 01:35:26.664 --> 01:35:28.777 knew basketball to help him make his predictions. 01:35:30.828 --> 01:35:32.868 But I want to tell you the story of Adam Garland. 01:35:34.115 --> 01:35:37.005 Adam is a developer in our Microsoft Office group. 01:35:37.494 --> 01:35:41.863 Now, unlike Satya, Bing and Google, Adam didn't have any experience using machine learning. 01:35:42.634 --> 01:35:44.181 He's just a developer, like you and me. 01:35:45.782 --> 01:35:50.136 And we had a hackathon going to predict March Madness using Azure Machine Learning. 01:35:50.643 --> 01:35:55.510 Adam thought it would be a great way for him to learn machine learning, so he came 01:35:55.881 --> 01:36:01.722 to the tool, he played around with the data, and within a few hours, he was able 01:36:01.837 --> 01:36:06.285 to create a model to predict who will win in a basketball match. 01:36:06.687 --> 01:36:09.884 And then, after a few days, he got to be really good. 01:36:10.371 --> 01:36:16.366 And not only Adam, but many of the people who participated on that competition ended 01:36:16.484 --> 01:36:20.520 up predicting much better than the automated algorithms that Google used 01:36:20.763 --> 01:36:22.804 and Bing used and was pretty competitive. 01:36:23.816 --> 01:36:28.284 So these Azure ML tools are very easy to use for developers, so I thought it might 01:36:28.516 --> 01:36:34.091 be actually fun to replicate what Adam did, to show you how Adam built that app 01:36:34.484 --> 01:36:39.884 right in front of you and actually publish it, publish an API live and show you 01:36:40.117 --> 01:36:44.715 a real demo, calling it with a basketball match. 01:36:46.059 --> 01:36:50.055 So let's see how easy it is to create an experiment. 01:36:50.421 --> 01:36:52.207 To predict the outcome of a match, 01:36:54.024 --> 01:36:58.473 whether team one will win or team two, we typically need something called a binary classifier. 01:36:59.249 --> 01:37:03.602 There are experiments in the Azure ML Gallery, so let's take this particular one. 01:37:03.872 --> 01:37:06.773 This is an experiment you'll find in gallery.azureml.net. 01:37:07.617 --> 01:37:09.673 So we just brought it to the studio, 01:37:10.915 --> 01:37:12.923 and this experiment is a workflow. 01:37:13.224 --> 01:37:14.610 It's actually really simple. 01:37:15.227 --> 01:37:16.596 So it reads and cleans the data. 01:37:17.764 --> 01:37:20.089 You split the data into training and test data. 01:37:21.016 --> 01:37:25.452 So you train a model with that training data, and then you score the test data 01:37:25.649 --> 01:37:28.851 that you then provide to the model and see how well it did. 01:37:30.150 --> 01:37:33.963 So let's actually run this and see how it performs. 01:37:39.917 --> 01:37:46.317 So we are going to now connect the March Madness data set -- so to just get Adam's 01:37:46.575 --> 01:37:51.372 experiment, you take the data from the sample experiment, bring the March Madness 01:37:51.668 --> 01:37:54.120 data set, and now we actually go and run it. 01:37:55.336 --> 01:37:58.853 So that experiment is running, and while that's running, let's look at the data. 01:38:01.570 --> 01:38:05.493 So you'll see, this is data about 12 years of matches. 01:38:06.123 --> 01:38:11.520 We have the seeds for each team, and the final column there, team one wins, true or 01:38:11.767 --> 01:38:13.206 false, is what we are trying to predict. 01:38:14.318 --> 01:38:17.070 So let's go back and see how well the model ran. 01:38:17.851 --> 01:38:19.273 So let's look at a performance. 01:38:20.251 --> 01:38:24.120 So that graph actually shows what's called an auto-seeker, but the most interesting 01:38:24.375 --> 01:38:25.373 number is the accuracy. 01:38:25.766 --> 01:38:29.925 So you'll see that simple model performed at an accuracy of 66%. 01:38:30.857 --> 01:38:32.732 Not bad for a minute or so of work. 01:38:33.625 --> 01:38:37.820 You just put in new data into that existing sample experiment environment. 01:38:38.052 --> 01:38:39.873 So now let's see what Adam did. 01:38:40.824 --> 01:38:42.288 Adam did a couple of things. 01:38:42.624 --> 01:38:46.576 Adam brought in an algorithm called Two-Class Decision Jungle. 01:38:46.920 --> 01:38:49.469 That's one of the algorithms in Action Machine Learning. 01:38:49.871 --> 01:38:51.216 He used something called Sweep Parameters 01:38:51.663 --> 01:38:55.569 to sweep over the parameters of that machine learning model and find the best one. 01:38:56.567 --> 01:39:01.220 He also found that bringing in team performance data would be very, 01:39:01.467 --> 01:39:08.475 very helpful, so he took the historical data, married it with team stats, and then let's 01:39:08.745 --> 01:39:11.492 look at that data. When you look at it, you'll find that he 01:39:12.343 --> 01:39:17.967 brought in data such as the field goals made, rebounds. 01:39:18.325 --> 01:39:23.295 He even found, for example, that team performance from away games was actually 01:39:23.424 --> 01:39:26.894 a really important predictor of performance, of match performance. 01:39:27.519 --> 01:39:29.820 So let's see how that model ran. 01:39:33.371 --> 01:39:35.495 So you'll see the performance went up dramatically. 01:39:35.775 --> 01:39:37.692 It's just about 74% accurate. 01:39:38.910 --> 01:39:44.093 And now that we have a model in Azure Machine Learning, we can create an API 01:39:44.425 --> 01:39:46.017 with it, a cloud-hosted API. 01:39:46.457 --> 01:39:49.693 Let's do that now. To do that, we create a scoring experiment. 01:39:50.473 --> 01:39:53.851 And this is a workflow that will run inside of a web service call. 01:39:54.521 --> 01:39:56.296 And here's the magic. 01:39:57.126 --> 01:40:01.523 It automatically identifies what the predictive model is, identifies the web 01:40:01.626 --> 01:40:03.271 service input and web service output, 01:40:03.615 --> 01:40:06.074 and we are now ready to publish this as an API. 01:40:06.620 --> 01:40:08.122 We can run it, and when you publish it, 01:40:08.373 --> 01:40:11.096 you'll get an API that is live in the cloud. 