#615 – Augmented Engineering

Chris Gammell: This is The Amp Hour Podcast. Released January 16th, 2023. Episode 615. Augmented Engineering.

Dave Jones: Welcome to the Amp Hour. I'm Dave Jones from the EEV Blog.

Chris Gammell: And I'm Chris Gammell of Contextual Electronics.

Dave Jones: 2023.

Chris Gammell: Oh, yeah. Yeah. New Year. Happy New Year.

Dave Jones: Yeah. Funny, though.

Chris Gammell: Dave's off in the middle of nowhere. This is a very first. Not a first. This is a first in a while. Dave's got these clandestine audio sounds behind him. He just had a twin prop plane go overhead.

Dave Jones: I did. I had it called. He's out in the bush. Yep. Yep. I'll go and walk about. Go and walk about.

Chris Gammell: Yep.

Dave Jones: There you go. Although not far enough that there's no internet here. I do have like satellite internet or something. Thank God.

Chris Gammell: I mean, yeah. Yep.

Dave Jones: Yep. Yep. So I had a satellite through Wi-Fi. So this could be our worst nightmare.

Chris Gammell: You know, vacation with Wi-Fi, it's what separates us from the animals. So, you know.

Speaker ?: Right.

Dave Jones: Oh, no. He goes and play it again.

Chris Gammell: Oh, yeah. They're just circling. You know, they found you, Dave. You got to watch out.

Dave Jones: They geolocated me.

Chris Gammell: That's right.

Dave Jones: I do play that game. I do play that geolocation game on Twitter. I actually post photos of, you know, myself. And I'm amazed. Yeah. People get it right? People actually find me. Yeah. Yeah. People do find me.

Chris Gammell: All right.

Dave Jones: So, yeah. It's quite amazing.

Chris Gammell: I guess if you go to like popular enough spots. Oh, yeah. Yeah. Yeah. You know, they know you're still in Australia.

Dave Jones: So, you know, if I post a photo of some grass or something, then, you know, they're not going to get me. Yeah. Yeah. Yeah. Yeah. But I can remember somebody, one of my viewers sent me a, because I think I could have, I put a call out, maybe like a sort of a half jokingly call out. Oh, send me, send me a photo and I'll try and guess where it is. You know, cause I was trying that, you know, there's geolocation software and stuff.

Chris Gammell: Sure. Yeah. There's like almost like a wordle, wordle for a geo spots. Right.

Dave Jones: And I can remember I spent like half a day on it and I couldn't, like, I thought I'd narrowed it down. Like I thought, you know, but I couldn't get the exact, like, I thought, like I did actually, I thought I had it, but then I couldn't find a matching photo that was similar. Like it was overlooking this city at night or something. And I thought I had it. Like, cause I actually tracked down like the person of where they lived and start like the person who sent it to me. So I tracked down where they lived and then I figured out, well, you know, like they were like, where would they like holiday around there or something? Or where would they drive to? And I narrowed it down, but I couldn't get it in the end. And then he appeared on one of my live shows once. And he said, Hey, I'm the guy who sent you that geolocation photo thing. And I said, and he said, do you really want to know where it was? And I went, yeah. Okay. And it was sure enough. It was one of the places that I had picked. Oh, no way. It was actually a universal studios, Hollywood. Yep. It was universal studio. It was like up on the, up on the top of the, you know, there's a lookout or something up there. Like, and there's like this train station or like, you know, one of those, you know, trains they take you around on, you know, on the tours and stuff. Yeah. And I had it nailed, but I couldn't get a photo from that location. So that was actually one of my, one of my pics, which I spent hours on. That was half of my day. Half a day of life. Ways to waste Dave's day. Yeah. There you go. Yeah. Yep. You can get really good. Yeah. Yeah. I've done the rabbit hole.

Chris Gammell: I've, I've tried those games before. Like where you try and like guess it based on just like random Google maps, you know, they drop you in the middle of a feet, you know, like a, the farmland. Yep. I've tried the best you ever get is like, maybe you get like the language on a sign.

Dave Jones: Oh, something like that. Yeah.

Chris Gammell: And you can kind of like seek around a little bit. It's, that's a tough thing though. Some people are really good at that.

Dave Jones: Oh yeah. No, they, yeah, they're great. So yeah, I've actually watched quite a few videos. There's actually a entire channels out there devoted to. Oh, no way.

Chris Gammell: Okay.

Dave Jones: Yeah. Solving those things. There you go. If you want to waste more of your life watching.

Chris Gammell: Yeah. Like other people do it. Yeah.

Dave Jones: Instead of wasting your life, listening to the amp hour. So. Well, Hey, Hey, Hey, Hey now.

Chris Gammell: We are the premier news source. Yeah. Right. Well, what are we going to talk about if you're on vacation? I mean, I guess, uh, what, what you ended up the year with, what I ended up the year with me, that sort of stuff. Or what are you thinking here? What do you, what have you been up to? It's been, it's been a while since we chatted. We, I took, I took a, a hell of a break.

Dave Jones: You took a sabbatical.

Chris Gammell: Yep. I, yeah, basically, uh, I, I had meant for Dave and I to record, but I was then.

Dave Jones: No, you were in a blizzard or something, weren't you?

Chris Gammell: I was expecting to be back home, but I got stuck in Buffalo where I grew up. And yeah, it was like a huge, the 2022 blizzard, basically like a meter, meter and a half of snow plus like sustained winds above, you know, like that's how you define a blizzard. And so, yeah, even though the snow started to clear up the airport, just like, we're like, nah, we're not going to open today when I was supposed to fly out. So I got home much later. So always bring a microphone with you. Where are you going? That's, that's the, uh, that's the lesson here, folks.

Dave Jones: Well, I'm, I'm currently using my old, old school Samson CO1U and it's in a shoe, sitting in a shoe. And the shoe is, uh.

Chris Gammell: That's the tip. That's the hot tip from, from the former traveler, you know.

Dave Jones: And the shoe is currently sitting on my laptop keyboard. So I've got, yeah, I'm sitting in on a deck chair. So. Yep. Yep. Relax. It's classy with a pillow. Yep. That's good. Oh, there you go. Hear some noises. Yeah.

Chris Gammell: The kids screaming in the background. That'd be good. You said you're on a farm where there's, uh, kangaroos as well, right?

Dave Jones: Yeah. Yeah. Yeah. There's kangaroos. Can't see any at the moment though, but, uh, no, no, but it's, uh, it's, what is it? 10 AM in the morning, 930 AM in the morning. So they're usually a, uh, they usually come out at, you know, dawn and dusk. So it's kind of past that. They're kind of hiding. So. Yep.

Chris Gammell: Makes sense.

Dave Jones: Yep. You saw a bunch of kangaroos while you were here. You, you ticked that one off. Yeah. On the Canberra trip. You saw about 50 in a paddock.

Chris Gammell: Yeah. That's right.

Dave Jones: For those who don't know, we went on that road trip to, uh, Canberra to the deep space, uh, tracking complex. Yep. And we put the, and we put it in the GPS. We put the deep space tracking complex in the GPS and we turned up and it was this like visitor. It was this other like visitor center or something.

Chris Gammell: Yeah, we were in the wrong spot.

Dave Jones: Yeah. We were totally in the wrong spot. And that was the whole paddock full of kangaroos. And there you go.

Chris Gammell: Yep. Yep. Check, check, check. There we go. All done. Yeah. That was like, uh, going on six years ago now. That's, it's been a while.

Dave Jones: Yeah.

Chris Gammell: Yeah. Time for you to come to us, man. Time for you to come to the US.

Dave Jones: Yeah, sure. It's going to happen. Yeah. Drag the family to the US. They're not going to let me go there on my own.

Chris Gammell: No, of course not. No, they should come. They should come and hang out, you know?

Dave Jones: Yeah. Yeah.

Chris Gammell: Well, last week we did a impedance matching episode. It was great. We brought up your, uh, video about, uh, AI and GPT got taken over the world.

Dave Jones: Oh, it's right.

Chris Gammell: Uh, I'm not sure I agree with you, but you know, I'm also not sure if I got enough credit for showing you how to use mid journey, but, uh, it's okay.

Dave Jones: Thank you very much. Yes. I do enjoy mid journey. It's great.

Chris Gammell: Yeah. Yeah. It's been, it's been, uh, fun to play with.

Dave Jones: Yep. Well, what, what, what don't you agree with? What don't you agree with? Well, like, like, you know, I think you. I was being a bit hyperbolic, you know, I'm like. Of course. I'm not.

Chris Gammell: That's kind of your thing.

Dave Jones: Like, um, when I say art is finished, right? Like, like it's over. I don't mean it's over.

Chris Gammell: Right. It's like. Augmented. It's changed. It's another thing.

Dave Jones: It's changed. Like there are many jobs, which I would have gone to and a, uh, graphic artist before. And now I just, I just wouldn't. Yeah.

Chris Gammell: So I think that's the thing. Like, and that's, that's that I do agree with. I think it's like sweeping up kind of the, the low hanging fruit, not the low hanging fruit, but like the, the menial, the menial work of certain industries. Right. And not to say that that's not people might be earning a living from it, but there's always that kind of like job creation, destruction type of stuff. You know, I think about like certain automation activities generally, right. If we move away from AI and, you know, chat bots and stuff like that, and you just think about general, like menial, menial tasks that become kind of automated either by software or like IOT type stuff, or, you know, it's like, yes, it is in theory taking a job, but it's also usually moving people up the value chain and they have more coverage area to like, okay, instead of going to check, you know, the pool water temperature on. 20 pools in a day. Now they become the pool manager for a region. And it's like, you know, is that destroying a bunch of jobs? Yeah. Maybe if you count those, you know, the five people that replaced with one person as like destroying four jobs, but like, also like there's, there's nothing that's going to stop the march of that progress because that, because there's always the drive towards efficiency, I feel like.

