#349 – An(other) Interview with Jon Oxer

Chris Gammell: This is The Amp Hour Podcast. Released June 25th, 2017. Episode 349. Another interview with John Oxer. Welcome to The Amp Hour. I'm Chris Gammell of Contextual Electronics.

Jon Oxer: And I'm Jonathan Oxer from Superhouse and Freetronics. And just because we don't have Dave here, I've got to say it. What's up, nerd?

Chris Gammell: Not much. Not much. Classic Aussie sound. I guess we have a replacement Aussie this week.

Jon Oxer: Yeah, I can be the token Aussie. You need one.

Chris Gammell: Yeah, well, let's hear your best Aussie-ism.

Jon Oxer: My best Aussie-ism. I'm actually not very Aussie at all. In fact, my secret is that I was born in London.

Chris Gammell: I didn't know that.

Jon Oxer: No. I was born to Australian parents who were living in England at the time. And I came back to Australia when I was really little, like not even one year old.

Chris Gammell: Did you get dual citizenship or no?

Jon Oxer: I think I qualify for it, but I haven't gone through the process of applying for a British passport.

Chris Gammell: Yeah. Well, these days, too, with Brexit, who cares? Yeah.

Jon Oxer: Anyway. Yeah, it doesn't give us the benefits it used to. That's right. Yeah.

Chris Gammell: Well, welcome back, man.

Jon Oxer: How you been? Yeah, thanks. I'm really good. Yeah. We were just saying before the show that it was 2012 when I was last on, and it doesn't seem that long at all.

Chris Gammell: I'm looking at the picture, too. And you had hair in this picture, and you don't have hair anymore. Yeah, not anymore. I'm more streamlined now. Yes, that's right. You're aerodynamic.

Jon Oxer: Yeah. Yeah, there's actually one of a strange reason why I did that, among others. One of them is that when I've been filming videos for Superhouse, I've been recording them in little snippets. And the funny thing was that I did a video where I started the recording, I think it was back in like 2010. And then I finished that episode in like 2016 and edited it together. And partway through the video, I make it go blank. And I just have a thing on the screen that says six years later or whatever. So you can't rush some projects. But I looked so radically different between those two. And I figured, hey, if I just keep my hair clipped all the time, then I'll look the same. And no one will know if it takes me six years to finish a project.

Chris Gammell: Yeah, man. The only thing that'll be different is if you have a tan from the summer or whatever. Yeah, that's right. Right. Well, yeah, that's a good plan. That's a good plan. Yeah. So yeah, you're filling in for Dave today. Thanks for doing that. Dave's a slack ass like usual. Oh, of course. Yeah. But you have some news, huh?

Jon Oxer: Yeah, I do. So let's hear about it. Sure. So as listeners may know, I've had several businesses. There is Freetronics, of course, and Superhouse. But I've also had a software company. And just recently, we sold the company actually to our biggest competitor. And just because of the way the sale worked out and the circumstances around it that I won't go into in detail here, the end result is that my net position is negative. So we ended up with a very big tax bill. So like seven figure tax bill to pay, that sort of thing, which is quite scary. So even though I've just sold a company, and technically you can say, oh, I've sold a company for millions of dollars. It's not like I'm sleeping on a pile of cash. So I've had to start looking around at other work. And just recently, I think it was only about four nights ago, I was sitting on the couch with my laptop. And I thought, I'm going to see what jobs are around for engineers. And I went to one of the local job seeker sites. And I just started looking down, you know, entry level positions and things. And there were all they have all these requirements, like they'll say, five years of Python experience or whatever. And they're all quite focused on a particular area. So they have a specific need, they want someone that is going to fulfill that specific need. And that's totally understandable. But my particular circumstance is that I didn't actually finish university. And my experience since that time has been very broad. I've worked in hardware and in software and associated things as well. So as I was looking down this list, I found that even for entry level positions, I couldn't tick any of the boxes. Like I couldn't say, yeah, I am a great Python developer or I can do, you know, kernel development or whatever. So I got frustrated and I thought, I don't qualify for any of these positions. And I've been doing this sort of work for so long. What's going on? And I ended up writing a Facebook post that was a little bit tongue in cheek. Basically, what I did was I listed a whole bunch of different things that I've done over the years. So I said, you know, I've worked on satellites and I've written five books. And I've designed boards that have been used by hundreds of thousands of people and, you know, all of these things. And yet when I go and look at entry level positions, I don't qualify for any of them. Right. And it was basically just, you know, it was a throwaway post. It was me venting. It wasn't meant to be anything really big and deep. But it seems to have struck a chord. And I've had a few different types of reactions to it. There are many, many comments on that post now.

Chris Gammell: I also read the Y Combinator responses. Yep. Those were less friendly than the friends on Facebook. Oh, yeah. Yeah, definitely. Yeah.

Jon Oxer: So I know the reality is that once you are in the industry for a while, you're not a recent graduate. People aren't going to be looking at you and saying, you know, what was your score in this particular class? You know, that sort of thing isn't relevant anymore. It's much more about the sorts of jobs you've done, the experience you have. And I'm well aware of that. So those checkboxes on job ads are really a bit of a filter that are used by recruiters to try to match up people to positions. And they don't necessarily understand the roles themselves. They're acting as a conduit.

Chris Gammell: Right, right. I would say the other thing, too, that's interesting about all this is, like, you know, they do this, like you said, it's a filter mechanism and that kind of thing. But really what people want is, like, it's kind of indicative of the job industry or the job market, rather, too, that they want something that could solve a problem right now. And that's fine. Like, that's what they're looking for. But oftentimes, like, that inflexibility means that they just keep the thing open forever, whereas someone who could actually go in and figure it out, you know, like, without all the experience, then they don't necessarily – they miss out on a lot of people.

Jon Oxer: Yeah, so they're missing out on opportunities for people that could solve the problem but not necessarily in an optimal way. Right. Yeah. So I've had a few people give me some snarky responses saying, well, you totally don't understand the whole job hunting process and that sort of thing.

Chris Gammell: That's probably true, though, right?

Jon Oxer: And it probably is because the reality is that I haven't worked for anyone else for something like 25 years. I haven't had to go and do interviews and people have said, oh, so send me a resume. It's like, I don't have a resume. Like, I haven't – my resume is all online. It's the projects that I've done.

Chris Gammell: Right. This is kind of terrifying to me. And I'm sure that a lot of people in the audience can relate to this, too, where it's just like this is – a lot of people are like this now, you know? It's not like – it's not like – a lot of people don't have this defined path that, like, these jobs look for. And so I think the only difference is that normally you would just go and start another thing, right? I mean, that's really the only difference.

Jon Oxer: Yeah. And in fact, I do have other things. Like, I've got Freetronics, which is still going on in Superhouse. But the reality is that those don't have big margins in them. They're things I'm really doing for fun. Like, Freetronics doesn't pay me a wage, so I can't just sit back and live on that. And I need some other way to bring in income. And so, yeah, it is an interesting position to be in. It's kind of daunting. And I've definitely gone through the whole imposter syndrome process as well. And I think that's partly because I've been – anyone who works in engineering probably has this to some degree.

Chris Gammell: Oh, yeah.

Jon Oxer: I mean, there are probably people out there that have an overblown sense of their own abilities and big egos and things like that. But I think for most of us, we look at our own work and measure it against what we see other people doing and think, no, I can't stack up to that.

