Embedded - 100: Unintentional Radiator

Episode Date: May 6, 2015

Star Simpson (@starsandrobots) and Jen Costillo (@RebelbotJen) catch up with Elecia and Chris, discussing how hobby projects have changed over the last two years since the show started.  Jen's website: RebelBot Star's website and weekly drone newsletter The Buzzer. Star works at Orion (formerly OnBeep). Novena board and Star's project Balboa ODROID Open Cores Crowd supply and What it took to make the Octopart reference card

Welcome to Embedded, the show for people who love building gadgets. Happy 100th episode. I'm Alicia White. My co-host is Christopher White. Today we have Jen Castillo and Star Simpson. We're going to talk about tinkering. And if that sounds a little familiar, it's because this is the group that did the very first show, back when we recorded the Start Tinkering session at the 2013 Embedded Systems Conference. Hi, Jenna and Star. Thanks for coming back.

Thanks. Thanks for having us. So, Star, it has been a long time. Can you tell us a bit about yourself for people who don't know you? And basically, who are you and what do you do? Sure thing. So I think, as we probably covered on our first episode, I like to— That's a long time. It was almost two years. Yep. I have done electrical engineering. I like making stuff, especially out of circuits.

I like building things for other people to use on my own kind of as a hobby. And so I end up tinkering and building a lot of projects. I think that's why we talked about that before. What about your career? I have sort of, I think since we last spoken, moved on from doing electrical engineering as my job to more what my current role is now doing technology strategy, which both involves understanding how specific technologies work, but also understanding industry and understanding

what we should do with the technology as it develops. And when you say hardware, you mean both circuit boards and FPGAs? Yeah, well, I guess an FPGA is a little more toward where the water meets the sand, but I do have a recent project experimenting with building for FPGAs, too. I have some questions about that, but Jen, you've been around a few times since we got this party started, but for new listeners. Yeah, I'm still a firmware engineer engineer still do embedded

occasionally have to get my hands dirty um at the hardware level um and i've been working primarily in wearable startups for the last two two and a half years and you're not at bia any longer but you can't really talk about your new place no No. More wearables. Bia, unfortunately, closed shop about a month ago. Ouch. It's hard to get funding. Yeah. So we did this session.

It was really called Design West, the Embedded Systems Conference in 2013 because they couldn't choose a name. What do you remember about it? I remember us being on the panel and giving an interview. I don't know how I got there. Yeah, not sure what else to say. What do I remember? It was like a dark night a long time ago.

There was one guy that had a lot of questions. He sat up front and I remember he's like, well, have you done this? Or what about this? Or I can offer to help you with this. That's what I remember. There's that one guy that asked a billion questions. Well, weren't you talking about Taco Copter? Yeah.

Was I? I had your sticker. I had your sticker. Yeah, yeah, yeah. Actually, funny, speaking of stickers, I was just showing you both the new sticker on my laptop where I note that I'm building wearables as I use them. Perhaps wearables is a good theme. So actually, let's go with that. What does Orion do?

It used to be OnBeep. It used to be OnBeep. We're now Orion. I think most people are pretty happy about that. We're building a wearable device to let people communicate with each other who aren't as we are actually still all in the same room. So people who are spread out all over the, you know, all over the globe, really. And so is it like a cell phone? It is unlike a cell phone, although that's pretty reasonable. Some people like to say it's like a Star Trek communicator badge.

The idea is that you and the group of people you'd like to communicate with all have these badges and can use them to instantly in real time speak with each other. Especially where you need your hands for some other thing. You're driving, you're skiing, you're off doing something active. You can't have your phone out. You can't be sending a text message. It sounded like you almost said skiing. Have you guys tested that particular phase? In fact, yes. Well, perhaps it was snowboarding, but Andy Sherman, one of the people who I work with, wrote a good blog post about his experiences having them out on the slopes. Is there a security aspect to it? Security aspect?

What's that? Well, I mean, I could see, I have my phone always, I'm not sure. I would want another gadget to communicate. But if they were secure and I could give it to a few number of people so that I was more confident communicating with them than I am with my cell phone, who I just assume everybody listens to my cell phone conversations. I might as well start recording them and posting them as podcasts.

I mean, I think, okay, so firstly, I think it's not a slippery slope in terms of communication availability. I think, you know, in the last, perhaps even since our last time speaking, you know, we've certainly learned a lot more as a nation about robust and secure communication. Thank you, Mr. Snowden. Thank you, Snowden. But that said, I mean, there is still a difference between recording your phone calls and posting them as podcasts. But that said, security is actually pretty important to us at Orion.

We think about it quite a lot, and I don't know what we're communicating about it publicly right now, but it's front of mind. Right, and we may have you back to talk about Orion once we make sure we've read the website and know what we can and can't say. So, TacoCopter. Now it seems like that idea is coming to fruition.

How much is Amazon paying you in licensing? Is it in tacos? You guys are so optimistic. You didn't patent it? She's like, no provisional was filed. One of the best things about having created TacoCopter is getting a lot of, you know, my friends every now and again will demand to know where their tacos are. They got lost. We've got to write them off. No, I'm not in any way receiving anything from Amazon.

Did you get a lot of press as a result of it? No, no, really no. You got a lot of press with the Google Glass teardown. That was a couple of years ago. And boy, I saw that all over. It was really very cool. That was a lot of fun. What did you do? With the press? No, with the Google Glass teardown. Oh, right. Yeah, me and a good friend, my friend Scott Torborg, and I, we got our hands early on a pair of Google Glass and immediately decided we need to take it apart and see what makes it tick. Because I think it's probably hard to remember, but at the time we had no idea what was going to be inside. And that was really exciting, really interesting to understand what possibilities it contained. When we got it, we actually weren't sure whether, I don't know, the entire board inside was going to be potted and epoxy

and impossible to extract. We were just really curious about what it was made of. It's come out since, I think, that it's some ridiculous stack, like 16-layer, that's not the actual number, but some really intense number of layer board, and it's got a lot going on. This is not the actual number, but some really intense number of layer board. It's got a lot going on. So when you did that in 2013, the Google Glasses were really expensive.

