Embedded - 177: Boba Fett Fell Down

Episode Date: November 22, 2016

Chris and Elecia answer listener emails on-air. Patreon Embedded.fm blog SparkFun Tinker Kit BB8 Sphero Jewelbots (from #173) Superforecasting: The Art and Science of Prediction by Philip E. Tetlock and Dan Gardner  The Wild Robot by Peter Brown

Welcome to Embedded. I'm Alicia White here with Christopher White and it is just us this week. We're going to be answering listener questions and making general show announcements. Christopher, you go first. We're going to make announcements about other shows? We could. You said general show announcements. I don't know what the general show is. Moving right along, let's go straight to our normal announcements, which is about a contest. The SparkFun contest is still going on.

What would you use a tinker kit for? I forgot last week to put an end date so the end date is december 9th give me your response by then and we will choose the answers from random number and it'll all work out so we're not choosing based on the cleverness of their answers no the cleverness of your answer is only because I find it hilarious. All right. And to differentiate from people just emailing with random things. Which is better. So far, the suggestions have been good. Yes.

I enjoy reading them. Entertain us, listeners. That is your job. Well, actually, we did have something happen this week that changed a little bit so that you can support the show without entertaining us. Although we sort of like the entertaining a little better, maybe. Whichever. Whichever. What changed? What happened? What are we doing? Why are we doing it?

Okay. So, Tony from SparkFun, I should have sent her a mic all of our mics were out we normally send them out and then people send them back and they were just all out and I kind of mumbled to Chris Gamble about it and he said well we just send mics every time we don't even ask or they really just always ship the mics. And they're 40 or 50 bucks.

It's not a huge amount, but it was one of those, I've been shipping stickers and we do a lot for the show. And sometimes I just get overworked and the show just has to be put on hold for a little while. And I was like, no, not one more thing. Yeah, and 50 bucks an episode adds up. I mean, we pay for the hosting for the blog and for Libsyn, which hosts our audio. And it's not that it's hugely expensive. It's just one more thing. And so he said, why don't you have a Patreon?

I said, because we never wanted to do a Patreon. Or even Patreon, or however you pronounce it. We don't know how to pronounce it. Listeners, tell us how to pronounce that. But while you're doing it, go to patreon.com slash embedded and consider donating. You don't have to. You really don't.

The show is not going to change. You don't get to. You really don't. The show is not going to change. You don't get a lot of benefits from doing that. I might post contest announcements and end dates since I don't like to put it in the show notes. But, yes. We'll see how it goes. We might do something else.

We had some other ideas. Bubbling in the back of our minds that we aren't sure about. But, yeah, there's no guarantee of rewards and it's just a way for and people have asked about it in the past and we've always said nah forget it so now those people have an opportunity to support the show in that way and you can certainly support the show just by listening and telling your friends about it that is perfectly fine and we are not pressuring anybody to give us money. It's just another option.

And there is a benefit if you do donate right now. There is? Yes. What is it? Pretty much if you donate anything, you jump to the front of the sticker queue. Okay. You have to send in your address, though, to be part of the sticker queue.

You don't automatically get stickers, but if you do... Well, and a side benefit for everybody is it helps us get better audio quality. A lot of our guests are, you know, all across the world, so we pretty rarely get somebody actually in the studio these days. And so having better mics and being able to pay for that helps us get better guests and better sounding guests and helps me not spend three hours with the EQ and the iZotope audio repair plugins. Yes. Trying to make it sound good. Yes, exactly.

So that's something new. And it is on our website, embedded.fm, under the support ticket page, click, link, whatever. And that's enough about that? But there's more about embedded.fm. There's not enough about that. No, there's enough about that. Oh, okay. So embedded.fm slash blog

one of you wrote to me last week and said i had no idea you had so much stuff on your blog it's really a lot and it's pretty cool and i i wanted to write back to you and say yes that's what i've been telling you isn't it neat but then i think i just you know said thank you went on uh so andre and chris and i chris veck have been putting up posts most of us do it every week some sometimes every other sometimes twice a week occasionally twice a week and uh and yeah so there's a lot of stuff there we started last spring and right now in november it's NaNoWriMo, which is National Novel Writing Month. And the idea is you write 2,000 words a day, and at the end you have a 60,000 word novel. As if 2,000 words a day is a small number.