Chris Gammell: And so – Well, usually we also compare it to the finished product, which is what we see on the internet. Yeah, that's right. That's the worst thing. Exactly. You don't see the process. You see the end result.

Jon Oxer: Yes, that's right. And so we judge ourselves against the cream of the crop, like the very best polished result, the one that actually made it out. Not a lot of the failures along the way. So – and because I've been – I haven't been working in industry like for other companies over that time, I haven't seen a lot of that process. So, yeah, I've had a whole lot of people come back to me and say, you don't know your own value. And that's probably true. And so I've got to overcome some of my own internal resistance as well. And this is also just partly my nature because I'm kind of a shy person and –

Chris Gammell: Oh, yeah, super shy. Yeah, yeah.

Jon Oxer: And I don't know why, but I've always been down on myself in general. So I will do something and other people will say, hey, that's really cool. And I'll go, oh, well, it took me like 27 tries to get there. So I focus on the 27 failures and not the one that got there at the end. Right, right. Yeah, I think it is a thing that a lot of people have to overcome. It's about valuing ourselves. Of course.

Chris Gammell: It's an interesting conversation in 2017 too of just like – well, so like a lot of things you're looking at are like these kind of – these specific job listings, right? And what your post basically says is like you're the – not the everyman, but like you can do a lot of different things. And that is necessary for a lot of posts, but a lot of people won't list that kind of thing, right?

Jon Oxer: Yeah. And one of the things that we're often told when applying for jobs, like the general advice is if you look at a job that you're going for, then tailor your application to highlight the particular attributes they're looking for. Yeah, right, right. And that's the exact opposite of what I did in this Facebook post, which is just lay out, oh, I've done all of these wildly different things. And therefore, I can address myself to a whole lot of different areas. So I think if you're applying for a specific job, then yes, you definitely need to tailor your approach to it.

Chris Gammell: But in my particular case – Right, because then like you're saying here too, it's like you've – you know, you design boards. So if you're applying for an EE position, right? It's like you could talk about that one thing, but then it's like that's, you know, what, 5% of your time, 10% of your time, whatever you were doing it for. And so even just in a time perspective, it's crazy. That's right.

Jon Oxer: And it also means that if I'm applying for a job designing boards and I was going up against someone like Dave Jones who has spent years with that as his sole job –

Chris Gammell: Dave does not design boards anymore. Let's be honest. Not anymore, but he did.

Jon Oxer: Like when he was in previous jobs, that was his sole job was laying out boards. And someone who has done that job solely full-time for a couple of years, they're going to whip my ass basically.

Chris Gammell: Of course.

Jon Oxer: But I can probably do other things that they can't do. So yeah, it's this whole tension.

Chris Gammell: That's the other thing that's interesting to me is that like – so like I think about like an electrical engineer, especially because you say you're looking at entry level, which is kind of silly I think. But like so thinking about electrical engineering. So you spend 5% of your time building boards, let's say. Like that represents the amount of time that you've spent on Freetronics and everything else. Like that's pretty close to how much time an EE will spend building boards most of the time too. Right? Because it's like most jobs are mostly meetings and emails and talking to distributors and relationship management, all that stuff. But again, that doesn't ever get highlighted. It's not like show me a portfolio of your emails, please, when you come in for an EE interview. Yes. And yet that stuff is so important. Like knowing that you're supposed to go and design – or sorry, go and talk to a vendor, you know, six months before you're actually putting a thing out into the world, right? They never ask that in an interview, but yet that's a skill.

Jon Oxer: That's a critical piece of knowledge, yes.

Chris Gammell: Exactly. It's something you learn over time that is like – that can literally save the company millions of dollars.

Jon Oxer: Yeah. Yeah. That's right. So, I don't know. These are the sorts of lessons that you learn the hard way once you actually get experience and work on real projects.

Chris Gammell: Right. So, it's almost like an indictment of job listings more than anything, right?

Jon Oxer: Yeah. But I don't really know how to fix that problem. It was – and you're right, I probably shouldn't be looking at entry-level jobs. But it just came about from this evening, that particular evening, sitting on the couch watching TV, pulled out the laptop and thought, oh, what jobs are there? Yeah. Yeah, exactly. I think this was just my reaction to that. But the reality is that I know a lot of people. And I suppose one of the interesting things to come out of this is – that can be useful for other people to know – is that you and Dave have talked in the past about showing real products or projects that you've worked on. Yeah. And doing things like maintaining an online portfolio. So, if you work on a project, if you can, depending on your employment arrangements, then talk about it online. Make it public. And the thing is that because I've worked largely on open source for all of this time, and I kind of like to show people what I do, if you start looking around at the things that I've put online, there is quite a big body of work there. Now, it's not necessarily any bigger than anyone else. It's just that it's visible because I've been talking about it. Yeah.

Chris Gammell: Exactly.

Jon Oxer: And so, getting into that habit of posting about projects that you're doing along the way, it's not necessarily going to give you any immediate benefit right now. But five years down the track, when you're looking for a job, you can go to a prospective employer and say, look at this track record I've got. I've worked on all of these different things. Right. And they can see it.

Chris Gammell: It's not just a LinkedIn page with years you've worked at a place. Exactly. It's actual real things. Yeah. That's right.

Jon Oxer: And it also means that you end up known by many more people. They start to recognize your name. Right. That's one of the other – one of the strange comments I saw, I think it was on Hacker News, was someone commenting about, oh, I've seen his work and I know his name. That's all the qualifications I need or something like that. Right. And it was just because I've been visible. Excuse me. Because I've been visible about the projects I've worked on.

Chris Gammell: Right. Exactly. Yeah. Yeah. That's – yeah, I wasn't actually worried when I saw your post. I don't think many people who know you were, so. No. Yeah. Don't give this man a job, guys. I mean, like, don't – yeah, he's going to be fine.

Jon Oxer: He's going to be fine. Yeah. I'll be fine. And the reality is I know I'll be fine. It's – yeah. And actually, one of the other interesting aspects to come out of this is the issue of relocation. So, this is probably a big topic for a lot of people as well. Oh, that is interesting. Yeah. My initial reaction – Are you guys thinking about that? No. And that is a problem as far as job prospects are concerned. So –

Chris Gammell: Well, you are in the backwaters of the world. I don't know if you know that.

Jon Oxer: I'm right at the bottom, yes. I'm in Melbourne, Australia.

Chris Gammell: Right. So, I'm not just – You can go a little further south, but you have pretty much one place you can work. That's right.

Jon Oxer: I can go to Tasmania, which is a beautiful place.

Chris Gammell: Oh, actually, I meant the Antarctica Science Station.

Jon Oxer: The research base, yeah.

Chris Gammell: Yeah, right.

Jon Oxer: Yeah, I've known people who have done that. It does sound interesting.

Chris Gammell: Yep.

Jon Oxer: So, my initial reaction to these job postings was the issue of meeting those checkboxes. Like, can you say you've got five years of Python experience? And that's one aspect to it. But the other aspect to it is a lot of engineering jobs require you to be located in a certain place.

Chris Gammell: Yeah. Now, over the years – And not only is the nicest place either. No, that's right.

Jon Oxer: And so, over the years, I've spoken to many companies about possibly working for them. And they're all, you know, the big names that everybody would recognize. But they've all required relocation. For example, if you go and work for Google, then you're probably going to end up in California or maybe Dublin. They do have a Sydney office here as well.