Dave Jones: Yep. But there's also like the art of, like, as we've discussed before, the art of actually making it do what you want. And most people aren't going to learn that. I did agree with that. Right. Most people aren't going to put the effort in to learn that. So they're just going to go to an artist who is just then going to use, is they going to know how to drive the engine to generate what they want.

Chris Gammell: Maybe you should explain that for people who haven't seen the video as well, what you were saying there, because I think that's a salient point.

Dave Jones: Oh, you want me to explain?

Chris Gammell: Yeah. Yeah. You explain what you mean by the driving the engines, right?

Dave Jones: Right. Because the, like, you can't, well, you can, you can just say, like, I used an example on Twitter. I wanted a, what was it? A bunny rabbit with a, wearing a top hat while eating a banana or something. I don't know. And it like, and it does it. No, it was a rabbit sitting on a plant while wearing a hat or something like that. And it actually did it. It actually generated that. But in a lot of cases, probably the majority of cases, you can't do it that simplistically. Like you really have to get the, like, you have to tell it that you want to render instead of a painting type, you know, cause it can do different styles of art and stuff like that. So you have to, you know, say, oh, I want a realistic render or something like that. And then you've got to, and yeah, you've got to put the things in a certain order. Otherwise it's going to prioritize, you know, certain things first before, you know, so it's going to think about, oh, okay, I'm going to generate this first and then add that later over the top and it's back to front and stuff like that. So there's an art in, in doing the script to generate exactly what you want to actually tweak it. Otherwise you're just going to end up with something random, which will probably look funky. Right. But you know, it's not exactly what you wanted and that's the point.

Chris Gammell: Yes.

Dave Jones: So. Yeah.

Chris Gammell: I have found specifically, so I use a lot of mid journey stuff like Dave's been doing too. And like, I find there are certain things that I am absolutely incapable of getting it to, to, to actually output. So like, here's an example. I have a webinar coming up where I had this idea where I was like, okay, we're talking about like low power. And for some reason my stupid brain's like, oh, that's like sipping power. Like, oh, I can be so clever about this. Okay. And like, and yeah, so it's like sipping power. So I made a, so I was like, okay, I want to have like a, like what looks like a Duracell battery, but it's actually like a, a can. And then there's this curly straw coming out of it for some reason. I don't know why I tried. I spent probably an hour and a half trying to get it to render a straw.

Dave Jones: Wow.

Chris Gammell: And I could not, it'll do a straw, but it won't do like the curly Q straw that was in my head. And for some reason it just, I just, I got so frustrated. I spent my whole lunch hour doing it. I was just like so frustrated. And at the end of the day, what I did is I just generated a, actually I'll send you the post that you have it here too. I just generated most of the art. And then I said, you know what? I can just do the old. Photoshop it in there. I get, I can just Photoshop it in. Yeah, exactly. And like, and, and like that to me is a, you know, it's the hard part is definitely done already. Right. The hard part was to generate. So I like rendered like a, what kind of looked like a D cell battery with like kind of was looking like a Duracell, you know, just kind of like the copper top style, uh, coloration. And I wanted to have, be in a kitchen and I wanted to have like light coming in through window, like all of that stuff. It was awesome with, but then everything else I had to like edit. And so like, again, so I can see it now.

Dave Jones: Yeah.

Chris Gammell: Yeah. So like I was trying to, and people should sign up for this webinar too. Uh, I was trying to do all this stuff and, and get this in there, but, but, uh, or sorry to take it to the electronics analogy here. I would say this is like, if I was trying to build a switch regulator, now the AI generator engine basically is like moving me up a level where now I go and buy the switching regulator, like the finish. I'm not building the switching regulator from scratch. I'm not sketching the switching regulator out on a piece of paper. I am buying that finished good and implementing it. And now there's a whole other set of problems because now I've moved up a level and I need to, I want to just trigger this one little detail there, right? I need to implement the, you know, the soft start and it's not in this chip set. Okay. Well, I got to go and cut and paste this thing in there. It sounds trivial. I know. Okay. But, but like there, there isn't the value that I previously may have had being able to draw a battery on a table and have it in there. I, I've now moved up a level and I, that's just kind of done now. And I can go and focus on some other stupid stuff that I've implemented for myself here, like putting a curly straw into it. You know, that sounds stupid, but, but that's, that's kind of how I like imagine a lot of this, this similar like augmentation is it's moving up a level, like letting someone else do the silicon piece. That's, that's kind of how I, I, I think about it in terms of electronics.

Dave Jones: That's what a lot of people have tried to argue with me on Twitter and elsewhere in, in the comments. They've tried to argue that, oh no, it's not, you know, it, it just generates this garbage, but you don't understand that like I was not even able to generate garbage before. Right. I am not an artist. Right. I cannot produce a photo realistic photo of a rabbit. Okay. Now I can, I've gone from zero artistic skills to like the, the level is actually enormous. Right. Yeah. And, and then, yes, I can use my feeble, uh, earthen, uh, view or, um, you know, Photoshop skills to actually, then I, then I want to, you know, overlay a transparent PNG over the top of that or something to kind of like finish it off. Yes. It's like, you have no idea what capability this is just given a hack like me. I've gone from nothing to, you know, at least semi-serious ability to actually produce stuff.

Chris Gammell: Right. And I think, I think the other thing that too is, is that it becomes a sliding scale now too. Right. And it's like, now this is a tool that is available to everyone. So again, to take it back to the electronics analogy, right. Whereas maybe only a Jim Williams could implement this crazy ass all, you know, silicon level, you know, single transistor isolated thermocouple circuit. Right. And now TI comes up with that same kind of circuit that's as good, or maybe even a little worse, but still accessible to everyone.

Dave Jones: It's close. Yeah. Yeah.

Chris Gammell: Yeah. It's, it's now accessible. Okay. Well, now that problem's kind of solved. There's a cost to it, but it's kind of solved and we've moved on. Now we might bemoan the lack of knowledge when we have, you and I have bemoaned the lack of knowledge that that person that knows how to, to do that circuit. But more than likely that person goes and works at the company that now is producing that circuit, like the TIs of the world. And, and in doing so it becomes accessible to everyone. I feel like I'm trying to like promote AI here. I'm not really trying to, but I, what I'm really trying to say is I'm terrible at art, just like Dave's terrible at art. And I need art.

Dave Jones: It's like everything in the engineering chain, right? You can, you can do phenomenal engineering without any, using anything more than Ohm's law, right? You can do phenomenal engineering without not knowing how to lay out a PCB because there's, you know, companies that do that, or you can just, you know, wire modules together, or you can, you know, you can develop these hugely complex systems, right? Without any of the deeper underlying knowledge, you can do a lot of practical stuff. If it doesn't mean you're the world's best, you know, like you can slap together, like all this cellular, right? All this cellular stuff that you're dealing with, right? All this internet of things, cellular stuff. You don't need to know. There's so much stuff under the hood. Yeah.

Chris Gammell: There's so much stuff under the hood that I have no visibility into and I don't want to, and I just want it to work.

Dave Jones: Exactly.

Chris Gammell: I'm like, Oh my God.

Dave Jones: And, and you're a decent engine. You're an experienced engineer and you still don't have a clue how that works under the hood. Yeah. Right?

Chris Gammell: You would just never do anything though, right? Or you would have to work at such a massive company that is so vertically integrated that you have, you have a silicon designer all the way up to someone writing software on a cloud server and you're building all the servers in between. Like it would have to be the biggest company in the world, right? It's just, this is just part of interoperability.

Dave Jones: Yeah.

Chris Gammell: I'm sure there's still arguments against this stuff, but.

Dave Jones: Oh, well there is, but now it's given anyone with any half decent Photoshop skills, right? They can combine that with mid journey or some other, you know, art, AI art generator thing and they're warm. They have leveled up so much. It's crazy. Right? Sure. The exact finished product from the AI is not perfect, but then you combine that with your Photoshop skills to like finish it off and boom, you've opened a whole new area instantly. Like without having to learn anything more than a couple of command line, how to drive a command line. Right?

Chris Gammell: It's just, it's nice. I think really it just, it brings up the entire, like, I think we're going to be more discerning as consumers of art, if you want to call us consumers of art. I think we're going to be more discerning generally. I think I'll be able to look at some things and be like, oh, that's probably generator versus, you know, maybe not. I don't know. It doesn't really matter. But I feel like there's just going to be more in the art area, there's going to be more stuff out there. And so we're going to have more expectations that art is attached to a thing and that it's, you know, and I'm really going to say blog art versus just art as a category, you know? So like there's just more expectations that it's possible to do these sort of things. So it'll become less special, right? It will actually grab our attention less, I'm guessing. Right. Unfortunately, you know, for better or worse.

Dave Jones: Yeah.

Chris Gammell: And I don't know. I think the more interesting stuff is on the code side of things, like all of the AI. Oh, yeah. Yeah.

Dave Jones: I only touched on that in that video. Apparently it can do almost any language. I'm not going to say it can do any language, but it's trained on a whole bunch of them. Sure. Yeah.