It's five pages. It's not a hideous amount. But if you apply that to months instead of days, if you write 2,000 words a week for a year, you get a lot of words. And Swek and Andre and I edit each other and we work to make each other's writing better. And we talk about what we're going to write about. So yeah, there is a lot on the blog and I know those of you who listen to podcasts don't necessarily read blogs, but if you want to know more about the MSP430, if you want to know more about microprocessor architecture, you want to know more about the discovery software architecture, you just want to know what my goals are in life because i've been slacking lately and that

was my post last week yeah it's pretty cool check it out so embedded fm slash blog all right uh all right i guess as long as we want transitions uh betty speddy emailed last week and wanted to know what podcast we listen to and so we all wrote a gigantic spreadsheet that we posted in the blog

well it was really nice to have Andre because apparently he listens to everything yeah yeah it's impressive how much time some people have to listen

to podcasts my whole like three quarters of an episode every week doesn't mean I get to get through a lot i think producing a podcast you know cancels out listening to maybe five well you listen to a lot yeah not as many as i used to i saw accidental tech podcast ended up under chris vex list and

not yours i'm surprised because i know you listened to that one. Yeah. It's very Apple-y. It wasn't really what the listener was asking for, so I cut that one out. Plus, Speck already mentioned it. Yes. Speddy wanted more technical ones,

and he listens to The Amp Hour and Spark Gap, so that wasn't a surprise. But also 99 invisible engineering commons titanium physicist and he liked all those and he just wanted more in the same vein i don't know if we actually helped him probably not most of most of my suggestions were not technical at all or in the same vein but we we took it as an opportunity to share about podcasts to other people, not necessarily the questioner. So thank you for asking the question,

and we used it to answer the question we wanted to. That's what makes us journalists. Or politicians. Media? I don't know. So yeah, if you are looking for new podcasts there are little descriptions of everything we liked hey and feel free to add your own in the comments yes and i i will say that there was an art podcast i really wanted to listen to and then i couldn't because someone has totally spoiled me with such great quality sound that I can't listen to people who apparently record in their shower. I blame you, Christopher.

I can make it sound worse. Let's go back to the first few episodes. Yeah, I seem to recall there was one recorded in a restaurant. That one sounds terrible. Really bad. Okay, other questions. Peter wanted to know about working in the embedded industry without a degree. Without a degree of what? Well, he started out on

Arduino and he fell in love first time it blinked. I mean, that's such a familiar feeling. I understand that. It's a bait and switch, I'm telling you. And he wanted to know what was really going on inside and that led him to more fundamentals and just learning about systems. And he wanted to know if there was a future for the autodidact, which is, I believe, the word for self-taught. That's a word for self-taught. Right.

And he wanted to know what an employer was looking for when hiring. And is well-documented projects enough? How will it compare to a college diploma? I mean, I don't have a lot of experience with this because I didn't ever get a lot of applicants without college degrees. So I can't really say what I would have done if confronted with one. And I have a college degree, so I didn't experience that from the other side. Did you ever interview somebody who didn't have a degree? I've interviewed people without degrees.

And I, of course, have a college degree, so I place a high value on it, whether it's correct or not. We've had lots of guests who don't have college degrees. And I've worked with people who don't. And I don't have an innate problem with it. But I will say one thing. I had a friend without a college degree who was very good at his job. And he was always the first one laid off. He always had a little bit harder time finding a job than other folk.

And it was hard for me to see that because it seemed like a pretty strong bias. And I think he was competent in roles above his long before he was given the title. So he went from tech to super tech to junior EE to EE. And he was competent as a EE the first time I met him. And yet it was 10 years over that span. So I'm not saying it's impossible, just that it's harder. Or it looks harder to me, given that I value a degree. And whether it's right or not, that's a realistic thing to realize,

is that you are at a disadvantage. Yes. Because people who are hiring have college degrees usually, and they value them for whatever reason. Sometimes it's a network effect where you have a college degree from some place and you know people from there, and then that snowballs to getting better and better jobs.

And that's a separate issue from being kind of discriminated against, or a different way of looking at it, I guess. The assumption about a college degree that people have, and I think that I have, is that college isn't necessarily about learning specific knowledge, although that's part of it. It's about learning how to deal with stress, learning how to accomplish tasks in a finite amount of time, learning how to prioritize, quickly and how to filter it to what you need to acquire to solve a particular problem. It helps to learn when you're out of your depth. I remember working on some problems that there was no way I was ever going to solve. And I started to learn the feeling of that, where I had to learn how to ask for help. Yeah. where I had to learn how to ask for help. And that was useful, has been useful in my career to learn, oh, right, this is what it feels like when I'm out of ideas.

I need to go find an outside thing. And I think a college degree helps you learn a little bit about drudgery. Sometimes it's just about finishing what you were asked and moving on. Yeah. And without that, if you have a great project, if you have lots of novel things, if you have a good background, if you're willing to prove yourself over and over again, then yeah, you can do it. You can totally do it. I know people who do it and it works out. There's several high profile people. Oh yeah.

Who have done it and it requires you, I think, to make up for that discriminatory part with something else exemplary. Like, look at this thing I did. It's gotten publicity, you know, this project, maybe you made a product on your own and got that sold. I think it requires, in order to move yourself up to the level playing field, like we were talking about, where you're not kind of, yeah, you can get a job, but maybe you're behind everybody else for 10 years, not because you're not doing as well, but because there's some discriminatory aspect there.

