Embedded - 272: Stick ‘Em on Whales

Episode Date: December 28, 2018

Chris Gammell (@Chris_Gammell) of The Amp Hour (@TheAmpHour) joined us to talk about the state of the industry, listeners, guests, and life in general. Embedded’s accounting episode (150: Sad Country Song) Contextual Electronics Consulting forum (requires you to apply) Remote work 250: Yolo Snarf Excellent video on how prototype PCBs have improved over the years Quickly falling cost of dev boards Elecia worked on learning and building robots and happily got a related job Chris W is building IOS apps Object oriented Prototype to Product: A Practical Guide for Getting to Market by Alan Cohen (Emebdded.fm interview) CircuitPython Visual Basic as a prototyping language ESP32 and EXP8266 longevity and use in products WiFi provisioning Electric Imp, Particle.io Azure IoT Hub, AWS IoT, Google Cloud Iot, Ubidots, and IoT App Story (the one Chris G remembered later) Wallet.fail Anki Vector robot Genuine People Personality (from Hitchhiker's Guide to the Galaxy) Genuine people personalities are coming to our gadgets (ArsTechnica) LoRA and chuckable sensors LoRaWAN and ARM Mbed OS Telepresence and mirroring others The Amp Hour ToorCamp episodes Sourdough (a novel about robotics and AI) and Embedded’s interview with the author Energy and Civilization: A History by Vaclav Smil Jeri Ellsworth spoke about the demise of CarstAR in The Amp Hour 394 The Stone Monsters music products Llama and the IoT zines Related Oatmeal comic Supporting Embedded Patreon leads to a link to their slack channel, mentioned in this show. Supporting The Amp Hour Patreon is also a great idea.

Welcome to the M-Bauer. I'm Chris Gamble of Contextual Electronics. I'm Alicia White of Embedded FM. I'm Chris White of Embedded FM. Not a fan of the show names still. I think we could do better than that, but I think it's what we're stuck with now. Now that we've done this, what, two, three times now? We could call it something else.

We could call it the Amp. No, we did Amp Edd once, didn't we? Yeah. We did Amp Edd once, yeah. Well, I mean, we could be like the Avengers, the greatest crossover movie of all time. Yeah. How are you guys doing? It's been a while.

It's been... A year. Well, we talk a lot on Slack, but it's been a year in the talky talky, so... It's a year that feels like 10. Yeah, it's been a long year. But i haven't accomplished a damn thing so yeah you have chris you've been doing a lot of things i've been doing some things yeah i left my job uh i'm now consulting i announced a conference that i'm running so yeah a couple things it's been good though and conference is the kai cat conference key cat conference i'm just gonna keep kwee cat is kwee kwee kwee cat as brian So, yeah, a couple things. It's been good, though. And conference is the KiCat conference?

KiCat conference? That's right. I'm just going to keep... KweeCat. KweeCat, as Brian Benchoff says. Yeah, the KiCon is going to be in April of 2019 in Chicago, where I live. And I'm really excited about it. You guys going to come out for it?

Huh? No. You know, I would have predicted that, but I'm still a little sad. I mean, if you get Harrison Ford to give keynote or something. To keynote it? Okay. We should call it the keynote. I've never actually opened said program, so I'm not your target demographic. That'll also do it, yeah, right.

If people are interested, I'll link it in, but it's probably not as many of the embedded listeners as I'd hope for, huh? I don't know. Probably more than... Our listeners are probably more interested or more targeted than we are. I mean, we don't even go to San Jose anymore

to go to conferences. Oh, okay. We just go over there for croissants. Yeah. That's a good reason, though. I mean, what kind are we talking here like chocolate croissants yeah yeah chocolate almond chocolate almond yeah okay that's that's worth the drive and consulting uh you've been doing that for six months now uh no only like four and a half four and a half five months um back now i guess august so but it's going well yeah it's it's great i mean it's um you know someone had so we basically you know we had been putting you guys are always so

good about show notes and i saw one of the the questions from the embedded uh slack was asking about skill building and stuff like that and i'd love to talk about that kind of stuff but that was one of the things that i was most excited about is that my skills i was doing a lot of marketing stuff and my skills were just falling off you know and it i don't like that and now i'm just kind of back in doing 3d modeling and doing you know circuit design and doing some firmware stuff and it's it's not easy by any stretch but it's it's great to be back into it and i really like that no that's great the The great thing about consulting, too, is you end up getting exposed to a lot of different things. Even if you went back full-time,

I'm a double-E full-time at megacorp.com. You might find that you still feel like you're not learning anything or your skills are stagnating because you're doing one kind of design over and over again. I really liked consulting for that because it was like, oh, these people need this and i have no idea how to do that but i better learn really quick well you've both been consultants for a long time or you have done a lot in the past so like what what should i keep an eye out for on the uh when's it going to turn like what made you go

back to the chris i guess you're back at a full-time gig right i can't really say what made me go back without sounding like an ass so i'm not is it money yeah uh yes let's say that um well yeah this is probably something i shouldn't discuss okay that's fine that's fine well at least you've gone back and forth before right yeah i i was at i was doing uh consulting and then I went to ShotSpotter because they needed my skills and it was a really interesting company. So they didn't want contract because they wanted to develop algorithms in-house. And so I did. I went back to full-time and then went back to contracting and have applied for a full-time job in the last year. Not because I don't like contracting, just because I wanted that job.

I see. Okay. And it didn't work out, although now I have a contract with those people. So we'll see. Okay. I've got an answer now. Okay. Go ahead.

Part of it was because I was made an ultimatum. It wasn't stated as an ultimatum, but it was probably an ultimatum. And I had a choice whether to go do something else or continue what I was doing with the people I liked working with and go full-time. So that was a large part of it. And, you know, it's been good so far. Things are getting a little full-timey.

Yeah, I mean, that's the thing. There's more meetings, there's more politics, there's all the stuff that I've experienced, I suppose. Yeah, and the big thing for me, which is kind of personal and idiosyncratic, is I found that when I was full-time, especially at startups where I was, where I had a lot more responsibility of a smaller team, I was way-time, especially at startups where I was, um, where I had a lot more responsibility

of a smaller team, I was way too emotionally connected. So any decision that people above me made that I thought was ill-advised, I would get kind of angry. Yeah. And you know, I get angry in meetings and stuff and I really didn't like that. Um, and that was one of the reasons I went contracting the first time, uh, which is now almost 10 years ago. So, yeah, going full-time, all those meetings and things that when you're a contractor, you can kind of say, well, I'm hourly, and you guys can talk all you want. You don't want to pay for me to be there.

And if you don't want to listen to what I have to say, then it's your dime anyway. Now it's like, wow, the company is, you know, there's some emotional investment there. So now decisions like that or arguments like that become more personal. So there is a different kind of psychology that I find, that I have when I'm working full-time versus contract. And how do you define contract versus consulting? I always keep getting those intermeshed. You have a i do um contracting is where you do the implementation and the work consulting is where you usually work

more with the managers to create a plan that somebody else implements uh so if you're doing architecture work without implementation then that's consulting or mentoring or doing the reviews that's more the consulting but when you are you know sitting down to do the technical work that's contracting and i enjoy doing both um sometimes it's really fun to come in and be their sounding board or to solve bugs that they're having trouble with. But sometimes you come into a situation and you don't get all the information you should. And so you spend a lot of time consulting

for things that don't make sense. Yeah, I'm not sure it's a formal definition, but it's a good, I like it. People use it interchangeably. I've always used it's a formal definition, but it's a good... I like it. People use it interchangeably. I've always used it kind of interchangeably. Yeah, consulting, you tend to be dropped in for a shorter period of time, maybe. We have a specific problem and we need your expertise, and here's our code.