Chris Gammell: Well, and I was talking to someone about this recently too, and I think, so then like, okay, so now it can write code. Okay, cool. So I'm very, I'm not great at code, but even the code that I'm writing these days, it's actually just kind of like understanding code that's out there and pasting it together. I'm pretty sure that as of right now, right? So I could probably say, hey, chat GPT, can you write me something in Zephyr? And it would probably be like, but I would put like something that would work in Zephyr, the real-time operating system that I'm often working on, and it would work. And then if I said like, hey, can you go and write me a new driver in Zephyr for this new part from, from analog devices? I don't think it's going to do that yet. And so like. Right.

Dave Jones: No, but that would, that'd be powerful. Wouldn't it?

Chris Gammell: That would be super powerful. But you know what? That's, that's work that I want to get off my plate anyways. Like, so I don't see, I don't know. I guess there's a lot of concern about that sort of thing, but like, I see that, I see all of these things as like supplementing solo engineers. Yeah. And like. Same here.

Dave Jones: Yeah.

Chris Gammell: Yeah. I don't know. I, I've, there are definitely downsides to like, you know, bias and AI engines and all of the potential there's pages and pages and pages of why certain AI things are bad and that's fine. But like, I see it as a tool for like doing more stuff myself and like reducing barriers to like communication between myself and other engineers. If I can like prototype something before I go talk to a professional firm or engineer and be like, Hey, it kind of looks like this and I can like help, help it to formulate my ideas. I'm still going to use that professional firmware engineer to make it really good. And so.

Dave Jones: Yeah. Yeah. To polish it. Of course. Of course. But to get something off the ground and, you know, at, at least it doing a hollow world or something like that.

Chris Gammell: Yeah.

Dave Jones: You know, which is half the challenge, especially if you're doing a complex system. Then. Yeah, definitely. Yeah.

Chris Gammell: Yeah, definitely. And I think, so like, I see a lot of this stuff is like, you know, we, we've talked about on the show before we talked about like layout and like how AI might play into that stuff at some point. Like, I think it's all going to be supplemental for the next many, many years. Right. Yeah. Yeah, totally. It, it might, it might move faster than we've predicted and you know, better or worse, fine, whatever. But like, there's still like system capabilities. Like I, I still needed to define what I even want in the system and being able to interact with people that don't know what they want in a system and understanding systems. And like, so personally, I'm not personally a very, very worried about this sort of thing. You know, like that's where usually where people start is like, oh no, what about my job? It's like, look, there's a lot of work to be done. So I'm not, I'm not super worried about that sort of thing, but it will be interesting to see kind of that. I think we're, we're very rapidly entering the age of co-working with these AIs. And I think then it's a question of, that is co-working, right? It's a collaboration. It's collaboration. It's, it's augmentation.

Dave Jones: Well, like people use the example, like, well, is it, is it AI if I use a Photoshop's, you know, cloning tool, you know, like it's got all those really advanced tools, right? To like filling backgrounds and stuff. That's like kind of like, you know, how is that any different?

Chris Gammell: I mean, you don't know how many photos I've edited of myself to just like have fuller lips and just smooth skin. Yeah. Put on Instagram.

Dave Jones: So you can think of it as, so if you're worried, you can think of it as just a more advanced version of, you know, a tool that just allows you to do stuff.

Chris Gammell: I think it is. I think it's, I think we're entering that age of augmentation and then, you know, what does that mean then moving forward? And really when do we exit that age of augmentation? That's, that's what people are really talking about.

Dave Jones: Augmented engineering.

Chris Gammell: I think that's right. Is that the name for this episode? I think it is. Yeah. Yeah.

Dave Jones: Yeah. Yeah. It's pretty good.

Chris Gammell: I like it. Dave's going to be a cyborg. Yeah.

Dave Jones: Oh boy.

Chris Gammell: Here's, here's, here's how crazy this stuff has gotten folks. I would trust Dave to make blog art now. Like that's, that's a big deal.

Dave Jones: Yeah, I know. That's it. And people, and people think it's a nothing burger. It's taken me from a nothing to, to Chris Gammell thinking that I'm, I'm, I'm a capable blogging artist. Do you know what I mean?

Speaker ?: Yeah.

Dave Jones: It's nuts. Like, like, seriously, I had like.

Chris Gammell: No, there would not have happened.

Dave Jones: Zero ability. Right. The only ability I had before. Okay. I can take a picture of, I can steal a picture of a rabbit from Google images. Okay. But I've got to make sure it's transparent. And then I've got to overlay it on the background that I want. And then I've got to add like a hat on top. And then I've got to add a, you know, like, you know, right.

Chris Gammell: Right. No. Yeah. No, I think, I think that's, that's a very realistic look at kind of how I do it too. Sometimes I would trace over top of things as well in, in Inkscape.

Dave Jones: Yeah. No, I'm, I'm real. Like my, you have no idea. Like I'm, I'm still earthen view level. Right. I'm like, yeah, that's how poor I am.

Chris Gammell: So I am excited. So I will say for people that are listening that if you're worried about Dave and I talking about this, I think we're gonna keep talking about it. In fact, I have a guest that's supposed to come on at the end of this quarter. That's kind of doing some augmented layout type stuff. I, I was kind of skeptical about it. I talked to them and they said it wasn't right to talk to, but they wanted to wait a couple months to showcase what they're doing. But like by the end of like a 20 minute call, I was pretty sold on the concept because again, it was like a measured thing, right? If someone came in and they told me like, I'm going to do your, your, I'm going to, you put in, you know, two data sheets and I will output a layout for you. It's like that, that doesn't make any sense. But if it's like, I'm going to augment it, augment some of your design stuff. Yeah, that, that is a good start. Will that company maybe do more and more of layout for me over time? Yeah, that's possible too. But I don't really, you know, I think maybe one thing is that we, you know, I, I, I have held on to the identity of electrical engineer a long time. But like, if I just think of myself as like a person solving problem with electronics, then I really don't care if I'm doing firmware or hardware or layout or, you know what I mean? Like, it's just like, if you're just trying to get to that end product, I think you're going to be more effective. You're going to, you're going to use more tools in your tool belt. And some of them might be some of this, these engines that are, are coming out.

Dave Jones: I agree. Can we talk about something else now?

Chris Gammell: Yeah, totally. I found out what trade shows I'm going to be at this year. I'm going to be at Embedded World again, which is in March, and I will be at Embedded Open Source. Is it? Is that the... That's right. That's Nuremberg. Yeah. I really love that show last year. I mean, aside from getting COVID. Right. I think it's, it's one of the best trade shows I've been to. And I would recommend, you know, if people are in Europe and travel is not too tough. I would love to go to it.

Dave Jones: Yeah, but yeah.

Chris Gammell: Yeah. And then I will be at the Embedded Open Source Summit, which is, I went to the Zephyr Developer Summit last year. This is kind of like a combination of that plus Embedded Linux Conference and Automotive Linux Conference. And so that's in Prague in June. So I will be at those two shows if all goes well, and hopefully without COVID afterwards, if all goes well, we'll see how that goes.

Dave Jones: Surely you're immune now.

Chris Gammell: Yeah, right, right. I would love to hear about other, you know, I found that other shows in the US have been kind of on the downside. If people know about other shows, I'd love to know about them. I'm always on the lookout for that.

Dave Jones: Well, I just found out, I think John Oxer mentioned it. There was a, there was a Maker Faire in Melbourne. I thought Maker Faire was dead. Yeah. Yeah. I think there, he just went to a Maker Faire in Melbourne and it looked like pretty decent size. I didn't hear a thing about it. So I don't know. That's bad.

Chris Gammell: Yeah. Maybe next time.

Dave Jones: But they went bust, right? I'm sure my memory isn't that bad. I think. Like the actual brand went bust or something.

Chris Gammell: No, it's not, it's not like bankrupt or anything, but the, the big Maker Faire has not come back and I don't think they'll come back. So that was like, the big ones were in the US at least. And like, they were centrally managed as well. So it's kind of like a, I think of it kind of like a.

Dave Jones: Yeah. Yeah. I had to like license the name, right? You had to end up.

Dave Jones: Like what's the McDonald's model?

Chris Gammell: Like franchise, like a franchise model. I kind of think about it like that.

Dave Jones: Right.

Chris Gammell: And the centrally managed ones, I believe are gone. And those were like New York and San Francisco, I think. But I think.

Dave Jones: So how's this little tin pot one in Melbourne happening?

Chris Gammell: Well, that's a community based one. So the community based ones were kind of all over the States and all over the world.

Dave Jones: But you still need approval for, to use the name and everything, right? And the logo. And I'm pretty sure it was a Maker. It was branded as Maker Faire. So.

Chris Gammell: But the company's not gone. I don't think. I think it's just, they're not running the conference themselves anymore. This is all bad memory on my part. So.

Dave Jones: Yep.

Chris Gammell: I do believe. Cause I think make.com, make.co is still there. I think it's still.

Dave Jones: All right.

Chris Gammell: A site. Make.co. Yeah. Yeah. So make.co is still there. They're still, they still have a publication, but I think it's just significantly scaled back. Okay.

Dave Jones: Interesting. Yeah. I just like, it didn't even pop up on my radar until, until it had happened. Yeah. That's too bad.

Chris Gammell: Yep.

Dave Jones: Oh, well. Otherwise I probably would have gone, I think. So, you know, it's only Melbourne. Yeah.

Chris Gammell: Yeah. That's not, that's a pretty short flight from you, right?

Dave Jones: Yeah. Yeah. It's not far. I could do it as a day trip. Like, you know.

Chris Gammell: Oh, there you go.

Dave Jones: Yeah. Yeah. Yeah. Hmm.