01:40:12.171 --> 01:40:13.719 So let's see that API now. 01:40:14.626 --> 01:40:17.096 Well, there. This is the page you get when you run 01:40:17.297 --> 01:40:19.571 and publish it. You see that API key at the top? 01:40:19.768 --> 01:40:22.417 That's the API key you will use on your REST call. 01:40:22.969 --> 01:40:26.627 We can test this. We can test this API live in the cloud. 01:40:26.870 --> 01:40:32.226 So now, let's take that historic match between Duke and Wisconsin that just 01:40:32.337 --> 01:40:38.615 happened and see what the API predicts. So there you go. 01:40:39.167 --> 01:40:45.076 The API predicts Duke will win over Wisconsin with a probability of 63.2%. 01:40:46.468 --> 01:40:48.168 And that's exactly what happened. 01:40:49.624 --> 01:40:54.409 Now, you can predict the future and build this intelligence into cloud apps. 01:40:55.974 --> 01:40:57.893 It's very simple. 01:40:58.425 --> 01:41:01.293 You have C# code. You have Python code. 01:41:01.920 --> 01:41:05.265 Those are things that you can use to just call that API. 01:41:06.520 --> 01:41:10.618 So now, let me move from the demonstration to something else that's 01:41:10.874 --> 01:41:14.625 pretty exciting about the language of data. 01:41:15.669 --> 01:41:20.221 Now, machine learning is a field of advanced statistics. 01:41:21.274 --> 01:41:26.691 And over the last 20 years, a revolution has happened here with the development of 01:41:27.025 --> 01:41:28.325 open-source R. 01:41:29.276 --> 01:41:34.025 Statisticians all over the world have now started contributing their latest 01:41:34.265 --> 01:41:36.778 innovations and research as packages in R. 01:41:37.667 --> 01:41:42.694 If there is one language that is truly about data, something that puts the power 01:41:42.921 --> 01:41:46.334 of data at the fingerprints of every developer, it is R. 01:41:47.071 --> 01:41:52.670 In IEEE Spectrum's rankings of programming language popularity in 2014, 01:41:52.923 --> 01:41:55.528 R placed ninth, after Ruby. 01:41:56.177 --> 01:41:58.780 That's an amazing feat for a domain-specific language. 01:41:59.379 --> 01:42:04.619 So if there's a single language that you choose to learn today to tap into the 01:42:04.874 --> 01:42:06.895 power of that cloud, let it be R. 01:42:08.718 --> 01:42:14.264 Now, and that's why Microsoft acquired Revolution Analytics, the company that 01:42:14.695 --> 01:42:20.774 made R scaled to big data and brought open-source R to the enterprise, making it 01:42:20.896 --> 01:42:21.864 enterprise grade. 01:42:22.184 --> 01:42:26.491 For example, American Century Investments, one of the largest mutual 01:42:26.621 --> 01:42:30.702 fund companies in the world, uses it to build their quantitative investment platform. 01:42:31.823 --> 01:42:36.191 Data Song, a local company, a big data marketing company in San Francisco, 01:42:36.580 --> 01:42:41.291 uses it along with Hadoop for their marketing performance management and attribution platform. 01:42:42.742 --> 01:42:49.094 Now, to bring the story of R and big data to life, let me tell you another story. 01:42:51.869 --> 01:42:54.009 We are all big data. 01:42:55.723 --> 01:43:01.531 There are billions of letters of genomic data in every one of our cells. 01:43:03.029 --> 01:43:05.031 It turns out that each human genome 01:43:05.778 --> 01:43:08.288 is about 2 gigabytes of information, and with 01:43:09.182 --> 01:43:11.933 the power of R, we can analyze all of this data. 01:43:13.268 --> 01:43:17.381 And when we plot this set of information against the general population, we can 01:43:17.747 --> 01:43:20.017 actually see some incredible things about ourselves. 01:43:23.276 --> 01:43:27.736 So the first human genome was sequenced in the year 2000. 01:43:28.715 --> 01:43:33.225 It cost about $200 million and over 10 years of work. 01:43:34.060 --> 01:43:38.290 Since then, the cost of mapping the genome have come down much faster than 01:43:38.667 --> 01:43:39.625 Moore's law. 01:43:40.342 --> 01:43:45.460 It's about $1,000 to get your entire genome, and you can get the most important 01:43:45.732 --> 01:43:50.061 parts for about $100, and this is driving an untold revolution. 01:43:51.189 --> 01:43:52.735 In the future, when you go to a doctor, 01:43:53.741 --> 01:43:57.491 the first thing that they will do will be to map your genome to understand your 01:43:57.745 --> 01:43:58.693 disease risks. 01:43:59.315 --> 01:44:04.982 This will let them treat you based upon your individual genome, personalized medicine. 01:44:06.314 --> 01:44:10.100 You will even know the risks you should anticipate in the future and change your 01:44:10.225 --> 01:44:11.188 lifestyle because of it. 01:44:13.716 --> 01:44:16.695 So let's show that to you with a demo. 01:44:17.472 --> 01:44:21.060 Joining me is Mario, who was at Revolution Analytics. 01:44:22.033 --> 01:44:24.696 So genomes are analyzed using the language of R. 01:44:24.903 --> 01:44:27.329 It has one of the richest libraries for analyzing genomic data. 01:44:28.474 --> 01:44:32.613 1,000 people worldwide have contributed their genome to scientific research. 01:44:33.387 --> 01:44:36.305 It is public data called 1,000 genomes data set. 01:44:36.670 --> 01:44:40.087 It's about 2 gigabytes of data per person, and we're going to analyze all of 01:44:40.298 --> 01:44:42.622 that data and see their disease risks, 01:44:43.475 --> 01:44:46.161 and Mario is going to actually walk through the demo. 01:44:46.497 --> 01:44:51.331 Now, it's going to take only a little bit of code, but it's a tremendous amount of communication. 01:44:52.126 --> 01:44:55.337 So it really needs big Hadoop clusters, 01:44:55.982 --> 01:44:59.986 so if we could show the Hadoop clusters that we are actually showing, the slides, 01:45:00.183 --> 01:45:05.781 we're going to use about eight HDInsight clusters in four Azure datacenters, two in 01:45:06.035 --> 01:45:12.072 the US, one in Europe, and one in Southeast Asia for a total of about 1,600 01:45:12.307 --> 01:45:13.543 cores in the calculation. 01:45:14.917 --> 01:45:16.668 By the way, they are Linux clusters. 