Chris Gammell: They have a Chicago office too. Yeah. Yeah.

Jon Oxer: But if you start looking at these big companies, they have almost no presence in Melbourne. So, Melbourne is basically a desert as far as tech companies go. There are a couple of exceptions. There is a company like Blackmagic Design, which is very – Oh, cool.

Chris Gammell: The camera people.

Jon Oxer: Yeah, exactly. So, they are based right here in Melbourne. And they've got manufacturing facilities physically here in Melbourne, which is incredible. Right.

Chris Gammell: They can afford to manufacture there because they charge $5,000 for camera. That's right.

Jon Oxer: So, that's right at the top end. So, there are a few exceptions like that. But almost universally, if you start talking to, say, Google or Facebook or someone like that, they will say, yeah, we can hire you. And we will relocate you to so-and-so. Now, for many people, that's fine. Particularly, if you're earlier in your career, that is a brilliant opportunity to go for. And it's also really good to travel. You get to see new things. In my particular circumstance, I have two kids that are in school. So, they're in year 7 and year 11. Oh, yeah. That's right. And it's a critical stage of their education. So, we really can't afford to take them out of school right now, move them around the world. So, it doesn't matter how much they would offer to pay me. It's just not going to happen.

Chris Gammell: Yeah. It's just a lifestyle thing, right?

Jon Oxer: Yeah. That's right. And so, what we've seen is for many software engineering positions, it is actually possible to work remotely. But for electronics, it's much harder.

Chris Gammell: Well, and that's another interesting point. It's just hardware in general is hard to work remotely. Right. It's like, unless you're a prototyper kind of like out on your own, doing your own thing, it's like, even that, it's like, oh, I guess I'll ship you the prototype. Yes. That is, there is just not, it's not like there's like a band of digital nomads who are EEs. Yeah. As much as I love the idea of it, and I've talked about portal labs forever, like it's just, there's nothing like a lab. You know, you got to have a lab with you. That's right. Well, you got to have a lab somewhere.

Jon Oxer: Yeah, you've got to have that. And so, in my particular situation, because I am super lucky in terms of my home lab. We did a renovation a couple of years ago, and I took the opportunity to build what is essentially a four-car garage. And the front half is for two cars, and the back half is my workshop. It's for nerds. It's, yeah, it is the ultimate nerd cave. So, I've got a fantastic setup here, and I've got, so I've got all the basic tools, but I've also got things like a laser cutter, a 3D printer. I've got the little pick and place machine that I made. I've got a little photo studio here, so I can take photos of products. Basically, I can fabricate anything I like, just sitting right here in this room. But you really need to be physically located with someone for a lot of jobs, and that is definitely a barrier.

Chris Gammell: Yeah, well, what about like, I mean, consulting and stuff like that? I mean, I would think that would be a pretty natural thing. I do know some people who do that. I mean, like, well, our sworn enemies over at Embedded FM, they're consultants, even though they are in the embedded realm. I think that there's other hardware people that do that same kind of thing. So, what about that kind of thing now?

Jon Oxer: Yeah, yeah, that's true. That would be a good possibility. So, that may well be what I end up doing. Part of that as well is learning about the soft skills around consulting, and that's something that I haven't done before.

Chris Gammell: I get to mention one of my few favorite pages on the internet. Oh, please. Which is unixwiz.be. I will always remember that. Yes. There's, it's like, no, of course, that's actually wrong. It's like, basically, this guy's actually a, he's a server-based consultant, right? So, he's obviously doing Unix-based administration. Mm-hmm. Mm-hmm. But, but basically, he talks about like the, the warm, fuzzy feeling. What's, oh man. Oh, sorry. Unixwiz.net. So, you want to be a consultant. And yeah. So, I'll put that in the show notes here. I guess I can send it over to you, even though it'll, it's going to tell, our lovely mumble, telling, telling John that there was a link being sent through. Oh, yes.

Jon Oxer: I wonder if there's a whack and mute that probably is. Oh, it's fine.

Chris Gammell: It's fine. Yeah. So, basically though, it's, why work eight hours a day for someone else when you work six, any 16 you want for yourself?

Jon Oxer: Exactly. Yeah.

Chris Gammell: So, this is, this is just a great, this is a great document for anyone. I've recommended this before on the show, I think. Mm-hmm. It's about, you know, how do you give that warm, fuzzy feeling? How do you, you know, like, billing hours, being consistent, like all that kind of stuff. Yes. And, and obviously, you know, a big part of that is marketing yourself too, right? Obviously, you've talked about, you've done a lot of that over the years of just showing that you can build things and you can build things for other people, but then you also have to get the word out that you're building things, right? Obviously, if you're on a podcast, maybe it would be a good, a good way to do that. Yes.

Jon Oxer: But you are known as the place to go to get a job.

Chris Gammell: Oh, is, oh, I didn't know that. All right. Yeah, well. Well, send, send the, send the $30,000 recruitment checks over to Chris Gammell, not Dave Jones. No, no, no. Yeah. In Chicago. Yep. U.S. dollars only. That's right. Yes. None of that funny money. Yeah. Yeah. So, yeah, I mean, that, that sounds like a great option though. Yeah. It sounds like something that would be normal for someone in your situation where you've, you've had the experience. It's, you know, like, like past guests too, right? So we were, we were talking about Jack Ansel before the show, but like Jerry and Sean and like people that are non-traditional educational routes where the, that stuff was shut off to them, but they still were doing interesting work. Right. So, yes.

Jon Oxer: Yeah. Um, and, um, I do know a few people that have done, that have been consultants for many years. That's their primary mode of operation. And, um, one of my friends who does that has said that he'd like to have a chat to me next week to talk about some of the, the process around it, how you go about pitching for work and how you, you know, how you report your work back to them and that sort of thing. And so there are a whole lot of things that, uh, I probably don't have the experience that I need, but I can pick those skills up along the way. I think so. Wrap those around my technical skills.

Chris Gammell: I have a couple of friends who do it too. And it's, it's, it's really interesting hearing how it goes. Like, I think that, you know, like you talked about imposter syndrome and not valuing your work, but there's actually like a very specific version of that where people don't charge enough for their hourly. Oh yes. And I've watched my friend do that. And basically the only time that he ever raises his rates, uh, is when he just has too much work. But that is, that is, that is the pressure valve, right? It's like, well, I got people asking for 80 hours of my time. I've only got 40 billable, you know, like that's probably still 60 hours of work. Yes. Um, and it's like, yeah, you got to put the rates up, you know, and it, and it stinks because sometimes you want to help people that are doing interesting work. You want to charge them less, whatever, but it's, it's simultaneously like you need to value yourself and then you also need to stave off future work, right? By saying, well, this is my time. And he's, he's devised some interesting ways around it.

Jon Oxer: Okay. And I think one of the other pressure valves is, uh, fudging the hours and not charging for all the time you spend.

Chris Gammell: Right. If you're learning something you're saying.

Jon Oxer: That's right. So I think that partly comes into the imposter syndrome as well, because you might spend 20 hours working on something and think someone smarter might've done this in 10 hours. So you build 10 hours instead. Right. And, uh, you effectively. Well, that's not always the right call.

Chris Gammell: Yeah. That's not always the right call. I don't think, but actually that's covered in this Unix was thing too. No, I don't think it is the right call. Yeah.