Could you look at it and tell us what's wrong? Here's our board. Could you fix it? Less about, okay about okay well for the next six months we'd like you to work on this part of the system and finish it for us i was thinking about the contract being like an actual like chunk of time like oh it's only three months but you're here for that whole time and we get you for you know half of your time or something like that it felt more that felt more like a contract kind of thing obviously there's a contract in both cases but whatever um versus like yeah here and there kind of stuff like a contract kind of thing obviously there's contract in both cases but whatever versus like yeah here and there kind of stuff like a consultancy kind of feels like so how

are you finding stuff are you marketing yourself are you working with your existing network say i'm marketing myself right now chris call today to talk to chris gamble for three easy payments of 1999 are you doing it under your own name or under analog life under analog life yeah yeah i'm uh i'm back to use i mean i've had an llc for a long time but um it's uh i think i'm gonna probably try and switch to the i was thinking about your accounting episode a couple weeks ago um and i'm probably gonna switch to an s corp next year and do that kind of thing and it works out fine i don't know i i right now i'm just like a you know a 1099 and a schedule c employee not even an employee i guess i'm just no you can't say that

yeah yeah not i'm not i'm just i'm just a guy waiting here to do circuits for you you know and uh yeah it works okay for now but i think next year it'll be a little bit more a little bit more formalized and payroll-y and stuff like that and so i really and if people haven't listened to your episode about that stuff i i always recommend that i think that was a great it was a very unexpected but a great episode about like understanding how the accounting stuff works and you know it can eat a lot of your day if you if you do it right or wrong so much of your day especially if you make a mistake as complicated as electronics and and and software is yeah a uh an untraceable error in your books is often much harder to solve.

Yeah. I haven't done our books yet. Don't curse it. All year? Or just for the month? I usually do it by the quarter. Oh, okay. That's not bad then.

Yeah. You know, yeah, it's like silly things for us. So we have very generous patrons and, uh you know so we get donations as you do i'm sure and people should support embedded fm's patreon um uh and amp powers yeah sure sure but it shows up as like a payment is like you know so like we get a you know say we get a five dollar donation you know paypal sends of four dollars and 28 cents once a month and then that is like a journal entry and that's like a thing you have to deal with you know and it's like i'm very grateful for the the donation but it's like it's just like

these small things that i just need to clean up my process and stuff that's it and patreon doesn't do that they send it all as one chunk and and that's nice. Yeah, right, right. Yeah, we still have people from the old days as well. But it's, yeah. We had a, me and Dave talked about business a couple weeks back, and it was two people who don't know anything about business just complaining about business pretty much. But it was, it definitely changes the focus. You know, you get to do less technical stuff.

And I think, the thing I always think about is, as you get bigger and bigger as a company, that stuff must just take over your life, you know? Like, that just must, like, having actual employees and dealing with larger invoices. Yeah, and you need to hire people to deal with that. Yeah, exactly. Right. I mean, I think the bookkeeper is one of the first five people you hire,

because it does take up so much of your time if you want to do something other than play with QuickBooks. It's true. So other than that, other than like the business-y stuff, it's been really good. And I should mention, too, if people on your channel haven't heard on the Amp Hour, I do have a consulting forum now. And that's been very helpful. It's semi-private, you know, so people share their names, you have to apply for it. I'll put a link in for the application. But people that

are consulting or contracting, that's the main focus. If people are getting into that space, they can also apply. But it's been really good just for putting stuff out there of things that are having issues with and best practices and things like that. And it's been really helpful for me and hopefully for other people that are there. So I like that. Because it's just like putting on Twitter would be a little bit different, I think. Well, it's hard on Twitter to say this part of the contract looks a little odd to me can anybody tell me if it's ever been used against them or or what they think of it and on a smaller forum where you know everyone it's it's easier yeah i mean i've even put out a couple

of uh a couple people come to me with with gigs that i just couldn't do uh either because the skills or because of timing and so i've already put a couple there too and to me with, with gigs that I just couldn't do, uh, either because the skills or because of timing. And so I've already put a couple there too. And, you know, I think one of them has worked out out of the three that I put there. So that's encouraging. I don't know if that'll work longer term.

They all have to be remote, obviously. And, uh, but I do wonder, like, I mean, both of you work remote,

right? Yeah. So, I mean, where do you see remote, remote? Yeah. So, I mean, where do you see remote work going? I mean, like, do you think that that's still a consulting thing that's going to keep happening? Or is that increasing, decreasing? What's the trends you're seeing? I don't see it decreasing, really, because the demand for talent is so high right now.

Still, a lot of small places are looking for whatever they can get. the demand for talent is so high right now still. A lot of small places are looking for whatever they can get. And some of the small places are in remote places. You've had contracts come across from Denver. Denver and Boulder always seem willing to do remote. Yeah, so it's like, well, you're not going to get as vast a pool of talent out of Colorado, maybe. But if you are willing to work remote, then you can pull in Bay Area people or Boston people or what have you. I don't see it's decreasing.

You're going to say Chicago next, I think. Yeah. Sorry. It's okay. Yeah, I don't think it's decreasing. I do think there's a tendency for larger companies to go through this period of everyone needs to sit next to each other. And we're getting rid of all remote workers and we're consolidating. I've seen that happen a couple of times and doesn't tend to fix any problems.

Yeah, of course. If you're not communicating next to each other, you're not going to communicate. If you're not communicating remote, it's not going to help if you're sitting next to each other. But yeah, I don't see it decreasing. Do you? No. No, i don't see it i don't see it decreasing do you no no i don't um the tools are getting cheaper so it used to be having your own oscilloscope or your own logic analyzer was too expensive so you had to go in and use theirs a shared resource oh yeah in a whole company now, you know, I have my gear. I have a soldering station that's not bad, and it exceeds my abilities. So if I break something so bad it needs more soldering than I can do here, I would have had to bug somebody at work anyway.

And it's amazing how fast the postal system will work if you're only going a couple hundred miles now at least at least for software the collaboration tools have gotten better too like most people use github you know for a small company and gets reasonably good for collaborating and has all the code review stuff you need and it's you know it's in the cloud so you don't have to have infrastructure at your company that somebody needs to share i don't know is that true for electrical stuff i mean these days i'm using github for all of my uh all my repos for uh for like hard like all of my hardware repos are on github and similar sometimes bitbucket or or what's the other one gitlab i guess too but yeah

who cares right i mean like it's all the same on my end. And, yeah, I mean, one time, I mean, I think it's interesting seeing when and if the hardware people that I'm working with know how to use it. I think that that's definitely a skill that hardware people should be working on if they don't already know how to do it. Command line Git is really important, and understanding how and when to do commits and all that other stuff. But I think that is much, much more prevalent than it used to be.

I'm going to take a plug here for one of our shows, and that is Yellow Snarf, where we talked about git and version control as a theory. The idea was to have a basics episode where we we just talked about it and we started to find a fake uh version control system called yellow and then we would just make up commands to describe the general process and it worked out really well for me because i don't love git maybe it just isn't my the way brain works. Maybe I haven't used it well. No one loves Git except Linus. I kind of like it.

I like it. Linus and Chris. But the YOLO episode, I started thinking about, okay, so if YOLO was a real thing, how would I explain it? How would I do it? And then I would type Git and whatever I wanted, whatever I would have put into YOLO into Google, and suddenly I got a lot better at git because

I got a lot better at thinking about what it is I wanted to do instead of just doing git reset remote head purple penguins. And that never, that command never worked for me. That command is very dangerous. You should never type that command.

Yeah. I think the big thing with hardware is that it's, because you can't really merge stuff. Right. Well, you can, but it's a mess. Not easy. Yeah. But most hardware people aren't working like that anyways, right?

It's a lot of, you know, doing a chunk of work, handing it off to someone, doing a chunk of work, handing it off to someone, you know, back and forth like that. So I think from that perspective perspective revision control can work great for that that stuff um but yeah i love it i don't know i sure as hell love it when i when kyke head crashes you know it still happens so yeah lose an hour or two or a day is a work instead of weeks. Right, right. So you did a show recently where you were talking to Jeff Kaiser and Dave about your view on the industry. Was it last week?

Yeah. Yes. I'm so behind on podcasts. I got into an audio book and then it was a trilogy and now I'm just like, okay, it'll be another three months. We'll take a pause. We'll take a pause on Jeff andff and dave what's what's the audiobook oh it's um elona andrews uh it starts with clean sweep and it goes through three others and it's kind of silly but it's also really fun yeah okay i'm down science. I was just doing, I just did the second book in the, what's the sci-fi show?

They made it into a sci-fi show, but I haven't watched it yet. It was Leviathan Wakes and then Calibre Expanse. Calibre Expanse, yeah. Expanse, yeah. Oh, man. And the book, the audio book was amazing for the second one. I thought the book for those were a lot better than the sci-fi shows. Yeah, I haven't watched the show yet. okay well i'll check that out the quick quick quick interlude so that's good

alona andrews that's what was the question uh so they were talking about oh what's your view on the industry and yeah and chris thought up his answer many hours after they recorded. So I figured we should give him a shot. So Chris, what is your view on the way the industry is going? You know, that's a great question, Alicia. I'm glad you asked it. You ever notice that people do that in interviews? Radio announcer voice. Yeah. That's always like the stall that people do. Oh yeah. That in media trading, that's what they tell you to do. That's a great question. Or I really,

really glad you brought up your point. That's just waste two seconds while your brain answers it. I've forgotten what the question was, but yeah, I'm glad we went down that interlude. So the, the thing we were, we were,

we were talking, we ended up just kind of divulging into, not diverging into, you know, the, the hype curve and stuff like that. And that was silly just because we were floundering. I was floundering.