I think you have to push yourself ahead and make people not be able to ignore you by saying, this is this thing I did, and this is how good I am. Yes. And people are going to look at you, and they're going to wait for you to make a mistake. It's annoying. Yeah, the college degree thing is hard, because I think there's been way too much of a push to get everybody to get one. On the other hand, in the highly technical field, I think it's still valuable. Because there's times that I fall back to the math or having understood the math i can identify likely errors there's definitely times i've come across problems

you know as a software engineer on some new project okay this is exciting oh i need to know how to do this particular algorithm or this particular signal processing thing. Without that background, boy, would I have been in trouble because you have to go all the way back to, okay, start from calculus and move my way up to Fourier. So there are things like that that can be quite difficult. It depends on the kinds of work you're going to end up doing. But I think in general, the college degree is a shorthand for people hiring. It says, okay, this is a minimum level where I can assume this person has competence instead of they have to prove it to me. Yeah, that's it. It's the, you've already proven a form of confidence

and now we have to teach you something else. So I'm not sure we're making this person feel good. I don't know. I mean, but it is a lot about the biases we have. Chris and I went to a fantastic school that completely trashed the both of us. And we paid a very high price for it, both emotionally, physically, and financially. And so we have to have value for it, or that was all a waste. I mean, just cognitively, we have to have value for it. So don't let us discourage you too much, because there are fantastic people who do great jobs. Yeah, probably the best people to ask this question are people who have succeeded without

college degrees. So, you know, if who have succeeded without college degrees. Yeah. So, you know, if you have succeeded without a college degree, maybe write in and we can share that with Peter also and get a different perspective than ours, which is, again, right into the amp hour, folks. They have college degrees, but they definitely don't care as much about the math and the low-level stuff that we do. So, you know, different perspectives. Get a different one. I'm not sure we're the right people. On the other hand, we have a listener named Chris who asked for some pretty different advice. I have too many college degrees. He wanted to ask us,

he's finishing his master's degree in physics and he wants to spend some time outside of academia. He enjoys working with hardware and he got into Arduino in college. And embedded systems are great and hardware and software is great and he's dabbled heavily in electronics he's even taught the electronics lab for the undergraduate physics majors

and he minored in cs and he wants vocational advice so but he hasn't had a job yet he wants to know if he has the basic skill set for a job in the embedded field. He doesn't know what to put on his job application since he doesn't have an engineer in his resume. He wants to know, would someone with his background fit into what companies are looking for in an embedded developer? And there was the note that Chris has a master's degree in physics, so he might have some good advice. So I got mine after working in industry for, well, I started mine after working in industry for six or seven years. I finished it, you know, sometime after that. You didn't start out with an engineering degree.

Oh, right. That's true. I have a math degree. Right. My undergraduate degree is in math, but I did take a lot of CS courses. And I did intern. Yes. I mean, I went straight into industry in a software development job before I even graduated.

So, not quite sure it wasn't an applicable degree. Especially back then, because CS, believe it or not, was still relatively new as a degree back then. So most software people, a lot of software people got math degrees. So there's that aspect of it. But... And so by the time you had a master's in physics, you had already had an engineer degree,

so that wasn't the same as what he's doing. Yeah, and I'd already been working in industry for a long time. And I worked while I was getting my physics degree, too, so it wasn't like I was out of the picture for very long. So, just from a general impression, that background seems really good for... It's going to be fine. Well, no, I'm sympathetic to the worry about getting that first job, though, because what does he put on his resume?

He can't really say engineer. Well, I guess you can. Who cares what you put on your resume? Yeah, I don't know what country he's from. He gave college letters, but I don't know if they're in the States. In the U.S., you can put an engineer behind your name no matter what. If you're in Canada or a number of other-

Oh, LSU? Yeah, it's the United States. Okay. I don't know. It's Louisiana State, right? It could be Lithuanian something university. Anyway, you don't have to have engineer on your resume you have taught electronics lab so that is very valuable and you probably have projects that you can talk about and you

need a couple of well you know what I would do? I would go to Craigslist and I would look at other people's resumes. Well, that's a good idea. That's always a good idea. Because you want to know what they're saying. I mean, you want to know, this is always so helpful to me. I like looking at other people's LinkedIn. What are they saying about what they did?

And what sounds better to you if you were hiring them? So you TA'd undergraduate electronics lab for physics majors. Awesome. What does that entail? Are you familiar with soldering techniques? Are you familiar with? I don't know. There's just so much here that you could really apply.

Well, okay. I mean, I don't know really. I don't know how I'd write a resume straight out of a master's degree. So I think you're going to have to acknowledge that there's no experience other than academic and kind of hands-on and lab experience besides personal projects. It's the same kind of problem as the person without a degree. And school projects.

Yeah. But, you know, it's a new grad. It's an advanced new grad. Yes. And so you're looking first at the internship aspect if you can. I guess I should have looked to see if he was graduating with his master's degree right away.