Jon Oxer: That's right. So, but it's a trap that I think I personally would fall into just because of, you know, what I know about myself.

Chris Gammell: Right. Well, one of the things this document talks about is like, if you are going to do that, if you're going to give away, if you're going to give away hours for free, because you were learning something, you still mention it. Right. You say, I did this thing. I was learning. I'm not going to charge you for it, but you still mention it. Right. Because that is a value you provide and it's just being transparent. Right. It's like, so I think there's something there, but it's just, you know, you gotta, yeah, you gotta kind of feel out that situation. Yes.

Jon Oxer: Yeah. Very good point. Um, yeah. So. All right.

Chris Gammell: We're witnessing the birth of a consultant here. Maybe. I don't know. So this, here's the real question though. Okay. You've been your own boss for 20 plus years, you said, right? Yes. How the hell would you work for someone else? And I know that consulting is working for someone else. And even when you have your own company, yes, you're working for someone else as a, you know, as a service provider, as a, as a products provider, but still. I don't know, man.

Jon Oxer: It is a big psychological change. In fact, I've already seen that within the context of my software company, because I am continuing, um, just on a limited number of hours during the transition period under the new owner.

Jon Oxer: Yeah. Um, we had a situation come up just a few days ago where we needed some assistance with a particular database problem. And in the past, what we would have done is just, uh, called someone that I know and said, can you come and, uh, spend a few hours on this and bill us and get it done. Um, but I actually don't have the authority to say that anymore. So I keep coming up against these situations where I think, uh, I'll just put that on the credit card or hang on. No, I can't. So I'm having to change my thinking totally about that. Um, I think it would be very difficult for me to work, um, directly under someone and have them, uh, tell me what to do and what I'm not allowed to do. Um, and that's, that's a big change that I have to go through. Well, or not maybe figure it out.

Chris Gammell: I mean, I think, well, I mean, there's certain, there are certain roles like, so, you know, I, I do product management now and they, and they, they, like, I've read a lot of stuff about it where it's like, oh, you're supposed to be like the CEO of a product, which is a bunch of BS if you ask me, but in general, you are in, I mean, you're still, you're still in charge of that thing. So that's good. Uh, but there's, you know, you're still going to have a reporting structure and yada, yada, yada, but you were kind of at the top of the company. So that's the real difference, right? I mean, the buck stopped with you. So that's any, any, uh, any solace in that, that the buck might not stop with you?

Jon Oxer: Oh, yes. Um, this change is both good and bad. So in some ways it's very freeing. Um, I can think, well, it's actually not my problem to solve that. And, um, when you're at the top, ultimately everything is your problem, even if you're not working on it personally. Right. It doesn't, yeah. So that is a tricky one.

Chris Gammell: Hmm. Hmm. Um, I, like I said, I'm not worried about you, man. You're going to be fine, but it is a very interesting situation, I think. And people will obviously be able to get in touch with you. We will give all your contact info at the end. So, yes. So what about, what about the status of the other things though? So you got, you got Superhouse, you got Freetronics. What's, what's going on with those? Yeah. I haven't seen a Superhouse episode in a while, have I?

Jon Oxer: No, it's been about five weeks since the last one came out. Okay. And that is, uh, they come out in bursts, which is unfortunate. And, uh, when you're running a YouTube channel or obviously with a podcast, you want to be on a regular schedule. And I just have- Regular-ish. Regular-ish. Yeah. Yeah. Yeah. Yeah. So people expect a certain cadence. And whenever I go a while without uploading a video, I start getting messages and backlash. And then I publish one again and people say, oh, welcome back to YouTube. And it's not that I've ever actually been away. I just took a while to get the last episode out. Yeah. Um, and I would really like to be able to spend a bit more time working on that. Most of the time, because the projects that I feature on Superhouse and, um, just for the benefit of listeners that have no idea what we're talking about, um, Superhouse is my YouTube channel where I do videos about home automation and hacking your house. So it's DIY home automation. Lots of Arduino and Raspberry Pi and those sorts of things. And, um, because I want to be able to show people how to do things, I don't want to just plug something in and say, oh, look, these are all the features. It takes quite a long time to produce each episode because I have to think about the project and plan for it. And, um, quite often what I have to do is design a circuit board for that project and then send off a fabrication, et cetera. So in the most extreme case so far, as I said earlier, one of the episodes took about six years, but it's quite common for them to take two to three months. That's normal.

Chris Gammell: And, uh, sometimes you got to inject stuff in your arm too.

Jon Oxer: I mean, oh yeah, that's right. Yeah. Yeah. Yeah.

Chris Gammell: When you can listen to episode, uh, what was it? One 23, one 20, 23, one 23. Yeah. Oh, one, two, three. One, two, three. It's easy as one, two, three. Yeah. And shoving a RF pill in your arm. That's right. Blech.

Jon Oxer: So the result is that I end up with multiple, um, projects all running in parallel at different stages of development. And then, uh, sometimes it just happens that I finish off a few of them close together and then there'll be a long period where nothing is done. So it's very patchy. I have episodes that come out. Yeah. In every now and then, um, I'd like to commit to a regular schedule. Like ideally I'd like to be putting out an episode say every two weeks, but at the moment that's just not practical. It hasn't been practical. That's right. It hasn't been practical with my previous circumstances. Now that things are changing, everything is up in the air. Maybe I'll be able to get into that.

Chris Gammell: Yeah. That's great. See another benefit right there.

Jon Oxer: And, um, so what about free? Yeah.

Chris Gammell: What about free? Yeah.

Jon Oxer: So free. Electronics is still going along. It's been quite consistent for quite a long time. It's one of those businesses that, um, that hasn't fallen in a hole and it hasn't, um, become, you know, the new biggest tech thing. Most businesses tend to stay fairly flat and even and consistent and that's what it's done. So unfortunately there's not, it's not a big margin business because, um, selling, you know, Arduino related stuff. Electronics. Yeah, that's right. I think we all know that one. Yeah. So even if you.

Chris Gammell: Especially from Australia, let's be honest. I mean, that's right. It is.

Jon Oxer: Um, we have currency issues and we have shipping cost issues. Probably shipping cost is one of the biggest barriers that we have. Yeah. Everything. It costs us a lot to get parts in, costs us a lot to ship products out. So, um, yeah, it, it's a nice business. It's still going on. It's consistent. Um, I ship orders every day. Like they'll, I'll go down to the post office and take a big pile of boxes and padded bags. Um, but it's not enough to, um, provide me an income. So what would be really nice would be to spend my time on freetronics and super house and not have to worry about anything else. But it's just not there in terms of the income.

Chris Gammell: Of course. Yeah. Well, that makes sense. I remember, uh, I think you, you had sent me a board. I mean, I think you sent me one of the first boards that anyone ever was nice enough to send me. So I, I, I do appreciate that. Yeah.

Jon Oxer: Yeah.

Chris Gammell: I should send you some more. No, I'm not. Believe me, since that first board, I've collected. Yeah. I've collected a few and I've made a bunch of my own too. But what I was going to ask about is that, uh, I remember you sent a power over ethernet board. Yes. And like, you don't really hear about that anymore. I mean, is that still, I mean, like, I know that like it's a thing, but it's not like, I don't know. Are a lot of your customers buying the, the POE stuff these days?