It's true. And exclusive. And I think they were invitation only. Yeah. Hats off to SparkFun right now, who sponsored the teardown, which is why we were able to do it. Well, that's what I was going to ask you. I was like, did you guys go like, if we can we can get our hands out we're going to tear it down as opposed to like let's check it out for a little bit and then decide if we want to tear it down it was really a lot more opportunistic than that it was like i whimsically entered the competition to

win the opportunity to buy a pair of google glass and did not expect it to come through it came through and then we were like well what do we do now right okay truly there's something interesting we can do with a pair of glass so like what would we do and then the idea to tear it down came through ran it by um some spark spark fund folks who were hugely in favor and wanted that to happen so um they basically you know we made it go from there so my follow-on is did you feel like you one wanted to put it back together to play with it you know i did you did put it back together so it is possible to put it back together it was possible we were very delicate because again we had no idea what we're going into um it was possible to put back together but the plastic was uh deformed so um to wear it was never

gonna look and the optics might be a little weird perhaps uh optics i mean it was all still deformed. Tough to wear at that point. It was never going to look professional again. And the optics might be a little weird, perhaps? Optics, I mean, it was all still functional, but it was duct-taped, basically. And then later, Google actually issued a return merchandise recall-ish thing where they, I think, upgraded the hardware and wanted the people doing

Explorer stuff to have a hardware rev. So they basically sent a box and put all the parts I could find in it. It's like, oh, I got punched in the face. All the parts I can find. No, and then so they've since replaced it. I no longer have the parts I have. Did you write in the corner like, sorry about the state of this.

I got punched in the face or we had to do a teardown. They didn't ask, I think. They probably do. And did Google comment on it? No. Oh, okay. they didn't ask um i think they probably do comment on it no okay and then after that you went on and did the octopus card octopart yeah octopart octopart yep octopart uh chip search engine um so uh yeah that was fun gosh was that also since we last spoke yes it was 2014 you've done a lot of stuff i don't think you realize. No, I really didn't. No, I didn't realize.

That's amazing. Octopart card. That was great. That was actually, that's an interesting lesson for me. And it was kickstarted. It was a run on CrowdSupply. CrowdSupply is a crowdfunding website that specializes in quality open hardware, especially. And yeah, they were great.

They also do fulfillment, which is why I decided to go with them. Basically, as soon as we got the cards back, they distributed them. All the orders were filled in three days. That would have been me on the floor with a stack of envelopes. That was me on the floor at VIA. So I could totally appreciate that. So that was really good um but yeah it's funny because I thought that was totally going to be just a two-week project I was like

oh great I'll design a circuit board it's going to be beautiful it won't have to you know pass any functional tests it'll just work and it'll be great and a year later a year later whoops we've delivered I have a panel from that on my wall as a reminder of what it's like to take on a two-week project. So, and I know that you, Elle and I have discussed this a couple of times about how the crowdfunding thing is kind of nebulous. And I've definitely had, since working at BIA, there were a number of people that I was housed with, different companies were housed with all together and they all did a lot of Kickstarters and they, the consensus is, if you think you're going to make money or it's going to be easy to do a Kickstarter, think again. Yeah.

Is that basically your takeaway was the crying, the public, you know, the fear of public humiliation. Oh my God. I mean, every asshole. So like, uh, for me, it was, it wasn't just like that. It takes a month to prepare for a month long campaign, but that like, you know, I was going to have to watch a video with myself in it. And I found that very distressing and, you know, having to write the script for it and like do all these other things that you wouldn't normally consider as part of the work of getting a circuit board shipped. But if you were looking for funding,

if you're trying to go for funding, you effectively are doing that over and over again in front of a VC. So might as well make a video of it. Post it. And yet I don't listen to very many of these podcasts. I'm sorry, listeners. I know some of you like the podcast, but listening to my own voice, it hurts. Yeah.

Yeah. And it was out of scope for what I thought, you know, again, two week quick. Quick turn thing. Yeah. That said that I think you had mentioned recently also that you had had no luck with Indiegogo. Right. I have never gotten anything out of my Indiegogos, one of which was a dance project that I just owed a T-shirt.

And I never got that paid out either. In their defense, I backed a Kickstarter, not Indiegogo, a Kickstarter for a climbing wall with LEDs embedded in translucent handholds. Oh, cool. It's like night climbing at Burning Man or some such thing. And my reward was just going to be a video of someone climbing the wall and then falling and yelling my name as they fell. That sounds so awesome. Sounds pretty simple.

Was it like $5? What was it? I don't remember. Two years later. Two years later. Poor creator ran into him at the climbing gym and buttonholed him. And we filmed it on the spot. Did you push him off the wall?

No, he fell of his own volition. But, you know, it's hard to deliver rewards, I guess. I mean, I guess like if you're doing everything customized, but if you're doing a batch of t-shirts, come on. No, I mean, sometimes it just gets away from you and you think it's going to be two weeks. And then a year later, you're still looking at it going, oh, I never did get that answer on whether that person needed a large or an extra large. And so I can't really start the run until I get answers and these people didn't pay me. And there's just all these things that go beyond what you expect.

Yeah. I mean, so when I did a Kickstarter for one of the dance companies I was in, we had just a random set of shirts already that we needed to get rid of. So we just made the reward, like, you want a small? Here's your reward. You want a pink small? You get this one. So that kind of took care of that. But now there's so much third-party backer lit and all this other,

this whole cottage industry has come up around, after you close your Kickstarter or Indiegogo, that you have all this additional help to kind of take care of stuff if you didn't plan appropriately. Yeah. Quotes. Quotes on the help. You probably also get a lawsuit.

You get a whole bunch of people saying, you have to pay me this in order for Kickstarter to release your money, even though that's not at all true. It's the same kind of scam that happens when you incorporate a business in California, I found. Is that why we get those funny mails now? Well, this is what I found in the process of starting a corporation and putting it into a nonprofit, was that suddenly you get mails that say, hey, you need to get these documents done.

Otherwise, you will be out of... You have to post the labor posters. No, but she's talking about the fake government documents. The fake stuff. No, no, the articles of incorporation and other stuff. It's all fake. There's a whole industry. I mean, yeah.

Oh, right. Language. My mouth? Just out of curiosity, it sounds like you both have some background with crowdfunding. Did I miss a campaign? Oh, no, not for me. She's just an observer of other people's crowdfunding. With the podcast, I've gotten to speak to a lot of people at various points in the campaign. And it has been very educational, such that it makes me never want to do one.

But that said, every single Kickstarter I've funded has paid out. No matter how small, how artsy, they have taken care of their stuff. I'm tempted to lick right now. It may take a while. It may take a while, but they get it done. I think I've had one Kickstarter that didn't come through. I think I've had a few, maybe one or two that came through or sort of came through.