Oh, he's finishing, so yeah. Yeah, yeah. No, but I would apply for jobs, but I would apply for, here's what I would do after seeing the industry for a little while and startups. I would look for, with physics, if you have any interest in medicine, medical device companies, that background is very useful. Optics. Optics is very useful. So much optics. or, you know, more general architectural engineer doing research kind of stuff

for these medical device companies and startups where they're looking at somebody to solve, you know, be in a lab, be hands-on, but have a broad range of background, you know, in electronics or optics or... Or CS. I mean, he's got a broad range. Right.

That would be where I would look and where I would target because that's a really interesting place to work and at a startup they're looking for people a little outside the box a lot of times they're looking for people who have a broad background and comply themselves in all sorts of areas and that's what jumped out to me reading his background is, okay, this person is a systems engineer, which means, yeah, you might not be sitting there writing embedded code, but you might be doing a little bit of that, a little bit of electronics and a little bit of lab work and bench work

and building prototypes, you know, a little bit of tech work, sitting in meetings and influencing how things are going, running experiments. You know, it's kind of three quarters of the way out of academia into the working world. But, and those are, I don't know how easy those positions are to find,

but I wouldn't go to some big company is I guess what I would say is don't apply to the Googles of the world. The ones that have really strict pigeonholes although google does have a lot of places like that yeah google so alphabet i wouldn't write off the alphabet conglomerate um the other thing i would look at you didn't say what kind of physics you were right specializing in but you're probably using some interesting tools and the makers of those tools, start looking at them. You know, if you're, if you're into lab equipment, that is a place where being a user and being a developer can really help you. And it's again, a place like

startups where multidisciplinary knowledge is useful. And that is a word that should be on your resume, by the way. Multidisciplinary. Because if you can write code, you can do electronics, and you have a physics background, that qualifies. You probably can do some mechanical, too.

Most of the physics grads I know can draw a box. Yeah. And be aware that you only have a master's degree and the PhDs at your company are going to lord it over you. Not that I have any experience with that and not that anybody has completely ignored everything I've said, even though I'm completely right just because they had PhDs. I'm not bitter.

All right. So I hope that was helpful. Robotics would be a good... You know, just look at multidisciplinary engineering and you might find something in one of the job sites that fits you. And you just read the description. I don't know what else there is kind of in the academic world too. There's lots of research institutes and things,

robotics research that aren't academic necessarily, but they're adjuncts to universities and things. Well, it's kind of like going to the tools site where you're working on the tools for inventions and research. It will let you go back and forth. You said you wanted to get out of academia for a little while. If you want to go back, there are easier places to go to and come back from.

Yeah. All right, so let's go on. We actually got a lot of email. I guess we haven't done one of these shows in a while. No. From Giuseppe, what is your best advice for someone who's longing to get started We actually got a lot of email. I guess we haven't done one of these shows in a while. No. From Giuseppe, what is your best advice for someone who's longing to get started? I think this was, I think there was context.

Longing to get started with home hacking. Home hacking. Or with doing things outside work. Oh, sure. But doesn't know what to make. I have zero advice because I'm in the same situation 90% of the time that's not true you had a new idea I have lots of new ideas I just don't do any of them

well you were talking about making after we talked to Tony about electronic drums you were looking at that and you were like I could probably I didn't want to make electronic drums but I wanted to make some sort of weirdo instrument yeah that's still in the back of my mind so I guess what I wanted to make some sort of weirdo instrument. Yeah. That's still in the back of my mind.

So I guess what I would suggest is what kinds of things do you like? What do you want? Yeah. I mean, I'm going to make a little UV sensor, even though they do exist, because I want a little UV sensor. And do you want a little robot to show your eight-year-old? Do you want a lighting system that works just the way you want? Do you want to know how to hack your car? There's just so many options. Are you a ham radio person?

If you're a ham radio person, there's tons of projects to build radios and filters and things. Yeah. build radios and filters and things. And yeah, so I would take, it's definitely easier to do something you're interested in. Yes. Cause there will be a time where you get bored or disgusted or you're frustrated and you want something that is amusing enough to come back to. It's the same advice I give people when they want to learn a programming language.

It's like, well, okay, do you have a project? Do you have a project that's going to motivate you that you know is a reasonable reason to learn this programming language because otherwise you're not going to do it unless you're you know freakishly self-motivated person who can basically go to school without ever talking to anybody or you know it's really hard to learn on your own if you don't have

something pushing you. And if you can find a tutorial from Adafruit or SparkFun or any place, you know, Hackaday is a good place to look for tutorial sorts of things. And, you know, if they're building something that's almost exactly what you want, you can follow their tutorial and then change it to however you want. you can follow their tutorial to the end and then change it so yeah do something you want do something as easy as possible get the taste for it before you start in on the making your own 6502 uh out of resistors project i think you need transistors too. I think if you're good enough, it could just be resistors. That's not true.