Chris Gammell: Well, Dave's still on the lookout for the best Aussie tech conference. I still think he's going to have to start one, but.

Dave Jones: Yeah. I know. I'll probably have to start one. Yeah.

Chris Gammell: Yeah.

Dave Jones: Yep. Someone's got to, someone's got to take the bullet. Yep. It's going to have to be me. Yep.

Chris Gammell: Yeah. I'd say you just, you bankroll it and then let someone else set it up for you.

Dave Jones: Right.

Chris Gammell: You got tons of money. Come on, man. Oh, yeah. Some of that DMM money. Right. That DMM money. Yep. All that, all that cryptocurrency I keep hearing about. Oh, boy. Not going there. Sorry. I didn't say it. I didn't say it. It's doing just fine. Don't know what you're talking about. Good. Good. Good. Good. I just meant I didn't want to talk about it. Oh, boy. Yep. Yeah. What else is, what else should we talk about then? I guess we have a list, a list of things.

Dave Jones: Which I have not looked at. I haven't either.

Chris Gammell: Oh, actually, I mentioned this. In the last show, but it still blows my mind that it's a thing. So, Qualcomm Snapdragon satellite enables two-way messaging from the Iridium network. So, basically, like, these new chipsets are, like, talking. And Alvaro Prieto was on the show last time, and he works for a company that talks to satellites regularly with, like, actual big Iridium modems. But, like, phones talking to satellites. For some reason, that blows my mind. I don't know, like, if that's other people, too. Yeah, it is.

Dave Jones: But it's, like, hundreds of bits per second. It's, like, you know, Voyager Space Probe-level data rates, isn't it?

Speaker ?: Yeah, that's great.

Chris Gammell: Yeah, text messages and no emojis. What's better than that? They should make us do a T9. You know, that's the real measure. A T9? T9 was, like, the old Nokia. That's how you, like, input text using a keypad. Was T9? No. No? Was it called T9? I think it was called T9. I think so.

Dave Jones: It doesn't ring a bell.

Chris Gammell: You know, like, you remember how you used to put text messages into your Nokia phone using the keypad? You know, it was only, like, one through zero. Oh, yeah.

Dave Jones: Well, they have the three letters on each number of the keypad. Is that what you're talking about?

Chris Gammell: That's right. It's called predictive text. So, that was, like, that was the system where it would look. Yep. It would be, like, you know, you type in, you hit, what was it, 888 or 8, and then it got to T, and then it's, like, then you put in an H, and it's, like, oh, he's probably trying to say the. Right. He's probably, yeah. And then it's, yeah. Show you stuff on the screen. Yeah. Yeah. So, you don't have to type out everything, but just most of it. So, that's T9. Yeah.

Dave Jones: Oh, okay. There you go.

Chris Gammell: So, I feel like that would be, if you're going to talk to satellite, you know, you got to do it through T9.

Dave Jones: But this is one of the things where people think, oh, we're just, you know, six months, a year away from having, you know, massive data rates directly from phones to satellites. No. Yeah, no. Yeah. It physically is not possible unless you have, like, massive power levels. Like, it is simply the power, you know, it's a power level problem. Yes, you can do text messages. Yes, you can do, you know, voice. Mate, we have a loaf line.

Chris Gammell: Oh, wow. Nope.

Dave Jones: Nope.

Chris Gammell: They're coming for Dave, folks.

Dave Jones: Yep. They've geolocated me. But, no, it's, like, it's just one of those things that's just, it's, you can confidently say it's not going to happen because they've just got massive power requirements. Sorry.

Chris Gammell: Yeah. Yeah. Yeah.

Dave Jones: So.

Chris Gammell: Yeah. Yeah. I think, and antenna requirements too, like, in that.

Dave Jones: Oh, yeah. No, totally. Yeah. Exactly. Like, it's, you're not going to be able to pull your shoe phone out of your pocket and start streaming HD video directly from a satellite. It's just, no. No. Yep. Yep. I mean, how many, how many watts is the Tesla Skynet thing take? It's like, what is it? 800 watts or something? It's enormous. Oh. Power requirements. Just, just for that transceiver dish.

Chris Gammell: Oh, really? The one, actually the ground, the ground station for it. You're talking about.

Dave Jones: Yeah. Yeah. The ground station thing. I think it's like 800 watts or something. It's enormous power. I've been, maybe it's not now. I'll try and get it live here, but yeah. Power requirements are enormous. God. What's it actually called? Tesla.

Chris Gammell: Starlink.

Dave Jones: Starlink. Yeah. That's it.

Chris Gammell: I couldn't remember either. Dave and I are increasingly get into like the, oh, I can't remember. And they're like, yeah, I can't remember either.

Dave Jones: Starlink. Maybe we can ask chat GPT.

Chris Gammell: There you go.

Dave Jones: Yeah.

Chris Gammell: It could just be like our supplemental during, during the show. Be like, oh no, here we go.

Dave Jones: It's a bit over 110 to 150 watts on average. Right. I mean, come on. Right. There's another thing that says 45 or 75 watts. Right.

Chris Gammell: Yeah. I mean, if people go and look at the Shariar teardown of the, I guess, the older generation, because it was circular. Yeah. But he looks at the beam steering with that thing as well. And so, I actually hadn't seen this. Wow. It's 110 bucks per month. That's crazy. Oh, is it? US. It's crazy that this is accessible. So, there's a price on the US side, at least. It says 110 bucks per month with a one-time hardware cost of $600. That's pretty bonkers. Like, you think about-

Dave Jones: Which they're making a loss on that, apparently. Apparently, they're making a loss on the hardware cost. Yeah. They're eating that. Oh, interesting. I don't. Yeah. Long-term bet-wise, I don't think they're going to stay in business. But, you know.

Chris Gammell: I don't know. Like, you think about how much this enables, like, in terms of- There's so many remote areas in the US. Like, there's just so much- Yeah. I mean, I guess I don't have to get coverage on the Outback yet. But you think about, like, all the people in the Outback? Yeah. True. I'm sure it will.

Dave Jones: Oh, yeah. Like, it's a great option, you know. But it's, like, it's not cheap. And, you know. How many people actually use it? There's more and more people move into the bloody cities in the big smoke where there's coverage. You know, even cellular. Once you've got cellular coverage, there's no point having startling. Right?

Chris Gammell: True. True.

Dave Jones: So, you know, because you can already get your, you know, your 5G whatnots.

Chris Gammell: Sure.

Dave Jones: So, yeah.

Chris Gammell: Soon 6G. Why not? Bring it on. Even 4G.

Dave Jones: Even 4G gives you just as capable, if not more. Capable than Starlink. Right?

Chris Gammell: I don't know about that. But I don't know the details.

Dave Jones: All I know is that I can upload YouTube videos direct from my phone on my 4G connection. Like, in minutes. Wow. So, it's like, you know, it's pretty quick.

Chris Gammell: Yeah. That's nuts. Yeah. Yes, we have a wealth of connectivity options. Which is why it's interesting in the next story that we're going to talk about here. That a well-known connectivity provider, known as Espressif, decided, you know what? Actually, we're not going to put anything, any connectivity on this upcoming part. The ESP32 P4 was just announced. And it is a dual-core, 400 megahertz, with a secondary 40 megahertz, all RISC-V processors. And there's...

Dave Jones: Why that secondary processor? Is that a low-power jobby? So, then you can have that running... Yeah, low-power.

Chris Gammell: That's right.

Dave Jones: Yeah, right. And then that powers up the bigger beast of a core. And then... Right.

Chris Gammell: Yep.

Dave Jones: Okay.

Chris Gammell: Two cores. Two beasts.

Dave Jones: Right.

Chris Gammell: But the crazy thing is that it's an Espressif part, known for their low-cost Wi-Fi.

Dave Jones: That's the only thing they were known for, right? That's like... Yeah.

Chris Gammell: That's right. Right. Well, increasingly, they're getting into the Thread game as well. And so, they're doing Bluetooth Wi-Fi, right? Thread. But now, this is nothing. No connectivity.

Dave Jones: Is it really cheap?

Chris Gammell: Well, there is no pricing on it yet. But, yes, it's going to be really cheap.

Dave Jones: Like, how much silicon area does that save? Like, how much cost per chip does that save? I'd really love to know.

Chris Gammell: Without the RF components? Yeah, without all the RF stuff. Yeah. I mean, pretty significant because you do have to have some inductors and stuff.

Dave Jones: Well, that completely changes the node, doesn't it? Completely changes the process fab node, right?

Chris Gammell: That, I do not know. But, I mean, they're not lacking on other stuff in here too, right? So, if you look at, like, the system diagram, it's got every, you know, it's got DMAs. It's got codecs. It's got MIPIs for doing display stuff. It's got all the usual embedded DACs, ADCs. I don't like the ADCs for Espressif personally. But UARTs, you know, all that kind of stuff. And then, it's like all this stuff. And then, I'm guessing at some point, they will also have a version when they tack on a, you know, Wi-Fi 6 or something as well. So, I don't know. It's interesting. It's interesting to see what they're going for in this game. I'm guessing they're going for low-cost, high-power. I'm guessing they're kind of going after STM32 with this sort of thing. Maybe even, like, more like the NXP side of things because it's, like, pretty beefy. You know, this almost looks like an infotainment kind of, like, if it's got encoding for video, it's got, like, display drivers, things like that. Like, maybe a car manufacturer came to them and said, like, hey, if you can make this 50 cents and, you know, we'll buy a gajillion of them for all over our infotainment system sort of thing. But we don't want any connectivity because that'll drive the cost up. And then, you think about the fact that it's risk five instead of tensilica. And so, then, the licensing cost drops off, too. And it's, like, wow, this, I feel like this is a leading-edge kind of component for me because other manufacturers are going to look at this and be, like, oh, yeah, they don't pay tensilica or they don't pay ARM any of those licensing fees. That's, wow, we can really get cheap with this stuff. And, like, I feel like this might be a leading edge for other chip makers to keep dipping toes into the risk five. Still do it. The risk five pond, I think.