But one thing I actually was thinking about, and then there was a video that was posted today on your Slack channel that actually hit this perfectly, was just how PCB prototyping has changed. And so on your Slack channel, one of the people had posted a video about actually showing showing the different uh the the quality of of prototyping pcbs over the past eight years and um it is pretty stark so it was user andy brown on youtube i'll link that in um but not

even just the quality i mean, the quality has definitely definitely changed, but but it's even more so just the cost, it's just so cheap. And so I've come to this idea of like doing what I call Lego prototyping. And the idea of like, you could just take modules off the shelf, quickly connect things together in your CAD program of choice, put a bunch of headers on board, and then just, you know know instead of wiring things point to point now it's literally faster and cheaper to just do it in a repeatable way using a two layer four layer pcb and i just think that that's changed a lot of things um obviously very specific to the electrical side of things but that that is uh you know that has changed the cost basis i think for even assembly and modularity and uh a lot of things i don't know

i don't know if you you've both seen that i haven't seen the video but i understand where you're going because things have changed so much it used to be you couldn't dev boards cost ten thousand dollars yeah right and you never bought them you were only given them by the reps for like two or three weeks. Right. And you couldn't get an accelerometer board just to try it out. And now you just go to SparkFun and there's all this stuff. A couple bucks. I was kind of appalled to find out that some of the dev boards, like the Nucleo ones from ST, you're not supposed to OEM them. So if you want to build something with that as a core,

you're supposed to make your own board. I was a little disappointed. Yeah, but some of those things are huge, right? Yeah, except I was doing something that was, you know, we only needed about 20 of them, and it wasn't size constrained, and it was kind of like what you were saying. It was Lego blocks put together for a fairly looks like works like prototype. Oh, okay. And yet I was like, this isn't right.

We shouldn't do 20 because they're saying we shouldn't do this. But on the other hand, making our own boy to do exactly this is dumb. I wonder why they do that. I mean, I doubt they're taking huge losses on dev boards i think because dev board quality control is not as important as oem quality control if your dev board doesn't work you just get frustrated if your product doesn't work yeah but i mean they don't like going to give you any warrant warrant here they can just say this is not warranted if you use it. I don't know. But yeah. I think it's just the, I think the margins are definitely lower, and they don't want it to extract stuff out of the supply chain either, right?

So if you think about, well, we're going to make 5,000 dev boards this year, because we expect the demand to be 5,000, but then one user, for some reason, the cost basis, even if they're making money on it, whatever, right? But then the cost basis is good enough that they're like, oh, well, we're just going to design 4,000 of these into our product, and then they go and buy out the supply chain. Oh, sure, whatever, right? But then the cost basis is good enough that they're like, oh, well, we're just going to design 4,000 of these into our product. And then they go and buy out the supply chain. Oh, sure, sure, sure.

Then their part's not in the hand of what they expected to be 5,000 engineers or on the shelves of 5,000 engineers as the reality usually is, or maybe the shelves of 2,000 engineers. The idea being that I think it's just because it's a marketing vehicle, they don't expect it to,

they don't want it to be part of a product. makes sense i mean you you do give your dev boards a lower profit margin than your production boards and i'm sure it's like internally too i'm sure that it's charged as a marketing expense versus a development expense regardless so they they don't want it to be like well our you know our division is doing $2 million a year in boards. Obviously, this is ridiculous. And our marketing department is doing $20 million a year in boards. But in theory, that could get skewed in a certain way. So going back to prototyping faster. So one of the things you were talking about was that's changed in the industry.

That was one of the things you talked about last year as a resolution was learning to prototype yeah yeah yeah is that are those related uh they're not related actually that was actually but you you had written that on the list after the fact um but i think that i think that just doing consulting and and such is is actually kind of just knocked me into that mode of like well i don't have time i have to kind of just knocked me into that mode of like, well, I don't have time. I have to kind of glue this stuff together. But like as an example, I did a project a couple months ago where I didn't have time to spin up an entire new board, but I took a dev board off the shelf. I made a sensor board that was custom for my application. There was a standard pinout on the dev board.

I just hooked into that. I had a screen that hooked into that and then i 3d printed a 3d design 3d modeled a case around it and at the end you know it was a thing that had a screen it did sensing it did it had rf on board like all of these things were something that it was basically a proof of concept but at the end of the day it was a thing that i needed to get done and out the out the door it was probably i don't know 10 20 times what it would cost in a final in a final version of it but it worked out great for for what i was talking about so in that case like the i think the pcb cost two bucks to to make you know i think the and then the prototyping i get to 3d print prototypes at my workspace for the cost of materials, which is amazing.

Really, it's just the need to quickly turn things around kind of drove me towards doing faster prototyping. And then the cost of materials makes it kind of a no-brainer at that point. The cost of materials. I'm glad you brought that up because it used to be when you made one or ten of things, the cost of the materials was as much as the cost of the NRE. But now the cost of material is so cheap and NREs is still expensive. So, yeah. It is kind of, it's kind of tough because it's like, well, like, so we were talking about consulting it's now it's like yeah that thing's gonna cost you know a hundred bucks to make but it's gonna cost you five thousand dollars for all my time to get it to the point of working and i think someone looks at that as a as a customer

and they're like uh really and it's like at the end of the day the thing still costs fifty one hundred dollars right but uh but they say well, it only costs $100 in materials. Well, okay, that's a pile of atoms until someone puts them all together. So what are you going to do about it? It's just, it's tough from a sales standpoint, I think. Well, the good news is the second one also costs $100. That's true. Yeah, right. You're amortizing $5,000 each time you build another one, I guess. What are some skills that both of you are trying to learn? I've been on this robotics and AI kick. I'm still on it.

How's it going? And I am fortunate enough that I did a bunch of work on it on my own because I had some free time, and then suddenly I had a contract that was using every single bit of it from robot operating system to image processing to deep neural networks and it was really strange to fall into this project that I had it was like my hobby became my job but in a good way yeah usually that's not said with the uh with a positive spin on it yeah I mean it wasn't like I was doing ceramics and I tried to make that my job it was always related to my job but getting

to work on large autonomous robots is pretty cool I didn't expect to have that to pay off so soon was that because you talked about it as you were going along is that how you found the gig i think that the person who recommended me is someone i know in real life or knew before have known many years um and maybe he knew i was working on it because of the show, or maybe he just recommended me. And I talked to the guy who was looking for the person, and I guess maybe I answered all the questions right. You know, it's like feeling out. Well, do you think this processor will work?

And I'm like, well, let's talk about this. We want to do this and this and this. And here are the risks. And here's how we'll figure it out. And so I think I just, you know, the connection was made. And because I had been thinking about that and working on it, the connection went forward. So, yeah. I don't know if it was show related.

He definitely didn't know about my typing robot before we had the full conversation. You're like, I'm typing you right now on a robot. You didn't even know. My dirty little secret is I have not been an embedded developer since October. It's only a few months. Yes, I am learning iOS development. Are you going to change the name of the show or what?

My job now is iOS developer. Wow. I'm working on iOS application code and learning Swift and Objective-C and all that good Apply stuff or bad Apply stuff. When are you going to make a game that makes us millions? Never. Does that still happen?

Do app developers still make buckets of money like they did at the beginning? It's pretty damn rare. It's winning the lottery. It's pretty damn rare unless you've got something really special. It's the whole viral thing, right? I have a viral app.

But there's a lot of money being spent on development. So to do something that competes well, you have to have a real company. It's just become a normal software industry. Yeah, except the... It's a gig now, yeah, right. Okay, it's a marketplace, whatever. Except the average price is super low, so that's an additional problem.

But yeah, I've been learning a ton of new codes a new language um new methodologies well sort of new i mean things i knew about but wasn't proficient in at all it's way different it's so different it's almost it's almost the difference between sometimes what i think electrical engineering and software is but just going to the because because there's a lot you can do with very little code, but knowing what the very little code is, is a huge challenge and reading it back is a huge challenge. Like one line of code can be this huge expressive thing that, you know,

goes through your entire array of things and they could be random things and it does something to all of them and it does them on different processors and then it all comes back in an asynchronous something and that can be one line of code whereas in c you you know you'd be spending days coming up with just stepping through infrastructure to make that happen yeah yeah uh but to read that you know that's that's the challenge of learning is is okay here's some sample code or here's some code from the application I'm working on. It's like, what? Okay, I need to spend a half an hour reading this line of code because I don't quite get it all.