Okay. So from Fernando Cola, what are your thoughts on the differences between arm cortex M and arm cortex A? What are the pros and cons of each? Jesus. We should get somebody from Armond to answer that. I mean, that's kind of like compare and contrast to the 8051 to the 8386.

I was going to go with a Geo Metro with the Honda Civic. More like a Geo Metro to a Ferrari. Really? Is there that big of a difference? Oh, yeah. I'm totally unfamiliar with the Cortex-As because they are things you run operating systems on, big operating systems, Linux operating systems. At the high end of the Cortex-M,

you have something that's probably equivalent to a... Okay, I'm going to butcher this completely, but... I don't know to a PC CPU from 20 years ago maybe not even because it doesn't have a memory management unit and at the high end of the Cortex A's you have something equivalent to

a really expensive laptop from like 2 years ago or 3 years ago I mean the Apple iPhone the newest Apple iPhone, runs a Cortex-A variant, and it's comparable to MacBooks from just a few years ago

in terms of some performance benchmarks. So it's a really high-performance CPU with modern architecture and all the tricks they do with caching and pipelining and all that stuff, and high power, relatively. It's really a computer CPU. Cortex-Ms are suited for the real embedded stuff,

and they're stripped down. Only recently have they had floating point. I mean, you know, Which boggles my mind. Quote recently, but it's, you know, it's only been a few years that they've had floating point. And you want floating point. Who wants floating point? No, I get it. I do. I just, it's just Everyone. So weird to me to have floating point. And I'm like, get rid of the floating point. Oh,

wait, nevermind. You can keep the floating point in this algorithm. It doesn't slow anything down. It just really is strange to me. But the Cortex-Ms are low power. They're usually quite a lot cheaper. Small. And they're small, physically small. They have lots of peripherals on them. They're microcontrollers.

So they have, you know, usually you get one packaged and on die is i squared c and your serial stuff yeah cortex a is you know there's all kinds of variants of the cortex a because they're usually custom and sometimes they're bare cpus and all that support stuff is outside it and sometimes there is part of a system on a chip with a gpu and um you know and with a Cortex-A, you're interfacing with real RAM, you know, double data rate, dynamic RAM or the Cortex-M, you've got some RAM on chip and then you've got spy flash. You maybe have some spy flash or some spy RAM or maybe, maybe you have a RAM interface. I guess we're struggling to make a good distinction here because to me, they're completely different

worlds. Usually on a Cortex-M, you're writing code bare metal. You might be running an RTOS. Little RTOS. On a Cortex-A, you're usually running Linux or Mac OS or whatever, iOS, or some high-powered real operating system, Unix variant. Okay. Yeah.

Cortex-A is for computer-like applications, and Cortex-M is for embedded-like applications. All right. That would explain why I know nothing about Cortex-As. Nothing. I have a note here about Adobe and doing this show entirely written and having you impersonate us what what are you talking about there was some adobe plugin that you just send it audio but i don't remember what

it was called now i know i'm out of my depth oh okay so the thing is you send it a bunch of audio like 20 minutes of audio project voco is that Is that right? Yeah, yeah, here we go. Project Voco. And once it has 20 minutes of audio or whatever, it can impersonate you. Yeah, so you can just type out a new script and it'll read it as you. And so you have no idea whether or not right now I'm saying this or typing it. Or if Chris is typing it. Rhinoceros.

Rhinoceros. I think that seems like an autocorrect. Yeah, it's kind of disturbing. I mean, well, it's kind of like Photoshop for audio. So, you know, Nixon probably would have loved this. I was thinking about all of the sci-fi movies that breaks. I think it's probably still detectable. I think, hopefully, it's still detectable in some way.

For a little while. Like there's still ways of detecting when an image has been doctored. To me, it's useful because sometimes I lose a piece of audio due to Skype. You actually use it? No. Oh. Sorry. No, I don't think it's available, is it?

But it would be nice if we could fill in Skype bits. Yeah, fill in Skype bits. Or, you know, if somebody says a word wrong, it's, you know, sometimes I have to do this by hand and go find little, you know, if somebody says a word wrong, it's, you know, sometimes I have to do this by hand and go find little, you know, the phonemes and edit them together. I have done that. All the times you put the not in front of enjoy the show. I deleted a whole piece from the last show that I didn't tell you about. Anyway, yeah, that's kind of funny.

Okay. Don't trust anything. Yes, exactly. Let's see. Robert Baruch. Sorry. Yes, sorry. Recent podcast made him catch in fourth enza.

There's a vaccine for that. He wants us to discuss pipelining, speculative branch prediction, and efficiencies of the if-then instruction. On a Cortex-M4? On a Cortex-M4. Well, if you'd asked about any other processor, I would be able to. I'm lying. I have no idea how to discuss that stuff. I know what they are, but I'm not a CPU microarchitecture person.

