Embedded - 110: Happiness Is a Warm Puppy

Episode Date: July 22, 2015

BeagleBone's Jason Kridner (@Jadon) returns to tell us about his new book. Jason co-authored a new book: BeagleBone Cookbook: Software and Hardware Problems and Solutions (or at O'Reilly). His older book is Bad to the Bone: Crafting Electronics Systems with Beaglebone and BeagleBone Black. Previous Embedded.fm episode 60: Fun Things You Can Make out of Beagles BeagleBoard.org's Google Summer of Code page (including BeagleSat and underwater drones!) Some information about putting Xenomai on a BeagleBone Black for real time response. Chris mentioned Brillo, an alternative Google supported OS that isn't on the BBB. Project Ara: an open source smartphone Ardupilot: Autonomous drone piloting.  Dronecode: Drones in Linux OpenROV: Underwater vehicles Mars lander Beagle 2 (the Apollo 11 Lunar Module was the Eagle despite some comical confusion). [UPDATE: Listener Mark Stevens pointed out that the Apollo 10 Lunar Module was named Snoopy who was a beagle.] TI's E2E Forums BeagleBone Green

Welcome to Embedded FM, the show for people who love building gadgets. I'm Elysia White, here with Christopher White. Jason Kreidner has returned with a cookbook for dogs? Well, I suppose since we talked last time about fun things you can make out of beagles, this may be an improvement. Unless it is a cookbook like To Serve Man. Then maybe not. We'll see. Hi, Jason. Good morning, good afternoon, whatever it is for you.

Midday, right around lunchtime for me. Could you tell us a little bit about yourself? Sure. I'm Jason Kreidner of Texas Instruments and the BeagleBoard.org Foundation. And I work with the BeagleBoard.org Foundation. And I work with the BeagleBone Black quite a bit, and that's an open hardware computer for makers, right? From, I like to say, kindergarten to Kickstarter, right? So everywhere in between and beyond.

So it really lets you do all the things that you can't do with a bunch of the single board computers that are out there today. It's focused on the fact that it's real open hardware, and it's capable of doing real-time stuff at the same time as running Linux. It's got a nice big community out there of other people developing to kind of help get going, and I think it has the best of the out-of-box experiences for getting started with some of these single-board computers. It's got a bunch of add-on boards. Everybody comes up with clicky names for their add-on boards. You can't just call them add-on boards or daughter boards anymore.

You've got to come up with a short, pithy name. So ours are Capes. Lots of people have done clones. We get started real quick with the Cloud9 IDE, and I think it's the easiest way to kind of get started with programming and building with electronics. That's kind of what it is.

Cool. And you, I mean, it's often compared with the Raspberry Pi, except the BeagleBone Black is open source. That's probably the biggest difference. There's a lot more things that comes with what it can do that are different. Just kind of its raw capabilities

and the experience when you get started with doing things with it. It's much more of an embedded system type device. Sure, you can put a keyboard, monitor, and mouse on it, but it's really geared at making electronics and programming rather than, oh, it's just another desktop computer that's cheap. So that's where we're really different, right? It's open hardware, so you can actually change it all to be all exactly what you want and understand all the little bits of it.

But it's really about making things rather than just having another screen. Do you think that having it be open source helps your adoption rate? I do. Certainly, there's all sorts of reasons why it does. These things tend to self-select the type of people that have interest in them.

And that means that our audience is a bit more of the kind that wants to do more things with their hardware. They're more fundamentally big believers in openness. So those people become more attracted to our platform. Or they want to actually do something professional out of it. So they don't want to have a lot of barriers to doing things that somebody else decides. So we certainly see, I think, a big increase in adoption. And we also can't measure a lot of our adoption because a lot of our adoption isn't just straight board sales. It's not all about just the almighty dollar of buy a board because a lot of the people that are in the BeagleBoard ecosystem

are ones that are actually using derivative boards or variations of our boards, but they're running some similar software. And also because a lot of things are Linux, we can share back and forth to a lot of the other single board computers. So our ecosystem is bigger than people that just have BeagleBone Blacks.

Do the derivatives get in touch with you and say, hey, we're using your design so you can keep track? No, some of my clients have said, we're using BeagleBone because we're going to use it in the future, and we want to use our own board, and we're going to strip off all this other stuff. So I don't think… Well, I just meant out of kindness. Yeah, we do, but it's a small percentage of the peopleiki pages, like there's a TI processors wiki that has a list of all these AM335 modules or different derived boards that you can get.

And a lot of those guys are really leveraging the software that's coming out of the BeagleBoard community, a lot of the documentation, a lot of the experience that's coming out of there. And it's probably in the order of about 5% that actually bothered to contact me. I should be part of the license. You're required to say thank you. I mean, that's always an interesting debate, right? Do you want to have licenses that push you back or are you just trying to make it more freedom? So we're a little bit more on the BSD sort of side of that argument. Well, yeah.

And I mean, B-Goals are TI things. And the goal is to sell TI parts. Well, it's a... Part of the goal. The foundation is independent. And most of our board members are not part of TI. Our only full-time staff member is not an employee of TI. So there's certainly a tie-in to TI, and a lot of what BeagleBoard is is coming from that relationship with TI. But it's not just driven by TI.

Yeah, I mean, it's a lot of the sponsorship that comes from TI is because of that promotion of the TI processors, but that's not what it's all about at the end of the day. No, because at the end of the day, you have a good single-board computer that's flexible and relatively inexpensive. So TI did a good job launching you that's flexible and relatively inexpensive. So TI did a good job launching you, but now you're on your own. Yeah.

So you wrote a book or collaborated on one. I did. This is actually the second book that I'm listed as a co-author on. The first was Bad to the Bone and this one's The Beaglebone Cookbook. There's a good reason I'm listed as the second author in both cases. The professors who were listed as first author did most of the heavy lifting and really a vast majority of the work. So with Bad to the Bone, it was Stephen Barrett. And this new book, it's Mark Yoder. How did you meet up with him? So Kathy Wicks, the TI University Programs Manager, actually introduced him to me.

And Steve, Mark, and I have been to a bunch of training sessions together. We've actually taught university professors how to teach courses on embedded Linux and kind of brought them up to the speed and why embedded Linux is cool and what types of things you can teach the students and how you go about doing that and just getting them familiar with what you can do with embedded Linux. So we've taught those classes together and some of it was from lessons and teaching the professors. In some cases it was just a lot of our common community experience.

Mark has been actually publishing course materials on a public wiki for years now around the Beagle Boards, the older Beagle Boards, and had moved all that material over to the Beaglebone Black and has been teaching his class assistants. And he does it in a way that's really open. So a lot of the discussion, he has all of his students get out to the BeagleBoard mailing list and start to interact with people. He gets them to publish all their projects and open source repositories. So he's been having them not just learn about embedded Linux, but really learn about working

with an open source community. And that's allowed us to get additional feedback on the types of things he's teaching from not just his students, but from experts in the industry, other people within the Beagle community. And so it's a lot of those lessons learned that kind of help guide this book to try to have as broad an audience as possible, but not waste experts' time, just to get people up to speed quickly and kind of finding what they can do with embedded Linux.