Yeah. And for someone who's been doing software for forever, it's kind of demoralizing at times.'s like wait a minute i know i know this this is this is code but no how are the people that are that you're you're interacting with like how so like are you working with people that are coming into swift as like new developers who don't have that background and they're just like oh yeah this is just the way it is so they get it or are they also staring at that one line of code it's a mix so my team is mostly new developers who are at the same stage as i am and then there's an existing team that's that's been doing it for a while swift isn't that old so uh you know it's you need 20 years of

experience yeah i think it's i think it's maybe five years something like that five years old officially or something but it's also changed really fast like they had major revisions every year since it came out and they're starting to slow down on what those major revisions do but for a while it was like okay next language revision uh breaks all your code oh wow okay you have to rewrote you have to you know rewrite submit substantial portions of your code because we changed language so much uh so everyone's kind of learning and everyone has a different kind of opinion of how idiosyncratic to be so yeah you can do some really exciting things with very few lines of code and make it unreadable or you can write readable code and they probably

compiled the same thing but then you're not using yeah right stuff right so yeah so i get coded your comments you're like oh you should rewrite this as this and i'm like okay and i rewrite it it's that and then i commit it and then somebody says oh you should rewrite it back to the thing you had before i'm like hey you two you two get in a room and and work this out because i don't care i'm gonna put a couple baseball bats in there whoever comes out uh winning uh't care. I'm going to put a couple of baseball bats in there. Whoever comes out winning, that's who I'm going to pick. That's been interesting, and I've been happy to have the opportunity to learn something

basically completely new to me in a full-time job. Yeah, it's like getting paid to learn, pretty much. That's great. That's one of the advantages with Megacorp.com. Unfortunately, our team has been noticed, and people are like, oh, we need this feature, and they're putting stuff on us, and now we have some pressure.

It's like, well. Oh, my God, you have to work now? We aren't the experts, but we'll try. Yeah, I've been playing around with CircuitPython. I don't know if you guys have played with that at all. That's the Adafruit version of MicroPython, specific to the SAMD 21 51 chipsets and now nrf 51 52 it's a good system yeah i mean i i understood about micro python and then started looking at the ada stuff and

it's you can get a lot done with that yeah you can uh what chris was just talking about with like looking at lines of code and be like uh i'm like i'm a bit python challenged um and uh i'm i've been struggling uh as especially like because i'm basically even still just modifying someone else's drivers and some of the things i'm looking at you know it's just basically looking at object-oriented stuff that i'm you know i'd feel like i just got a better handle on c you know and not even not even that good and now it's like oh okay there's all this other stuff you know so it's uh there's a lot of hours there object-oriented is super super hard transition because you can look at you can look at a program and not be able to see at all how things flow it's like okay there's all these stuff side by side and

i thought where does it start and once it started where does it go and it's it's very hard to analyze statically a piece of code that's that's really object-oriented whereas with c it's like okay we start here it calls this it calls this i can map out the flow pretty easily start it started zero zero zero zero whatever the memory location is and go from there right and uh it yeah it's it's challenging i don't know so hopefully that'll get better over the years yeah i don't know how much of that will i don't know how much like python like all I don't know how much Python, all the way down to the low-level stuff. There is a lot of stuff in there, in between. So I can even see...

So this was like, I was doing writes to a chip, and I can very easily see how much slower the writes are to the chip. Oh, yeah. I agree it's much easier to do things, but I think there's like, there's going to be a lot of tinkering with now instead of, uh, just doing single rights, I'll need to do like bulk rights and bulk reads. And basically at the end of the day, I'm not sure how much more, uh, optimized it will

be, not even optimized the flexibility versus the optimization. I'm not sure how much that's going to matter. Whether or not you're building a prototype or a maker thing versus a product that has to be optimal and cheap fast and reliable yeah i haven't seen a lot of that i mean there's just this this divide between let's make it and then let's make it robust and i don't feel like we're addressing that well enough because it has gotten so easy to get to the point of let's make it yeah totally and you talked about that with the uh the prototype author right yeah alan cohen uh prototype prototype to production or maybe prototypeotype to Product.

I keep getting it wrong. But that was a good book about how to, the immense number and quantity of, the steps to go from here I made it work once to here I'm shipping a million of them. Oh, yeah. I've definitely, I've told this story, a million of them. Oh yeah. I definitely, I've,

I've told this story I think before in the amp hour before, and maybe even when I've been on your show, but like, uh, at one of my past jobs, we, we had like this three model. It was beautifully like in industrial design and,

and, uh, we got it like cut out of like a block, a solid block ABS in China. And like, just really, really gorgeous looking prototype. It was exactly like the drawings

you held it in your hand you're like oh man it's done and so my boss showed it to like one of the vps he's like this is great we'll we'll take 25 000 he's like uh that's a prototype you we know we can't do that uh and but it looks so real that it like that that you know to the untrained eye they think it's done. They think it came from a mold, and it's like, no, no, no. We're going to need another six months. And that's just to get the money to do the thing that's going to take six months. And software is the same way.

And CircuitPython is that shiny, pretty prototype. Well, there's always been this divide, though, in software. Visual Basic was always a thing that professional developers sneered at and people making products with Excel macros. I've been in that train. The easy path that people wrote stuff in and you'd come into a job and it's like,

well, here's our whole product. It's in QBasic. And it's really slow. We don't know why. So I don't think that's ever going to change because there's always going to be more accessible languages that you take a hit in performance

at the cost of, at the benefit of being able to develop and i think we've talked about on the show before too either a show we've done together or even just talking other times uh you know like the idea of like people coming to the world with an arduino thing it's like yeah that works actually like really great and then it's the you're going to hit a wall at some point it's like how soon until the wall until you hit that wall and when you do it's a then you have to start from scratch. And that's when you start seeing Kickstarter problems and workarounds and weirdness and not great products, right?

Not long-term products. Yeah. I have a question from our embedded Slack channel. Is ESP32 just a fad seems topical to this i don't think so i mean i think that they're coming out with a new product soon espresso is so in terms of is the esp32 as a standalone chip a fad um well i still haven't seen many production level devices with it maybe it's out there more than i know um but the cheapness really you know it's kind of like the pcb thing we were talking about the cheapness really changes the equation a lot and and when you really look at

what's in there you get a dual core with wi-fi and bluetooth and a crap ton of peripherals and it's like for four bucks in a you know a certified fcc certified module it's like okay well when something cheaper comes along or better or you know that actually has a sleep mode comes along then yes that'll that'll usurp it but i think until then it's like i think people are going to keep making stuff with it and i think that the development community that comes up around it uh will keep it keep it going because that's what got a lot of it started in the first place did i answer that question at all yeah i don't know how many products are made with the esp 8266 the older version i felt like it it came on the market and everybody loved it, including me.

But then when I tried to think about how would I put this in a product, I was like, this isn't secure. This isn't as configurable as I need it to be. This isn't the one. And I was hoping the ESP32 would be closer, but I felt like it wasn't yet. And I'm hoping they come out with something that's a little a little bit more robust, to use that word again. Definitely targets more

consumery, low-cost markets in the first place. I'm interested just because I like cheap stuff. Especially when it's over-specced and cheap. But like, I think there will be a place for things, chips like that for a long time. So I don't think that's a fad.

And I think that people will keep chasing low cost components because of how that ends up stacking up costs over time. So I don't think that low cost Wi-Fi connectivity is going anywhere. Not going away. What do going away, not for sure. Yeah. Oh, no, I agree. I mean, even trading off for a less secure, less well-documented, less everything else kind of solution, I don't think that's going away anytime soon. But I don't know, maybe you have other ideas.