We should have Nate back on. Actually, yeah, that'd be great. Because those are pretty important things if you want to write an assembly. You need to understand how pipelining works. Because if you have a good pipeline, and then you have an if statement, and you end up outside your pipeline, you've just trashed several instructions. Well, that's where the difference between the M's and the A's comes in,

because, you know... Do A's have a lot more pipelining? Probably. That would make sense. And then they have look-aheads that will pipeline both paths. They're much smarter about this kind of stuff and figuring out how to avoid branch miss and pipeline, getting knocked out of the pipeline.

So, yeah, I don't really have anything useful to say. I mean, we could sort of define what these terms are, but we don't have anything useful to say other We could sort of define what these terms are, but we don't have anything useful to say other than, yeah, cool. In general, unless you have a good reason, and we've said this before, unless you're into it for the fun, you shouldn't be writing

an assembly, because the compiler is better at knowing about this crap than we are, which is clear, because we're not saying anything useful. Robert was pretty... I defer to the compiler. All right, yes. I think he was doing it mostly for fun. Yeah, that's cool.

He also wanted to hear you, me, I don't know, ask what is best in life and then answer in Arnold's voice. This must be a movie thing that I'm not getting. Arnold Schwarzenegger? I. Arnold Schwarzenegger? I think Arnold Schwarzenegger. And I have to tell you, my ability to impersonate people

is phenomenally bad. Like really, really bad. Conan the Barbarian, right? Yeah. Conan the Barbarian, what is best in life? Yes, yes, yes. And there's that whole litany he goes through. Mm-hmm.

Yeah, mm-hmm, right. Okay, so movie quote identified. Moving on. Greg says that he is struggling with a topic and wants us to discuss it. Oh, this is from quite a while ago. Okay. He designs devices that are meant to be plugged into other boards, like universal debugging, settings terminal, or data collectors.

And he needs to have a good connection, sockets and a cable. He often uses 3.3 UART and SPI. So he has all these things he wants to connect. The cheapest connector that would do the job is an rj45 and he could have all of these signals go over to this rj45 common connector do it but they're not rj45 signals and if you plug in an RJ45 to that, you're probably going to toast something. What? So the question is, if you're designing... Label it. I've seen RJ45 used for all kinds of things. I mean, that's pretty much a known common connector that all kinds of signals can be put over these days.

I've seen serial put over it all the time. Well, yes, but not necessarily 3.3. Okay, label it. Not necessarily spy. No, I totally get label it, and I totally get having it be the cheapest connector. He also mentions the DB9s, although it's bulky and whatever. And so he's, do you use a common connector incorrectly,

or do you use something new because you have new stuff? So what is the law about RJ45? Am I missing something? I don't... I mean, what is incorrectly? law about RJ45? Am I missing something? I don't... I mean, what is incorrectly? What is RJ45, quote, designed to be used with besides telephone, which, you know, nobody does anymore? I thought it was Ethernet, too.

Yeah. Exactly. I mean, you don't... This isn't really a widespread cable anymore. Well, I mean, it is, but the precedent, I guess what I'm're little converters, so you can run nicely RJ45 or Cat5 cable for your video signals. I'm not as against that as other people might be, but only because I guess I've seen it abused pretty often. I guess my point is, somebody is going to plug it into a wall socket what is that going to do?

if it's just going to blow up your device, fine if it's going to burn down the house, then that's bad it's all low voltage stuff that's where my split is well then you're fine, because you can't burn down a house with RJ45 I defy you to do it really? you you can't burn down a house with RJ45. I defy you to do it. Really? You want me to burn down the house?

Really? I mean, it's no worse than what the phone companies do with it. They put all kinds of crap over it. Oh, yeah. It's a phone jack, but oh, it's ISD. Oh, it's, it's, it's, you know, whatever over copper. It's DSL. It's, you know, it's already been, that ship has sailed.

Okay. it's DSL it's you know it's already been that ship has sailed okay well then so you you have Christopher's permission Greg to some electrical engineer is probably screaming at his

at his thing right now but I I just yeah I'm looking here I don't I mean there are standards

about where to put signal and where to put you know ground and what have you. And I would try to follow those. But make sure that you're doing the right thing so that if you're using a twisted pair, you're taking advantage of that. All right.

Okay. My reservation is somebody will plug it in wrong and how badly will that go for them? And if the answer is it won't work, share, I don't care. If the answer is fire, then that's often bad. Not always. You know, he was saying DB9. DB9 is worse.

Oh my goodness. There's so many misseases of DB9. Everything's DB9. So I don't see that that's any better. You could plug that into all kinds of wrong stuff. Yes. But you can't plug it into your house. Maybe your house.

It's just the same house. Oh. All right, let's see. Oh, kangaroos. If there are kangaroos currently within viewing distance, please wave at them for me. Okay, we're going now. What?

James, he usually listens on, and I just wanted to say hello to the kangaroos. Gotcha. I so seldom get the chance. Next. Let's see. Oh, I was going to use, well, so I was going to use this as the end quote, but now I have a different end quote. It was about Galen, who was a second century Roman physicist. No, physician.