Well, and did we ever say the title of this book? I said it was BeagleBone Cookbook, but I'll go ahead and spell it out about three or four more times. It's BeagleBone Cookbook, Software and Hardware Problems and Solutions, not about cooking dogs. How tired are you of that joke? I'm actually not tired of it at all. Maybe I have a pretty warped sense of humor, but I like some off-color jokes.

I love beagles. Please don't think that I actually want to ever cook a beagle, but just those sort of contradictions are an enjoyment factor for me. And it's an O'Reilly book, so you have an animal cover. We do, and it's a beagle. I'm so jealous. I had to argue about my animal, and I bet you just got a beagle, and'm so jealous. I had to argue about my animal and I bet you just got a beagle and it was cute. Surprisingly enough, we did

kind of debate about it a bit and it's not because whether it's a beagle or not, but he's such a serious looking beagle. I mean, beagles are work dogs, right? Oh yeah, ours works a lot. Getting treats.

This is news to me They're supposed to be work dogs Supposed to be And this is a very You know, it's a very serious looking dog And beagles are work dogs And you know, it actually makes a lot of sense You know, relating it to beagle bones I mean, they can do very much serious work

But it You know, he didn't look as friendly and cuddly as, you know, what we wanted. So we were actually trying to look for a little bit more, you know,

push a Riley to give us a little bit more cuddly dog and that just didn't happen. They had a dog and they stuck to it and it's, I just love the fact that it's a beagle,

right? So I really shouldn't, shouldn't possibly complain. So don't take this as complaining, but there actually was a little bit of debate to get a more friendly, cuddly beagle. He does look like he's standing at attention for the Westminster dog show or something. He does. Not at all giving you the beagle eyes that are controlling in many ways.

Yeah. He doesn't even have his tongue sticking out, right? I mean, come on. But it is a cookbook style book. It's recipes and you come at it with a problem and it shows you how to solve it. And it isn't read through exactly. Yeah. You can read it cover to cover. I think it actually makes some sense to do it cover to cover. It wasn't thrilling read that way. No.

I mean, there wasn't really a story arc. No, there's not. No subplots. You could jump into it at any point and you're going to be able to understand it and follow it. It is, yeah, there's not a lot of drama as you go through it. But I remember last year we talked about me having trouble finding consistent sets of instructions. And that's what this was. I mean, it was, how do I do this?

And it just told me. Just work. And it's very much geared to that so that you can get some success on each of the examples that are provided. And it's the types of things that we see people want to do. And so that they can move beyond blinky light, right? So they can actually read a sensor, move a motor. And it's just all told from the perspective of interacting with electronics. I've got a computer in the center, and I want to hook up all these sensors and actuators.

How do I go about doing that? And there is some building on as it moves forward. And you might not want to do some of the later things until you do some of the earlier things. Oh, yeah, because it starts out with basics and sensors and displays and motors. And those are all pretty simple, and they're all using, is it Bonescript, the Java-ish thing? JavaScript-ish thing. Yeah, so Bonescript isn't a language, it's just a library.

It's a Node.js JavaScript library that can be run, you know, it's got a browser side extension so you could run it

from a browser as well just using socket IO commands or socket IO to transport essentially remote procedure calls. But,

so, all that making it more complicated than it needs to be. There's just a JavaScript library and that's what we use to build examples. It's not saying that, gosh, I'm going to build a real embedded system,

make a product out of this, that I'm going to use JavaScript to go and do this. I'm not saying you can't, you absolutely can. But it's about that first pass success. So move the language to try to not be too much of an issue or a problem.

And it gets you out of the business of oh, here's the tool set you need to install. Here's the compiler and here's what we suggest for edit. No, here, it's JavaScript. Yeah, and it's already installed on the boards when you get them and there's some tutorials on the boards when you get them, and there's some tutorials on the boards that you can follow.

It could be considered an arbitrary choice. I have a bit of a vision as to why I really push JavaScript to people as a first programming language to learn if you're new to programming. And that's because it's absolutely pervasive.

It's everywhere. Things are built off of HTML and JavaScript and CSS, and that's the language of the internet. You can be used by all types of professionals, even embedded professionals. I think it's really valid for embedded system design. The Bonescript library might not really convince you of that, but

it is. The idea here is just to make sure people can have some success talking to the different sensors, talking to motors, and then the hardware is working, and then they can start diving into some of the other deeper software components. And it's all open, so you can follow pretty easily.

If you read C, you can read JavaScript. It's different, but it's not so different that somebody who can read C can't figure this out. And for somebody brand new to programming, my gosh, there's some amazing tutorials out there for learning JavaScript. The Khan Academy stuff is fantastic. There's W3Schools stuff. I mean, there's just so many resources out there for learning JavaScript as a first programming language.

And Node is cool in the same way that Python is. There's all these packages out there. So if you're not super experienced, but you want a piece of functionality, there's probably an npm install. I just get this library and then I can call it and I can interface with this REST API for this thing. So you can build some Internet of Things devices

out of this probably with a few building blocks. Yeah, and I'm a big believer in kind of the asynchronous style of JavaScript, and that's one thing you're missing from Python. So the idea that it's just an event loop, and that's the way you make real-time systems. Okay, yes, garbage collection, forget about garbage collection. I hope I'm not getting too technical,

but not for you guys, I'm sure. But garbage collection is, yeah, that's not something you really want so much in an embedded system. There are ways you can kind of avoid the garbage collection problems. But setting up events and handlers is absolutely the way you want to build embedded systems. And JavaScript language is designed to do that really nicely. And Node.js in particular, everything's done in all the libraries. You've done it in an asynchronous manner.

I mean, you just set up events to handle, okay, a pin goes high, it runs this function. It's just that simple. And it can be structured nicely and cleanly. So you can have all the different types of events that might occur in a system and assign them handlers and have them talk to each other in a really clean way. When you're writing interrupt handlers and have them talk to each other in a really clean way. When you're writing interrupt handlers, you get this headache of, well, if you touch globals, you might have race conditions and all these other nasty things.

JavaScript makes that stuff go away. But you don't stay in JavaScript for the whole book. Absolutely not. We talk about C libraries, including LibSOC, LibSOC, for doing things in C. We talk about Python. And all the examples in the book, we've actually,

I don't know if we've completed this out. I have to actually go look at that because I think we're going to try to do all the examples in Python and C and all of them. I think that's a promise that we made that we might have forgotten about a little bit. But all this stuff in Python and C and all of them. I think that's a promise that we made that we might have forgotten about a little bit. But all this stuff is possible in C and in Python.

And we give examples on how to get started with those and point you to the libraries that you can use to kind of do the same things as we're doing in JavaScript. Well, in the Libsock and C, actually most of the C stuff seemed in the real-time I.O. Well, in the LibSock and C, actually most of the C stuff seemed in the real-time IO section, which was kind of a

you can have a real-time IO section in Linux? Oh, here we go. Oh, absolutely. So we take a couple different

solutions towards real-time, you know, because real-time just means you have to be fast enough to you know respond to the problem at hand right so it doesn't always mean you know real-time kernels or you know other other things that some people associate with real-time it just means that you're going to have a deadline. That deadline might be milliseconds away or even seconds away, and you just kind of have to scale your approach up.