Dave Jones: Can we actually call it now? Is it too early to call it that the few, like, everyone, like, risk five is going to be, like, the thing?

Chris Gammell: Oh.

Dave Jones: I don't think it's too early to call that. It's, like, everyone's doing it.

Chris Gammell: It's a little early, only because of implementation, right? So, I don't think, I think it's only early in that they're not out there broadly, right? So, like, in terms of, like, parts you can actually buy, which we've said on the show many times before, like, doesn't exist until it's on the shelf. Yeah. Even this thing, this P4 isn't out yet. However, you know, Espressif has the C3, which is available in large volumes, which is one of their cheaper, cheaper Wi-Fi parts, and it works well. It doesn't have as much memory. So, like, they're definitely, like, they're not using up floor space for that sort of thing either, right? They don't have a lot of, like, onboard flash.

Dave Jones: Yeah.

Chris Gammell: But it's out there, and people are using it. So, like, that's cool. And I keep seeing more and more RISC-V parts, like, pop up. You know, I think I might be buying you a bunch of RISC-V parts in a couple of weeks.

Dave Jones: I was about to bring that up. Yep, yep. You're going to have to concede.

Speaker ?: No, no, no.

Dave Jones: We'll get to that later. You know what? You're going to get to that. Okay. We'll get to that. We'll do that at the end of this.

Chris Gammell: That's future episodes. Future episodes.

Dave Jones: Can I actually shout out the PCB design in this article? This Hackaday thing, which you sent me.

Chris Gammell: Sure.

Dave Jones: This is a way scale built onto a PCB. So, they're using the flex in the PCB as the scale. They've got a little sensor on the bit of the PCB that moves. So, they've actually routed out slots on the PCB, and then they've put metal standoffs on there, and those metal standoffs would go to a plate on top. And then they're using the sensors on the PCB to actually get the strain. So, they're using the strain gauge on the PCB material itself to actually measure the pressure on the actual PCB material.

Chris Gammell: That's pretty cool.

Dave Jones: I don't know how linear that's going to be, but it's cool.

Chris Gammell: Yeah. And this is definitely using, I think it's using that cheapo 7-4, HX7, yeah, the HX711. That's like a really cheap, like, that's what's in all of your cheap scales that are out there.

Dave Jones: Oh, is it? Okay, right.

Chris Gammell: Yeah. It's not bad. It's not going to be super, super accurate on its own. But it's like, it's a 24-bit SAR. What's the other one? SAR and Delta Sigma. It's a Delta Sigma.

Dave Jones: SAR, for those who don't know, is a successive approximation.

Chris Gammell: That's right. Yeah, that's the faster one. No, this is a Delta Sigma. It's a 24-bit Delta Sigma, but, like, super cheap. And you can get them at, like, you know, you can get them online at, like, from China directly, but, like, also, like, SparkFun uses them. And usually any of the, like, the kind of the strain gauge cheapo weight scale type of things uses those HX711. So, yeah, that is really cool. I was a little confused at first. I wasn't sure what you were. Oh, sorry. What you were pointing.

Dave Jones: Yeah, no, I thought it was a Segway article. I thought it was a related thing, but it wasn't. I was just, yep.

Chris Gammell: Got it. Yep. Sorry. Yeah, no, this is very, very clever in that way, you know, to be able to, you know, and, like, being able to put those sensors directly in line in the PCB definitely can cut down on cost. Yeah. It's kind of like, how low can you go, though?

Dave Jones: Yeah. Yeah, it's just novel. Like, you wouldn't do this in a proper product where you're actually trying to, you know, you want properly. That's right. But in the arity and accuracy and stuff, it's just not going to do the business. But, no, it's kind of neat. So. Yeah.

Chris Gammell: That's cool. It's cool.

Dave Jones: It's not the first time I've added a strain gauge to a PCB.

Chris Gammell: Tell me about other times.

Dave Jones: For measuring flex.

Chris Gammell: Oh, really? Oh, yeah.

Dave Jones: Yeah.

Dave Jones: Yeah. So, you would, yeah, we'd superglue strain gauges to PCBs inside to see if any flex is getting, is transferring from the outside of the product to the inside. Like, is the PCBs inside like a shell? And, yeah, we need to know, like, when the outside product bends, because it's actually designed to bend. And then we need to know if any of that's transferred inside. So, yeah.

Chris Gammell: Right. Right. I always think about, like, parts that have, or sorry, products that have intended, like, plug-unplug type cycles where you might be flexing a PCB or similar, even if it's, like, built into the case. And then, like, thinking about, like, the, what's that capacitor problem? Is it, like, the cracking that happens where sometimes, because it's a rigid structure, but the PCB flexes underneath it, it can, like, crack the ends.

Dave Jones: Ceramic capacitor, yeah, that happens in multi-layer ceramic capacitors, yeah, because they've got, you know, dozens or even hundreds of layers, and they're all rigidly connected to the end cans of that.

Chris Gammell: Yeah, the metal piece. The metal piece.

Dave Jones: And then that metal is soldered onto the PCB, and then you flex the PCB, and it pulls on those little, you know, even the tiny amount of flex, it just pulls on those end caps, and then it's brittle. Yeah. I mean, you've got to remember, ceramic is brittle. Yeah. Right? And if it cracks, like, it'll reach a stress point where it cracks, and then you're either going to get a shorter capacitor when it catches on fire, or you're going to get an open capacitor, and, well, that could also ruin your day.

Chris Gammell: So, yeah. Right. Right. Yep. Crazy.

Dave Jones: Yep. Especially in those, you know, 100 mic in, 100 mic cap in a bloody 0402 or something. And apparently, AVX is a spoiler alert. It's sitting on my mailbag shelf. I haven't opened it yet, of course, because I open them on air. But AVX, Ron Demko, former guest, he sent me some of their latest, like, I don't even remember, like, 002, 001 parts or something. Like, that's not it. Oh, no. But, like, the smallest parts they make. Yeah, yeah. And it's like, you know, like...

Chris Gammell: Getting out the microscope. Exactly.

Dave Jones: I'm literally going to have to put it under my decent microscope to see these things. It's just, yeah, nuts. So that'll be...

Chris Gammell: I did think about that. My buddy who's in purchasing, he told me, like, how... You know, when you think about, like, the kind of the center of mass, if you think about, like, the center of mass of all the people that are ordering capacitors, that has to be shifting. Apple orders more capacitors than I will ever even conceive of using in my lifetime, right? And so when you think about manufacturing then, and, like, the manufacturers surely are going to have more of whatever Apple wants versus whatever what Chris wants. You know, if I want 0805 with, like, some massive value because I'm, you know, doing some large-scale capacitance on a board, Apple doesn't care because they just need to bypass as small and as cheaply next to the part that they've put down, right? And so, like, but you think about, like, the parts that are in the ecosystem and then even the surplus from an Apple that might be able to supply the rest of the ecosystem, that it just moves away from the technology nodes that you and I are used to, the 0805s, the 0603s, and that there are just more and more parts at the small scales that I just... I personally will never use, you know? And so then the costs go up on the 0805s and 0603s because they might still have to make them for some people, but they don't want to because the volumes aren't there as much. So we pay a premium for being old and stochial. Yeah, right. That's what I'm trying to say.

Dave Jones: Well, some people avoid 0603s because they're valid. They even go 0805 or 0402. Is it too small? No, it's kind of like in that median, like, it's kind of like the middle child everyone forgets, you know, because, like, nobody cares about. Like, if you want decently small... I was the middle child, Dave. Oh, well, you were. I was the middle child. Okay, there you go.

Chris Gammell: Yeah. I have so many scars.

Dave Jones: Right, okay, some emotional, yep, scarring. Come on, Chris. Let us out. Let it all out. Come on. I just want to be an 0603 capacitor. I just want to be my true self.

Chris Gammell: Yeah, that's right. Oh, God. Yeah, and actually, another reason not to use 0603 is because of the confusion with 0603, 0201, how the metric imperial lineup happens.

Dave Jones: Oh, well, that's confusion over the board, isn't it? There's several overlaps in the imperial metric.

Chris Gammell: Oh, what are the other ones? I guess the one that I usually do is the 0603. That's what I usually... I usually order 0603 capacitors, and I'm like, wow, these are really tiny. They're actually 0201s. Oops. Yeah. Metric gets me again. Oh, well. Were there other... There's other crossovers, though?

Dave Jones: Yeah, I do believe there are other crossovers.

Chris Gammell: Oh, okay. Okay.

Dave Jones: You have to look at the list to get them on.

Chris Gammell: We need more decimals. That's what we should push for, more decimals.

Dave Jones: No, we just need the world to stick to bloody imperial, please.

Chris Gammell: I believe when you were talking about 0805s and 0603s, you were actually talking to... Wait. You said stick to imperial?

Dave Jones: Yeah. Stick to imperial.

Chris Gammell: Oh, stick to imperial. Oh, wow.

Dave Jones: When it comes to component sizes, 0603, yes.