01:45:17.078 --> 01:45:20.898 We are actually running Revolution R on them, and Revolution R is going to 01:45:21.130 --> 01:45:25.268 coordinate a computation across those 1,600 cores in all of these regions across 01:45:25.380 --> 01:45:27.668 the world. We are actually going to take computation 01:45:27.909 --> 01:45:31.012 to where the data resides when we do that global computation. 01:45:31.638 --> 01:45:32.936 So now let's see the core. 01:45:34.981 --> 01:45:38.824 So only a few lines of R code. This is the code that's actually used to 01:45:39.186 --> 01:45:40.673 analyze the genomic data. 01:45:41.037 --> 01:45:46.501 It leverages some extremely powerful functions in an R package called Bioconductor. 01:45:47.182 --> 01:45:50.190 And let's also see the code that coordinates the computation. 01:45:50.585 --> 01:45:54.859 So this is the piece of the code that coordinates computation on HDInsight 01:45:54.985 --> 01:45:58.828 clusters across eight clusters across four datacenters. 01:45:59.348 --> 01:46:04.785 Let's kick that run off, Mario, and let's see what it produces. 01:46:04.957 --> 01:46:06.324 It produces a heat map. 01:46:08.952 --> 01:46:11.071 This is a heat map of disease risks. 01:46:11.435 --> 01:46:13.180 Each row is an individual. 01:46:13.477 --> 01:46:14.931 We have fictitious names on the right, 01:46:15.556 --> 01:46:18.305 but each column is a disease. 01:46:18.830 --> 01:46:22.454 So let's take, for example, the second row from the top, Claire Hickman. 01:46:22.837 --> 01:46:26.531 Claire Hickman has a high risk of ulcerative colitis. 01:46:27.757 --> 01:46:31.780 And let's take, say, the third from the bottom, Sophie Albrecht. 01:46:32.316 --> 01:46:37.028 Well, it turns out Sophie has a high risk of multiple sclerosis. 01:46:37.871 --> 01:46:39.582 By the way, this heat map is an amazing 01:46:39.834 --> 01:46:46.186 visualization produced by a package called Shiny in R, and it lets you visualize 01:46:46.358 --> 01:46:49.181 things and zoom into these things. It's very, very powerful, as well. 01:46:50.271 --> 01:46:55.077 So now you've seen the disease risk of a population, of the population we analyzed 01:46:55.554 --> 01:47:00.873 on the Hadoop cluster, and almost happens in a few minutes, and it's pretty amazing, 01:47:01.356 --> 01:47:02.788 chunking through all of that data. 01:47:04.031 --> 01:47:10.023 But now what if we could take that R code, publish that as a web service API? 01:47:10.982 --> 01:47:14.992 So you saw me publish a web service API with Azure Machine Learning earlier. 01:47:16.184 --> 01:47:21.704 What if I could do the same and create an API that lets you submit data and see your 01:47:21.981 --> 01:47:23.108 disease risks? 01:47:24.556 --> 01:47:26.332 So let's see the scoring experiment. 01:47:26.558 --> 01:47:28.481 This is a scoring experiment in Azure ML. 01:47:29.954 --> 01:47:31.628 See that execute R script? 01:47:32.239 --> 01:47:37.780 That's where we can actually copy the R code and publish that as an API. 01:47:38.187 --> 01:47:39.536 So let's do that now. 01:47:40.834 --> 01:47:46.295 So Mario's going to copy some of the R code, first coding some genomic data, 01:47:46.928 --> 01:47:49.902 put that into the execute R script and run 01:47:50.730 --> 01:47:52.387 that experiment. 01:47:53.350 --> 01:47:58.507 By the way, as it's running, execute R script shows you how easy it is to 01:47:58.739 --> 01:48:00.471 operationalize R. 01:48:00.722 --> 01:48:04.529 Any R code can be published as an API in the cloud. 01:48:04.735 --> 01:48:07.940 And Azure Machine Learning lets you do the same thing even with Python. 01:48:08.476 --> 01:48:13.424 It is the simplest, easiest way to operationalize your analytics code as APIs 01:48:13.785 --> 01:48:14.505 in the cloud. 01:48:15.153 --> 01:48:22.075 So now, let's see -- let's publish the web service, and let's get that API in the cloud. 01:48:24.581 --> 01:48:28.099 Here, that's the API page. You got the key. 01:48:28.957 --> 01:48:31.108 There's even sample code to call that API. 01:48:32.589 --> 01:48:36.537 But to test this, we actually created -- yes, there's a sample code. 01:48:36.909 --> 01:48:40.430 If you click on it, you'll see C#, Python and R code are automatically generated. 01:48:42.238 --> 01:48:46.535 So now to call this API and show you the results, we created a mobile app. 01:48:47.323 --> 01:48:48.387 I'll tell you about that in a minute. 01:48:49.751 --> 01:48:54.304 Now, early in April, I sent a sample of 01:48:54.522 --> 01:48:58.739 my saliva to 23andme.com to sequence my genome. 01:49:00.359 --> 01:49:02.790 Last Friday, I got the results from that, 01:49:03.638 --> 01:49:06.795 and so for the first time, I was able to 01:49:07.038 --> 01:49:09.909 analyze my DNA and see my future. 01:49:11.240 --> 01:49:12.788 So let's do that now. 01:49:14.478 --> 01:49:15.908 So let's take the mobile app. 01:49:19.530 --> 01:49:21.333 Can you all see the mobile app? 01:49:22.929 --> 01:49:27.509 So I'm going to log into 23andme, get the data from their website. 01:49:33.474 --> 01:49:35.083 It's uploading the genomic profile. 01:49:35.922 --> 01:49:39.473 So you will see a graph that shows the risks. 01:49:41.184 --> 01:49:43.588 So the blue bar is a population risk, 01:49:44.329 --> 01:49:45.737 and red is mine. 01:49:45.988 --> 01:49:52.308 So you'll see that I have low risk for gallstones, low risk for asthma, 01:49:53.036 --> 01:49:58.623 certainly low risk for breast cancer, but let's look here at the top. 01:49:58.991 --> 01:50:00.714 Look at that very top line. 01:50:01.730 --> 01:50:04.231 That red bar shows 0.36. 01:50:04.474 --> 01:50:10.476 I have about 2.5 times the population risk for prostate cancer, and I just 01:50:10.733 --> 01:50:11.819 discovered that. 01:50:12.323 --> 01:50:20.679 That's my future predicted from my genomic data, R and big data analysis. 01:50:22.387 --> 01:50:26.845 That's the power of big data and advanced analytics. 01:50:27.789 --> 01:50:29.229 So if we switch back to the slide, 01:50:30.784 --> 01:50:34.997 big data and advanced analytics can change lives. 