No, I think that'll continue until it becomes too painful to do that. You mean security-wise? Yeah. Yeah, security. Come on. Well, there's going to be companies that are going to ride that wave of, this is cheap and I can get something out there. And then there's going to be other companies that are like, we need support and we need documentation and we need certification

and we need to be able to understand this chip before we're going to integrate it. And there's multiple kinds of companies out there. So I think there's definitely a market for integrating the cheap, the super cheap stuff that the expensive, some things you might not want to be expensed. But I mean... Yeah, I wonder if people will, not mistakenly, but like if someone gives it a shot and puts it out there and then just happens to be a breakout success product like that's very possible and that might also

lead to those security risks like you're talking about but at the end of the day that will mean that it's even more proliferation more people might be willing to give that a try as like a production level thing so i think if you really get into the high volume stuff people are willing to say well let's just start from the beginning because we'll get better pricing from a TI or a Broadcom or whoever, you know. But yeah, I'm excited about it. I think it's still fun to play with. So that's enough to keep things going. The thing that I'm surprised about, though, is that like, why isn't there a better way to hook into Wi-Fi networks still?

Provisioning. The process of putting in your Wi-Fi name and password, the provisioning piece, it still sucks. It's always going to suck. It just sucks. I hate it. I've hated it for more than 10 years, and it hasn't really changed. No, it has. If I hold my phone up to my brother's phone and says, would you like to send your Wi-Fi credentials?

And you push the button and it goes. Because he's in your contact list. Yeah. No, that was cute. And there are other ways to do it now if you've got Bluetooth as well. I think that's one of the reasons the ESP32 has both. So that can be easier if you've got your phone with an app and then your device and then it connects over Bluetooth and then they negotiate and then it sends over

the credentials. That could be easier. But there's no... I don't think there's going to be any kind of Wi-Fi IEEE protocol improvements. I'm not sure it's really solvable because anything you try to do to solve it probably

is a security hole. Yeah, right, right, right. The button you push on your router. Yeah, yeah. That didn't last. Yeah, that didn't, yeah. I was actually asking about, so I've been thinking about this

for some product ideas I have, but I really care about it because when you really start thinking, if you get all the BS terms out of the way, the Iots the you know smart homes or whatever right at the end of the day it is useful to have a thing on wi-fi and uh it is useful to like be able to manage a big bulk amount of them and so i started looking around like all right so if i was going to design something in with an esp32 if i was going to take that plunge uh how do i then go and update firmware on all of them?

How do I manage all of them? How do I make sure all the data is coming back, assuming I'm in control of this fleet of gadgets? And I don't know if you saw that Twitter thread, but I was asking about it on Twitter just because I haven't seen all that much about it. And I've got a couple irons in the fire but have you have you seen how to do that yeah so there's there's people people there's companies that do that for a living uh so electric imp that was their whole deal even though they sold devices it was the management of the devices in the back end um particle does the same thing um or but there's both platforms they're both platforms they both want to sell you the hardware

in order to make you pay for the software. Microsoft Azure has a whole IoT management platform. Not hardware, but they're on their servers and things. I think this stuff exists. AWS IoT? Yeah. I think we have another question on here somewhere. What do you think is coming in 2019 and what do you think is overhyped?

And those two, I was thinking IoT management. It is overhyped, poorly implemented. And underhyped. And it's going to be very important, but there is no solution that looks like it's going to win. I mean, if... There shouldn't be a solution that wins. But there should be a pattern. It's not a protocol thing.

Nobody's going to have an ERRORC for IoT management. I don't need it to win the whole market. Yeah. I just need it to win for me. I just need a solution. Like, I need something to buy, honestly. Exactly. I need to pay 50 grand

and three to choose from you know like that's fine but i still actually haven't seen that i found a couple things um but i was curious if you had seen them otherwise because like so you had mentioned particle and electric cam both of those require that you buy their hardware yeah so they're running custom firmware on a so like the p0 p1 from particle which is you know a great solution for some people and they got their new stuff too which looks cool um it's a you know it's a bespoke piece of chip piece of hardware that runs bootloader firmware and has like a wrapper thing and then you write firmware that runs within that wrapper i think that's as far as i understand it and then you pay

a platform fee for each device and however much data you send, and that including for Wi-Fi. And then you have to buy the hardware for them. Cool, fine. That's like a very specific solution. I think a lot of people like that. I was talking like, okay, now I want to design with just an arbitrary piece of hardware, specifically in this case, the ESP32. Is there a way to get an arbitrary piece of hardware, you know, that has maybe some starter code and then has that wrapper for firmware updating, OTA-type stuff.

And then on the website, then it's like, okay, well, now you have 30 devices in the field. Here's all the names you gave them. And now you want to push firmware updates to 20 of them? Okay, click Go. You know, and it tells you which ones work. I think you need to look at Azure. UbiDots was doing some of that.

UbiDots? UbiDots? Microsoft Azure has IoT Hub, is what they call it. I think AWS has something similar. AWS has IoT. I don't know what the costs are. So IoT Hub is... So that's what...

A lot. Yeah, a lot of them are a lot there's a homebrew one that's i think called home automation.io something along those lines okay it does seem like the sort of thing you should be able to spin up on aws instance for your well and there was a spark fund thing that was doing that that they abandoned is it rhino or something i don't remember so the things that i found so far are one is the amazon's version of free artos that was i think you guys talked about and i also talked about on the empower i didn't quite understand it but that actually is one thing that they're talking about so it's actually compiled for esp32 uh and it can handle some

of that stuff like so like esp32 hasn't OTA bootloader, whatever's internally. And then Amazon wrote it so that basically it can handle that. I receive a piece of a new chunk of firmware. I'm running my old firmware. I try to swap them out and do all that stuff. So that's part of what that is. And then I think on their cloud side, they can do it. And then the other one was Andreas Spies, one of the YouTubers.

Him and his business partner started up a new one too. I'm trying to find the name of it, but I'm failing miserably on that side of things. This is what show notes are for. Yeah. So there is another one, and it's not professional. It's like there's no professional pricing plan yet, but they're working on that kind of thing something maybe in the new year this is this is going to be one of those areas that i think next year we will see a lot of uh shuffling but we won't see any resolution for two or three years or we'll see the iot bubble collapse oh please no because then i wouldn't have a job wait no my jobs depend on that

hey hey all right uh yeah it's it's very fuzzy and i think there's money to be made but you are dealing with lots of people who don't really want to pay anything or it's every company that does this that needs it spins up their own thing i don't think that's going to continue to happen. It's just too hard. When you're managing 30, it's fine, but when you're managing 300, you need somebody to have thought of all the problems.

When you're managing a million, you don't want Microsoft doing it. I don't know if Google has a manager. I mean, we said Amazon and we said Microsoft, so I feel like there must be a Google one. The great balance point of the Fang world. Google Cloud IoT, fully managed

IoT services. Try it for free. Yeah, everybody's got this. Yeah. Use cases. I think the tough thing is that, in either case, you need to have some kind of hooks in the firmware, right? You need to be able to have something that is actually modifying the firmware, and then... Or, like, self of hooks in the firmware, right? You need to be able to have something that is actually modifying the firmware

and then, or like self-modifying. For firmware update. Yeah, exactly. And then there needs to be a cloud component piece. So like, I don't feel like I'm capable of writing the cloud component piece, even if I was able to write the OTA piece. It's like a lot of, it's a lot of things in that stack

that end up being a hassle. Well, and you're focused on updating the firmware, but most of these also need status and verifying things are awake occasionally while still sleeping or taking small amounts of data. 90% of it is I have an app for my thing, like your weather station in your house, and I want to see what the temperature is at my house. Yeah, there's some piece of data.

It goes to the server and says, well, the IoT thing checked in 20 minutes ago, and here's the temperature. But that's a very webby thing, it feels to me. Well, it's all webby at some point. It's all HTML endpoints, right? Yeah, exactly. But I mean that

an Azure team could spin that up without knowing anything about the hardware. They're making a door, and then a device gets itself onto a network, goes and knocks on the door, and it says, hey, I have a key, and I can get through that door and dump my data in there. That's what they really want to do

because it's super low. It's very hands-off. I think the other problem is that when now you're going the other direction, you're saying, well, now I have a door, and I'm going through the door, and I'm now going down to your device, and I'm dropping a package off. It's like the Amazon ring, or what was that called? The Amazon thing where they're going to drop packages off in your house?