So we deal with it by degrees where we kind of baby step you along a little bit faster and faster and faster ways to go or more predictable, right? So it's about predicting latency. Exactly. And the, like, xenomai is a… Is that how you say it? That's how I say it. X-E-N-O-M-A-I, and I was just like, hmm.

So, yeah, the X-E-N part, I say zen. Yeah. And then O-M-A-I. I don't know how to do Zen-O. Zenami, I think, is valid. Ooh, that actually sounds nice, too. Zenami. Okay, we're going to go with that. But it's A-I-M-A-I, so Zenami?

I don't know. I don't know. Is it? I don't know. English, let's go back to engineering. It's a kernel that actually sits in parallel to Linux. So you patch the Linux kernel so that you have a way to talk between that RTOS and tasks running in Linux. So it gives you a way to bridge between the two.

But essentially you're not running the real-time stuff in Linux. Essentially it just runs in a different RTOS that just knows how to talk to Linux. And you just capture higher priority interrupts with it. And so you have a timer tick that's higher priority than the stuff that gets assigned to Linux. And you just run stuff at that higher priority. And then you talk to Linux.

And that approach has been done by a number of different solutions. We just have a really nice community around the Xenomi, Xenomai, whatever it is. Mostly this toolkit called MachineKit. Hopefully I'm not getting too deep here and throwing out too many different terms. But MachineKit is a fork of Linux CNC. So that's software that's used to control mills and lathes, 3D printers,

all sorts of really, really cool stuff. And it's actually one of these things that the BeagleBone is magically suited for almost. I mean, it's wonderfully fitting for doing these things. And so we have a nice community of people using that particular Xenomai, you know, RTOS and parallel to Linux. So they put just a small number of threads on the Xenomai side in machine kit. And that allows you to have a nice fancy GUI for controlling your machine at the same time as driving all the stepper motors and everything in the system in hard real time.

They also make use of another thing we talk about in the real time chapter, which is those PRUs or the programmable real time units. Those are a couple of microcontrollers that are sitting on the BeagleBone that are completely idle and open for you to use. They have single cycle latency to the IOPins, to a

number of the IOPins. And so you can use those PRUs to handle some of the real-time tasks in parallel to the main Linux processor. And they're running at 200 megahertz, so you can actually sit there and do toggle high, toggle low on IO pins at 5 nanoseconds. So if you wanted, you can generate 100 megahertz waveform out of those PRUs. And people have used that on the capture side

to actually build a 100 mega sample per second logic analyzer. It's doing 14 channels. Okay, working too much off, but the idea is, so you've got the real-time kernel, but you've also got these microcontrollers in there, and that machine kit software will actually use both so that the stepper motor pulses are generated by the peer use and the math calculation, the path planning stuff is done in a real-time task

on the ARM Cortex-8 that's running at a gigahertz, so you can do some really fancy path planning, and they put it all together in this system. And we talk about some of those individual components right in that real-time chapter. So we talk about the Xenomai, how to get started with some of that. We don't go very deep. No, you really don't.

But you do go through how do you patch the kernel for it and how to configure Linux to use it. And that's meant to just kind of introduce you to these key technologies and kind of let you know which are the ones that you should really know about and hopefully give you a little hint as to why. And then there's a lot.

You mentioned in some of the prep that, wow, it seems like we're sending people off a lot to websites. And we really are. But the idea here is to make sure that they have a solid background onto why we're sending them off before, so they even know what to go look for. Exactly. So just to wrap up the real-time thing. Some people feel that if you're running Linux or a Linux-like operating system, that you're no longer a real-time or embedded system. I think it's clear that this is designed

to do all kinds of things in that whole space and that those people might be wrong. Quit pointing at me. Well, okay, so A, it is using a different kernel, the Xenomai, and B, it is using two ARM microcontrollers. I'm not saying it's not. PRUs. They're not actually part of Linux

at that point. Oh, okay. Fun debate here. So with the 4.1 kernel, there's also this thing called there's a few different levels of real-time with Linux. There was somebody at the University of San Diego

that actually took, they're sampling an IMU, inertial measurement unit, over I2C in a user space task, right? So not in the kernel, not in some real-time thread. And they're relying on, I think it's a two millisecond,

actually, heartbeat. So two millisecond heartbeat in order to create a balancing robot. And they're using a user space task to go out and grab values off of this IMU to tell if it's starting to lean one way or the other. So kind of tell what its orientation is. And then drive DC motors, right?

It's actually capturing a little bit of feedback using the quadrature encoders, so it knows where the wheels are. So there's actually some hardware quadrature encoders on the BeagleBone Black. So reading those, they're reading the IMU, and in a user-based task, they're able to actually make a robot

that's actually really hard to knock over. They call it the BeagleMIP, or the Mobile Inverted Pendulum. And they've published their software on this and kind of how to do it. It's all just a user space Linux task. So when we talk about not being able to do real-time stuff in Linux,

I mean, you're able to change the priority of threads, boost them up, and even create real-time threads within Linux. Now, there is still an element of unpredictability within an unpatched Linux kernel. Right, and that's where we come into real-time, like real-time with a capital R and a capital T.

Well, this robot, I mean, it's soft real-time in this case, but this robot doesn't fall over. Well, no, it doesn't fall over because the BeagleBone is blindingly fast. Right. Not because Linux, I mean, Linux has never starved for anything. So, of course, everything gets handled. It's what happens when things get starved. When you start adding I.O., right?

When you start adding different parallel tasks that might cause something to sit in an interrupt handler or some thread for an extended period of time. And Linux does have a preempt capability. So that means that you can actually, if you turn on preempt as a build flag in the Linux kernel, you can actually go and interrupt the kernel that's running. So you can't just have a kernel thread come and steal all the time. And then you're only kind of stuck with the interrupt blocking. So if you have some long time where interrupts are disabled, and there are some drivers that can be kind of badly formed.

But if you make the kernel preemptible, that solves a vast majority of your real-time problems. And then the next kind of level beyond that is actually making your interrupt handlers themselves interruptible. And that's where you become a real-time kernel, right? That's the RT patch. So if you've been around Linux, you've heard of preempt,

and maybe you've heard of preempt RT. So preempt RT is a patch to the Linux kernel to actually go and enable those. And that's not yet mainline. But a lot of the people using the BeagleBones actually do apply that preempt RT patch. And we've been trying to get closer and closer to mainline. We have a very nice 4.1 Linux kernel.

And just recently, the preempt RT patch has been released for 4.1. So if you do that, that makes Linux a real-time operating system. It's not yet the mainline Linux, but you can actually patch Linux to be a true real-time operating system it's not yet the mainline linux but but you can actually patch linux to be a true real-time operating system and not just have this funny parallel kernel thing and that's i mean that's not only fast because you can get around a lot of latency issues with just being fast that is truly the capital r, capital T real time. Where you have a fixed latency, you can say that this interrupt will never be handled later than this amount of time.

Yeah. And that's short of some catastrophically stupid bugs and software bugs in the system. But if you understand most of your system, you can still, without the RT kernel, guarantee real-time if you kind of understand what the software and worst-case latencies are that you're going to build into the system. I mean, RT is really just a workaround for not understanding what your worst case interrupt handler time is.