01:50:46.280 --> 01:50:50.429 My next story is about the connected grid. 01:50:52.261 --> 01:50:57.907 For a century, the management of power has been by analog systems, just like the 01:50:58.147 --> 01:50:59.516 media we were consuming was analog. 01:51:00.049 --> 01:51:04.310 But now, utilities are discovering the power of big data analytics and the cloud. 01:51:05.678 --> 01:51:11.340 And the hero of this -- the hero of this story is a small startup from Norway, 01:51:11.951 --> 01:51:14.842 born in a small town of 30,000 people, 01:51:15.074 --> 01:51:19.547 and they're reinventing the future of energy management, and they're going global with it. 01:51:20.042 --> 01:51:23.782 They're reinventing the grid with the cloud, creating great efficiencies through 01:51:24.074 --> 01:51:25.394 data and advanced analytics. 01:51:26.357 --> 01:51:30.506 To tell you that story, let me welcome Erik Asberg, the head of development of 01:51:30.612 --> 01:51:33.796 eSmart Systems to the stage. Erik? 01:51:35.404 --> 01:51:37.310 Welcome. >> Thank you, Joseph. 01:51:38.929 --> 01:51:40.826 So let me tell you about eSmart Systems. 01:51:42.672 --> 01:51:44.350 We accelerate energy systems. 01:51:45.794 --> 01:51:51.401 We optimize energy investments and we minimize carbon footprints through better, 01:51:51.623 --> 01:51:53.876 faster energy decisions. 01:51:54.849 --> 01:51:57.880 Like many of you, we are in the business of disruption. 01:51:58.652 --> 01:52:02.281 We're taking an industry that has been operating in the same way for over 100 01:52:02.527 --> 01:52:07.475 years, and we use technology to radically rethink the way we approach it. 01:52:08.781 --> 01:52:13.323 Our disruption is directed at the business of energy, and we started with a pilot 01:52:13.481 --> 01:52:15.394 project in Norway two years ago. 01:52:17.874 --> 01:52:21.823 So when we started, our goals were to reduce power grid investments by 01:52:22.249 --> 01:52:26.745 increasing the utilization factor of existing grid capacity. 01:52:27.278 --> 01:52:29.347 We needed detailed predictions of load 01:52:29.600 --> 01:52:34.601 and power generation, and we needed to take advantage of IoT and big data, 01:52:35.451 --> 01:52:36.879 since smart meters and sensors are 01:52:37.179 --> 01:52:39.606 a prerequisite for our predictive analytics. 01:52:41.174 --> 01:52:45.123 So the big question was, how were we going to go off and do this? 01:52:45.446 --> 01:52:50.192 Were we going to spend millions of dollars and thousands of person hours in datacenters? 01:52:51.453 --> 01:52:54.557 For us, this was an incredibly easy decision. 01:52:55.281 --> 01:53:00.112 We leveraged Azure from the very beginning, and what makes this so amazing 01:53:00.329 --> 01:53:05.347 is that we made this journey of changing the business of energy without acquiring 01:53:05.672 --> 01:53:06.750 a single server. 01:53:09.279 --> 01:53:12.396 So here is a look at the eSmart Connected Grid application. 01:53:12.731 --> 01:53:15.727 All the houses that you see here are equipped with smart meters and sensors, 01:53:16.282 --> 01:53:20.753 and we use machine learning to identify potential congestions in the power grid. 01:53:21.878 --> 01:53:29.230 We control consumer flexibility to avoid substation overloads, and we use both 01:53:29.347 --> 01:53:31.077 short-term and long-term predictions to 01:53:32.374 --> 01:53:36.681 control heated floors and water heating through home automation. 01:53:38.282 --> 01:53:44.682 eSmart and Azure Machine Learning use real data to learn for each specific area. 01:53:47.931 --> 01:53:52.228 And we take this information and we make it available on your phones with consumer 01:53:52.334 --> 01:53:53.877 apps managed by Azure. 01:53:56.373 --> 01:53:59.080 Every part of the power grid is unique. 01:54:00.181 --> 01:54:03.728 There is no system today that is able to adapt to this uniqueness. 01:54:05.276 --> 01:54:09.226 eSmart Connected Grid and Azure ML totally changed this reality. 01:54:09.377 --> 01:54:11.259 This is truly revolutionary. 01:54:13.181 --> 01:54:16.831 So we started 2.5 years ago with no employees, no revenue. 01:54:17.203 --> 01:54:21.327 By the end of this year, we'll be over 40 employees and over 50 million Norwegian 01:54:21.477 --> 01:54:22.625 kroner in revenue. 01:54:22.980 --> 01:54:26.933 I couldn't be happier to be partnering with Microsoft and Azure, and I want to 01:54:27.128 --> 01:54:30.125 say a big thank you to Joseph for having us here today. 01:54:30.482 --> 01:54:31.926 >> Thank you, Erik. 01:54:33.373 --> 01:54:34.802 It's an amazing story. 01:54:38.575 --> 01:54:43.532 The cloud amplifies startups such as eSmart Systems. 01:54:43.881 --> 01:54:46.076 It gives them the power to go global. 01:54:46.446 --> 01:54:50.730 It gives them the power to build applications that revolutionize entire 01:54:51.128 --> 01:54:53.682 industries in a very short period of time. 01:54:54.178 --> 01:54:59.531 In the future, all of the planet's power generation, power grids and power 01:54:59.799 --> 01:55:03.575 consumption will be optimized with real-time data and data analytics. 01:55:04.627 --> 01:55:06.643 And so with the power of the cloud, 01:55:07.377 --> 01:55:12.524 even really small companies like eSmart systems are able to make that dream come 01:55:12.736 --> 01:55:14.620 true at a speed that no one else can. 01:55:15.474 --> 01:55:17.902 That agility, by the way, that agility 01:55:18.071 --> 01:55:20.848 is a hallmark of the cloud, and that's its magic. 01:55:22.947 --> 01:55:26.424 The cloud turns hardware into software, 01:55:28.218 --> 01:55:35.213 software into services, fully managed services, and data into intelligence. 01:55:36.577 --> 01:55:39.496 It makes you smart and it lets you free. 01:55:39.979 --> 01:55:47.246 Thank you. >> Nice job, man. 01:55:50.086 --> 01:55:54.047 I don't think anybody does a better job of the notion of data as a platform 01:55:54.384 --> 01:55:57.924 and bringing intelligence to data than Joseph, a great storyteller, great at really, 01:55:58.226 --> 01:55:59.019 really bringing this to life. 01:55:59.367 --> 01:56:03.386 So I think we're on our final run. We wanted to bridge from Joseph and pick 01:56:03.515 --> 01:56:04.941 up on where we were a few minutes ago. 01:56:05.334 --> 01:56:08.922 So last year, when we were talking about the web, we were talking about the 01:56:09.143 --> 01:56:11.188 Babylon.js project. John mentioned that earlier. 