It's more like that. And nobody liked that idea. Well, I mean, when we go back to security, you mentioned the NRF-52. They have a pretty good, pretty secure method of firmware update, except for the fact it's Bluetooth. But their methodology is very reasonable. And yet, the last time I did one of those, every device was keyed differently

because that's the smart security way to do it. And how am I going to update all of them with different keys? No, well, yeah. We're going to get into details about how to implement this, but, I mean, you do it with certs, and you have a public-private key relationship for the site that has the firmware update and that delivers it,

and the device is the thing that should be authenticating it at that point, in that direction. You want to authenticate to the cloud when you have a device attaching, but you don't need to have a separate key to get firmware updates. But if you need to change every key, is that

what you were saying? Every device had its own key. Yeah, then you just burn them. You come out with version 2. It's a very complicated problem and it requires quite a bit of intelligence on the device and so right right to make the devices more manageable they are more expensive and then that's why you write in python yeah well yeah i mean on one of the uh one of the projects i was working on we were doing a python thing and it it was way easier to do that kind of stuff um so i can at least i wasn't actually doing the python writing but i was working with the person doing the python writing way easier but i think it's also like understanding the stacks of how that stuff works like i i don't know i'm starting to learn more about the the key

system and certificate certificates like you're talking about and i think until you learn that that's even a thing would be it's difficult to to know where to look next right and it's super easy to make a mistake with that stuff is the other problem that that's one of those areas where if you don't understand it completely you can just subtly mess it up and have a big problem. And if you understood it completely last year, it's already changed. Yeah, that's true. You're a security engineer and you're not doing what you're doing. So aside from that,

do either of you have any wild speculations for 2019? New tech, new methods? I don't see anything super exciting with new tech besides what we already know i boy yeah i'm kind of a downer on next year i was a downer on this year too so then you know i i think somebody will do a cross-country self-driving car trip that'll be cool um i think i mean i guess the high level stuff i i look i look at like those those high level things like self-driving cars too and it's so far from anything i'm doing that i don't have any visibility it's not super far from what what elise is doing so okay giant

autonomous robots sorry um but yeah you know, just in terms of processors and stuff, I mean, everything's going to proceed as it has been. There'll be some new cores that do something slightly better. Christopher's barely managing not to take his finger in a circle and do the big whoop sign.

I don't know. I mean, a lot of what Tech has been... My favorite version of that is is when in little rascals when when the kid that plays buckwheat goes whoopee and then his mom's played by whoopee goldberg classic just classic yeah i don't catch that sorry i i feel like uh everything that's exciting now is just a culmination of the last 20 years the internet was really cool in 1995 and now it's on everything so what does that mean well it's on everything um i mean the protocols are the same i'm just jaded yeah i think oh i i think uh i think there's

going to be some absolutely stunning uh security holes found in every process of architecture. Like this year with the meltdown stuff, and Spectre was just the beginning, and it's going to get much worse. How's that for possible? Thanks. Chaos Communication Camp is going on in four days. Sorry, Congress. Chaos Communication Congress is going on in four days.

And I know one of our past guests, Dmitry Nedespazov and uh josh uh datco who was just on our show uh they're basically talking about holes in crypto wallets and uh yeah they're gonna they're not working much anymore yeah that's true that helps a lot i guess but uh the wallet.fail is the uh is the site that they they've basically done marketing for a talk that they're going to do so wallet.fail is the site that they've basically done marketing for a talk that they're going to do. So wallet.fail is the site to check out if you're interested. It's going to be very fun.

All right. How about you two? Any wild speculations? Positive, maybe. I have some. Alicia, you go first. Okay. Yeah, please go first.

So Christopher showed me this little robot that was a desk robot. Oh, the Anki. Anki? Oh, yeah, I've seen robot. Oh, the Anki. Anki? Oh, yeah, I've seen that. With the little front loader shovel thingy. And it's called Vector. Vector, yeah. It was called Vector.

And it was something Echo or Google Voice-like. Yeah, it had the word Echo. And so it could go to the internet and it could respond to voice commands and it had this adorable emotive affect about it yeah and so my prediction for 2019 is that we are going to see an explosion of adorable robots that act as assistants. And they're not going to be very functional, but they are going to be darned cute. Where did I just see this?

I just read an article that somebody was lamenting that the, what was it called? A Hitchhiker's Guide where every single thing had a personality. Oh, yeah. Like the doors, the personality doors. Somebody was lamenting this is what's going to happen. Everything's going to happen. Everything's going to have a personality.

And I think that's true. I mean, I think your water pitcher is going to be cute and have a personality and have, when it's empty, it's going to be sad. And when it needs its filter changed, it's going to look, I don't know, constipated. I just, it's... Genuine people personalities. That's what it was called. That's right. G don't know, constipated. I just... Genuine people personalities. That's what it was called. That's right. GPP. That's right. I do. I think that's coming and I think that the

internet of things and the prevalence of enough processing power to make basic neural networks work on this sort of thing is going to just kill us with kawaii style cuteness that is my prediction for 2019 is everything is going to blink at you with giant eyelashes

i think that facebook is going to be broken up oh i hope so well yeah that's my eyes you said for wild predictions there's a wild prediction that's a wild prediction. That's a wild prediction. Chris, what about you? Yeah, I don't have much. I'm still, like I talked about on the Empower, I'm still pretty excited about Laura Wan. I know that that's like very in the dirt kind of like protocol level versus higher level like robot emotiveness, but emotivity.

I just wanted you to explain LoRaWAN. I mean, I think we've talked about it, but never for more than a minute. So could you give the two-minute version? Oh, sure, sure. Okay, so I got this real confused at the beginning. LoRa is long-range or low-power radio or something like that.

It doesn't matter. It's basically how the chip talks to other chips. That's the easy way to think about it. LoRaWAN is then a protocol layer that's on top of that that basically uh it works kind of like the internet so now uh a device is in the middle of a field and it's saying my temperature is four because i live in chicago uh and for what's for c sorry for c uh or f yeah that's that's coming too uh it's tough guys i might come visit i have to say uh so the device is sitting in the middle of the field it's it's saying my temperature's four and then there's a bunch of listening devices that are called gateways and basically uh any it receives all of all of them receive the packet you know

maybe there's three in the area and they all receive that the thing is four and then also there's like stuff in the header that says where that package should end up going and basically it's it's certified to be on the gateway and then the gateway says oh i know where to send this and it sends it back to like a application server that then the application server receives from three different gateway servers hey you know node 23 out in the field just said it's temperatures forward gets that three times it says okay i get that it's you know that i definitely got this packet and now i'm going to hand that over to the aws iot or whatever whatever your graphing charting interfacing system is so the idea is there's like software layers on top of the

protocol layer um both of which are interesting but it's stupid that they give them the same name so yeah very very low power and very very low long distance and long yeah yeah yeah so it's like 600 bytes per second or bits per second rather in the lowest power mode. And you might do it once every hour or day. But you can have a thing that runs for 10 years on a battery. So that's kind of cool. So I'm

personally very excited about it. I think that's super cool. I'm going to be doing more LoRa stuff for... I've already done a couple of professional projects. I'm doing a personal project right now. I'm going to be doing more Laura stuff, uh, for profession. I've already done a couple of professional projects. I'm doing a personal project right now and, I'm very excited about it, but I don't know what that means for other things, but I think that that's a,

that's a winner in my, in my mind. Um, there's definitely a lot of, just like anything else, a lot of security questions there too. Do you have to pay for your bites to go over? Uh,

if you use the things network, it's an open source. That's that conference I went to last year in February in the Amsterdam that's coming up again. And that is an open source implementation. It's basically like a shared

network of network nodes where everybody processes packets that are allowed to be on the Things Network. And you do not have to pay for that. Comcast right now is spinning out uh basically every cable box in the nation will have laura wan receiving capabilities um and so you'll have to pay to be on that but at the same time you'll have uh imagine now every cable box in every

suburban household that comcast is in which is a lot uh basically we'll have these chipsets in there and then it'll hand it right back to the network yes and no what band is that on in the states uh it's 433 and 915 both ism band um okay in europe it's 868 and in australia it's 902 so well it's nice you can use the same radio for all of them not Not. You can't. You cannot. Oh, okay, you got that. But a lot of the chipsets switch over. Basically, it's just a configuration bit. So there's a SX1276 is one of the Semtech chipsets,

and that's 76? I don't know, whatever. One of the Semtech chips, basically, you just set a register and you go. And it basically swaps between 868 and 915 really easy. So that's nice. And the antenna's kind of close. So I'm excited about that personally, but I don't know, you know, like, who knows how it's going to work out. I just think that a lot of the times people are going to think that they're going to do a lot more than they actually can do with that.