Chris Gammell: Yeah, I was arguing against you. No. No. Uh-huh. Got it. Yes. You're a traditionalist. That's what you're trying to say.

Dave Jones: Same with PCB grids and... Yep. Yep. I'm happy to go metric with whole sizes and PCB dimensions and stuff like that, but nope. Otherwise, bugger off. Imperial. Thank you very much. All right. Yep.

Chris Gammell: I will die on that hill. Did you see there was a guy that developed a PCB tool and he was a US guy and he absolutely refused to do any kind of... He would only release it with imperial measurements and he had all these snarky responses about having metric measurements. I was like, oh, come on, man.

Dave Jones: Oh, yeah. No. No. It's too entrenched. I don't know. You know, the metric components. Yeah.

Chris Gammell: I think once all the boomers retire, that's what'll do it.

Dave Jones: Yeah. Because you know it by the number. It's not like, oh, it's... You don't think of it in terms of its actual dimensions because you already have the footprint. Right? The footprint's just like it's done. Right? And it's like, otherwise, you don't think about it. So, yeah. It's just annoying. Yep. I see no advantage to actually specifying them in metric because you get like your footprint is different. Right? And go and specify your bloody footprint in completely in metric if you want. I don't care. I just want a consistent numbering system for the actual component. Right?

Chris Gammell: Yeah. I agree with that. And also metric is fine with me. I prefer it for footprints. I switched to mostly metric for even trace widths and stuff because a lot of the PCB manufacturers are starting to do that. Not all of them. Right. But like the Chinese ones are doing metric and I'm like, all right, fine, whatever. You know, like, and I'll make the switch over, but yeah.

Dave Jones: I've done it too, but I still prefer. Yeah. My fourth hour tracks. Thank you very much. Yeah.

Chris Gammell: Yeah. Yeah. Yeah. Yeah. I think it's, it's, for me, it's, it's acted as a good, you know, I have to like stop and like ask myself if it's actually what I want. Right. When I'm going to order. And that actually is a good, it's a good slow down step. Any kind of step for that that slows me down is actually good. Cause then I'm like, oh, I should check my work.

Chris Gammell: Yeah. I found a bunch of that. Exactly.

Dave Jones: Yep. Yeah. Yeah. Is there a tool for that? That can like force your design to slow down every, you know, like once a day or something like that. So it forces you to go, oh damn.

Chris Gammell: I believe that's called a boss.

Dave Jones: That's called a boss. No, they usually come in and just give you a different task.

Chris Gammell: Well, you know, or they say, Dave, we need to talk.

Dave Jones: Oh, it's been a long time since that's happened to me.

Chris Gammell: Yeah. You haven't had a boss in a while, huh? You're ruled by the algorithm.

Dave Jones: No, no. I'm talking 12 years now, 11 years. Wow.

Chris Gammell: Wow. Wow. He's off the chain, folks. Yep. Totally. He can't be controlled.

Dave Jones: No, but for a good 25 years there, that was the thing, you know? So, yep.

Chris Gammell: Yeah. We are in the season of, what's it called? Yearly reviews. So.

Dave Jones: All right.

Chris Gammell: Any tips for people that are doing yearly reviews that are getting reviewed? What should they should? Oh, as in you're talking about job interview yearly reviews.

Dave Jones: Yearly review.

Chris Gammell: Yeah. Like job progress and like, you know, your boss is reviewing you for the year, maybe for bonuses or getting a promotion.

Dave Jones: Oh, the snarky side of me wants to say, if you get a job performance review and they start putting wank words in there, just leave. Sorry. But, you know, it's like.

Chris Gammell: Well, that works great unless you, you know, have a house payment, right? Yeah.

Dave Jones: You were hoping for something a bit more optimistic there.

Chris Gammell: No, not optimistic. Actually, I was just looking for stories. Oh, right. You know, just thinking about my own past reviews. I, you know, like when I think about, I never, you know, this may come as a surprise to some people, but I'm not that great of an engineer. And so I never like, nobody was ever like, oh yeah, you know, you did so well this year, Chris, we're giving you a promotion to, you know, next, you know, you're skipping a level and we want to give you so much more. Like I hear about that sort of thing. And, you know, that's not me. I'm getting a solid three out of fives across the board, which I don't know about in Aussie culture, but in like US culture, like it'd be like, you'd get kind of like the middling grades on, on reviews and then you, and then you just know that you're not, you're not moving up that year. Okay, fine. Whatever. Yep. Yeah. Is that how it was there?

Dave Jones: Well, you can probably guess my performance in my job reviews.

Chris Gammell: Dave does not do well with authority. Yes.

Dave Jones: That section was well known. Lacks tact.

Chris Gammell: Should be seen, but not heard. Exactly.

Dave Jones: That was pretty much the gist of all my performance reviews. Yep. It was like, yeah, that's why they locked me in the dungeon, you know? Yep. Like, and I'm not joking. Like that is a literal thing. I have photos of me in the dungeon. Like.

Chris Gammell: Yeah, I believe it. Yeah. You're basically a Milton. Yep. Yeah. Yeah. Yeah.

Dave Jones: I'm Milton in office space. Who gets. Yeah. Yeah.

Chris Gammell: Yeah.

Dave Jones: Totally. And then there's a photo of me surrounded. Like I was down in the dungeon and it's, there's a photo of me surrounded by like 20 chairs. Cause this is where they stored all their chairs as well. Excess chairs. And that's what. And that's what. Uh-huh.

Chris Gammell: Uh-huh. Yep. That's great.

Dave Jones: Yep. Oh boy. They tolerated me is the answer is, you know, they, okay, Dave, we'll tolerate you for another year.

Speaker ?: You know?

Chris Gammell: Try and do better.

Dave Jones: It's like. Yep. Oh God.

Chris Gammell: Yeah. I, so like when I talk about like, you know, trying to, you know, sell a solution to someone or like, uh, you know, in my current role, I'm trying to convince people to use the thing that I'm, I'm working on. I always talk about like making people look good for their boss. Yeah. And that's like a common thing that I, I mentioned because when I think about like those design reviews or sorry, those performance reviews, like that's really where the rubber hits the road. Like if I could walk in to a performance review like that and I could say like, Hey, I did this demo and that allowed us to, you know, make a hundred million more dollars last year or something like that. And then I say after that, I'm like, and I would like a 10% raise. Like if they say no, then you walk out the door because like, that's a very easy. And so like, when you think about like, that's an easy sell.

Dave Jones: Yeah.

Chris Gammell: Yeah, exactly. So like, but that's the tough thing about that is my experience in like hardware is that you're always so far away from the, you know, being able to prove that what you did, did a hundred million dollars in sales. If you can, if you can connect those dots and they don't give you that money.

Dave Jones: That's incredibly rare.

Chris Gammell: Yeah. No, that's, you know, it's, it's tough, right? Especially if you're in a big company where like you're doing a component of a component that gets exactly.

Dave Jones: And that's most engineering jobs. Let's be honest. Right. It's just one small cog in a giant, giant wheel. Yeah.

Chris Gammell: But I think about like, I think about in that case, like, so, okay, you can't prove your end result that you spent, you know, three months, like three solid months doing good work on, you know, like, or like you did all the testing so that you, you prove that, you know, the meantime between failures is going to be like a million years or whatever, you know, whatever you tested, you know, and you did a good job as like, that's, that's a true engineering job, right? Like long-term testing and, and, and proof and reporting on that sort of thing. So you can't, you know, you can do what you're expected to do on that kind of job. But then if you can somehow like walk into the meeting and be like, but now I can show that I did my job and I was more efficient at doing it. If you could have some kind of like demo to show that sort of thing. And like, you know, like, however you can show that like you provided value, maybe not at the end, a hundred million dollar, you know, increase in business, but like even doing your job a little faster, like, Hey, I pulled in mid journey images, so I don't have to go and talk to the art department. Right.

Dave Jones: I just made your art department redundant. Yeah.

Chris Gammell: Yeah. Unlikely, but you know, like that kind of thing. That's what I feel like you kind of have to do in order to, to grab attention during those, those time periods. It's just, it's tough. I mean.

Dave Jones: One tip I've got is to try and be like, you know, you, you've got your job, right? So you're in your department or whatever. Try and like, instead of sitting in the cafeteria at lunchtime, yapping away, right. Go to some other department and see if there's something that you can help out on or something like that. So when it comes around to review time, you can say, Hey, I was actually, you know, I've, I've helped out on all these other different departments. I'm like, you know, I'm, I'm like very useful across many.

Chris Gammell: I'm an invaluable connector between, between subgroups. Right. Yeah, exactly.

Dave Jones: I'm the glue that binds them together, you know, on the, yep.

Chris Gammell: That's right. Yeah. Yeah. I think the other tip that people, especially like young career engineers need to hear is you should leave your job. You should switch companies often because that's how you're going to get your biggest pay bump.

Dave Jones: If you're waiting for that 10% pay bump or 20% pay bump, the best way to do that is to

Chris Gammell: move companies. And it sucks because it means you have upheaval and, you know, you might not like your new

Speaker ?: job.

Dave Jones: Don't do it too frequently because it looks bad on your resume. Some people say two, two to three years is like.

Chris Gammell: I'd say three years. You got to take the first year to really know a place and like get in there. But like after three years, like, yeah, the best way to move. It sucks.

Dave Jones: If you've got, like, if I'm looking at somebody's resume and they've got, and I know we want somebody for a project, like a five year long project, which some of our projects were right. You know, because they like, they've got long tail stuff, right. We want someone to stay on for a long time. Then if I look at a resume and they've moved every two years, then I know that's pretty consistent pattern, you know?