So it's still just kind of, it's making it hard real time by saying, well, okay, well, I don't have to understand all this stuff. If you really, really, really want to make a real time system, you just have to understand everything that's in the system. Like what is, what are these elements that are going to cause me to get into states where I don't respond in time? And you know, having that RT patch just reduces the number of things that

you have to know about in order to make that true. Exactly. Making it easier for people, which is what the Beagle Bone seems to be about. Very much so. Yeah. For all types of problems. And for people that don't want to run

Linux, I think Linux does make things easier in that it brings this huge community around it where people are trying to solve this huge variety of problems. And that's what makes Linux interesting to take as an approach. But if you want to take all this stuff on yourself,

I mean, running without an operating system, you'd save a lot of money because you don't have to buy that stupid memory chip that sits out there if you take Linux away. Well, and there are other operating systems that run on the BeagleBoard. We've had the RTEMS, Joel from RTEMS. Yeah. That runs on a BeagleBone Black.'ve had the RTEMS, Joel from RTEMS. Yeah.

That runs on a BeagleBone Black. So if you want something that's been to space, you can have that. Absolutely. Yeah. No, RTEMS is fantastic. And they're doing some more work on their BeagleBone Black port this summer as part of Google Summer of Code. We've tried to support their students a little bit on that. So, ArtTems is fantastic. It's open source. It's got a good community. But there's a number of other ones. I'm going to do a controversial shout out

right here just because, did you know that Windows Embedded, like Windows CE or Compact Embedded or something like that, is actually a true real-time operating system? That was my head hitting the desk. I appreciate that. I haven't heard anything since you were talking about

RTEMS. Did you get lost in RTEMS? I think you just didn't want to. Oh, I see. But there's a bunch of other open source ones, right? So BSD runs on BeagleBall, and I'm very happy about that. While we're on the topic of operating systems and Google, I was wondering if you've been following,

they've been talking about an Internet of Things operating system called Brillo for a while. Is that something you're following or that could become a possibility? Uh-oh. Hadn't heard of that. I should have.

It's supposedly a cut-down version of Android for use on embedded systems. Maybe I have heard of it. I don't know if it's available yet or if they're still just talking about it. But yeah, you should check it out. I like that Brulow was the highest search.

What you're talking about was the first thing on the search page, but all the pictures are for this scrubby thing. It's a terrible name. I know Google uses BeagleBones a lot for their development. I was actually really impressed with a lot of work they did for Project Aura. I don't know if you've seen that.

It's the open source phone that they partnered up with. Somebody did a big Kickstarter for an open hardware, open source phone. And there's been a lot of open source phones built around Beagle. The OpenMoco, they did a version. They took it,

so the GTA04, they actually took an original Beagle board and kind of merged it into the older OpenMoco and made the GTA04. So there's been a lot of efforts of doing open phones with Beagle.

I mean, used the same processor for the original Beagle that was used in the Nokia N900, which is one of the nice open phones, and then the N9. Anyway, so we have a lot of connections

to people doing open phones. And in Project Aura, they did a lot of the development work for that interconnect bus with BeagleBones. They built a lot of the test infrastructure, a lot of the examples with BeagleBones. And that one's a pretty fun little project. I don't know if you've seen the interconnect. You guys like to play with hardware, right? You should check out the interconnect bus that they built for Project Aura. Cool.

You mentioned Google Summer of Code. You've got some of those students too. We do. So we've got eight projects going this summer. We had, I think we ended up with six last summer. And that's absolutely one of my favorite parts of working on BeagleBoard is the GSOC students. That's a program where Google sponsors the students to write open source code, open source software over the summer. And they get like $5,000.

It's a full-time job for the students over the summer, so we do see a lot of people doing that in other countries around the world. They have to be enrolled in a university to participate in Google Summer Code. But they do it from their home location. They don't come to you. They do. They just do it right where they're at. And we interact mostly on IRC and on our mailing lists.

We occasionally will do Google Hangouts or some other things like that to kind of interact with each other. But mostly it's all just good old IRC internet relay chat. So we like that. And the projects are fantastic.

You have a Beagle set and a Beagle Pilot that's an underwater drone. Those are my two favorite of your list. So I'll start with the Beagle Pilot 2.0. Because last summer we had a Beagle Pilot where they ported ArduPilot, which is an autopilot software for Arduino, for the Arduino boards and microcontrollers based off of that. That's used to essentially do flight planning. So there's this APM, there's this flight planning software that you can run on your computer and talk to the boards and set them to essentially fly completely autonomously.

And they'll do all the keeping the vehicle level, sending it different directions, reading the GPS values, and kind of dealing with all that stuff. And they ported that over to BeagleBone last summer. Since then, we saw the Linux Foundation actually pick up and do drone code. So a lot of that work in porting the RGPilot over to Linux is that basis for doing the drone code project for the Linux Foundation. And so you see the same people kind of leading up that effort. And so we've got a bunch of drones built out of BeagleBone Blacks that people are using.

The guy that was one of our GSOC students went out and started a company called Early Robotics. Now, one of the reference platforms for Ubuntu Snappy, which is this kind of container stuff. And I could just keep dropping crazy technology terms and stuff. It's not really what I want to do. They're doing some cool stuff with drones, with Beagle and all that work

is continuing to influence a lot of other projects. Now they've turned their... Those guys have turned their attention to underwater drones. Just a little bit. They're still doing a lot their attention to underwater drones, just a little bit. They're still doing a lot of times in air drones. But they said, okay RGPilot code that's running on Linux and moving that into an underwater vehicle. And OpenROV is the underwater exploration robots and it's remote operating vehicle.

Yeah, that's the ROV term is remotely operated vehicle. And here I guess they're trying to turn it into a remotely autonomous vehicle or something like that. Yeah, so it's all open hardware. They came out of a

Kickstarter. It's one of my favorite BeagleBone projects. As I was quote prepping for the show, I came across a shark named Poor Beagle, which is a portmanteau of porpoise and beagle. I just wanted you to know about that for when they're naming their projects. See, I'm starting to think I was born at the wrong time because when I was a kid, I

remember trying to design a submarine that would live in Dana Point Harbor and so nobody would know it was down there and it would go around autonomously and spy on the boats and stuff. And that was completely impossible that way. I'm just thinking, well, that would be pretty easy. No. Okay, that would be doable. Don't discourage him. Don't discourage him.

Don't discourage him. I spoke with an underwater, with the OpenRV, one of the people who works on it, a couple of years ago. And he's like, it's really not very hard in my bathtub. Don't have a lot of places to go. Yeah. Dealing with those environmental conditions, I imagine, is a little tricky. I would like to try it out, though. Yeah. In some place like

Lake Tahoe that's so clear. I don't know about going in an ocean. There's a lot of potential for poor beagles. You can't be nefarious in Lake Tahoe. Don't drown them. Don't mistreat them. But going into space with the BeagleSat, it's sort of a harsh environment too.