01:56:11.536 --> 01:56:13.709 And one of the things we had done is we actually took a web experience, 01:56:14.189 --> 01:56:19.198 used Babylon.js, hooked it up so we could do virtual reality with it, took advantage of 01:56:19.433 --> 01:56:22.114 Oculus Rift at the time. This time, we've been working further on 01:56:22.344 --> 01:56:25.538 Babylon, connecting it into the data side, trying to make it more intelligent. 01:56:25.941 --> 01:56:27.029 So in the city of Pompeii, the government 01:56:27.191 --> 01:56:30.785 has been using drones to fly around and map the city, and I'm going to fly around 01:56:31.139 --> 01:56:33.886 using what they've created from that on the web and let John tell you about it. 01:56:34.592 --> 01:56:39.225 >> So in partnership with the Italian government, we took these drones that were 01:56:39.482 --> 01:56:44.223 flying around, about 50 hours of drone time, and we took over 30,000 pictures, 01:56:44.729 --> 01:56:49.432 high-res images of Pompeii, and then the very interesting thing is we moved that 01:56:49.793 --> 01:56:54.483 information up into the cloud, where we used Azure and machine learning algorithms 01:56:54.943 --> 01:56:58.431 to go and generate the 3D mesh that you're seeing in front of you. 01:56:58.719 --> 01:57:01.389 Now, the thing that's pretty incredible about this is you're actually looking at 01:57:01.557 --> 01:57:04.861 an image that's about close to a trillion pixels. 01:57:05.088 --> 01:57:09.231 And as Steve's able to do, you can wander around in this thing. 01:57:09.482 --> 01:57:10.458 You can see any aspect of it. 01:57:10.643 --> 01:57:15.028 But I think this really illustrates the power of when we can take great innovation 01:57:15.391 --> 01:57:19.031 in the client, couple it with the machine learning, the big data in the cloud 01:57:19.316 --> 01:57:23.548 and bring experiences to every platform that we wouldn't have been able to do before. 01:57:23.787 --> 01:57:26.118 >> Flew us around. I ended up in the basilica. 01:57:26.523 --> 01:57:29.030 So this is a good bridge and transition. 01:57:29.334 --> 01:57:31.484 A couple of things here, which is obviously the cloud side. 01:57:31.937 --> 01:57:35.918 I also have a machine up here, a laptop with two graphics cards, two GPUs. 01:57:36.194 --> 01:57:39.143 And one of the things I want to bridge into is gaming, DirectX 12. 01:57:39.391 --> 01:57:42.088 One of the things that DirectX 12 allows us to do is actually take advantage of 01:57:42.317 --> 01:57:44.753 multiple GPUs, so I can have some client-side assist here. 01:57:45.236 --> 01:57:48.725 And then on the gaming side, look, like music, gaming is an area that I'm pretty 01:57:48.920 --> 01:57:50.892 sure everyone in the audience has some connection to one way or the other. 01:57:51.135 --> 01:57:58.627 There's lots of conversations around gaming from the mobile side to the Xbox 01:57:58.886 --> 01:58:00.628 and others. The real thing there is when you think 01:58:00.756 --> 01:58:05.392 about high-end graphics in gaming, it's not actually the Xbox or any of the consoles. 01:58:05.644 --> 01:58:07.002 It's actually back to the PC. 01:58:07.245 --> 01:58:10.353 And so if I can do that with a two graphics card system, we brought along 01:58:10.724 --> 01:58:14.625 just for fun a box that has four graphics cards. Digital Storm lent it to us. 01:58:15.172 --> 01:58:17.344 It's the very highest end on the hardware side. 01:58:17.934 --> 01:58:21.793 And then we asked our friends at Square Enix to take the work they do, which is 01:58:22.115 --> 01:58:25.126 a mixture of science and really the computation side and graphics quality 01:58:25.520 --> 01:58:29.346 and art and bring those together and show us a little bit about what you can do with DX 01:58:29.627 --> 01:58:32.997 12, with Windows 10 and sort of where this whole technology area is going. 01:58:33.169 --> 01:58:36.799 So first off, I'm going to have their chief technologist give us a little introduction. 01:58:37.682 --> 01:58:38.955 Let's go ahead and run the video. 01:58:39.196 --> 01:59:15.152 >> [Spoken in Japanese]. >> Domo arigato. 01:59:15.373 --> 01:59:18.792 So Square Enix has built Final Fantasy, over 100 million copies sold. 01:59:19.193 --> 01:59:22.479 This is an art and science technology demonstration they put together. 01:59:22.753 --> 01:59:25.524 What you saw was DX 11. What I'm going to do now is run through 01:59:25.798 --> 01:59:27.589 a similar scene in DX 12. 02:00:14.881 --> 02:00:18.289 Now, you might assume, looking at that, that that's also a video, but the truth 02:00:18.417 --> 02:00:21.196 is, that's in-game DX 12. So what we're going to do now is I'm 02:00:21.481 --> 02:00:24.032 going to go between manual and automatic, so I can shift back and forth, and when 02:00:24.278 --> 02:00:26.437 you see the word manual up there, I'm going to drive. 02:00:26.679 --> 02:00:28.730 So let's let this go through auto, and then I'll change the lighting, and you can 02:00:28.850 --> 02:00:29.754 see this real time. 02:00:31.683 --> 02:00:36.431 So as it pauses here, I can take over, so now I can scroll around, I can look down. 02:00:36.652 --> 02:00:39.125 I can change the lighting effects as they come through. 02:00:41.332 --> 02:00:42.752 I can move on to the next scene. 02:00:46.135 --> 02:00:47.554 Again, let it run in auto. 02:00:47.937 --> 02:00:49.727 And now I'm going to take it and drive around. 02:00:51.081 --> 02:00:52.675 Again, I can adjust the lighting. 02:00:53.276 --> 02:00:54.721 We can keep going through in this way. 02:00:54.879 --> 02:00:57.287 Let me bring the shadows in, take them out. 02:00:57.885 --> 02:00:59.359 All right, let's keep going through. 