Jon Oxer: Surprisingly, yes. Really? Yeah. And we get, uh, customers that order them in reasonable numbers with a lot of our products because we sell to hobbyists. And, um, so we have a different, a couple of different classes of customer, uh, essentially there are hobbyists. There are professionals who are using them for prototyping purposes and there's education. So in education, they will typically buy things in like class sets. Hobbyists will buy one or something. Engineers will buy, you know, whatever they need for their project.

Chris Gammell: Yeah. Like three or something. Yeah. Whatever.

Jon Oxer: Exactly. Um, but we also have, uh, customers particularly for the POE equipment where they'll often buy 50 or a hundred of a particular board because they're putting it into some small production run themselves. So they're designing our board into their product.

Chris Gammell: Okay. And so like system level design. Yeah.

Jon Oxer: Yeah, that's right. So one example is that there's a company that makes, um, energy monitoring systems for solar panels and they designed our Ether 10 board into it, which is basically an Arduino with onboard Ethernet. Um, they designed that into their product. So they just plug the Ether 10 in using its headers and that becomes...

Chris Gammell: Which we do not condone normally. We'll let it slide this time. But that's, from a sourcing perspective, that's actually pretty risky. But like... But I mean, Freetronics quality is so great. That's right. Yeah. Yeah. But yeah, that can be risky from a supply chain perspective. But yeah.

Jon Oxer: Yeah. And obviously once they scale up, that sort of thing doesn't scale because it's just not economical to take a whole board, plug it in. Right. Right. Yeah. You're better to design that functionality into your product. So once people get up into the thousands quantities, they go for a much more optimized solution. But we do find that for power over Ethernet in particular, a lot of people are buying either our PoE modules or the boards with PoE support on them like the Ethermega. So that's probably one of our strongest sellers. I shouldn't say that because that'll give away our secret sauce to the competition. But if you look across our product range, we've got about 60 or 70 products. The Ethernet-based boards are always the strongest sellers.

Chris Gammell: Yeah.

Jon Oxer: So even in this age of Wi-Fi and ESP8266. So you can put...

Chris Gammell: I tell you what, most days I kill for a LAN connection. Well, I asked you before the show. Obviously, we're both on LAN right now. Like, man, non-deterministic connections, they can... You can keep them, you know? Oh, yeah. They can. Yeah. Yeah. I don't... I'm not a fan of that, so... No.

Jon Oxer: Well, talking about the economics of scale and optimizing as you go up in quantity, it raises an interesting topic that I've been thinking about recently. And I was just reminded of it yesterday. So, I mean, as you often talk about on the show, engineering is all about compromises. You've got to balance cost versus efficiency and size and power budgets and all of these sorts of things. But it's always really interesting to look at the extremes of situations. Well, I find it interesting anyway. So if you look at a scale of, say, number of something that you're going to produce, you can produce a very small number of it or a very large number of it. And a few years ago, I was involved in a project that was really at the small scale, which was the RGSAT project. So I designed the payload processor module that flew on RGSAT 1 and RGSAT X. And when you're dealing with that sort of project, basically, cost is meaningless. You really don't care. So we made a total of five boards. So this was like final production boards. Total production run was five. And two of them were for me for bench testing. Two of them were to go in the actual satellites. And one was in a functional satellite replica, which was set up at the headquarters so that they could run tests on it before actually uploading code to the real satellites. So it gave them an exact replica of what was in the flying satellite.

Chris Gammell: Right, like an Apollo 13 when they were switching all the switches, right? Yeah, that's right.

Jon Oxer: Just to simulate what's going to happen on the real thing.

Chris Gammell: That's right. Yep.

Jon Oxer: So when you're dealing with that sort of scale and the ramifications of failure are so high, I mean, if a part fails on that board and it's in orbit, then that's the end of your mission. You really can't afford to scrimp on anything. And so in that particular sort of situation, if there's a part that'll do the job and it costs 10 cents, and there's another part that will also do the job and it costs $10, but maybe it uses slightly less power or it's got some other attribute that makes it just slightly better. I'll go for the $10 part. I'm not even going to think about it. It's just give me the good stuff. I don't care if it's plated with unobtainium. So, and the thing is that that's an extreme situation. What about lead times? Well, with that sort of project, lead times usually aren't too much of an issue either.

Chris Gammell: Okay, so you're already designing with stuff that are in stock at a distributor anyways.

Jon Oxer: Yeah, that's right. And you really want to make sure that your system is not going to fail after it's been accelerated up to 28,000 kilometers an hour or something and thrown out above the atmosphere.

Chris Gammell: So details, details. Details matter.

Jon Oxer: Yeah.

Chris Gammell: Yeah.

Jon Oxer: But the thing is that if you look at the other extreme, and this is what came up in this conversation with a friend of mine yesterday, he's an engineer who is working for a large US company, and they make a whole lot of different things. But they're essentially a toy company, so their scale is enormous. They make things, you know, quantities of a million, not quantities of five. And we were talking about an interesting problem that he was facing. And so once you're talking about a scale of, say, making a million of something, if you can save one cent on your BOM cost, then it's worth spending up to $10,000 of engineering time to optimize out that one cent. That's right. So in this particular case, say you've got a tantalum capacitor that might cost you six cents, and you want to swap it out with a one cent ceramic. So that's going to save you five cents, which means you can put $50,000 of engineering time into making that change. Right.

Chris Gammell: And most of that $50,000 is testing, right?

Jon Oxer: Partly. But this is the interesting thing that came up with this conversation in the front with my friend. In this particular case, they had some noise coming from some motors. And so these motors were affecting a magnetometer and they could put cost into the hardware and filter out the noise, or they could let the noise through and then put a lot of development work into software filtering to remove it after you get the value out of the magnetometer. And yeah, it's one of those situations where it's worth spending $50,000 of coding time to write the software filters to deal with the noise. Rather than use a six cent tantalum in place of a one cent ceramic. So it still fascinates me when you look at that very high end that a few cents here and there is worth tens of thousands of dollars of effort just to achieve that little saving. So most of the time, of course, when you're working on a project, it's much more balanced than that. You might be designing something that is going to be made in the thousands or tens of thousands quantities, maybe up to hundreds of thousands. So it's quite rare to work on a project where the ultimate goal is you're going to make one board and that's the board that's going to go into space and it can't fail under any circumstances. Or you're going to make five million of this thing and every fraction of a cent matters. So those are extremes, but I always find it's fun to look at extremes and get my head around what the constraints are. Because it means that as you are designing your solution to the problem, the constraints you're working with can be totally different.

Chris Gammell: Yeah, again, it drives a different set of skills, right? I mean, like the people I know who do work at toy... Well, I guess I had this a little bit as well where I just... I went from like a test equipment company to an industrial company and it wasn't even that big of a shift, but it was just a super high end mid-range. And it's just... You can't... You know, when you're going from that super high end to mid-range, it's like you don't worry as much about every little thing, right? Mm-hmm. It's more about speed. It's more about whatever.

Jon Oxer: Yes. So...

Chris Gammell: Yeah. What is... I guess most of the stuff you were doing in the past has mostly been cost optimized? I mean, it's low volume though, right? Lowish volume?