Chris Gammell: Right. Well, the history is going to change now. I'm sure. Yeah. They're going to settle down. Yeah. Yeah.

Dave Jones: Yeah. And a year is like, no, that's like, that's red flag kind of stuff. You know, if you're moving every year. Yep. In fact, I would say like, if you've got a job that's like less than a year that you spend at, don't put it on your resume. Like, unless it's like, you know, like you got retrenched or something like that. And you can put the reason right up the top. I was retrenched, you know, or something, you know, the company folded or whatever, then that's a fine excuse. But if you simply left.

Chris Gammell: Right. Unless you're the reason that the company folded, then you shouldn't put that on there. That's it.

Dave Jones: Oh boy.

Chris Gammell: You know, that resistor value was too low. That company folded.

Dave Jones: We relied on chat. We relied on chat GPT. It said it could do it. Chat GPT. It's like, yep. Yep. Yep.

Chris Gammell: Turns out, turns out it cannot write pipe on very well.

Dave Jones: And I've got a beautiful segue for this. If you, you know, engineers being valuable. I had one too. I had one too. In that. Yeah. This is what thousands of Twitter engineers said. I'm so valuable. I'm so valuable.

Chris Gammell: Oh man. I knew you were going to go there. I know you're going to go there. You can read me like a book. Before we go there. I do want to go there.

Dave Jones: You have to acknowledge the quality of that segue though. Thank you very much.

Chris Gammell: That was a fantastic segue. You know what, Dave? Five out of five on your performance review.

Dave Jones: My performance review. Excellent. Yeah. Segway performance review.

Chris Gammell: Yeah. Five out of five. Before we do that, because we're talking about jobs, I, if people want to work with me, there is a job opening that would work directly with me. I would submit my resume tomorrow. Sorry. Your segues are great. Everything else. Yeah. Nah. Sales engineer and head of marketing. Probably the latter is not going to be listening, but I think a sales engineer might be. So I will link it. If you'd like to work with me.

Dave Jones: Oh, so the job is sales engineer. What?

Chris Gammell: It's like a FAE basically. It's listed as sales engineer, but it's like an FAE. Yeah. If I, if my job is.

Dave Jones: Dude, you, you, you, you totally, you almost totally lost everyone there by saying sales engineer. I know a sales engineer. You should have said FAE.

Chris Gammell: It's tough. It's tough.

Dave Jones: All right.

Chris Gammell: Yeah. FAE is, is close. Honestly, it is closer. In this case, like I'd be like an, Oh, sorry.

Dave Jones: There's a magpie attacking another magpie. Yep.

Chris Gammell: Australia.

Dave Jones: Yep.

Chris Gammell: Yeah. In this case, I would be like an application engineer and they would be like a field application engineer.

Dave Jones: Oh, okay. Does that mean you're getting away from the fieldy stuff now?

Chris Gammell: No, no, no. Like just that I'm kind of making references. Right. So like if you think about an application node, it's kind of like a reference design. That's what I'm doing. And then this would be like taking reference designs and then like deploying them to real world locations and like helping people actually make them into real things. Cool. Yeah. And it's still like feedback and like designing. There would still be some room for like designing reference designs as well, but probably more implementation of those things in the real world.

Dave Jones: So do they bypass HR and just send their resume straight to you? Of course they do.

Chris Gammell: Everyone knows that. Sure. Yeah. Everyone knows that. Yeah. I will put my, I will put my email address in the show notes. How about that?

Speaker ?: Woo.

Dave Jones: Last time I, I did a job. Last time I put up a job ad, I got like 500 resumes.

Chris Gammell: Oh my God. Yeah. But that was working directly with you and they didn't know what they were getting themselves

Dave Jones: into. Yeah. Yeah. Yeah.

Chris Gammell: Yeah. That's, I hope not that many. All right. Let's talk about how I owe you a bunch of chips probably.

Dave Jones: You owe me a bunch of chips. When was the date? Was it the 15th?

Chris Gammell: It was, I think actually the date is the release, the release date of this episode. Oh, okay.

Dave Jones: Right. Yeah. Is that, I don't think Twitter is going to fall.

Chris Gammell: The 15th is the day we released this episode. No. I don't think, I don't think four days is going to do it.

Dave Jones: No, no. Sorry. My Twitter, my, for those who want to know my, I've been gaining followers like crazy. So I've been, I've been gaining followers and the engagement is as solid as ever. So, and it hasn't, hasn't collapsed. Good for you. So, sorry, dude.

Chris Gammell: Twitter has not collapsed. I repeat, Twitter has not collapsed yet. I can still hold out hope. Right.

Dave Jones: Oh boy.

Chris Gammell: Yeah. So let me know. I will let you know. Remember it was a hundred dollars in chips. Do you want me to buy you these CH32s, the RISC-5s? Not yet. Because I am. How about any, any RISC-5s of your choice? Does it have to be a RISC-5? I think it should be a RISC-5. But you know what? It's, we'll just put, we'll just tag it out a hundred bucks.

Dave Jones: Well, I'll try and use a RISC-5 in my latest design, but it's kind of like overkill. I think a RISC-5 is probably overkill for projects I'm currently working on. But if it's 10 cents, you know.

Chris Gammell: Yeah. I think it's really, yeah. I think that's the real thing is the novelty of it in trying it out.

Dave Jones: Yep. Yep.

Chris Gammell: I do think a lot of people are, I saw Greg, our old friend, Greg Davil is working on this part and he had linked me some resources. I got pulled away to something else. There's still like the, the tool chain is still kind of underdeveloped. Versus like, you know, looking at other, other chips that are.

Dave Jones: Is that that Chinese one? Cause I just got that before I left. I was going to do a video before I left on it. You know, that little demo. Yeah, that is that one. Yeah, the CH32-003. Yep. Yep. I finally got it.

Chris Gammell: CHV. Oh, there it is. CH32V003F. Yep. So that's the, that's the part number.

Dave Jones: Yep.

Chris Gammell: Yeah. So you and I have that same thing.

Dave Jones: Yep.

Chris Gammell: Have you actually? No, that's, I got pulled off on other things.

Dave Jones: So yeah, I'm, I'm going to use, that's probably the first thing I do. When I get back is have a play around with that. Yeah. Okay. Going to install it. Cause I actually downloaded all the tools, but I didn't have, they've, they've, they've got their own compiler and stuff like that. It's well, it's from, it's on a different website, but they're affiliated with it or something. So.

Chris Gammell: That's right. Yep. Moon, moon, moon, moon, moon, moon, moon, moon, moon, moon, moon, moon, moon, moon,

Dave Jones: moon, moon, moon, moon, moon, moon, moon, moon, moon, moon. Yeah. Yeah. Yeah. Which is kind of cool name. You know? That's not bad.

Speaker ?: Yeah.

Dave Jones: Maybe they asked our chat, uh, GPT, what is a cool English name that would Western English name?

Chris Gammell: Oh man, that would be much better. I swear. The more I look at like Amazon products, you know, that are just coming straight from China. It's just like, they're just choosing the smallest number of letters they can get away with. Right. And then they, they have no, like, they don't even make sounds. It's crazy. Like I've seen some real losers. Oh boy. Let's, let's see though. Let's see what chat GPT says. Oh, okay. What is a good name for a, a compiler company?

Dave Jones: For a compiler company. Yes. Ask it. I don't have a tab. I don't have a chat GPT tab open. Hang on. Chat. There we go.

Chris Gammell: GPT. Okay. Try. I don't, it doesn't, uh.

Dave Jones: Oh, hang on. I have to verify that I'm human first. Yes. Ironically, I've got to verify that I'm human.

Chris Gammell: Yeah. What is a good name for an embedded compiler company? Oh, here we go. A good name for an embedded compiler company would be something like firmware forge or embed compiler technologies or micromate compiler, which effectively communicates the company's focus on embedded systems and compilers. Not bad answer chat GPT.

Dave Jones: That's all right. Yeah.

Chris Gammell: Did you see my, my, uh, my screenshot? I asked, I asked about how to do a layout in KiCad and it referred me to my own video series.

Dave Jones: Oh yes. I saw that. Yeah. I thought that was funny. Yeah. Yeah. Yeah. You've, you've, you've finally come full circle. Yep. Yeah. I made it. You've made it.

Chris Gammell: Ma, look, ma, I made it. Yeah.

Dave Jones: The, the, the AI knows me. The AI recognizes me. That's right.

Chris Gammell: Yeah. Yep. Yep.

Dave Jones: Yep.

Chris Gammell: All right. Well, we should probably end on AI, right? Or do you have other things? Yeah.

Dave Jones: Well, that's, that's one of the things is, um, because one of the, one of the reason that spawned my video on the topic was that a, um, yeah. A, uh, satire company were now taking their content behind a paywall because they don't

Chris Gammell: want to feed the AI engine.

Dave Jones: And that, you know, like will other, I'm sure there will be other companies that will do the same thing. They'll go, well, I'm not going to feed it for free. Like, yeah, they can try and steal it from going around the paywall and all that sort of thing. You know, it's not, you know, it's possible. Right. But at least I'm not going to give it up for free. Does this mean a lot of people won't be giving up their source code? They won't want to, you know, put their source code on, uh, you know, open source anymore. Like I'm sure there are people out there who will be thinking that. Right.

Chris Gammell: I'm sure there are. I don't think it'll have. I don't think it'll have a big impact.

Dave Jones: No, but you know, it's like.