Whatever it was was not so great. Yeah. I don't think I've had any that chip nothing, though. But I'm pretty picky, too. I mean, because it's... Yeah, I'd like to look at number of Kickstarters-backed numbers for the folks here. I have several that I think are maybe still undelivered.

I try to limit myself to one, maybe two electronics, technology ones at a time, because I know that the leeway maybe. But I found these playing cards that I needed delivered by a certain time. And the guy delivered them to me personally before everyone else. But it was very sketchy. that was just supposed to be a pretty easy print job he was very confident i said i i spent time with him and you ordered him like six months before he needed him and yeah i was very i was very like are you sure because i know what it's like to be on the inside things can go wrong very quickly now one of the things with the podcast

is we thought well if we ever want to do a Kickstarter, now at least we'll have people to talk to. Because that whole initial funding period is so important. But instead, I talk to Dharma and Light Up and Carrie with her Skyneminder, and I can't even think that that would be a good idea. That sounds like not much fun because it isn't very much engineering. At that point, it is building a startup.

Yeah, for sure, for sure. Which can be fun, but it also can be not fun. I don't know. But let's be honest. I mean, Kickstarter has basically opened hardware startups back to being in the game again when they were not getting funded. There was a whole lull between 2009 and almost 2012 when hardware VCs were, it was really hard to find them.

Well, and now with Tindy and this crowd supply thing looks pretty cool. Yeah, it does, definitely. I just, I don't think I would go that route for me personally, because I think I would want to focus on the product and the engineering and try to spend less time with the walking around talking about it all. Well, you don't really get a chance to do a small run with Kickstarter. I mean, if you're successful, you're probably going to have hundreds, maybe thousands of things to make.

Where with Tindy, you can put up 15, 20 things and start out with that way. Okay, if there's some interest there, then maybe go on beyond that. And I tend to build things that way. I want to do little to big, and I don't need to follow a big dream right away. And we're lucky enough we could bootstrap if we wanted. And Tindy does seem fairly true to their name in that being a tech indie marketplace. On the other hand, I think for when you do want to do a run that's on the larger side,

for example, I got Novena Bunny's open hardware laptop through CrowdSupply because they do so much work to make you know, make sure their fulfillment is solid and that they, I think they actually try to prevent campaign overrun and like keep expectations scoped and keep people to a realistic schedule because they have background building hardware. Yeah. Yeah. Yeah.

I mean, that's definitely the one thing that I learned from Kickstarter was you have a sense when someone hasn't made an update. You know when you need to make an update when the comments, you see the comments building up, building up, building up. It's like, it's overdue for having an update to go, where the heck are you? Speaking of Novena. I was going to ask. Yes, how is that? Speaking of Novena, I hear there's an FPGA project you've been working on.

There is. And it was kicked off, actually, by Novena existing. Do either of you guys have Novenas? No. I don't. I mean, it's Bunny's project, and he's made it to be a very cheap, very modifiable laptop. Yep. And it's pretty cheap, like $600?

I think the board is $500. And it's a neat way to think about computers not being monolithic things. And if you think about it, that's like what? That's five Raspberry Pis, but can stand in for your oscilloscope, can do all of these cool like lab test equipment type things does it have all that on board it has a high-speed DAC on board I don't know if they ended up designing a board to do a oscilloscope conversion but I think that was part of the original goal and it also has gigabit ethernet so it can be your router and can it's you know

very very versatile you can very if you have the, you can replace that really expensive 60 plus grand oscilloscope and logic analyzer with... Which one is that? At least that's what I remember. The one that had the Ethernet on it, and then you could just remote log in,

and it was running XP. Wait, but doesn't that measure up to 60 gigahertz or something? Yeah, but if you can effectively replace that with this, with your sweat equity and this, that would be pretty cool. I feel like it replaces a tech. I don't know if it goes up to that high of a signal sampling rate. Let's see. Can you still buy a Novena?

I think you can. You can't on crowd supply. Okay. Yeah, I don't have an actual... We'll find out and I'll post if we can. I think they had an update where you might be able to buy Nuvenas again soon or something. Maybe we can have a special embedded FM Nuvena edition. He's on my list, yes.

Benny is definitely on my list. It looks like you can buy just the board, at least. And it's $5. Yeah, that's what I have. But you did not build a laptop with it or an oscilloscope with it? Well, no, I guess not. I mean, technically speaking, it could be my laptop, but I haven't actually booted the UI yet.

I've been using it just as a dev board for the FPGA that they've added. What kind of FPGA is it? It's a Spartan 6, I think. Yeah, that's right. Who makes the Spartan 6? Xilinx. Yeah, thank you. Xilinx.

Xilinx. He always tells me to talk in the mic, and then he just shouts from across the room. Thank you, Professor White. So what are you doing with it? Do you have a lot of FPGA experience? I have zero FPGA experience previously. This was, again, a case of getting access to the hardware and then deciding what would be cool to do with it.

And I'm collaborating on this project with, well, we've got a couple of people who've expressed interest in joining, but it started out as being just me and Andy Isaacson. And we realized, well, first we set out to do something cool with the FPGA and learn Verilog together through doing a project. I'm so sorry. Oh, come on. No, I think it's cool.

I don't know. I like VHDL. Okay, well, we looked at both and we decided on Verilog. Yeah, I'm not surprised. It's much more usable industry-wide, the Verilog. Plus there are a couple of neat projects like Chisel looking to replace Verilog and VHDL. What's Chisel? Chisel is writing hardware in Scala. What's Scala? I sound stupid. What's Scala?

If we get back to Rust, I'm going to be impressed. What if we get back to Fortran? We're not at Fort? Can't get to Rust from here, I think. Wait you get to kevin you know what rust is haskell oh yeah i spent all last week compiling rust what is rust i suggest i look at it and i was like it's oxidized metal i'm the one who brought it up on the last show oh i thought really yeah what's rust it's a language it's a that's sort of designed, I think, for more system-level stuff. Beta programming language right now.

And it's specially designed to not let you shoot yourself in the foot with memory and whatnot? Theoretically. Have you gotten to use its store? So it's a managed memory version. I haven't even read the Wikipedia page, and I bet he has. I downloaded, I ended up downloading someone else's compiled nightlies for ARM, and then the thing I wanted to build compiled failed, so I don't know what happened yet.

I have to go figure it out. Midbug. We should talk to you later about it. Sounds great. We can do Haskell for hardware, though. Personally, very excited about that. I thought Haskell was something webby, but I guess not. It's very difficult to talk about why a Haskell person likes Haskell.