Jon Oxer: It's... In the past, surprisingly, it hasn't been cost optimized because the volume is fairly low. And in the early days of Arduino, an Arduino board would cost... I remember paying 40-something dollars for one of the early... Like a genuine Arduino in the early days. And of course, things have changed now, but I think this is possibly the bit of the middle road problem that Freetronics is in right now. So we're still trying to make the good quality boards, but compete against the really budget manufacturers who have trimmed every cent out of it that they can. Right. So it's disheartening sometimes when I look at... You know, you go to eBay and you look for a cheap Arduino coin and their retail price is less than, you know, half of my bill of materials cost. Right. Exactly. I just scratch my head and say, how the hell is that happening?

Chris Gammell: Yeah. But I mean, you never really want to compete on price. Even the ones that are, quote unquote, competing on price is more like the industry they're in, right? So like the toy industry is just... There are almost ceilings on toys, right? Like certain classes of toys can only cost so much, right?

Jon Oxer: Yes.

Chris Gammell: So...

Jon Oxer: Yeah, there are these price points that everybody has to hit, like the $99 and $149 and $49.95, that sort of thing. So there are those constraints.

Chris Gammell: I was really surprised by that recent... Did you see that car that Sphero made? It's actually... It's beautifully designed. And it's... So the Sphero folks, they made a car, like a promotional toy for the new Cars 3 movie. And so it's like the Lightning McQueen. And it's like $300 though. And it's like... So that's what double that BB-8 was. And... But it's got like six... So it's got like six motors on it. It's got like a... Like the cars... Where the cars' eyes are usually in the screen. It actually has like a mini LCD and stuff like that. And it's just... And it's all Bluetooth controlled. And I mean, like everything about it is just really nice. But then I saw the price tag. I'm like, wow, that is... That's barely a toy at that point, right? And it's like... Yeah. I'm not getting that from my nephew. I'll tell you that much. Yes. If I get that toy, it's for me, even though I don't like the movies.

Jon Oxer: Yeah. It sounds like the sort of thing where... Someone like us would pull it apart and be fascinated by the engineering that went into it. A bit like that ridiculous juice press that we probably shouldn't get into now.

Chris Gammell: No, definitely not. We've talked about that. Yeah. And yeah.

Jon Oxer: It is actually interesting looking at products like that. And because it's an example of an extreme in engineering, it's one of those don't cut any corners, do everything as well as you can situation.

Chris Gammell: Yep. Hmm. Yeah. Actually, so Fictive did a teardown of the Ultimate Lightning McQueen teardown. So it's... Yeah, it's... I mean, kind of like the exploded view of everything that's in there too. Even just like having more than one circuit board in a toy. Like that alone is like, whoa, wait a second. You know, like too fully blown. Yeah. So anyways, that's a nice little toy. So many words. So yeah, somewhere between there and super low end, right? I mean, like that is... Those are the things that you have to deal with. And it seems like... You know, I was talking to a friend about industrial stuff the other day. I think it's actually really interesting in the industrial space of like accessible tools that are usable in a factory, right? So like ruggedized electronics. I still think that that's an interesting space for people in, you know, your position, my position, whatever, right? Where it's going to be low volume. It's going to be high mix. It's going to be, you know... It could even be slow or kind of expensive. Like all of these weird corners there, but it's because it's super customized for the needs that they go for it, right?

Jon Oxer: Yes. Yeah, exactly. And in fact, many of our boards, it's kind of parallel to the industrial application is farming.

Chris Gammell: Oh, yeah, right. I found that... And I can't put those in the same bucket, really.

Jon Oxer: Yeah, yeah, that's right. They've got a lot of the same constraints and the same similar environmental issues. Right, exactly. They're going to be, you know, hot and cold environments and stuff gets wet and all of those bad things. So you really have to make things that are able to withstand that. So every now and then, a lot of the time, we just ship boards and we don't hear anything else. We've got no idea what happened to them. But every now and then, we'll get people coming back to us and saying, oh, use your board in this particular application or whatever. And that sort of thing is always fascinating to hear. I think that's one of the things that's really interesting about providing general purpose boards. It's not a finished product. It's something that people build into other things. So we see all sorts of crazy applications.

Chris Gammell: Right, exactly.

Jon Oxer: Yeah, so I've had a surprising number of people come back to me and say that they've used them in farm automation. Things like monitoring environmental conditions inside a chicken shed. And then, you know, displaying the data in real time up on some big status board and those sorts of things. So I think that's a very interesting thing. That's how many things it is. It's just like there are four eggs. Yeah, that sort of scale. It's more like there are 40,000 eggs.

Chris Gammell: Yeah, right. Exactly. Well, yeah. And that's the crazy thing too. Like that's why it's almost industrial where it gets big pretty quick on some of these farms. They're the big operations.

Jon Oxer: Yeah. And often they're doing it with surprisingly few people.

Chris Gammell: Yeah.

Jon Oxer: There was one very interesting conversation I had recently with an electrical engineer who is working on equipment for milking cattle. Yep. And so I've seen a number of systems set up for automated milking, but they still involve people and they involve the cow being put into like a very confined space and they're stressed and that sort of thing. And so what they've done is build a system.

Chris Gammell: Cow massage chairs.

Jon Oxer: Almost. It sounds almost that luxurious. Yes. So, and the thing is that in a typical milking situation, because there are humans involved, you want to get them all through in a short space of time. So there's always this time pressure and they do it early in the morning and rush them through one after another. There's lots of stress involved. And then the farmer can go off and do whatever else they need to do. And in this particular system, what they've done is there are three farms just outside Melbourne where they have set up a system which has a totally robotic milking system. And the actual suction cups are applied automatically by the robot using vision processing and all sorts of things. The system scans the ID of each particular cow.

Chris Gammell: Right. They get like a tag in their ear or something.

Jon Oxer: That's right. It uses RFID or whatever.

Chris Gammell: You better watch out if you go there, buddy.

Jon Oxer: Oh, no. I don't want to be milked. You're going to get milked, yeah. And the way the system is set up, it's all done basically on the cow's schedule. So they can be milked at any time of the day because there are no humans involved. The cow walks through the mechanism that does it. The food is provided to them and it's all done very gently and slowly. And the machinery doesn't make noise. So they're not stressed out. And the interesting thing is that these cattle can go months without seeing a human. And the machine just takes care of them automatically. And it also does things like automatically check the quality of the milk. Right. And it also dishes out the feed individually. So every cow, as it's being milked, it's been identified. It's also been weighed and various other things have been checked. So the quality of its milk has been checked automatically by the computer. And so they can be given a tailored feed that gives them enough food, gives them what particular supplements that specific cow needs, regardless of anything else in the herd.

Chris Gammell: I'm imagining like feeding a chocolate for getting chocolate milk. Is that what I'm hearing?

Jon Oxer: I don't think it works that way. I know. I know.

Chris Gammell: I also saw this really sad statistic the other day that said like 30% of Americans think that chocolate milk comes from brown cows. I know. It's probably a fake statistic. It's fine. But even like there's one person that thinks that. That's sad.

Jon Oxer: Yeah. I find it really interesting to hear about projects like this because it's such an interesting engineering problem. It is.

Chris Gammell: But is it economical? That's what I always get back to, right? I mean like I can make a robot that does anything, but it's...

Jon Oxer: Yeah. Is it give you enough money and enough time?

Chris Gammell: Yeah, exactly. Right.

Jon Oxer: Yeah. Well, I don't know about the economics of that particular project. And...

Chris Gammell: I also should be upfront. I actually can't make a robot that builds anything. I have very few robotic skills.