Chris Gammell: I think the other thing too, is that you think about like the things that are feeding back into these engines are often, you know, so like every time you use an AI model, you're training an AI model as well. Right. So you think about like the, what's the, the GitHub code mate or whatever it's called. It's like the copilot. That's it. So GitHub copilot, you're basically, it's like kind of doing like almost like autocomplete based on whatever you're writing. Right. And, you know, mixed reviews, but like generally pretty amazing stuff. And that like, it's, it's a, it's a, it's an enhancer, right? It basically, it helps you to go faster. You're still kind of creating the ideas, but like you're, you know, you're not as dependent on format and structure and because it's kind of just auto completing for you. Well, you're also training it. Right. And like, so if you want to utilize these things, you're also training these things. And so, yeah, I mean, people are going to, you know, the source material that was out in the public domain might, might shrink over time, but I think it's really the, the usage of it that will, and the feedback mechanism around it that ultimately will drive the most growth into it. I think.

Dave Jones: Yeah. I just used the engine wrong. I said, what is a good compiler company name? And it told me there are many reputable companies that produce compilers. Some well-known examples include GNU, LLMV. Yeah. Never heard of that.

Chris Gammell: Yeah.

Dave Jones: All N, C++, Clang. Clang.

Chris Gammell: More than C++. Nice.

Chris Gammell: More than C++. Clang. Okay.

Dave Jones: Yeah. Yeah.

Chris Gammell: Yeah.

Dave Jones: Yeah. There you go.

Chris Gammell: Did they, did it say GNU is a good name? GNU is a terrible name. Yeah. Yeah. Yeah. It's terrible. Totally. Yep. Yeah.

Dave Jones: But see, it didn't understand what I was, what I actually wanted. I needed to, you know, prompt, I needed to do some prompt engineering, which is the term.

Chris Gammell: Prompt engineering. That's right.

Dave Jones: Prompt engineering to, you know, sorry, sorry. Now I have to use the word generate. So I'm going to generate a compiler company name. No. See if we can do that. Okay. Generate a compiler company name. Progressive compiler technologies. Streamline compiler systems. Innovative compiler works. Oh, it's just thrown in the wank words. Optimized compiler solutions. Precision compiler dynamics. Yeah. Efficient compiler creations. Nimble compiler innovations. Oh, this is, this is wankery. Futuristic compiler core. This is, this is like Douglas Adams kind of like.

Chris Gammell: And you know what the thing is though? Like they're going off all of these terrible names. Yeah, it is. It is. Yeah. Yeah. What's the, what's the name of the robot company? The, the robot company in, in Hitchhiker's Guide. The one with the doors that go genuine people personality.

Dave Jones: Oh, don't, don't put me on.

Chris Gammell: Ask the robot Dave. It's right there. It's right there.

Dave Jones: Next gen compiler technology. Serious. Serious cybernetics. Serious cybernetics. Cybernetics. That's it.

Chris Gammell: Yeah. Yeah.

Dave Jones: Yeah. Advanced.

Chris Gammell: There's a, there's a. This is terrible. There is a Pete Holmes thing that I'll link in here. Pete Holmes is a comedian. And he talks about like back in the days when we didn't have Google in our pockets and now, you know, chat GPT in our pockets. Right. Yeah. And just like the amount of joy that it would bring when you like remembered where a musician was from. It's a really good standup bit. I'll, I'll get into.

Dave Jones: I'm going to, apparently chat GTP is good with follow-up questions. So I'm going to put more please, but less cringey. Yeah.

Chris Gammell: Okay. Yeah.

Dave Jones: Okay. I'm going to put less cringey.

Chris Gammell: Dave, we could just, we could just do this all day. We could just start up a new podcast, live chat, live. New podcast. Dave, Dave and Chris just put crap into an AI engine and see what comes up.

Dave Jones: Sure. Here are some more compiler company names that might be less cringey.

Chris Gammell: Just like our audio as well, right?

Dave Jones: Here we go. It did actually do it. That might be less cringey. Here we go. See, see if you agree. CodeWrite compilers. Cortex compilers. Don't mind that one.

Chris Gammell: Nope. Cogent compiler systems. That's going to be.

Dave Jones: Vertex compiler technologies. Blue river. We're getting close to the moon river. Blue river compiler solutions. That's all right.

Chris Gammell: I like the Cogent one. Cogent wasn't a bad name.

Dave Jones: Yeah. Yeah. Cogent compiler systems.

Chris Gammell: I mean, when you go through like how people actually name things anyways, this isn't that different. No, no, no, no. Exactly.

Dave Jones: You throw random shit at a wall and see what sticks. Yeah. And then you sort of mix and match. Oh, I like that word, but I don't like it mixed with this other word. Exactly. Yeah.

Chris Gammell: Yeah.

Dave Jones: Yep. Hyperion compiler industries. Zenith compiler solutions. Oh God. Yeah. Radiance compiler technologies. OptiMax compiler. See, you know, it like.

Chris Gammell: I mean, Hooper starts a company just based on a compiler though too, you know, like that's a bad business.

Dave Jones: Oh boy. Anyway, what it can't do is company logos. I have tried. It cannot do company. Mid journey cannot do company logos. I have not been able to get it to even come close to even given a simple logo. Have you tried like, uh, try like line art.

Chris Gammell: So like give it a. Okay. Line. Okay. I can't do that. Line art.

Dave Jones: Yeah. Line art. Okay. Yeah. Yeah. Yeah.

Chris Gammell: Although otherwise it thinks it's a vector vector drawings. Okay. So like simple shapes, vector drawings. But like, again, that's like, if you don't know.

Dave Jones: That's your prompt engineering skills. I mean. Yeah. Yeah, exactly.

Chris Gammell: Yeah. Yeah.

Dave Jones: Yep.

Chris Gammell: It's, it's, uh, it's a funky new world. We're living in Dave. Indeed. But for now people are stuck with the meat robots known as David, Chris. So our stupid brains, whatever we can output.

Dave Jones: But we have some prompt engineering skills and people are laughing at us for building up our prompt engineering skills. I don't know. I think it opens a lot of doors. So, yep. Even if you don't, you know, if you're a, you know, if you're a freelancer or you're just a Joe blogs engineer at work, having access to, you know, being able to have those killer prompt engineering skills might, might make you look like a genius.

Chris Gammell: Might. Yeah.

Dave Jones: Can I give you a, I think I might've mentioned this way, way in the past, but, uh, I can't believe like all management at my company thought I was an absolute God when I started to take images like photos with a, uh, Sony, you know, the floppy drive, the Sony Vika camera, which I've done a video on. I started taking photos and then you got a six 40 by 400 JPEG out of that. Right. It was, that was the highest res you got out of it. Right. But then I started and then I put the floppy drive into the computer and then I'd insert images into word documents as part of doc as part of the, uh, documentation for like, you know, test, like, you know, test procedures and stuff like that. They thought I was a God and all I was doing was like, you were just, you were just, yeah.

Chris Gammell: And that's, that's a really good point. Actually. You were just leading edge by what? Like a couple of years, but like that's leading edge does make you look really good. Yeah. And you know, you say that they put you in the basement. It was just actually probably just to make sure you didn't make them look bad. Right. Exactly.

Dave Jones: But yeah, like seriously, like, oh my God, the pats on the back. I got, I was like, yeah, they were bowing down. Like, you know, Dave is the document God. Right. You know? And yeah. Right. Unbelievable. And then eventually it became normal. Eventually. That's a great, that's a great example. Yeah. I mean, this was in the age when the company was still doing stuff on a grid paper and we would actually print out something, get the glue. We'd actually stick it on the grid paper and then we'd photocopy the A3 grid sheets and then we'd have backup documents. You know, it was all. Awesome. It was all still manual. Yeah. Yeah. I was God-like. Yeah. Yep. Those were the days.

Chris Gammell: Those were the days.

Dave Jones: Yeah. But you can be God-like now. I'm telling you. That's right. Right?

Chris Gammell: Mid-journey.

Dave Jones: Seriously. I could just like set up a new Twitter account and just, I could picture myself as an, you know, I could market myself as an artist and I could put up all these funky, you know, art and people would think I'm a, you know.

Chris Gammell: I do wonder, you know, like there's a lot of like street fairs around here.

Chris Gammell: Like in the summer when it's nice out. Walking around.

Dave Jones: You could sell art. You could stand outside and you could sell all this art that you've generated, you know.

Chris Gammell: I've printed one. Oh, no. Really? I think I told you that. I printed a 40 by 30 inch like huge canvas.

Dave Jones: What are res? Because you've got the top pro account for mid-journey.

Chris Gammell: Yeah.

Dave Jones: What resolution does it give you?

Chris Gammell: The first feedback my wife gave me was, oh, it's kind of blurry, huh? And I was like, yeah. Yeah, right. It was way too big for that. I should have done some more. Yeah. But it was already kind of blocked. Like the style was kind of blocked anyways.

Dave Jones: Right.

Chris Gammell: But yeah, it's only 1500 by 1000. Okay. Right. Yeah.

Dave Jones: Yeah.

Chris Gammell: So yeah. Got to do a little post, a little post editing. All right, man. I got to go. I got to go generate some, some graphics, I think here.

Chris Gammell: Enjoy. Good show about AI.

Dave Jones: Yeah. We didn't intend it to. I don't know what it's supposed to be. I thought we were done with that. But no. No. Yeah. No.

Chris Gammell: It's coming back. It's coming back in a big way. 2023. All right. See you.

Dave Jones: Bye. Bye.

Speaker ?: ! x x x Thank you.