If you make them try not to use the word purity at any point in describing the appeal. Isn't Haskell one of those languages that's designed not to do anything? Oh, come on. Right. No side effects. So you can't actually do anything. Is this not Eddie Haskell we're talking about? Because he was always kind of a weasel. Haskell. Eddie Haskell?

Yeah, Eddie Haskell. He's the creator of Haskell, obviously. Curry Haskell. Never mind. Yeah. I don't know who Eddie is. It's from the beaver. Leave it to beaver. Really?

You guys know? How old are you? Nobody gets my jokes. We're the old people here. Let's just make that clear. You and I, mega old. I'm sorry. Did we have a point?

Okay. Wait a minute. You guys are talking about Haskell. So Haskell. So you said it was pure. I don't know what that means. Is it pure bits? Is it just pure bits?

That's what you you on the spot. People who get excited about Haskell, I think, are literally excited about being able to prove things about their programs in the mathematical sense before or without having to run the programs. So it's kind of like Mathematica, except with an H. No, it's a functional programming language, right? It is a functional programming language.

So the same reasons that people like Lisp. Oh, I love Lisp. There you go. You should check it out. It's my favorite. I was not allowed to whisper. I got in a lot of trouble last week for whispering. But I love Lisp. There you go.

I think people for swearing to whispering. LearnuaHaskell.com. She's trying. LearnuaHaskell.com. Okay. Apparently English is not a priority. It's a great website for learning. Okay. LearnuaHaskell.com. Okay. Apparently English is not a priority. It's a great website for learning. Okay.

Where were we? Well, it's one point. We were talking about our FPGA project, and then we got onto Scala, and then it fell apart after that. Then we went to Rust. No, we didn't mean to go there. Let's go back to FPGAs.

FPGAs. We had Jack Gassett on the show, and we talked about the Papilio board. So Jack told us a little bit about FPGA, and Chris got to play with it some. But like many home projects, it gets tossed whenever we get busy, and lately we've been very busy. Why did you decide to use the Novena? Why not one of the other FPpga boards so the timeline goes basically like this we got a novena we decided we should do something cool with it and that it was

a great excuse to sit down and learn some verilog and just see what the whole world was about because neither of us had done any fpga stuff before and uh one interesting project because this was again about the same time that a lot of the Snowden revelations were quite fresh, was to say, look, maybe this is a way that people can run their own encryption algorithms. And so we sat down to write a core to compute AES to accelerate SSL on the CPU on the Novena.

And it was so bad. It was just such a tedious process trying to write any code for the thing. After I think two months, we collapsed in victory because we got the LED to blink off the IO. The FPGA was so terrible. And so we said, well, that was too hard of a project. We should pick something much more ambitious. And decided that we rebuild free and open tool chain for FPGA.

And that's basically, that's part of what we're currently working on. Free and open tool chain for FPGA. That is very cool. Yeah. And yet you're a hardware engineer. I know. I don't know where I went wrong. Where did you go wrong?

Yeah, I know. Truly, it's a good question. What do you consider a hardware engineer? Can a hardware engineer do FPGA work? Because every single hardware engineer I know who does FPGA work, that's... No, but she's building a tool chain. She's making compilers. It's like she works for Cadence. I agree, for what it's worth. But that's not all we're building. Balboa is not just an attempt to build a better way to write code for FPGAs.

And for what it's worth, Clifford Wolf, I believe, is a person who has built Yosis, which is a piece of software that we're working with. So shout out to Yosis for being great. Balboa is also meant to be a library and other software tools to make FPGA management, uh, FPGA core management better. And also a way to let people, uh, we think it's possible to resource share the FPGA. That's not really something people do right now. So to, you know, instead of say this FPGA will perform x function and then you like seal it in

and that's what your board does now to instead say like well we'll have this core and then we'll swap it out for that other core when we want to do something else with it like in runtime so that users can use the FPGA to do different things you guys are just giving me the most amazing looks right now this is cool it's like I don't even understand how all these tools fit together. I mean, the last time I even had to deal with an FPGA was over 10 years ago. And the last time I had to write anything that went into an FPGA was like 20 years ago.

Sure. FPGAs have become amazing. They're really, they're quite large. They do a lot. And I think timesharing, the advances that we got from being able to timeshare resources on computers in the 70s can now be applied to FPGAs perhaps. So when you say timesharing, do you mean, you know, I put my design on there and then I run my design for a little bit of time and then I release that resource and then you come along and put your design on. Yeah, and I don't even know what this might become, but we also are looking at whether it's possible

to share space on the FPGA. So to say, like, your core goes in the bottom left, and then this other core will go over here and run at the same time. There's also a lot of awesome security implications there. Oh, yeah. Especially because one of the best things to do with security, or with an FPGA is to do some hardware accelerators for security.

And so if you're going to share it, that makes it less secure. Maybe that's your key. Your key is the design, your custom design of the FPGA for decrypting and encrypting stuff. So let's say, you know, it's almost like you're making an Enigma part two. You know, you have a bunch of these Enigma machines, you give them all the same design,

you just say like, you know, trigger this on this message or on this day. With Balboa, Jen could write an app that's like Enigma, like she's saying. And I could write an app that used the FPGA to like do signal processing for that oscilloscope parts we had talked about.

Sure, yep. And runtime, we would be able to switch back and forth. Yes, that is the goal of the product. That's pretty cool. On the security side, there's also a paper from the 90s, which is really neat. If anyone's out there looking to read a cool paper about FPGAs, where someone went to evolve some computation,

I forget what the thing was, and ended up with a bunch of like unconnected blocks. And it was like, oh, what are these unconnected blocks? Delete, delete, delete. And when you deleted the blocks, it was like the junk DNA of the FPGA kind of like, you couldn't get rid of it, even though it had no obvious function.

And it turned out there was some capacitive coupling happening that made it possible for that computation to run as a result that they couldn't get rid of it. So you have, from that, the idea that you could evolve an antenna to spy on someone else's core or whatnot. Genetic programming was really popular for a little while, and then it's kind of gone by the wayside. I expect we'll see it again

because that sort of machine learning, deep learning sort of thing combined with computer languages, it seems like it will get there. So one of the things that's cool with FPGAs is the library of cores that you can get to do all kinds of stuff. So you usually get that

with the development environment from Xilinx or Altera. There's the OpenCores website also. So I was going to ask, how are you building a library of OpenCores? I guess that exists. Yeah, OpenCores is an okay website. It's a decent website where you can

find other people's freely published cores. But you're not necessarily getting IP from ARM. Did ARM release theirs? No, I don't think that's... I mean, you can get... We are not getting IP from elsewhere. Anyone, yeah. Period.