Jon Oxer: I'm sure you'd have a pretty good go at it.

Chris Gammell: Sure. Yes. It would definitely get near the cow. The cow might be in danger. Right. It may end up making hamburger more than anything, but yes. But either way, someone's getting fed. Yeah. Yeah.

Jon Oxer: So, I think that's one of the things that really fascinates me about this industry is that there are so many different types of projects like that for people to work on. I love talking to engineers because everyone I talk to, it comes out with some amazing story like this. Like, I built a robot that milks cows and nobody has to go near them. It all happens automatically or something like that. And, yeah.

Chris Gammell: Well, that kind of brings it back to the consulting thing, too, because it gets really... So, like, consulting is usually like, I need someone to build me this thing to go into the rest of the system where it's, you know, like, building a sub-circuit or subsection of a system, right? But that kind of, like, top to bottom, like a cow milking system, right? How often are electronics people talking to dairy farmers and then being like, oh, well, you should build a robot for this, right? Usually what a dairy farmer would say is, well, you should, like, you know, increase the efficiency of my workers or you should do all this other stuff. It's kind of like this... It's like these mental leaps where it's like, well, we might as well try this other thing. But there, you know, you have to also find not just someone who's looking for a solution to a problem but is also willing to fund and or try out something a little bit more unconventional.

Jon Oxer: Yeah, yeah. You really need that serendipitous connection. Someone within the particular industry or area of need that is thinking about the problem and someone else that comes up with a radical solution to it. Not just, can we make this thing 5% better? But, you know, can we do this in a totally different way? And it's really interesting when those sorts of things happen.

Chris Gammell: Yeah, right. Exactly. It's like the whole, if I would have asked people what they wanted, they would have asked for more buggy whips or whatever the Ford quote is. Yeah, that kind of thing, right? Yeah. So, all right. So, all we need to do is find you some rich farmers and we'll, you know, you can go make your milking robots or... Well, I guess that part's done. So, whatever kind of other robots are out there. I could spend all day on YouTube watching, like, especially with food. Like, food is especially, it feels like the packaging and, like, the cooking side of it, like, the really industrial side of things. Man, those YouTube videos are just magical to watch.

Jon Oxer: Oh, yes. Absolutely. Yeah.

Chris Gammell: The one, I remember the frozen pizza one. Have you ever seen that one? It's like, so it's inside, like, a frozen pizza factory. And it's just, you know, they, because it's all, it's soup to nuts, you know? It's like starting with flour and then making the actual dough and cutting out the things and applying all this stuff. And it's not just, like, one at a time, either. They're doing, like, 20 at a time, like, in, like, parallel fashion. It's just, it's an unbelievable process. Yeah. So, I'll see if I can link that in, actually. Yeah, I'd love to see that. Because that is a fun video.

Jon Oxer: Yeah.

Chris Gammell: Yeah.

Jon Oxer: Well, many years ago, the father of a friend of mine was a mechanical engineer and he worked at a local ice cream factory and worked on the machines for making the ice creams. Oh, man. It was interesting to hear things like how, yeah, how the machines are made. So, if you, if you think about this as an engineering problem, we want to make an ice cream on a stick. Like, how do you get it into the mold? How do you get the stick in the right place when the ice cream is freezing? How do you then get it out and into the packaging? How do you seal the packaging? Right. It's a whole lot of problems to solve along the way.

Chris Gammell: Yeah. Those, yeah, those are. And, and, and, and like I said, the parallel stuff too, it's like anyone can make one ice cream like that, but now you have to make a bunch and you have to do it like in time, you know, you gotta like, and you gotta, this whole thing's gotta move through the system. It's not like you can have someone reach into a freezer and pull it out. It's like, no, it has to like flash freeze as it's moving through. And yeah, those kinds of, those kinds of problems are.

Jon Oxer: Yeah. And if you change anything about the way the production line is running, then your timing changes on the freeze cycle and everything gets screwed up. So, yeah.

Chris Gammell: Oh yeah, that's true. Yeah.

Jon Oxer: I, yeah. I heard about bad days when, um, some machine would fail and they'd end up having to drop a whole batch of ice cream. So.

Chris Gammell: And it didn't like end up in your friend's freezer? Is that. No. Oh, that's, that's too bad. Yeah. You say, yeah, come home from the factory with like.

Jon Oxer: Yeah. We've got 60,000 boxes of ice creams.

Chris Gammell: I need, I need you to eat them before they melt. Okay, dad. I can grab a spoon. Yeah. Yeah. Yeah. Right. Yeah. That would be, that'd be great. So what are you excited about in the industry these days? I mean, I guess it's been, I mean, it's been five years since, not since we talked, I should mention too, that thank you for hanging out with me when I was in Melbourne. It was fun. Oh yeah, that was great. We went and flew drones and. Yep. Got food and everything was great. Yeah. Yeah.

Jon Oxer: I, yeah, I think I nearly killed you with one of them. It was a bit out of control.

Chris Gammell: The dogs were in greater danger than it was. Yeah. Yeah.

Jon Oxer: Dogs don't realize that drones are flying chainsaws. So they just think it's something fun to play with.

Speaker ?: Right.

Jon Oxer: They're just like, hey, funny, noisy thing. Yeah. Go for it. Yeah. Yeah. Yeah.

Chris Gammell: Flashing lights. Yeah. Yeah.

Jon Oxer: Yeah. That's been a lot of fun recently. A little while ago, I did some post on Facebook, I think, saying, you know, I'd really like to play around with drones. And I had a whole bunch of people contact me and saying, hey, I've got some spare parts. I'll send them to you. So I ended up with this big grab box full of random bits and pieces, like speed controllers and motors and props and things. So I cludged together a drone and ended up having to buy a few parts along the way as well. As you do. Yeah. Yeah. But that was a really interesting experience.

Chris Gammell: Is that thing flying now? The one that almost ran into us?

Jon Oxer: It's flying approximately as well as when you saw it.

Chris Gammell: Okay. All right. So it doesn't hit any solid objects in a while.

Jon Oxer: No, it hasn't. I just recently got GPS working on it, which means I'll be able to do position hold. So the problem when we were flying it out in the park was that I didn't have GPS working. It was just drifting around all over the place. And I'm not exactly a good pilot.

Chris Gammell: I think that's hard with anyone. Like, I think that's going to be a challenge as people get more and more used to these commercial drones too, where they kind of just stand where they are. They stay where they are. It's like, there's a lot going into making that stay.

Jon Oxer: Oh, yes. That's right. Absolutely.

Chris Gammell: You know, it's like all the control systems. It's got the GPS stuff. It's got, sometimes it's got ground sensors. Barometer. Right, exactly. Yeah. Yep. It's crazy.

Jon Oxer: And they're often doing image processing now as well. So they can have...

Chris Gammell: Yeah, like the look downs.

Jon Oxer: Yeah, the look downs to check whether they're moving relative to the ground. They can do... Some of them have forward-looking cameras for collision avoidance.

Chris Gammell: Oh, really?

Jon Oxer: Yeah. So I think the current Phantom 4 has forward-looking collision avoidance. So you can basically like fly it at a house and it will just refuse to run into it. It'll approach it and then just stop. So it's almost uncrashable. And that sort of thing is amazing because as you say, there are so many... Challenge accepted. Yeah. There are so many different inputs into the control system. So it's getting your input saying, go forward. And it's getting the collision detection saying, no, stop.