Right, right. Really low battery power. And it is, like, bidirectional, too, so you could push firmware back down and do a whole bunch of other stuff. I don't think you could push firmware down. It's too slow. You would have to do it in chunks. Over a year. Yeah, right. There was an ARM embed. They just they just verified that was on our

subreddit um they just verified that they they could push firmware back down and it does it in packets and there's a crc and all that other stuff it would take a long time though so i'm excited about it i talk about it a lot i guess sorry no it's cool. We don't talk about it that much, but it is something I've been keeping my eye on. So I'm happy to have you bring it up. I can't get past like, oh, what applications is that? Well, I could have a billion sensors

thrown through a wilderness reporting on... Yeah, that's actually what I think about. Yeah, you basically have a like... Imagine like... So the personal project I'm thinking about working on is like basically having an enclosed, waterproof, battery-powered thing. And then you imagine it is enclosed in a rubber ball, effectively. And this is not going to sound good.

It's going to sound actually very bad for the environment. But imagine you have a handful of these and you just chuck them. And then they all talk back to a node. Now, granted, they would stay there and you you have to go pick them up or whatever but like um you know the idea is that like like chuckable chuckable sensors that's kind of the idea um and i think that that that could help for environmental stuff and hopefully you pick them up when you're done stick them on whales stick them on whales. Stick them on whales. Yeah. Always a fan of sensors. You can just throw it away. Yeah.

You have to still find them, I guess, and get close enough. But then you're good. What was your favorite gadget of the year? Did you have favorite gadgets? My favorite gadget was something I got off of Chris's wish list for Christmas last year. And it has nearly no electronics and the worst beeping sound I have ever heard. The Annoy-O-Tron? It's an egg cooker.

It's a hard-boiled egg cooker. It's my favorite gadget, yeah. It's so embarrassing, but that's the one. But it does sound like a fire alarm when it's done. Yeah. Are you going to choose a synth or something actually reasonable, Christopher? I mean, I think I have to. Yeah, given the...

Although I'm not sure which one. Continuing. I only got two synths this year. Two. Only two. One isn't even really a synth. It's a sampler. That's not bad.

I mean, that's not bad. Yeah. And one of them I just got a few days ago, so I barely made it. No, I'm pretty sure that more electronics. Oh, electronics have arrived, not synthesizers. This is one of those things where you tell me that something's a drum machine versus a synthesizer versus a MIDI player versus a. No, I only got the Octatrack and the DeepMind.

They're all the same to her, Chris. Come on, man. They're just blips and bleeps and blurbs. Exactly. Thank you, Chris. That's fine. Get yourself a friggin' acoustic piano for the love of God, you know? Then you can't use headphones.

I've had one. They take up a lot of space. They do. They require weird people to come into your house and fix them periodically. Not weird people. It's one of my life goals. Fascinating people. To have a grand piano at some point. That's a life goal of mine.

We'll see. What was your favorite gadget, Chris? Personal one? Just got a Roomba. So that was fun. It's interesting interacting with that thing. Obviously not new. New to me, not new to the world.

What did you name it? Jim. Oh, Jim? Yeah. I find ours clean. Actually, I didn't name it. My girlfriend did. And she was really upset that all of the voice assistants

were all named females and that you're bossing females around. And I agreed with that. And she's like, you know what we're gonna boss jim around and i was like all right and also when he gets stuck on something we're gonna say he's dead jim which her jim's dead no it doesn't yeah whatever i did find that our roombo works quite a bit better and now that has googly's eyes on it so just a a tip there. I will definitely be doing that. I might take a pause here and order some on Amazon. Okay, let's see. What else do we have on our list?

Tons. Yeah, we have a lot because we asked the embedded Slack people and that's our Patreon reward. You get to be on the slack channel where sometimes we show up and talk and it's really useful and sometimes we don't really show up and people talk and it's really useful and it's just let it run itself yeah yeah it's nice it is i'm there i like it too many slacks though how do we how do we improve telepresence and i don't know slacks are just hey i'm adding it i'm'm adding it they're too much sometimes

everyone's getting rid of hip chat the hip chat's being discontinued so a lot of companies are going to slack oh boy there'll be more slacks i don't like it yeah i think telepresence can only be improved by uh those robots on a stick that you drive around no those are those are creepy those are so uh i just put that that in there for Chris, sorry. You know, there was a line in Snow Crash. You ever read Snow Crash? Yeah. Uh-huh, yeah.

And the one engineer she's talking about, like, her big thing was she basically added eyes to the avatars that are in the system. And, like, that was, like like the thing that made the whole system and i think that's it i think that like having talking about emotive robots how about emotive telepresence whatever's right but being able to uh there was someone talking about it recently on a i think on a podcast they're talking about like being able to see when when someone's eyes are dilating or when you're like mirroring movements you know like all of the things that are like

very very human you cannot do when you're telepresence you know it's just like you're just a floating head and if you don't have that like uh detail of like human connection then you're never gonna figure it out otherwise so we need to make emotive robots that actually mirror my face that's what i'm trying to say it's a vast uncanny valley to cross there i think yeah i think you're right mirror your face that would be weird it's true uh if you if you are a introvert or or on the spectrum or or just face blind and you are one of the people who have trouble connecting with someone mirroring like their head tilt and how they're doing their hands and their shoulder tilt and and occasionally making eye contact these make other people a lot more comfortable even though if you're like me it makes you much less comfortable uh well it's

comfortable until you notice it and then you're like oh why you do this but most people actually don't notice your strategies for meeting me but it's very effective um i mean, if you're really socially inept, practicing on the baristas as you get coffee is quite good practice. So step one, make a robot. Step two, take it to Starbucks. Step three, we'll find out. How many coffees can a robot drink? Oh, yeah. Let's see.

We should go into some of these sillier things. Speaking of silly things, I gave up caffeine. That's something I'm trying. It's very weird. I still drink a lot of coffee, but it's all decaf. Well, then you haven't given up caffeine. There's still caffeine in decaf. Not a lot.

Don't take away his decaf. My metabolism says there's still caffeine in decaf. Not a lot. Don't take away his decaf. My metabolism says there's still caffeine in decaf. Oh, that's true. Yeah. Yeah, if you're really, really sensitive to it, then yeah. I can get too much caffeine by sitting next to someone who has an espresso. Oh, got it. How's that go?

Are you finding that you're sluggish? No, no, no. I mean, after the initial thing, I'm definitely less anxious. Like, I haven't had heart palpitations in like weeks uh so yeah it's good it's good you know it's one of the things where i just wanted to break free of it and i think i am minus the i still like coffee well yes because coffee is excellent you've made a lot of changes in the last year you were already in chicago a year ago but you were just still, you were only barely settled. Last year, I had been there for a year.

So I moved there like August of 2016. So it's been two and a bit now, two and a third. But yeah, lots of changes. It's good. You know, got to keep moving. No, no. You can just sit still. It's nice. You should try it.

Chris has already been to our house. He knows how much we just lay around. You guys walk to the beach and listen to the whale sounds and it's a good life out there. You guys got it great. It's really good. Let's see. What are your favorite books for 2018?

Or movies or TV shows? Oh, man. I think we... Yeah, there's a lot of TV. I probably won't touch TV shows, but... Well, there's The Good Place and Mrs. Maisel. Those are amazing. I love Good Place.

Yeah, yeah, fine. I don't know. Movies have been kind of the ads here. The Rise and Fall of Dinosaurs was the best book about climate change I have ever read. Really? Okay. Putting that on my list now. It's paleontology and paleobiology. And the dinosaurs lived in a less stable climate.

And if you wonder why there are big sites where many, many dinosaurs died, it has to do with their unstable climate. It makes you want to go, oh, we really shouldn't screw that up. Good luck with that. Do you have a favorite book? Are you just going to say Lord of the Rings Because that's the book you read every year Or this whole set of books you read every year No I don't read it every year

It's like every five Nothing stands out that I read that was really great this year It's kind of been not a great year for books I'm trying to think I can't believe you're still reading the book you're reading You're like pick it up every day I'm like 90% done and it's boring But I have to finish it now.

Got to power through it. God, yeah, no, I feel bad because I don't really have any, I need to find some better books. We have many of them. There are many good books. I don't know why you keep reading the boring ones. Because they're the ones that, what? Same books you're reading, some of them. Yes, but I only gave it three stars.

I don't track your ratings. Three stars is pretty good for your ratings. I'm all nice on Twitter, but I'm really mean on Goodreads. Oh. Okay, Chris, what about you? He'll have something. He'll have something.

Energy and Civilization. I think I've talked about it a couple times because it's going real slow, but it's like every time I keep reading before bed, so it's like slapping me in the face. Yeah. But it's really, so it's basically about how energy has changed and energy consumption and food and how farming has changed everything. It's basically using energy as an excuse to talk about civilization

and how it's changed. So it's great. Backlaves smell. It's very using energy as an excuse to talk about civilization and how it's changed. So it's great. Backlaves smell. It's very, very good. I think I found it on a Reddit list or something. But yeah, that's a good one. All my other ones are like marketing books and lame. So don't listen to those.