02:01:07.534 --> 02:01:10.151 I'm going to pass through this one and go on to the next one. 02:01:10.531 --> 02:01:12.800 And we'll see as we get closer, we'll look at the skin tone. 02:01:20.975 --> 02:01:24.002 There we go. >> Now, the thing that's really incredible 02:01:24.264 --> 02:01:28.169 about what you're seeing is just the density of data that's involved in this. 02:01:28.419 --> 02:01:33.678 Each of these scenes is over 63 million polygons per scene. 02:01:34.046 --> 02:01:36.368 That's about six to 12 times more than we 02:01:36.522 --> 02:01:40.366 could do with DX 11. Just to give you an idea on the textures 02:01:40.601 --> 02:01:46.194 that you're seeing here, those are 8K by 8K textures, again, significantly more 02:01:46.462 --> 02:01:49.356 than we're able to do. I think where it gets really interesting, 02:01:49.595 --> 02:01:51.369 though, is look at the hair. 02:01:51.615 --> 02:01:55.367 Every piece of hair that you're seeing is actually being rendered as a polygon. 02:01:55.727 --> 02:01:59.928 This isn't surface map stuff, and as those polygons are running through the 02:02:00.160 --> 02:02:03.626 pipeline, they're running over 50 different shaders to go through this 02:02:04.064 --> 02:02:06.665 and to generate that level of clarity on this. 02:02:07.114 --> 02:02:11.404 I think it's an incredible example of just how far people are pushing the limits of 02:02:11.731 --> 02:02:15.985 big data and technology, bringing it together and really building experiences 02:02:16.216 --> 02:02:18.768 that would be hard to build almost any other way. 02:02:19.120 --> 02:02:21.598 >> And so our thanks to the folks at Square Enix for letting us show this off 02:02:21.708 --> 02:02:24.196 and for building it. Incredible piece of technology and art 02:02:24.322 --> 02:02:25.064 put together. 02:02:25.447 --> 02:02:29.026 Now, that's sort of the high end, incredibly fun -- let's give those guys a hand. 02:02:34.317 --> 02:02:38.288 Graphic intensive side of gaming, and now for something completely different. 02:02:38.386 --> 02:02:41.125 If you go to the other end of gaming, one of the acquisitions we made earlier in the 02:02:41.364 --> 02:02:45.522 year was Minecraft, and Minecraft is a community game, lots of people like to 02:02:45.860 --> 02:02:47.356 work on it, mod on it. 02:02:47.818 --> 02:02:51.399 It's a place we get a lot of feedback to not change the modding side of it, to not 02:02:51.503 --> 02:02:53.324 change the community. So what we wanted to do was actually 02:02:53.513 --> 02:02:56.475 contribute to the community, so what we want to do today is announce the 02:02:56.683 --> 02:02:59.130 availability of a modding toolkit for Visual Studio. 02:03:00.440 --> 02:03:03.528 I'm going to pass on this slide and come back to it, since I paused along. 02:03:04.410 --> 02:03:06.736 And so I want to invite Aiden Brady -- 02:03:07.011 --> 02:03:09.309 that's a junior out of Atlanta. He's actually one of the top modders in 02:03:09.648 --> 02:03:12.365 the US, and Briana Roberts from our team. 02:03:12.591 --> 02:03:15.536 They're much better equipped to talk about this and demonstrate it than we are, 02:03:15.846 --> 02:03:17.196 so I'm going to have them come on out, 02:03:17.716 --> 02:03:20.269 and why don't you guys show us what's going on with Visual Studio? 02:03:20.399 --> 02:03:24.080 >> Hi, everyone. My name is Briana Roberts. 02:03:24.462 --> 02:03:26.067 I'm thrilled to be here today. 02:03:26.643 --> 02:03:30.692 I'm a senior content developer in Microsoft Learning Experience's academic team. 02:03:31.333 --> 02:03:34.865 We sit with the Developer Experiences organization led by Guggs. 02:03:35.365 --> 02:03:38.876 So why am I here today? Well, I happen to think I have the best job. 02:03:39.043 --> 02:03:43.742 I get to create awesome content to get students excited about computer science 02:03:43.889 --> 02:03:44.562 and computer programming. 02:03:45.248 --> 02:03:48.540 You may be asking yourself, what's hot with kids these days? 02:03:48.680 --> 02:03:52.348 And the answer is Minecraft. It is one of the newest ways to engage 02:03:52.485 --> 02:03:57.019 students and teach them about coding using mods like the ones we're going to demo today. 02:03:57.321 --> 02:04:02.145 Minecraft is an awesome education tool. It's being used over 5,000 schools, 02:04:02.250 --> 02:04:03.917 libraries and after-school programs. 02:04:04.535 --> 02:04:07.949 All right, without further ado, I want to introduce Aiden Brady. 02:04:08.208 --> 02:04:11.646 He's a high school student out of Atlanta, Georgia, and let me just throw 02:04:11.807 --> 02:04:14.860 this out there, but he's a big deal in the Minecraft modding community. 02:04:15.224 --> 02:04:19.912 His very first mod, Mechanism, has taken off. It's huge. 02:04:20.409 --> 02:04:24.309 So, Aiden, will you tell us a little bit more about yourself and your love for Minecraft? 02:04:24.460 --> 02:04:26.682 >> Sure. My name is Aiden Brady. 02:04:27.141 --> 02:04:29.106 I'm 17 years old and a junior in high school. 02:04:29.932 --> 02:04:34.108 I started programming in Python when I was 10 but moved on to Java when I was 12, 02:04:34.341 --> 02:04:35.792 which I've kind of focused on ever since. 02:04:36.057 --> 02:04:40.112 A friend introduced me to Minecraft in 2011, and it was kind of the perfect fit 02:04:40.274 --> 02:04:42.663 for me, because I already knew the language and I just loved the creative 02:04:42.860 --> 02:04:44.226 potential that the game had. 02:04:44.480 --> 02:04:48.343 I hope to later on get a computer engineering degree and maybe an MBA, too, 02:04:48.588 --> 02:04:51.793 to do what I love, which is maybe to work for a company as awesome as Microsoft. 02:04:51.986 --> 02:04:53.412 >> You're well on your way. 02:04:53.735 --> 02:04:56.917 Well, I don't know if everyone in the audience has played Minecraft. 02:04:57.316 --> 02:05:00.631 Maybe their kids have. Can you tell them what a mod is? 02:05:00.881 --> 02:05:03.516 >> Sure. A Minecraft mod is basically an extension 02:05:03.737 --> 02:05:06.236 of the Minecraft game. It can add new content, and if you're 02:05:06.479 --> 02:05:09.631 familiar with the game, it's things like new blocks, new items, but in our case, 02:05:09.880 --> 02:05:12.314 we're just going to focus on one new block, which is kind of exciting. 