Well, one of the problems that I had, and this is probably because I haven't done it in a really long time, but did you have any particular stumbling blocks when you started working with fpgas like for me it was always mostly because i was doing schematic capture with the occasional writing of hdl it was glitching and race conditions if i didn't write it correctly that's why you need the genetic algorithm to tell you where to put those things

or i would fast track it onto one of the faster universal buses as opposed to one of the more local ones. You have no idea what I'm talking about, do you? I am so old. Timing is hard. Yes, timing is hard. Timing remains hard. Yeah.

And did you have to kind of brush off your digital logic skills as a result? Yeah, fortunately, I mean, not super much brushing i guess but like uh i really actually liked writing verilog because it was so much basic like you know digital logic and like assembly of blocks i think it was actually the biggest head trip for me about it was looking at um instead of like my usual issue with programming is i will tend to write things very imperatively like this happens and then this other thing happens and then everything happens exactly in order hopefully threads what are those i'm just kidding um but uh with the fpga it was like even more that like everything happens all at once right there's not even any that's where you end up with those race conditions yeah yeah so it was like even more of that. Like everything happens all at once, right? There's not even any.

That's where you end up with those race conditions. Yeah. So it was, it seemed mind bending to think about in the opposite direction than I'm used to. See, I thought hardware engineers would have that more innate, have a more gut feel for that because that's how hardware works. It's all relatively instantaneous. I mean, it was just a matter that the thing I usually have to guard against in myself was the opposite thing in this case.

I think you were born to be a software engineer. Ouch. I take it back. I take it back. I was just going to say, we will welcome her with open arms and no judgments, but apparently we will judge her. Look, I'm trying something new because I find it challenging and therefore interesting.

No, and it does sound really cool. And the idea that you're making it so that other people can more easily access the FPGAs, I think is awesome because it's still got a pretty high barrier to entry. Yeah, that's definitely true. So back popping up the stack the esc session in 2013 we never really did get to what do you remember other than the guy who has a lot of questions i remember the index cards oh yeah i remember we talked about uh toolkits like what tools are essential and how you prototype and we talked a little bit about

boards we play with and i think that and the tools at spark fun adafruit we probably both we all recommend those still um yeah i don't think that's really significantly changed no if anything they've gotten better and And my notes say Arduino and Raspberry Pi were on our list of things to recommend. Turns out. Turns out that hasn't changed much. That was a safe bet for us. Right.

Good job, guys. I mean, Raspberry Pi and BeagleBone, those should just be in the same breath all the time. We'll just call them Raspberry Beagles. Well, I don't know. I think Raspberry Pis were certainly previously the more flaky hardware. They had a lot of hardware problems with

earlier revs of Raspberry Pis, and I think they've improved that a lot. And they've gotten it cheaper. Both true. And they've made more different kinds. And I would add, I think there's actually a new dev board to pay attention to.

Which one? Have you seen the Odroid? It's designed in Korea. And instead of, it is ARM-based, but instead of running sort of like Raspbian or like embedded-ish Linux, you can run like actual full Debian on it. I've been impressed. I ran the GIMP on it,'ve been impressed i ran the gimp on it so just fyi it worked 100 1.2 gigahertz quad core processor two gigabyte ram this is a computer yeah whose chip is this they didn't give me that sorry well i mean it's not in my do you know what oh arm seven arm seven okay you can also run android i guess guess. Yeah, I would imagine. I just was wondering who, I'm not surprised it's ARM.

I was just wondering if it's like a Broadcom or, probably not Broadcom, given what I keep hearing, how tightly they hold their data sheets. I mean, Logic? Amlogic? Cortex-A5? Yeah, I don't know. I'm not going to continue to speed rate that but I would

one thing I was only a little aware of then was the embed boards and now they're on my list they come after Arduino if you're tired of your Arduino you're running out of space you are sick of those training wheels

here are some those training wheels. Here are some better training wheels. Here's embed. You get higher processing speeds, much more options. I mean, the BLE embed, I'm in love with it right now. So you're saying that now that you have your white belt in Arduino, you want to move up to your yellow belt in the embed. And the embed isn't going to hurt, but then once you get sick of your training wheels

and want to get down into debugging stuff with an on-chip debugger, you want to really get some efficient stuff going on, you have a board that is ready for you. I mean, the same board that I use that's embed compatible with the online compiler, I can use it with Kyle or IAR. And so I really like that. If Arduino continues to battle Arduino,

I may stop recommending them entirely. I'm not impressed with that. That whole thing happening? The depression, yes. It's depressing, definitely. Yeah. If anybody doesn't know, Arduino is fighting the trademark with Arduino.

And there are other letters that go in there, but who cares? They're supposed to be an educational, art, fun project, and instead they're just battling it out. And they may crush each other. It'll be sad. And so if your Arduino IDE wants to update, basically it's them fighting back and forth. It's probably going to update two or three times.

Really? Which is sort of hilarious, but also the sort of thing that makes me want to just whack people up the side of the head and say, what were you thinking? I don't know. Right now I have a couple different installs of Arduino because certain boards I'm working on, they just didn't quite jump the gap. Libraries just haven't been updated for those newer versions.

I have two. So I keep different versions. I keep a separate one for those newer versions. I have two. So I keep different versions. I keep a separate one for AT Tiny. I had an Arduino Due that was very powerful, but things like I2C just didn't work. And they said, oh, this will work eventually. We're maybe going to fix it. And then they ran off into the night.

Yeah. So it's like when fundamental things don't work, or you have to, if you have an educational thing that's supposed to be accessible to people, it's very frustrating when a major piece doesn't work and you have to go looking on forums to find out, oh yeah, that hasn't worked for a while. You didn't need that, did you?

You didn't need iSquared C, that's only really important. And like Arduino, Embed has huge libraries and lots of open source stuff. Maybe I will just start doing embed instead of Arduino. We'll see. Let's see, what else did we recommend then? DigiKey and Mouser

and future parentheses octapart. Can we take the parentheses off? I don't know. Boom. I don't know why the parentheses areapart. Can we take the parentheses off? I don't know. Octapart's pretty good. Boom. Yeah, no. I don't know why the parentheses are really there.