Yeah, there weren't too many Roman physicists. And the story was from Supercaster, which was a book by Tedlock. And I was reading it, part of my whole Bayes statistics thing that I've been trying lately. And Galen was untroubled by doubt. Each outcome confirmed that he was right, no matter how equivocal the evidence might look to someone less wise than the master. Galen said, all who drink of this treatment recover in a short time. That's great. I come across that attitude far too long this this is right and the only time it isn't right is when you're wrong i just it kills me i don't know i i don't really have a point with that it was only but please if you have that reaction and I do understand the reaction I do I am totally

somebody will bring me a bug and I'll be like nope that must be your fault and and I try to hold that reaction back and instead say let me look into it give it a five minute a period of just thinking before pushing back and saying, maybe the problem is between the keyboard and the chair. Oh, we got a nice email from Glenn Nelson in Santa Cruz. Remind me about the Santa Cruz thing. I forgot something. Gotcha. Right.

Okay, so Glenn Nelson said years ago he was at a homebrew computer club meeting at Slack, which is Stanford Linear Accelerator. And this was around 1979 to 1981, which would have been the time to be in that club. Wow. Lee Felsenstein was the esteemed and amusing master of ceremonies. He was the chief engineer at Osborne One and prior to that, Proc Tech Saul, which I bet I should know about. During Q&A, he asked, what is Forth? The answer is, Forth is a language that proselytizes. All right. Which I think is so true. Yes. Well well there's quite a few of those these days

i guess so but fourth is definitely infectious like once you once somebody uses it around you you do tend to use it yourself uh okay santa cruz part okay back back to Santa Cruz. So we live near Santa Cruz, California. We live near the Monterey Bay, and it's lovely and not going to gloat too much about that. But we have been talking about having a party. And we are trying to figure out if this is a, why don't we have 10 people over for beers at our house sort of party or if this is a 100 person mini conference which would be far too big are we talking saturday afternoon or is this a night thing it's a 40 minute drive from cupertino and that's with little traffic and no rain. So it's a bit of a trek for the Silicon Valley folk. And I know there are some local people in Santa Cruz, in Watsonville, who are interested in a meetup. But what do you want?

What's your actual, you know, what are your parameters? You just want to hang out? So far, my idea is hats or hacks. Where either you have to wear a hat or you have to bring a hack. We could change it to hats and hacks. We could. Requiring a hat and a hack. I wasn't going to require a hat. Or a hacked hat. Well, I did think that if it was hats

or hacks, it could be both. Oh, that would be hilarious. All hacks must have hats we did that for christmas one year all of our christmas ornaments had to have little um santa hats it was ridiculous even the little coal ornament got a santa hat uh so but we're not talking about before christmas uh probably january febru Would you be interested? Would you probably come? Would you rather come on a Saturday? Would you rather come on a Sunday?

Would you probably not Friday night? Because we live in Aptos and the traffic on Friday nights is awful. So, yeah. Email us if you'd come. Tell us what you want. Do you really want it to be a conference with lectures? Or do you just want to hang out? Whoa, whoa, whoa, whoa.

What's going on? We're escalating to a conference here? Well, that's the thing. If everybody wants a conference with lectures, then we shouldn't have a party at our house. That's true. But if people want something more formal, if they want me to acquire some embed boards

and to actually sit down and have a hack session. All right. Or a hat session. We'll all build UV monitors. Then we might do it at the library instead. I'm sure we can find someplace better. Yes, there's that nice resort. I'm sure we can find someplace worse.

Christopher's dialing in on this yeah so that's the last announcement and I'm out of questions and so yes party planning okay okay it's not true I do have a couple more announcements

first we had George on not too long ago to talk about software engineers' perspective on embedded systems. And the project he was working on was JewelBots, which is this bracelet for girls that they can program. I think it's middle school is the target age. And that has shipped, and they're shipping more so if you want a christmas gift for some girl that you'd like to get interested in programming you can either get the very expensive bb8 robot which is fantastic or you could get the jewel bot which is pretty cool and not as expensive reminds me actually got that question from a financial planner.

Sort of randomly. What could I get for a 10-year-old girl who likes robots and programming and stuff? And I didn't have a good answer for it, but you got the BB-8 answer, which was good. Well, I started to send him to SparkFun, and then he said that she had done some HTML programming. I'm like, you know, maybe it'd be more fun. And he had the idea of the Sphero, so we weren't pushing too far outside of what he was looking at already. Yeah, yeah. I was thinking Tinkerkit, too, but that's probably, that's too advanced.

They have so many tutorials that it's not. So, I mean, she might not be able to go outside, she might not be able to color outside the lines yet, but she could pick it up and do something with it. And there's LittleBits, which I don't know that much about. AgileBits? No, that's something different. AgileBits is software. Software.