Yeah, and that's actually a huge interest area right now around Beagle. So a lot of the people developing the CubeSat containers have actually used BeagleBones to kind of test their satellite, to test the payload connections, because they'll have different USB and serial connections to the CubeSat payloads. And a lot of them on the ground are using BeagleBones to test those. And they're looking to try to provide references for people to actually what they can

put in payloads into outer space. And we've had quite a few launches. I'm trying to find the documentation of how high it's been. And I've not been able to, if anybody hears this, it's launched a BeagleBone

black into space, please get back in touch with me. But I know a lot of people have been looking at that problem. And I think it's actually had a few launches, but I'm not finding good documentation of that. I've got documentation of high orbit balloons. People have taken them up and used them to take pictures, collect data. A lot of the microsats are just tossed out of the International Space Station.

So I don't know what you know about the CubeSats themselves. Yeah, that's what I was thinking about. The CubeSats are these 10 centimeters by 10 centimeters by 10 centimeter cubes. So it's kind of a standard format. And yeah, they attach them to the International Space Station, but that's not, I think that's not the only ones, but they launch up these ones. And getting

space on one of those is actually pretty affordable. I mean, affordable like a car affordable, but affordable. Not like, oh my gosh, I'm doing your own satellite launch, I'd say most of your listeners probably can't

afford that. If they can, they should contact us. We've had some local high schools, and they do funding drives, and every other year they end up with one of these little satellites. And so it's a big

community effort, but it's also something, if you have more money than cents, it's fun. I shouldn't say that because Stuart's going to be on the show and I think he's working on a CubeSat. So it's also really great if you're Stuart. Alright.

So the challenge with the BeagleSat guy, that project is working on is really compensating for some of the interference that comes from being on the satellite. So the satellites themselves actually generate some RF signals and stuff that can get in the way and they're essentially trying to mask out those um i guess it's a common problem when you're you're stuck really close you can't be put out on a far boom to get away from the other sources of noise and just kind of filtering out those in your in your data collection um well this is space hardened hardware either there's a lot more radiation and things that you have to deal with. There may not be a problem on these things because they're... They're not meant to last. Right.

Right. And so a lot of people are willing to put up this consumer-grade hardware because it's just like, oh, we'll collect some data, send it back, and when it burns up, it burns up. And so a lot of people are perfectly willing to get started with that, but the interest level is really increasing. What is it going to take to space harden a BeagleBone? We've gotten some pretty hard recommendations

from some of the space developers of things we can do with it. There's a lot of legitimate commercial interest in BeagleBones in space. Well, the heat dissipation or high, large range of temperature, that's got to be something you hear. Because BeagleBones in cars is another thing that I hear about, and that's a huge temperature range. Yeah.

I guess it's basically the radiation heartbeat. I don't know which one is a whole lot better. I guess more of the concern is, I think, about the radiation because in the containers, there's going to be a certain amount of protection. And when it's actually operating, it's mostly just really cold when it's actually running. So within a car, you're mostly dealing with kind of the absolute max heat sort of issues. No, cars have cold issues too because when they're parked, they get really cold. In Minnesota. And so you have to deal with the whole. Yeah, your hold time stuff gets to be the issue rather than the setup time for catches on the flip-flops, right?

So, and yeah, and a lot of stuff isn't built with the zero hold time specifications. So, a lot of them, there's definitely components in there for which there's expected to be a minimal propagation delay. Yeah, I think I remember reading an article recently about New Horizons where they were flabbergasted that it was running on a MIPS R3000 or something. And people are always amazed that, oh, it's running on this ancient hardware. Well, it's really expensive ancient hardware that's specifically designed to meet these requirements. Yeah. But I think that we're going to see some stuff that's like the BeagleBone Black. It'll be a while before it makes it to Pluto.

But I think there's enough interest in having a platform like that because it's so easy to get engaged in that desktop high school level. to have the skills to go and do interesting experiments in development. And now the trick is just how do we characterize it sufficiently so that we know what the failure modes are going to be in these harsh environments. And that's one place where we're kind of reaching out to people right now to actually give us some of that data. And it's vibration data. It's the heat performance, there is industrial spec parts, and trying to get the space-hardened parts is an interesting challenge. It's expensive.

It's absolutely expensive. Okay, so switching topics. What does the BeagleBoard Foundation do? I mean, it's not TI, and it's BeagleBoard.org Foundation, isn't it? It is. It's BeagleBoard.org Foundation. From a business standpoint, it's mostly local licensing and promotion, right? It's people that are in the ecosystem helping to get them known about kind of promoting the ecosystem altogether. So if you're making a derivative

and you want to say that it's 100% compatible with the software and it's really part of the ecosystem and supporting the ecosystem. And you want it to bear that logo. And you can pay a license fee, do all the appropriate testing, interact with the community in the proper way, and kind of get to that point of having that logo.

We're seeing a lot of people kind of take off to that. And that feeds back into the org and helps support, you know, the marketing side of things. But it's also very much focused on the education side of things, you know, making connections to various professors that are developing curriculum around the Beagle Bones, you know, occasionally doing donations, but really trying to partner with other people for doing donations and, you know, kind of pushing the priorities about, you know, solutions needed for developing curriculum.

And most of that's been geared at the university level as of now, but there's a real attempt by the org to try to get down into the high school level of education. And you work for both them and for TI? I do. I'm on the board for the BeagleBoard.org foundation. And my job at TI has pretty much become BeagleBoard. So I work for TI, but my non-BeagleBoard related tasks are mostly on consulting on open source and trying to teach people at TI how to interact with open source software and open hardware communities, kind of informing them about how people are using this technology. And that's justified because there's so much visibility of BeagleBoard that so many people

are choosing to pick up the processors and use them into different products. So supporting people and making that transition, making the connections back into TI. But I don't have a lot of non-beagleboard tasks at TI right now. And I work from home. What's your day-to-day job like? I mean, you work from home, you get up in the morning, none of us actually put on anything but our bathrobes and then we get to work, right? Yeah, well, there's a couple of windows near my desk.

So if I, you know, two, yeah, I do try to get dressed in the morning, but my commute is nice and short. My typical day, I'll go for a walk in the morning before I start having any calls. I've got a little bit of woods in back and kind of a nice wooded neighborhood, and we'll go walk around and kind of get started, talk about some of the things we have to do for the day. So my wife is actually heavily involved in the org. so we will spend the morning talking about the types of things that we want to go and try to solve

around the org for that day I'll get back I'll maybe work on a few of the coding projects maybe send off a few emails to the Google Summer of Code students and try to think about how we can best help the people trying to build the ecosystem.

I spend a lot of time on GitHub and IRC and email, but the IRC is kind of my window to the outside world. It keeps me from feeling isolated sitting at my desk. It's a lifesaver. Just trying to stay motivated, try to stay focused. Because a lot of times you run into engineering problems, just talking about them is the best way to get unstuck. Even if nobody gives you the answer,

just kind of spelling them out so you understand the problem a lot of times will lead you to the right solution. But I spend a lot more time talking and traveling than I think what a typical engineer would like. I'd love to spend more time coding. It is fun. Do you have any projects in mind that you're thinking about starting? Oh, yeah, so many. One of the things, I got a new home a few months ago.