I guess I have to say element 14 now because I work with them. And Tindy. Well, element 14 was around then. Well, I mean, it's also really good. Element 14 is also Newark. And in the UK, they're Farnell. Yeah, that's happened, I think, in the last two years. It's the Element 14 role.

Yeah, Element 14 came out, I think it was maybe a little right around the time. They're Raspberry Pi, too, right? For North America. They're very involved with Raspberry Pi. And they've got the Ben Heck show, which has got neat videos. And I was unaware of Hackaday at that point. I don't think it was nearly as big as it

is now. Really? I was so aware of Hackaday. I feel like Hackaday raised me, like, really, honestly. Like, back in Elliot Phillips. They were there when you got home from school. Yeah, Fabian Jerry. Making you a sandwich and milk. I mean, all that. I mean, they were my connection to, like, people are out there building things. I was often rural, wherever. I had coworkers who would get featured on Hackaday even before then, but I didn't necessarily. I didn't spend a lot of time on it, unfortunately. So I am super late to the party here.

I mean, it's a new thing now, right? Same name, but new program. Well, they did get bought by SupplyFrame, so they got a lot more money. And that lets them do things like the Hackaday Prize. Like, fling people into space. Or into a pile of money. Yeah. I think, so Hackaday was indie when I found them, then got bought by AOL and now SupplyFrame. And, let's see, what other websites?

Tindy, of course. Now, I don't buy parts as much. I mean, sure, SparkFun and Adafruit, I liked having their breakout boards because I'm not a hardware engineer. I kind of need somebody to put on the pull-ups for me so that I can just hook it together with jumper wires. But now with Tindy, I can do that.

And sometimes I feel like if I was going to build a board, I would just go ahead and build 15 and see if anybody else wanted them. Because there is that community and it's kind of cool. Are there other write-up places like Hackaday are there other community places where else do you recommend new people go or people who want to garage tinker i don't know it to me right now so much of this feels very mainstream to me that yeah because we're not going to the maker fair this year because it's too busy right there are too many people there so we're not you're telling me i'm not going it was a question uh i haven't gone since the second one because i thought it was already crowded then yeah so i am not 10 now and it's 10 are you going i don't believe i'm going there is still the bring a hack dinner

on sunday if you want some of the maker fair flavor without the extreme crowds with with with beverages beverages and pizza. Oh, adult beverages for sure. That was really fun last year. We said two years ago to avoid Alibaba, which is funny because they're public now. Yes, I'm not surprised that they're public. I'm not surprised that they're doing well. I'm not surprised you said that.

Yeah, I'm not surprised you said that either. I mean, mostly Alibaba is just spending their time buying. They've bought up a lot of companies. They bought up Quixby, which has, for a moment, seemed to take up an entire block of Castro Street and Mountain View. Quixby is, or Quixly, is a search engine for mobile apps. Sorry, it's not just iTunes? What? Isn't that just the Google Play Store?

There's a world beyond out there. Apparently, yeah. I don't necessarily understand how they made money. They bought up another company that my partner worked at and they basically sacked everyone so but they're at a six billion dollar valuation oh yeah of course i mean a lot has changed in mobile especially in asia since we last spoke yeah it's everywhere uh i guess the iphone is a thing yeah they let the iphone in several iphones uh one plus exists

now too what What is that? I don't remember how to describe it. I don't think I'm familiar enough with it to say. It's a phone. You put it up to your face and use it to talk to people. It's different from Star's product

where you tap your chest and then you talk into it which is also different from Nextel in the 90s when you would just talk to your friends, which is also different from having walkie-talkies where you'd just be two of them talking to your friends. But there are so many more processors now, and you can now get a cell modem for not that much. And Wi-Fi, you could do electric imp

if you want some nice things to surround it. Oh, right. Spark exists now. Yeah, that's what I was going to say. The Spark. The Electron. That's right. I backed the Electron.

I was looking at my list. I had to back out of that one because I had too many. The Spark is a cell modem. Yes. And it comes with some data program. Spark is more like... Electron is Wi-Fi and Photon will do cell. I thought Photon...

No, I thought the Electron does cell. Spark was just a processor board with Wi-Fi. Sorry, Spark. I don't know what Photon is. So Spark... They make a...

It makes it as a light. I thought it was called Spark. If I thought... I thought that's what it was. I talked to... Breadboard that lets your projects talk Wi-Fi? I thought the photon, they had one that was like a light bulb.

But maybe I'm confusing that with Misfits light bulb-y thing. Probably many projects going by the Spark name. But I know Spark.io. It's kind of a cool name, so might as well call everybody a thing then. But yeah, as far as I can tell, so I spoke with one of the investors in Spark.io and we talked a lot about some of the challenges we experienced at BIA with having the cell modem and the data plan management and all that jazz.

So I'm excited to see that because they're using exactly the same chip that Bia was using. Well, and now with Wi-Fi, if you want to go back to that, there's also this ESP8266. Like $5 you can be on online. It's been causing such ripples. I know. It's really cool.

Actually, one of the Hackaday features, CN Lore, Charles, is going to be on the show in a few weeks. So I will know a lot more about ESP8266 soon because it is really popular. What else? Have either of you done anything with it? No. I haven't gone on a date with it or anything.

I mean, I want to, but I also booked him for the show a little while ago, and I figured I would. That's your tutorial session, and then you're off and running. That's why we have the show, right? So people come over and explain these things to me. And then you can decide whether you're going to buy it or not, or whether you can get a free one. I usually buy it or not. Or whether you can get a free one. Ah, you should buy it.

So one of the things that I think has changed since we spoke was, at least it seemed like in 20 and 20. No, I thought we did in 2012. 13. I looked it up, I'm sure. It's been two years. We can tell because it's 100. And there are 50 weeks in every year.

Except for those two times or so. I must be on vacation those weeks. Sorry, guys. Can we backtrack? I just looked up the Spark thing. Correct us. Well, I got it backwards. Correct us.

He shook his head. Oh, we just got it backwards. You just got it backwards. I got it backwards. The photon does Wi-Fi and the electron does cell. Yes. As I think Chris was saying. Wunderbar.

So one of the things that I think has changed is that it seemed like in 2013 that we were definitely not giving a rat's ass about the privacy issues in the products that we were inviting into our house. And now it seems like, at least definitely in the last year, suddenly we're all worried about it. The Chip Whisperer guy sent me one of his prototypes. He's on Kickstarter, and I haven't played with it yet, but just watching the videos from Hackaday, oh my God.