02:05:12.736 --> 02:05:16.963 >> As Guggs mentioned, Microsoft recently acquired Minecraft, and we're so excited 02:05:17.196 --> 02:05:20.994 about it, and we really committed to continuing this vibrant community of 02:05:21.242 --> 02:05:24.398 gamers that we have out there. So what we're going to today is show off 02:05:24.636 --> 02:05:28.107 some Minecraft modding using Java in Visual Studio. 02:05:28.479 --> 02:05:30.914 >> Yes, and this tool is fantastic. 02:05:31.284 --> 02:05:35.307 I recently switched from using Eclipse as my primary modding development tool, 02:05:35.715 --> 02:05:38.943 and honestly, it's the perfect transition. It's easy, it's intuitive. 02:05:39.082 --> 02:05:41.311 I love it. I feel right at home. 02:05:41.565 --> 02:05:45.117 And right here, you can see on the new project dialog, we have a few little 02:05:45.477 --> 02:05:48.659 templates here, and I love this. These are Minecraft Forge mod templates, 02:05:48.914 --> 02:05:52.341 so by clicking this right here, we're able to make a new Minecraft mod just like that 02:05:52.630 --> 02:05:55.286 and get started. And there's also one that's near and dear 02:05:55.591 --> 02:05:58.030 to my heart right here, which is the Mechanism mod sample. 02:05:58.514 --> 02:06:01.394 So you can mess around with what I've done. >> This is awesome for all these Java 02:06:01.636 --> 02:06:05.112 developers, that they can just both play and change the game that they love, 02:06:05.596 --> 02:06:06.943 all within our IDE. 02:06:07.558 --> 02:06:12.015 Now, if you're familiar with Visual Studio, then you're familiar with IntelliSense. 02:06:12.244 --> 02:06:15.796 We are going to show off some of the key features of IntelliSense that really bring 02:06:16.036 --> 02:06:17.823 this modding to life for these developers. 02:06:18.089 --> 02:06:21.760 I want to make sure I throw it out there that this is not a replacement for Eclipse. 02:06:21.930 --> 02:06:25.416 This is an add in that's been built so students just like Aiden, who want to 02:06:25.684 --> 02:06:28.147 build and create, can do so in Visual Studio. 02:06:28.735 --> 02:06:31.118 We're going to get into the code just a little bit. 02:06:31.235 --> 02:06:33.835 You've seen this with other languages. Now you'll see it with Java. 02:06:35.084 --> 02:06:39.768 So as I hover over a keyword, I get the quick info, which displays the complete 02:06:39.883 --> 02:06:41.488 declaration of code that I'm looking at. 02:06:43.164 --> 02:06:50.208 Now, as I type a word, and I'll type that period, a valid list of members appears. 02:06:50.681 --> 02:06:54.766 All I have to do is hit tab. It'll add that selected line into my code. 02:06:55.252 --> 02:06:57.799 Aiden, do you want to tell us about parameters? >> Sure. 02:06:58.167 --> 02:07:01.083 Parameter info here gives you information about the name, number and types of 02:07:01.361 --> 02:07:04.715 parameters required by a method, and in this case, you can see we have a parameter 02:07:04.837 --> 02:07:07.235 listed in bold. This is the next parameter required by 02:07:07.516 --> 02:07:08.990 the series of parameters of this method, 02:07:09.441 --> 02:07:13.310 and it's fantastic for new developers. I wish I had this when I got started. 02:07:13.564 --> 02:07:16.083 >> Yes. I can't even imagine learning to program 02:07:16.199 --> 02:07:18.242 with this kind of help. This is awesome. 02:07:18.486 --> 02:07:21.722 >> All right, but on that note, I say enough coding, and let's play something here. 02:07:21.965 --> 02:07:23.685 >> Okay, so what do we have now? 02:07:24.154 --> 02:07:26.760 >> All right, now, this is a little project that I've been working on. 02:07:27.412 --> 02:07:30.605 If we open this up here, you see we're going to run this directly through Visual Studio. 02:07:30.964 --> 02:07:34.148 And as this is loading, I just want to say, when you think about Minecraft, 02:07:34.403 --> 02:07:37.211 you think about the game, like I already mentioned, the creative potential that it 02:07:37.481 --> 02:07:39.691 has, and that's why I got into the game in the first place. 02:07:40.080 --> 02:07:43.032 You're opening this world and you can do whatever you want, build whatever you want. 02:07:43.310 --> 02:07:44.288 It's so cool. 02:07:44.535 --> 02:07:47.921 But, okay, getting in here, I'll open up our little test world. 02:07:48.312 --> 02:07:51.804 And in here, I'm going to open up my inventory, and you see the normal 02:07:51.912 --> 02:07:54.689 Minecraft TNT block right here. >> Sure. 02:07:54.881 --> 02:07:58.681 >> But we have a fancy TNT block here that's going to do a little bit more 02:07:59.014 --> 02:08:02.360 damage than what you might be used to, so here, I'll place a few of these down. 02:08:03.765 --> 02:08:08.629 >> It's perfect, because I happen to have this trigger over here, so I think it's 02:08:08.885 --> 02:08:10.166 going to help us with this explosion. 02:08:10.489 --> 02:08:12.812 It's a little smaller than when I ordered it online. 02:08:13.110 --> 02:08:15.111 Okay, I'm going to hold this. You ready? 02:08:15.363 --> 02:08:17.685 Go. >> Whoa. 02:08:18.416 --> 02:08:19.636 >> That was awesome. 02:08:19.879 --> 02:08:22.761 This extension is available. Make sure you go check it out. 02:08:23.003 --> 02:08:24.321 We cannot wait to see what you build with it. 02:08:24.487 --> 02:08:26.835 >> Thank you so much. What do you guys think? 02:08:27.229 --> 02:08:28.078 Did they do a great job? 02:08:30.263 --> 02:08:32.012 >> Thank you. >> Thank you. 02:08:32.127 --> 02:08:34.417 >> Thank you guys so much. We've got to get him back to school. 02:08:34.567 --> 02:08:37.286 I think he's missed two days of school because of us, so we want to make sure we 02:08:37.398 --> 02:08:39.442 did a good job. Look, we've got lots of great partners in 02:08:39.551 --> 02:08:44.754 this area, as well, on the tool side with Unity and Cocos, Trivia Crack, Cut the 02:08:44.888 --> 02:08:45.726 Rope, social games. 