There's a lot of projects around the house. I'm looking at starting where there's a lot of Beagle tie-in. I've got a Beagle-controlled sprinkler system. I want to do some write ups on that I'm looking at doing a lot more home automation I'm hoping I can convince some of these guys that are doing some of the home automation solutions or just the home gateway sort of solutions around BeagleBone will ship me some hardware

so shout out if you've got ZigBee or even X10 controlled switches I can replace in my house and you want to get a nice write-up on a blog post, you can send me some hardware. There's too many switches. Oh my gosh, I looked at some of the, I don't know if you guys have looked at some of the ZigBee switches or Z-Wave switches. My gosh, they're at some of the, I don't know if you guys have looked at some of the Zigbee switches or Z-Wave switches.

Yeah. My gosh, they're like 30 bucks a pop. And you start replacing 40 switches and it's like, I don't know, I don't think I can justify spending org money on my own house. That doesn't work. Is it a licensing issue with Z-Wave that it's just, because everything else seems to get cheaper. Yeah, I don't work. Is it a licensing issue with Z-Wave that it's just because everything else seems to get

cheaper? Yeah, I don't know. That's an interesting, I really should try to find out what is the source of the

cost coming from. Yeah, I don't know. This home automation stuff is expensive and I think people really need some good

practical solutions so they can kind of take care of it and control it themselves. Some really interesting solutions out there, but they all seem way too expensive for me right now. Well, and I think that's why the Internet of Things gets a bad rap for consumers, in part. Because it's just been expensive and it's stayed expensive. Well, it's bimodal. Some of the stuff's really expensive and some of it

hasn't stayed expensive and I think yeah there's a little bit of cachet of you throw internet on something and suddenly you can drive the margin up 50%

so yeah well I think in general a lot of you know home components are just expensive too so it's not just the

if they're connected or not because I think that that, a lot of home components are just expensive too. So it's not just if they're connected or not. Because I think that adds pretty marginal costs. I think it's marginal cost if you look at the BOM. But if you are a light switch company, and then you add a whole electronics division, the NRA is going to be crazy. I agree. But it seems like the ones that have actually

launched solutions out there haven't done all that software integration engineering work either so they're charging you the NRE for it but they haven't actually done the work okay I have a couple of listener questions for you cool first from Andre

wow that is not really the nicest of questions alright Andre I'm just going to quote you. It is exceedingly hard to get a hold of the field support engineers. What is happening at TI slash Freescale slash AD analog design that took away the field support people? Sealed support? Field support. Field support. Field support. Oh. We're. Field support. Oh. We're not supporting seals.

Okay. We're cooking beagles, but we're not supporting seals. You know, I don't know why it's hard. I mean, I don't think it's hard to get a hold of me. And maybe some folks will tell me different. You can find my phone number out on beagleboard.org slash about and email address. And I'm on IRC and you can ping me. And if you're having difficulty getting support around something with Beagle, I mean, the thing I ask is send an email to the community because most likely you're going to get an answer from them and people on the mailing list and that's going to be better than an answer you can get from me most of the time. If they don't get a quick answer

then say, I don't mind people harassing me. If they've posted it to the community mailing list and the ete.ti.com forms, then start harassing me and telling me, hey, I'm not getting this response. And I'm sure there are cases that fall between the cracks, but I know that the TI support engineers at the factory are graded pretty heavily on responses to e2e.ti.com. And the team that supports the processor on the BeagleBone Black is pretty proud of their support work. As far as the field guys themselves go, I don't know. I think there's been a lot of focus in the industry towards

it kind of ebbs and flows with focus on larger or smaller number of customers versus broader number of customers. And there's just kind of that pain in the industry of kind of the ebb and flow of that. It's like we'll see vertical focus for a few years, then a mass market focus for a few years, then a vertical focus for a few years. I'm always focused on mass market.

That to me has always been what's interesting and all the innovation to me happens really out on the edges and enabling that innovation and being part of innovation is what excites me about my job. So you won't find me just kind of getting locked into one particular end system and solving the last possible nuance of something on just one particular system. So, it sounds like forums. And I hate forums, so I never go to them. I know.

But, yeah. And it makes sense because then they're solving the problem for multiple people. Yeah, it gives you that scale. And if you feel like you're missing that one-on-one engineer field support, a lot of times the sad truth is you just probably don't sound like that great of an opportunity. You're probably just not doing that much, something that sounds that interesting. And I hope that doesn't sound too assaulting to your listener. Right.

Always click the I'm shipping $100,000 in the next year box. Always. You know, but that doesn't mean that you can't get support. And I know, you know, forums can be

discouraging for some people, but that's why I'm trying to say that, hey, you know, if you need that human to interact with,

you know, as long as I'm saying is make sure that you try, you know, the broader community first, but then just contact me. And please make it about Beagle or the chips on Beagle. And I hope that other people are willing to stand behind what they do that much. And I'm

only able to do this because there is such a strong community behind it. I mean, I just don't feel that worry that you're going to, I don't know, make me look too awful. Because I think that when you look at the type of answers that you'll get through the community, it's kind of been done before. And if you can't figure it out, it's probably because you're not trying. I don't know. That seems...

Too harsh? Yeah. I mean, you're... I'm not going to use TI because it isn't quite fair. Although I have had similar interactions with TI. Another company whose name begins with A is very nice, very cool, very focused on lots of people. And yet their newest thing points to one of their older forums, which if you post there, everybody says, this is not the place for that chip. That chip is too new. We are only for the original version of this thing.

And it's like, okay, I went to the official page. I followed the links and now you've jumped down my throat. I have nowhere to go. And I don't think that this is a, I don't think this is just me. You end up, if you have all these giant product lines, and like where do you stop the, like the CC2000 line for TI? There's a great line. There's lots of neat stuff in there. But there's the very old stuff and then the new like piccolo motor controllers and they they have some of the same foundations and some of the same forums but you can easily end up in the wrong place because you are on the new thing, but posting to the old things forum because they both say CC2000.

Versus Piccolo or, you know. I wonder if it's about, I mean, I went to the TI website, to the E2E forums, and you have terms of service. And you have privacy policy. They're both lots of legalese. And we've been hearing about conferences and whether or not they have... What is the policy? Harassment? Anti-harassment?

Anti-harassment policies. Yeah. It would be nice... Well, to have those policies, they have to be heavily moderated, too. And now you have to have people who are... But these forums are for getting help. Yes, yes. Are you concerned that, like in the ETE, that people are too abrasive?

Or any forum, right? I mean, it depends on what type of forum you're going to. One of these company support forums. We're going to, we're going to stick to a company supportive support forum where it isn't the company people who are saying you're wrong. You're in the wrong place. You're an idiot,

which is always what it seems to be. Ah, okay. It's the, it's the other people. It's not the, yeah. Okay. It's the other people. Other people. Yeah. Okay. Because I think that generally, those of us at TI

would take it very personal if you're getting a bad experience from a fellow TI-er and they're being abrasive to you. I think that, you know, we would naturally be very upset with that, you know, that individual. And we have to give a little bit of lenience to

other people that are coming on the forums, but I don't think that we would tolerate them being, you know, something that would qualify as, you know, real harassment, right? And I think that you do have recourse towards that, but I don't think you're going that far And I think that you do have recourse towards that. But I don't think you're going that far. I think you're saying that it's just because you're...