I don't know that I care about privacy because it's clear that it doesn't exist. It's not that hard to crack things. I can't wait to get mine, mean i'm not i'm not seriously podcasting my phone calls but it just doesn't seem i mean like it seems like it's uh not being enforced right like if you think of me as being like an unintentional radiator i can expect that uh my whatever you know digital you know energy trail is being recorded somehow. But I think societally, there's still the idea of an expectation of privacy, which is what I think is the thing that's being eroded.

Well, and I think there's a confusion, too, between privacy from corporations, where you have Google looking at all the stuff you type or whatever, and Apple and whoever is managing your data, which is a choice you make. I know it's kind of hard to not make that choice these days. But that's still a contrast from the government, who have weapons and can put you in prison. Or deliver you delicious tacos, apparently. What?

Taco Cups are not a state, not even a proto-state. You know what I'm saying? Corporations taking your privacy is one thing, and they can do bad things to you, but they can't force you to do things. They can't take action against you or take your life. So that gets wrapped up in the privacy discussion too, and it makes things much more complicated.

And then there's the other aspect of criminals. I mean, people breaking into things. Black hat hackers, which don't have to be corporations or government. Oh, sure. Or corporations that put in back doors that allow the government to go in and look at all the data. Yeah, I don't know whether to blame that on corporations or government, but the third

group of people like me who want your data, except I'm not bad. And we're all pretty clued in. We're all pretty clued into this, and it's still confusing. And I think for the general public, it's probably just forget it. So what do you really think about the Internet of Things? Oh, it's so insecure. I mean, it's just ridiculous. And nobody's doing anything.

My Internet of Things are very promiscuous with probably everyone else's Internet of Things. I mean, we have all the things that should make me nervous. We have a nest. And so Google is watching us at all times. I'm just going to assume that. Well, they have a backup copy of this podcast too, so that's good. Thanks, Google. Our car is easily tracked and we determined recently that our data does go to their website.

It probably isn't stored, but it is stored long enough. I didn't think that acceleration data was being uploaded. How do you think it's going to work? How do you think they learned to make the product auto-drive you? Or to go that fast? Wow. Yeah, so I'm very skeptical that the Internet of Things will be anything other than a privacy disaster.

Did you guys see the IP browser for finding toasters that are connected to the web? No. We'll see that in the show notes, clearly. I don't remember what it was called off the cuff. Yeah, but going back to what Chris had said, there was the John Oliver show a couple weeks back where he interviewed Snowden. That was a really good one, too,

because it explained privacy in terms people can understand. Yeah, but also had people... Can I respond to that? Actually, I really didn't enjoy the John Oliver Snowden episode. I didn't like the show very much. I liked the interview beforehand. I liked the write-up of how... So I should just say a caveat.

Like I have enjoyed John Oliver. I'm not a John Oliver hater. In fact, I quite appreciate his work, but I really felt that the Snowden interview was like led and aggressively edited and cut. And it really, I didn't really enjoy that. I would agree. I definitely agree there. But for me, what was more,

made my blood pressure rise was how little the people that he pulled on the street seemed to remember who he was, what he did, or have a very clear opinion on it. Well, they had a clear opinion on it, but it generally was like, oh, it was bad. They didn't even know what he leaked. And that was just like... Probably people don't know as much as they should,

but those shows are very good at picking the stuff that will raise your blood pressure. Yes, yes, yes. But yeah, I mean, but it shouldn't be people's jobs to constantly pay attention to all this stuff. I mean, we shouldn't live in a country where that's necessary. Really? Well, if you don't take it as part of your civic duty to know what the heck's going on. Yes, to know what the heck's going on, but not at a level of detail where.

Help. Dig faster help dig faster dig faster we could drop it I totally respect what you're saying my point is the country shouldn't be broken enough so that it should be necessary for us to have to deal with revelations of that magnitude and because to understand it completely you really do have to

research a hell of a lot of stuff and so things should not be going wrong such that the average citizen has to go dig into that stuff that deeply other than the normal news consumption. I have a new idea. We have a branch-off podcast called Politics Now.

You don't want to talk politics now. I know I don't. I was going to remind you that I have cake. So there's some temptation to get some cake right now and eat it in the pot while we're talking. But Chris said that if I did that, he wouldn't. Exactly. Plus milk makes your voice sound weird. It makes you a little mucusy.

We actually have a podcast where Jen drank hot chocolate before. And if you listen to the hot chocolate episode and the one before she sounds like totally different yeah but the reason why I needed the hot chocolate was because my throat hurt I don't buy it no no I I don't remember that that that that was post editing in your brain not can we call this like the cake and milk revelations because it's blowing my mind there's so many things that can destroy a voice but uh so we've talked a little bit about how things have changed in the next in the last two years what are we looking at for the next two years next three years what what do you think is going to be different i think privacy and security would be on that list um and we talked about wi-fi chips and cell modems because those are becoming far more ubiquitous.

But is there something else that we haven't really seen that maybe we should be thinking about? Well, I think the battery technology is one of the last things that is really holding us back. I mean, we're shrinking chips pretty well, and we're making more powerful and more power conscious. But the battery technology, I don't think, has gotten there that we can really shrink things down. I think that's like kind of one of the next barriers that we need to punch through.

I have hopes for energy harvesting. Although I've had hopes for energy harvesting for like two years now. And it hasn't really gotten as close as I'd hoped. Do you have something? Good questions. I like imagining the future. And I do have a lot of high hopes for Star's project because I think that's also like the next area, the next wall to kind of break down

in terms of hobbyist electronics. FPGAs? Yeah. To me, it's very much a forefront because people are like, oh, I can spin a board very easily and I can master that. But for the most part, a lot of the board stuff has been just deal with the reference design and make very minor changes and connect the lines. I know that's a drastic simplification, but for the most part, that's what's been happening.

And with FPGAs, I think this is really going to be this chance to really get some deeper knowledge without a lot of the overhead and cost. So when she knocks down all those extra obstacles, I'm really curious to see where that's going to take us. No pressure. I should add that we are looking to make it more general and not Novena specific on a fairly quick schedule too. The lattice board, Ice 40, looks interesting. It might be our next target. I like the Papilio and I bet Jack Gassett would be happy to talk to you.

Wonderful. The other area of change has been boards. Two years ago, I can't really imagine that I would have spun my own boards. Well, just all kinds of fabrication. Right, because 3D printers and CNC mills and all of those things. But Osh Park made it easy to do 15 boards and cheap. And then Tindy made it easy to sell off your excess. Do you think we'll be seeing more in that?