02:08:45.968 --> 02:08:48.119 I'm addicted to Trivia Crack. I hate to say it. 02:08:48.683 --> 02:08:52.924 Monument Valley has got a new game coming out. So, again, we've got folks pushing us, 02:08:53.407 --> 02:08:56.418 working with us, making sure we do a great job in this area. 02:08:56.683 --> 02:08:59.919 Look, as we get through this, then, I wanted to come back around to 02:09:00.166 --> 02:09:02.955 conversations with developers. You saw a bunch of different ones today. 02:09:03.208 --> 02:09:04.423 One of the things that John and I do, 02:09:04.661 --> 02:09:07.437 and our teams around the world, is we like to get together and code. 02:09:07.913 --> 02:09:10.014 We run hackfests. They're not competitions. 02:09:10.127 --> 02:09:11.686 It's not a hackathon. It's getting together with you 02:09:11.927 --> 02:09:14.094 and building things. All the companies listed here and groups 02:09:14.365 --> 02:09:15.805 are groups we've worked with. 02:09:16.175 --> 02:09:18.837 We have space up in Seattle that we use for doing this. 02:09:19.232 --> 02:09:21.556 I'm happy to announce that we're going to open a space here in San Francisco under 02:09:21.764 --> 02:09:24.530 the Reactor label. That's why I changed my rugby jersey -- 02:09:24.687 --> 02:09:27.485 so that we have a space here. Obviously, I'd love to have space all 02:09:27.727 --> 02:09:30.239 over the world. We don't, but we are available all around 02:09:30.377 --> 02:09:31.947 the world. So we're going to make sure we show up 02:09:31.947 --> 02:09:32.689 wherever you are. 02:09:33.065 --> 02:09:35.951 The Reactor space will be here in San Francisco. We've got one in Seattle. 02:09:36.215 --> 02:09:39.722 And then as a team, we're going to take the content from Build, the things you're 02:09:39.874 --> 02:09:41.088 going to see the rest of the day and tomorrow. 02:09:41.368 --> 02:09:44.565 We'll be going to 26 cities around the world, try and show up where you are. 02:09:44.854 --> 02:09:48.723 In addition, we'll be either hosting or showing up at thousands of events around 02:09:48.963 --> 02:09:51.880 the world over the next year, to bring everything you saw from Joseph's work to 02:09:52.129 --> 02:09:56.082 Kevin and David to everybody else that works with us in the company, the Office 02:09:56.254 --> 02:09:57.207 Graph and all the other pieces. 02:09:57.937 --> 02:10:01.353 So with that, I want to thank you for your time today. 02:10:01.605 --> 02:10:04.009 We're going to have a maker video as our walk off. 02:10:04.279 --> 02:10:06.956 The thing I'd like to say, I think, which Joseph has teed up, if our meme this 02:10:07.159 --> 02:10:11.037 morning was more code, then in the words of Christopher Walken, I think our meme 02:10:11.325 --> 02:10:14.287 for the rest of the day and probably next year is more cow. 02:10:14.635 --> 02:10:17.166 With that, I wish you well. Have a great day. 02:10:17.652 --> 02:10:19.429 [Begin Video]. 02:10:20.170 --> 02:10:21.960 >> Making really is about sharing 02:10:22.177 --> 02:10:24.766 and about a lot of collaborators. 02:10:25.257 --> 02:10:29.288 >> What I love about being a Microsoft maker is what Microsoft does to enable us 02:10:29.520 --> 02:10:33.072 makers here at the company. >> You can come in and there are many 02:10:33.194 --> 02:10:36.636 different people from all different sorts of expertise, and they can all help you 02:10:36.765 --> 02:10:38.018 out in various sorts of projects. 02:10:38.365 --> 02:10:41.721 >> There's been a tremendous opportunity, especially in the last few years, 02:10:42.081 --> 02:10:45.481 of opening up beyond what Microsoft has traditionally focused on. 02:10:46.083 --> 02:10:48.530 >> B15 is a robotics platform that we 02:10:48.684 --> 02:10:51.231 built in conjunction with the HoloLens team. 02:10:51.457 --> 02:10:55.235 >> A lot of the IoT devices are headless devices. They don't have a display. 02:10:55.534 --> 02:10:59.636 But when you put on a HoloLens, that can become the display to not only view the 02:10:59.888 --> 02:11:01.633 rich data coming out of the IoT devices, 02:11:02.138 --> 02:11:03.999 but also to do command and controller. 02:11:04.218 --> 02:11:07.355 >> There are now increasing emphases on the Internet of things here in the 02:11:07.600 --> 02:11:10.045 company, and as part of that, we had to build a platform for us to work on. 02:11:10.269 --> 02:11:14.031 Micro Framework has become the Microsoft answer for resource-constrained devices, 02:11:14.611 --> 02:11:17.838 with Windows providing a solution on larger devices. 02:11:18.373 --> 02:11:23.123 >> When makers got a chance to experience Visual Studio 2015, they were delighted to 02:11:23.484 --> 02:11:28.490 see all of the capabilities and power of Visual Studio 2015 available for building 02:11:28.962 --> 02:11:29.819 IoT applications. 02:11:30.457 --> 02:11:35.422 >> My son did a picture of the sun and tree and landscape, and what virtual 02:11:35.685 --> 02:11:40.438 shields allows him to do is to take the current location, go to the web, get the 02:11:40.768 --> 02:11:45.124 weather forecast and actually turn on lights behind his picture on an LED strip. 02:11:45.528 --> 02:11:48.433 It feels good to tie that into the virtual shield work I've been doing. 02:11:48.686 --> 02:11:52.449 >> I've been building Tesla coils since I was a small kid, and it's been a great 02:11:52.698 --> 02:11:55.204 passion of mine. Because my Tesla coils are electronic, 02:11:55.699 --> 02:11:59.126 that means I can control them with some sort of electronic device. 02:11:59.498 --> 02:12:00.694 So I worked together with Jason who 02:12:00.853 --> 02:12:05.499 creates great music on his own free time, and we created this project where Jason 02:12:05.720 --> 02:12:10.129 will have his Surface Pro pumping out beats and music, and this could talk to my 02:12:10.364 --> 02:12:18.527 coil and make them sing. I'm a maker. 02:12:18.698 --> 02:12:20.127 >> We are makers. 02:12:20.557 --> 02:12:22.731 >> We are makers. >> We are makers. 02:12:23.018 --> 02:12:25.212 >> We are makers. >> We are makers. 02:12:25.560 --> 02:12:26.889 >> We can't wait to see what you'll make.