They're just being jerks, not being... Are you saying because it's a forum you're automatically hesitant? Yes, absolutely. Not because there's anything specific on that forum, but other forums have given me a bad experience.

Well, gosh, if there were just less semiconductor options, then you could, you know, if there were only 10 different chips out there, then all the people at these different companies can just focus on supporting these 10 chips. It's a hard problem.

You can actually get to know the person that's going to support you because he only has to worry about or focus on one chip. And so you can get to know that guy and it all just works out nice. And it's just sort of a complexity of the economy we have right now that there's so many different problems to solve and you're not able to go to just one person. You have to have this wide variety of people and solutions. And so to try to scale up to that, you're not

always dealing with the same person. So it's not this one-on-one personal relationship. And that's why solutions that really cover a broad set of problems, something like a Beagle, I think, are especially appealing. You kind of look for that lightning-in-the-bottle solution where you have a hammer, which is very general. A hammer is in nails, right? It's a very general solution that solves a lot of problems. You see a lot of that consolidation in the market where people don't – I don't need this huge variety of things.

I need one thing that's very flexible that I can't find anybody that's going to help me out on that solution. Because that engineer has gone off to build 10 other products, and there's not a big support staff around them. I think it's a very interesting dynamic. And you even got all these semiconductor companies popping up, state-sponsored semiconductors coming out of large

countries and building things for phones and tablets and just these things that go into high volumes. And then lots of people finding, oh, well, there's some wonderful engineering going into these things, but how do I really get the details and the support and everything else around them? And that becomes really difficult.

And I hope that people are experiencing TI as being kind of an exception to that, more so. I mean, we're certainly subject to all the same laws of physics, but I hope that people are finding TI to be at least a little bit of a breath of fresh air in that climate. I don't know how to compare. I'll keep that in mind and we'll think about it. Anybody has any particular things they'd like to bring up about the forums and better ways to use them? Chris and I are going to talk.

Even guidance on how to make, you can't hire an infinite number of field apps engineers, so that everybody has their assigned engineer. But how do you make the ones that are out there more accessible? So how, as a large company, do we make sure that our employees that are out in the field

are more accessible to the individual developer? I think those sort of guidances, because it's not all about directing people to the forums. I completely agree with your listener about that. I mean, it just can't be the answer to everything. But how do we do both, right? Increase the usefulness of the forms, make them more friendly. And then, you know, how do we make, you know, everyone in a large company more accessible? Because I think that's a huge benefit for the people in the field. And all the people that I

run into in the field, I encourage to go out to the maker fairs the local maker fairs because they'll find real opportunities for TI by going to these places and meeting with the people that are interacting, especially like a lot of the local hacker spaces

and I think TI engineers are always very curious and reaching out and learning about what people are doing in their local communities. I think that's one of the wondrous things about working in a really large company. It's something that keeps me at TI. Well, now the TI booth at the ESC is just going to have to be really nice now that you've set this bar nice and high. ESC is great, but don't forget to go to the Maker Faires too. Everything from pros to do-it-yourselfers. The Maker Faire is a great place to hang out with people.

ESC is just next week. Or actually it will be about the time this is posted. We'll be wrapping up. And I'm getting all these pings from people, my friends at ESC, which is one of my favorite shows, and that I won't be going this year. It's a lot smaller this year. We'll see how it works out. It's not a bad year to miss.

Maybe next year. So I think we're about ready to wrap up. Christopher, did you have any remaining questions? No. You didn't want the quiz question? The quiz question? Yeah. What's the quiz question?

How many definitions of the word beagle do you know? Okay. How many definitions of the word beagle do you know? I only know, I'd say a couple. Maybe three. But are there any that aren't a noun? Can I ask for hints? There's one that's a verb.

Ooh. Ooh. Okay, well, maybe I just learned a new definition. I would have to think if you're beagling, you're searching. No, you're hunting with rabbits. That's close. You're hunting rabbits with beagles That's close It's going to hunt with beagles

It's not just being Snoopy If you're beagling, you're specifically hunting rabbits? Well, you're specifically hunting It depends on the exact definition But you're hunting with a beagle horde Or whatever they're called Okay, cool Obviously the dog.

Yeah. I would think that at this point there's got to be a space reference when it comes to Beagle because it was the vehicle that first landed on the moon. Well, and it was one that missed Mars, which sounds very Beagle-like

it missed Mars? yeah that was the British and if you've ever had a Beagle they're just so focused on their noses they go everywhere that wasn't the units calculation that was the one that burned up god that kills everybody

so there was another so there was a space vehicle called Beagle that missed Yeah, it landed and they never heard from it

again. I think they got some pictures of it. That was the units calculation one then, right?

No, that was an American program. Okay. I don't remember. The units calculation one

landed and they have photos of it from the orbiter. I think it had some issue where it

didn't open properly so it had no solar power okay so we're adding a definition right now i'll go change word nick and say it is a class of spacecraft that uh mysteriously fail often due to unit calculations no but that's oh i think they're totally off on that. But the most famous beagle in space has to be the Lunar Lander. Beagle? Come on. Is it the beagle? Name of Lunar Lander. Oh my God, now we're fact-checking real-time.

Real-time fact-checking. This is what our listeners like. This is not... Okay, now, name of beagle. Name of beagle? The Beagle 2. Oh, you know what? I think I'm confusing it with the eagle.

The Beagle 2 is the Mars one. I think I've just said that so many times because it's the eagle. It was the beagle in the Peanuts cartoon in the comics. And I used to think it was a beagle too. Yes. We can just add a B though and it's done. Oh, I don't like where this is headed at all.

Okay. So for those of you keeping score, the eagle was the lunar lander. The beagle was the Mars probe from Britain that went badly. And I've been enjoying the rhyme for so long, and I'd never even heard of this British probe. So I've been enjoying the rhyme so long that I'd actually convinced myself that that was the original name. I'm pretty sure in the Peanuts comic that Snoopy landed on the moon

and that said the Beagle had landed. I too had that in my head. Yeah, it's something I tend to say to myself every time a plane lands. The beagle is landing. Okay, so having added a definition, or maybe not, you still have two more to go. Oh, two more.

Well, I have to think that the Beagle Board would be one. Okay, now you still have two more to go. Rats, rats, rats. Am I going to get one off the top? Are they still nouns? Yeah. You can give up. It's okay.

There's no prize. You don't get Alicia's voice On your home answering machine If I could think of Any horrible guesses That would actually Be somehow entertaining But

I really have nothing I can't even I can't even make one up I would be a horrible guest On Well one of them Is a type of dog No no no

He got that one Oh When is a person Who acts like Said type of dog Oh come on A person who acts like said type of dog. Oh, come on. A person who snoops on others. A detective.

That's actually in books. Okay. Sherlock Holmes was a great beagle. Legal beagle. Right. Legal beagle. Oh, yeah, yeah. And then-

Also eagle. There was one I hinted at earlier. The poor beagle, which is a shark. And so they call a group of small sharks and i think it's a canadian thing but i'm not sure it's a canadian maybe australian maybe somebody could tell me but a local colloquial name for several species of smaller sharks is a beagle yeah that's not going to be common knowledge pretty esoteric have to say. I did try to scale these as easy to hard, yes.