I might have read Sourdough this year. So that would be my favorite book of the year. Oh, that goes into... I thought I read it last year, but it might be this year. Into which episodes did you like this year? I know one of Chris's favorites was when we interviewed the author of Sourdough, Robin Sloan. Oh, yeah. He's a best-selling author and really had nothing to do with Embedded, and it was so fun.

Yeah. We had a lot that weren't really embedded related this year sunshine jones i've noticed that i've noticed that you you guys really don't uh don't don't stick to the embedded all the time but that's okay i think that people are there for the the variety of conversation right the recruiter is here too i hope so i mean otherwise they're all gone we've had a lot of good shows this year, and they're all connected to the experience of being in tech in some way. That's all it needs to be.

The theme of people who are on our show is people I want to talk to. Yeah. We should rename the show People Alicia Wants to Talk To. I've thought about that. .fm. That would work, yeah. I always point people at our... I mean, me and Dave talk about a lot of the same stuff over time

where we talk about the news and things like that, and that's still going strong, actually. But I always point people at the guest episodes to start with. I think that's the best place to start. I really enjoyed doing, I did 10 episodes in four days at uh camp this year and so yes and they were fun and weird and good i i really enjoyed that i was uh i was editing podcast in a tent by lantern light at one in the morning most nights so that was a little weird i was like whispering into a microphone like welcome to the amp hour this is chris campbell would you like to be scared yeah right right so that was a lot of fun i mean i don't know i just i this is the reason we

still do this you know however many years later is because it's it's very rewarding for us i think that hopefully people recognize that we're doing it for the audience but we wouldn't keep doing it unless it was fun for us too and uh so uh i think I just saw the Jerry Ellsworth episode was posted onto the Reddit a couple days ago. And Jerry's been on our show four or five times now. And this last one was talking about how working with naming consultants for the demise of Castar. Yes, I saw that. It sounded like so much wasted money. It was so painful to listen to. Yeah, funny i mean jerry's great well yeah it was awesome uh yeah there's

there's right exactly so uh yeah i don't know there's a lot out there it's hard because we do so many and you know both both of us we do so many and i i can't even remember you know when a particular interview was and sometimes people will say something like oh you remember what you said on episode 197 i'm like no not at all i don't remember but i'm it was on that show let me go search very flattered that you do remember that that's usually what i say is because yeah people saying that it's, very kind of them. It is very kind.

I actually did have that last week with Jeff and Dave. And like we were talking about signal integrity people and how there was like three big names. And I was like, oh, yeah, but we never got Dr. Howard Johnson on our show. And then I was like, oh, no, we did. We did. Sorry. I could not remember at all. Yeah. So when people ask favorite episodes, it's like, oh, I don't even remember what we've done.

And I don't listen to them except editing. So I don't have the same experience that other people do. I mean, I remember sort of ones. Oh, I really enjoyed talking to that person, but I don't know specifically why or what we said. So, yeah, it's like you know it's like asking musicians what their favorite stuff is like well i made that record i don't listen to it it is interesting too and it's like a song like i think i heard that from billy joel he gave his top five songs and one of them was like uh a song that no most people do not list listen as uh right top five and that's that's nice because it's like you know that it's something special about it but then you listen to it like yeah i'll go back to the hits thanks yeah there are these things that we've done this year that have sort of sunk like one of the good

things about podcasts about our podcasts is even when i'm like, okay, that show sucked. Somebody almost always tells me they liked it. And even if I love to show, usually people will say they liked it. But then there are other things, Chris, Christopher and I have done this year that we worked really hard on them. And they were really cool. We're going to go into this,

but nobody noticed. Okay. Well, no, this is the time. This is the time to do a callback and be oh it's exactly uh i don't know i know i mean if dave was here he'd probably talk about like the his uh whiteboard videos never got like they're like technical and take more time and i personally think they're better but like you know the mass audience doesn't care about them as much i don't we don't really do all that much special stuff here so what are you talking about so for me she's talking about my cover music project and my other music projects uh nobody listens to music or they don't like it either way yeah i the cover thing was and still is a thing to keep me engaged in making things and yeah i put it out there as kind of a uh milestone or deadline or keep me honest kind

of thing so i don't really care if nobody listens to it but then it's hard to not care if nobody listens to it right so if i don't get feedback i mean don't even get negative feedback like somebody if somebody came and said you know that was great but you could have done this i listened nobody listens to it. So I don't get feedback. I mean, don't even get negative feedback. Like somebody, if somebody came and said, you know, that was great, but you could have done this. That's all you really want to hear.

Yeah. Yeah. Yeah. Yeah. I listened or, you know, I didn't really like that even is fine, but so it's,

it's, it's hard to make things and not get any traction. And, uh, but that music has been like that for a while. And even though I know some of the stuff that my brother and I and our bands have produced is good because we played live and people come up to us and throw money at us and do strange things and, you know internet just doesn't go anywhere, and I don't know that there's a way to make it go anywhere. Yeah, that was my thing. And you had your zines?

The llamas. Yeah. Yeah, I worked really hard to make the llama drawings about different IoT technologies. different um iot technologies and i did it for supply frame for their magazine but it turned out that we didn't know how to print it properly and uh so it didn't get into their paper magazine it got on to electronic but even then i was like okay i'll put it on my website and it'll be really cool because i worked really hard on it and I think it looks neat and nobody cared. Nobody cared at all. Which is fine.

Not everybody needs to love everything I do. And I do it more for myself. I have to remind myself that. But that and some blog posts and it's just it's hard. The closest thing I have that you're talking

about is like i'll post a show and like you know one or two people will retweet it you know it's like i think a really good guest has been on the show and people would be interested in whatever and there's just you know a couple of people that are interested in it and then you know i'll post some offhanded tweet about silliness and it just resonates with more people it's like a mass market thing versus a specific thing and like i don't know it's it just feeds back into my nothing's harder than getting people to give a crap about something on the internet i think the easy way to do it is to you know uh copy what other people are doing just slightly vary it and be snarky with things and if you're gonna make something new it's gonna you know be tough it's but that's always been the case well

and you just gotta keep doing anything else and you have to thing i've come to realize you have to decide what you're doing it for because if you're doing it for adoration that's probably not a good approach but it is tough when you put a lot of work into it. But it's hard not to have positive feedback or any feedback. But you put a lot of work into a podcast or a blog post or even a project you're working on, and then you tweet about it three times and you get a couple of retweets and maybe somebody likes it, if you're lucky.

And then you post a picture of your dog and you get like 20 retweets and 20 likes. And you're like, let's talk about what I wanted here. I think this is just the modern conundrum. And the more controversial it is, too, the more it'll take things on, you know? Oh. Oh, yeah. Oh, see, okay, I've it'll take things on, you know? Oh. Oh, yeah. Oh, let's see. Okay, I've been picking the wrong cover songs.

Yeah, exactly. Now I've got some ideas. And I just saw Chris posted an oatmeal comic about this, huh? Yeah. Okay, we'll post that. That's a good way to end, I think, because I have to go. It's holidays here, and my nephews are about to come knock on the door, I think. Well, it's not holidays here.

Oh, he's in a different time zone. He's in a different time zone. I live in the future. You're in a different holiday zone. That's right. We're closer to cocktail hour around here, so that is a different holiday zone for sure. It's been really good to talk to you. I look forward to these.

Even though we talk a lot online, sometimes it's just nice to say hello sort of in person. Yeah. Well, hopefully we'll get to see each other in person in the coming year. Like I said, I might hop on a plane and just come escape to not Midwest weather in the winter here. It's not super great here right now. Don't tell me these things. I'll find out when I get there. It was sunny today and warm-ish.

It was not warm. I'm in Buffalo right now, so it's much worse. I see. Yeah. Yeah. All right. I don't know how you close your show. Happy New Year.

Yeah, and all that stuff. And now, a Winnie the Pooh quote, I'm sure. No, actually, I don't have any quote today. It was happy new year and go build good things. And thank you to everyone for listening. Yeah. Thanks, guys. Talk to you soon. Bye.

Bye. Embedded is an independently produced radio show that focuses on the many aspects of engineering. It is a production of Logical Elegance, an embedded software consulting company in California. If there are advertisements in the show, we did not put them there and do not receive money from them. At this time, our sponsors are Logical Elegance and listeners like you.