I mean, 3D printing, yeah, okay. That's been coming for a while, and now we're getting serious. It's really, really coming soon. 3D printing seems here. Seems arrived. Just me. Well, I don't have a printer. The last two,

at one point, two of the firmware candidates that I had sitting on my desk, both of which claimed that they had a 3D printer startup. Both. They weren't even working together. No, they're ubiquitous. I don't have one either, but I think that the technology has arrived in the sense that I feel very confident that if I wanted to get something made, like it's an accessible technology to me it's not lab grade it's not often you know wherever it's like I feel like I have access to it as an engineer so I feel like I could

could send something to shapeways and get it printed that that part's easy yes but I have a project I was working on this week and I didn't really like the container. And I didn't like it, and I didn't like it, and I tried like six things. And I have in my head what I want it to look like. But I don't have the skills to go from imagination to physical object yet. And maybe those are my skills, because I'm a software engineer and I'm a mechanical engineer.

I agree. I think the software engineer and I'm a mechanical engineer. I agree. I think the software for manipulating 3D is still hard and it's still really difficult to go from concept to physical, no matter what. What are you using, if you use it at all? Yeah, mainly, I guess the 3D stuff I've done recently was actually done using Autodesk's Fusion 360. I can't get it to install. I don't know what it is.

Autodesk has a web-based one that's supposed to be pretty easy to use. I've mostly been trained on SolidWorks, which is an expensive way to go. Yep, you've used SolidWorks. You know, I was also, this might sound funny, but I was a huge fan of Tinkercad. I think it still exists, but it was basically a CAD tool designed for children to use, and it was perfect for a lot of things. That would be perfect for me. What I like about all this is the prototyping is almost turnkey now.

All these physical things that used to be, you know, would take weeks if not months to get done. So now we're almost to the point where we could not quite, really not quite, but we have some sense that we could be really more agile in a similar way that web software and other software products are. We're getting there. We're not still quite there, but we can get like,

because we can finally turn things over so quickly if we have a 3D printer in house and we have a mill and so forth. It's getting there. It's just, you know, those are the things that end up killing us. And so we just take a lot longer to make it perfect before we send it out. That's always been the issue. That's what's always kept us being waterfall. So I'm wondering if we're actually going to make it to real agile for physical devices. And that brings us to replicators. I cannot wait for that. Also, teleporters.

And I am still waiting for my CastAR. That's one of the Kickstarters I haven't gotten anything yet. But soon. You need to have Jerry back. And you need to go visit her at when she's in mountain view at the new office yes she and i are exchanging emails i don't know when she'll be back because she's super busy take pictures i will ask before we move on from that topic do either of you have a other machine other mill yeah i I love it, but I don't. It's tempting. It was just a little too

expensive for what we do. And we don't make boards. So my desire to get the other mill would be more in the craft area. And it was not quite priced right for that, especially given my inability to use any

CAD programming. What if we went Tabsies on it? You guys, I would mill things for you, just FYI. In fact, I was cutting circuit boards as a service on my other mill and still would be open to that. So you do have one? I do. Do you like it?

I do. It is the only mill, and I say this as like I worked as a machinist in high school, like I would not leave any mill unintended except for I've come to trust the other mill that much that is great we did have Danielle Applestone on the show she is fantastic she was really great she made the earrings I'm wearing I love Danielle she's so wonderful yeah no they were They were cut on another mill. Yeah, see, I would use it for craft. It's perfect.

It's great. It's great for that, too. Well, we also have a tech shop membership, so if I need to mill stuff, I can do it. If you need to injure yourself, you can injure yourself. They don't have another mill. They have a regular mill. And usually if I'm doing something like that, I want it to be big. I like big machines.

I have lost it. It's time for cake. Your blood sugar must be dropped. Yeah. You're low on the ground. It's chocolate cake with chocolate chip ice cream. And you're all invited. And rainbow sprinkles.

Wonderful. Anybody's not here, you can just imagine. A slice of cake to all of the listeners out there in Radioland. Definitely. Yes. Radio cake. You can find it in your email after we invent it. Christopher, do you have?

No. He's just shaking his head like, I want cake. We've got to find an animated GIF of cake to put in the show notes. The cake is a? The cake is not a lie today. Well, it's a lie to everybody in there, but it's not a lie to everybody here. The cake is not a lie.

Star, do you have any last thoughts you'd like to leave us with? Interesting parting thoughts. Unprepared. We can make Jenga first. I'm not prepared either. We could just cut this section. I could say, I could say, Elle, you've made it to 100 episodes. I am so proud of you for sticking with it because I think after two months,

you were seriously thinking about dropping it. And think about all the awesome people that you've met in that time and all the new products you've introduced people to. In fact, I would say half the Kickstarter projects you've put on the show, I've invested in. Yes, me too. I would also say congratulations on reaching a hundred episodes. Quite the milestone. I really only meant to do a half a dozen. Here we are. And now it's two years old, and I'm amazed. I'm still interested. I'm still excited about doing it.

And I didn't expect the show to be like this. I didn't expect to get this many listeners. I didn't expect to meet so many really cool people. It's been a lot more than I imagined. There's a lot of work to go just to get some cake. All right. I have tortured them enough that they are demanding cake. My guests have been Star Simpson, technology strategy lead for Orion,

which is separate from the Balboa project. Absolutely. So you're also software? What title would you use for that? Balboa project. Absolutely. Okay. So you're also software? What title would you use for that? Balboa? Yeah. A title for my free time project? Yeah.

Oh, God. Free tour. Maybe no. Maybe no. Okay. And Jen Castillo is my other guest, firmware engineer extraordinaire. I'm happy we could get the band back together to celebrate.

Thank you both for coming. And this show would not be possible without Christopher. Jen has been our show's biggest cheerleader, but as producer, he has been here every week helping and doing and talking and being amazing. I also could not do the show without you listeners. Thank you for making it so much more than I really ever imagined. If you'd like to say hello to any of us, email or hit the contact link on embedded.fm. I like choosing my final thoughts, but this week I went ahead and let a listener choose. John, you wanted me to state an intention to continue on through Robert Frost's Stopping by Woods on a Snowy Evening. I'm not going to read the whole thing, just that last bit.

The woods are lovely, dark and deep, but I have promises to keep and miles to go before I sleep and miles to go before I sleep.