He got two points. I got one out of four. Well, we're going to make it two points out of five. That's... Well, it's our inaugural quiz corner section of the show. I'm not sure that was a browsing success. Alright, well, I think that we can be done now.

Jason, do you have any last thoughts you'd like to leave us with? You know, I'd prefer to talk about it, but, you know, I think that there's a lot to be done to make programming and electronics

more accessible to more people and more barriers. I think that we're, you know, in a world where we're still very much getting these amazing pieces of technology that the vast majority of people that we interact with on a daily basis have absolutely no ability to comprehend. And I think that's a sad state. And I think that if we do more to focus on enabling people to build things with programming and electronics, and I think there's some of that going on, but I think there's room for a whole lot more,

that I think we're going to end up with a better world where we're not slaves to our technology, but rather technology serves us. And there's so many things that we can go and build and I could just go off for weeks. I know you're wanting to wrap up, but I just encourage you to reach out to me personally

or Christy at BeagleBoard.org. If you're wanting to get involved with teaching people about programming and electronics, I think that we have a pretty unique experience with the BeagleBone Black and the Bone 101 tutorial system and a lot of the stuff that we're doing with the Google Summer of Code and a lot of the stuff that I can't yet talk about.

And really looking for people to get involved that want to change the world through making it easier to work with the programming and electronics. Well, and you've made some strides with that, in part from this new Beagle cookbook. I think it's a nice start. And if you enjoy it, I think there's a lot of ways to go. And I think we've already started work on Beaglebone Cookbook 2. I wouldn't expect to see that on shelves anytime soon, but there's work there.

There's a second edition of Bad to the Bone, which had incredibly mixed reviews. And so for all the negatives that were out there about it, there's a second edition that's already prepped. It's really being held up by me getting my good review time into it. But Stephen Barrett has already put a huge amount of work

into it, kind of addressing a lot of the issues that were raised. And a lot of them had to do with the same things, the types of concerns that you were raising, Alicia, that there wasn't enough on real time and dealing with more traditional embedded electronics in that book because it dealt with things at that Linux level. And a lot of people are looking to go deeper.

Cool. I look forward to it. Cool. And there is a new BeagleBone coming out. There is the BeagleBone Green coming out if that's what you're looking at that's a seed product, not a BeagleBore.org product

but it is BeagleBone, or BeagleBore compatible is the term that we're using and I'm super excited about it because it's it makes the real differentiation for it is these Grove connectors. And so if you want to wire up a bunch of different sensors and actuators to it,

you don't have to do any wiring whatsoever. You just plug these things in. Working on a lot of the software experience for that now and expect that to be something to iterate. But I think it's a great relationship between BeagleBoard.org and Seed Studio. I actually meant the BeagleBoard X15, the one that was

teased in the book. Well, that's not a BeagleBone, that's a BeagleBoard. We have two lines of products. We have the BeagleBoard line, which is kind of our high-end line, and the BeagleBones, which is the bare-bones BeagleBoards. So there's more

stripped down, focused more on real embedded type of stuff, and more headless, even though BeagleBone can have screens and monitors, keyboard, my stuff, stuff, but you can use it as a desktop, but it's less focused on the desktop experience.

So that's kind of the two lines. And the BeagleBoard X15 is a prototype. Right now it's in the prototype stages of a replacement for the BeagleBoard XM, which is still rather popular. So that's the extra MIPS, extra memory BeagleBo board from the original Beagle board. It's got full-size HDMI and USB hub. It's got the DSP acceleration for doing higher multimedia performance stuff. But the XM is getting a little bit long in the tooth. And so that's why we've been prototyping X15.

We've got a handful out in community members' hands right now. And we're continuing to work on moving it forward, but we haven't officially press released, announced anything. We just work very openly. So most of our stuff

that we're working on, we try to keep open to the community. So I don't have any big announcements on that, although it is a pretty cool little machine. It's got things like USB 3, gigabit, dual gigabit, Ethernet, more memory, a whole lot more performance, and a whole lot more, I kind of call it the, it's either affectionately called the beast,

or there's a core for that, I kind of call it the, it's either affectionately called the beast or, you know, kind of machine. There's a core for that because there's just a huge number of heterogeneous programmable cores in it for doing all sorts of crazy fun things. Wow. That does sound like quite a beast. Yeah. It's less focused on your beginning maker and much more focused on, I just can't find a technical solution for that. I need something much higher performance and much more capable. I'm ready to make a big step up.

I need a sledgehammer. That's what this board says to me. I need a sledgehammer for this problem. Yeah board says to me. I need a sledgehammer for this problem. Yeah, and it definitely fits that billing. Oh, cool. I'm glad we got to talk more about the BeagleBone Black because that is more our core area. But this is pretty neat too.

Yep, and we're going to see some fun demos on that. And stay tuned. But also stay tuned on all these different BeagleBone black clones and derivatives. I think that's one of the big things I'm really excited about in 2015 is just really seeing a lot more people take that design

and do some innovation around it and make it easier for people to do derivatives. So we're doing, we're focused a lot on making some of these designs a lot more easier, a lot more easier. Sorry about that. A lot easier for people to make derived designs out of. And that's good because they aren't as rare as they were when we talked a year ago.

Oh yeah. That was a huge problem last year. And fortunately it's a very solved problem. Availability of BeagleBone Blacks is excellent now. Just increasing our manufacturing partners and just doubling down on it. They're selling more than ever, and availability, it's real easy to go in and grab one. Oftentimes you can go and just walk into your local micro center and grab one.

Cool. Good to know. All right. I think we should all get back to our days. Really enjoyed talking with you. I love the, some of the, the deeper dives on some things and,

you know, hopefully I didn't throw out too many terms. I'd love to actually spend some more time talking about that next time we meet up at a show. Cool. I look forward to it. And we do have lots of links. And if you heard a term and you didn't get a link on it, just let me know. Drop us a line.

It is show at embedded.fm or hit the contact link on embedded.fm. My guest has been Jason Kreidner of Texas Instrument and BeagleBoard.org Foundation. Last week, Jen put her title as international puppy consultant. Since she only has a cat, the lovely and wonderful Beanie, I was a bit skeptical about that title for her. I think Jason deserves it. You can find his new Beagle Bone, Beagle Cookbook,

there'll be a link, on Amazon and O'Reilly and all your normal technical bookstores if they still exist. At O'Reilly, you can use the coupon AUTHD, that is A-U-T-H-D, to get 30% off the paper or 40% off the electronic book. It may still be cheaper on Amazon. I'm very sorry about that. We have no control. Thank you, as always, to Christopher White for co-hosting and producing. And thank you for listening and reviewing. Yes, now you're obligated to review.

Wherever you want, you know, you can review to your lunch companions. That's fine. Or on iTunes. We appreciate it. Final thought for this week. Of course, it's going to be from Charles Schultz. Of course. I just have to choose one of these many choices.

It's a long show, so it'll be a short final thought up to now. Happiness is a warm puppy.