No, LittleBits, remember when we had LightUp, the magnetic stick-it-together sort of thing? Yes. So Little Bits is sort of like that. It's connectorized. You just stick things together. It's hard to destroy it. And they have lots of pieces.

And it's pretty cool. I'm always a little ambivalent about those like it isn't that much harder to use a breadboard and it is a lot more obvious what you're doing yeah but the part you have to says the person who's so the jumpers are all over the floor when you're just starting there's a difference between parts and modules. Like, okay, here's a light sensor module versus, okay, here's a photo transistor and support circuitry that I need for it. And so, I mean, there's a big gap between here's a block that has all the electronics you need to do a function and figuring out how to build that block yourself. Fair enough.

I mean, just think of a motor driver. Yeah, oh, yes, it would be much easier to have a motor plug in and have all of its drive circuitry on its plug-in point. But I don't know. I mean, there's something to be said for learning from the bottom up, too. It depends on how excited you get about that kind of breakthrough. There's a good feeling when you piece something together like that,

but it's a longer distance between I have nothing and I have a working thing, right? Yes. So if you don't really get high off of that stage of piecing something together from first principles, then maybe that's not the way to go. It's a shorter distance to go from modules. I'm always frustrated when I don't understand how something works at the most minute level,

which is why I never get anything done. Let's see, what do you want for Christmas? Since this is the Thanksgiving episode that we don't really expect anybody to listen to. I cannot say what I want for Christmas because it's far too ridiculous. I see. It's the thing with the... Oh, the thing with the drum cassette. Yeah.

I want a way to play drums so that no one ever hears me. Because musical performance is all about nobody hearing you, ever. Yes, that is what I would like, but that's far too expensive. So I will probably get something else. Maybe I'll get you a plant that isn't dead. A plant that isn't dead? It's fine. I just watered it. It's much better than it was. Trust me. Is it pining for the desert? Is that what's going on?

This is much better than it was. This side is green. Give me your plant. No, it's fine. I'm taking it away from you. I'm going to take it over to the plant hospital. Plant hospital? So you watered this... Sorry, sorry.

Sorry for the scream, everybody. Just knocking things out. Boba Fett fell down So you've watered this once in the last year? Three times in the last year But those first two were like the week I got it I'm getting Christopher a new plant for Christmas That one's going to be fine It'll go to the plant hospital. Yes.

I'm going to give it some plant biotics. Actually, they do have succulent food. That's fine. I'll take care of your plant for you. This is important content. I know. Well, yeah. This is the kind of thing you can expect with your donation to Patreon.

If you like this kind of talk you can expect with your donation to Patreon. If you like this kind of talk, please give us your money, and we will be encouraged to continue to do things like this. I thought we were going to get into useful Christmas presents for people who wanted technology stuff, but we didn't. That's a whole show. I haven't thought about that. All right, well. We should do that for the Christmas show that nobody ever listens to, which I know for a fact because I have statistics in front of me.

Yeah, yeah. So, well, we did talk to a couple of people about Christmas presents, and I thought we would discuss those, and then I realized that he probably was going to make his son listen to our show, and if we gave it away, he would kill us. Ah, I see. I need a spoiler alert. The 13-hour drive to and from school means that the son often has to hear the Embedded Podcast whether he wants to or not.

Plug your ears now. All right. I think we're done. I think we've made him a little... We're done torturing our poor listeners. All right. But Joseph sent us done. I think we've made them done. We're done torturing our poor listeners. All right. But Joseph sent us a book. Ah, yes.

Not too long ago. And it is called The Wild Robot by Peter Brown. And it is a wonderful little book that needs to be read aloud. I mean, the writing is nice. And it is about a wild robot. I guess that's pretty self-explanatory from the title. And it's just, it's a nice book to read. I don't know.

I want to say 10-year-olds, but, you know, we kind of default to 4-year-olds, 10-year-olds, and teenagers. And I would put this at slightly above four. So, so let's see. Uh, that will be my final thought is the chapter four from that book, the entire chapter. It's a page. And so now we do the whole thank you thing. So thank you to Christopher for co-hosting and producing.

He's the one that makes us sound good. Well, I can. You often do. And thank you for listening, whether or not you do the Patreon, whether or not you go to the blog and say hello, whether or not you email for contests or party or whatever. Thank you. We do like it that you're out there. And now a final thought. As you might know, robots don't really feel emotion, not the way animals do. And yet, as she sat in her crumpled crate, Roz felt something like curiosity. She was curious about the warm ball of light shining down from above. So her computer brain went to work and she identified the light. It was the sun. The robot felt her body absorbing the sun's energy. With each passing minute, she felt more awake. When her battery was good and

full, Roz looked around and realized she was packed inside a crate. She tried to move her arms, but they were restrained by cords, so she applied more force. The motors in her arms hummed a little louder, and the cords snapped. Then she lifted her hands and pulled apart the crate. Like a hatchling breaking from a shell, Roz climbed out into the world. 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.