Chris Gammell: Yeah. Like all that. And that's the other thing. It's like layer on layer on layer too because like the IMUs that are in there, like the nine-axis sensor fusion output craziness, right? Like we're just spoiled these days, you know? Oh, yes. It's not putting like just like three coordinates or something like that.

Jon Oxer: Yeah. Yeah. Yeah.

Chris Gammell: And I have no idea how any of it works, let's be honest. Yeah. It's cool. It's a cool time. And like have you seen that fold-up one that DJI has now too?

Jon Oxer: Oh, that little spark. Yeah.

Chris Gammell: Yeah. My friend went to Ireland for a week and he just like threw it up in the air and he took like this 4K footage. It's just like, I didn't even know how movie makers stay in the business anymore.

Jon Oxer: Like what is left? Yeah. Well, whatever. Yeah. It's amazing. I mean, I was amazed the first time I flew a Phantom 3 and it really does just look after itself. So you take off, go up into the air, you can just take your hands off the controls and it will just sit there. It'll stay where you put it.

Chris Gammell: Yeah. Right. Waiting for next command. Yeah. Exactly. Yeah.

Jon Oxer: Yeah. But I've seen some incredible footage of first-person view flyers as well.

Chris Gammell: Oh, like the racing stuff?

Jon Oxer: Yeah. That's insane.

Chris Gammell: Yeah.

Jon Oxer: There's a local Aussie guy who has a YouTube channel called UAV Futures and he talks about all sorts of drones and things and does first-person view racing. And a little while ago, he put up a video of, I think it's the current Australian champion who's like a young kid. He's like 12 years old or something. Right. He's got those classic video game reflexes. And you see video from the point of view camera of the drone as it's zipping through these gates and things. And I can barely even follow it. And somehow he's controlling it in real time.

Chris Gammell: Right. Use the force, Anakin. Use the force.

Jon Oxer: Yes. Yeah.

Chris Gammell: Damn. Yeah. That's crazy.

Jon Oxer: But speaking of fast reactions, so that's human control. One of the other videos I just saw recently was the little Japanese sumo robots.

Chris Gammell: Oh, yeah. You sent me that. They are crazy.

Jon Oxer: They're so fast.

Chris Gammell: So, what about, so yeah, we posted this on the channel too. Are these pre-programmed? I could not figure it out.

Jon Oxer: I don't know. I assume because they're so fast that I was having trouble believing that anyone was controlling them. Yeah, exactly. Right. So, just to give listeners some concept, they have, it's a sumo battle between two robots and they have to push each other out of a circle.

Chris Gammell: Yeah, they kind of look like bulldozers.

Jon Oxer: Yeah, they're like little bulldozers, but like the size of half a shoebox, that sort of thing.

Chris Gammell: Yeah, yeah, yeah.

Jon Oxer: And they start on opposite sides of this disc and their job is to bash the other one out of the circle. And when I first heard about this, my concept was, you know, they'd move about as quickly as a radio control car sort of thing, which is fast, but not that fast. But these things are just like lightning, like blink your eyes and they're like zipping around all over the place and smashing each other.

Chris Gammell: Well, the thing that's about RC cars is like they might be able to get to a speed, but they're not like super like jerky, right? They can't like start and stop. These things are like tuned, like control system tuned so that they can just stop on a dime and rotate and like all this crazy shit.

Jon Oxer: Yeah, yeah. Because they're staying within this circle, which looks like it's about a meter or maybe slightly larger in diameter. It's not a very big area. So they're zipping around in this very confined space without running over the edge and bashing each other up. It's very entertaining to watch.

Chris Gammell: It really is. Did you notice too that there's, so there's the, there's a referee or whatever, but he's wearing shin guards. Did you notice that?

Jon Oxer: Yes. And in one of the videos, there were some crash barriers around and yeah, bits flying off. But right near the end of the video, someone is holding a remote control and that may be one that maybe they are manually controlled, but it could be that they just use that for arming it. Like once they're ready to go, the referee says go, then you hit go. But I have trouble believing that a, that a human is in the control loop for these things because they are so fast.

Chris Gammell: Right. Well, and the other question though is too, like, so, okay. So say they are programmed. Is it just like, they're like open loop, just do this, do this pirouette and then, you know, hope that you kind of.

Jon Oxer: Yeah. Hope that you're going to crash the other one somewhere, somehow.

Chris Gammell: Or is there, or is there some kind of feedback where it's like, oh, ping sensor. Well, not even ping, that'd be too slow, but like some kind of sensor that's like proximity, whatever. And then react based on that.

Jon Oxer: Yeah. Well, some of them did seem to be targeting each other. So I don't know. We're talking about something that we don't know the answers to, but it's very entertaining. Welcome to the amp hour.

Chris Gammell: This is the amp hour. Yeah. Yeah. Super fun project though. Yeah, definitely. So what's next for you? I mean, what are you, what are you going to be building just for fun? Yeah. I'm always working on a few things.

Jon Oxer: What are you building right now? I suppose. Yeah. A few things. So I suppose if you look at what I have at PCB layout stage, but haven't yet sent out for fab, there are a couple of things. One of them is a board that I'm working on to retrofit wifi into electric window motors. Now this is related to a project that I did a little while ago for super house there in my house. I have electric blinds and motors to open and close windows. And the motors themselves have an interface on them called LIN, which is, it's commonly used in vehicle, electrical systems. Local interface network, right? Yeah, that's right.

Chris Gammell: It's like there's CAN and LIN. LIN is like the car window.

Jon Oxer: LIN is like CAN light.

Chris Gammell: So it's like the ones you don't care as much about, right?

Jon Oxer: Yeah, exactly.

Chris Gammell: Right.

Jon Oxer: It's much lower speed, much simpler implementation. And so these motors have LIN built into them, but it can be a bit of a pain to set up. And I did a super house episode a while ago about making a LIN interface and then controlling them from the home automation system. But I've just been playing around. One of the things that these motors have is a built-in battery compartment. So you can either plug in an external power source or you can just run them off four AA batteries. And the batteries themselves need to be replaced from time to time. It's a bit of a pain. So if you're going to do it properly, you'd want to run external power to it. So you never have to care about the batteries. So it sits there with this empty battery compartment. And so what I've designed is a circuit board, which is elongated and it's a bit of an odd shape with two slots in it so that it can clip in place of the four batteries and then a jumper cable that runs around and plugs into the LIN interface so that it's basically like a piggyback control board. And the idea is you can pop the cover off the existing motor, plug this little board in where the batteries used to be. And now you've got Wi-Fi control of your window motor. So yeah, that's just a nice little project that I'm hacking around on.

Chris Gammell: What are you using for the Wi-Fi? Are you doing ESP?

Jon Oxer: ESP, yeah. I've got ESP8266 on these. I'm wondering whether I should swap it out and put an ESP32 on it.

Chris Gammell: I think the sourcing would be a little tougher for that if you get to that. I mean, like right now at least, maybe in a couple months it'd be fine.

Jon Oxer: Yeah, that's right. It's at that changeover point. That's an interesting process as well. One of my friends works at Espressif, which is the company behind the ESP. Right.

Chris Gammell: I think I met him. Oh, yeah.

Jon Oxer: You did, yes. Yes. Dinner together. That's right. So shout out to Angus. Yes.

Chris Gammell: And thanks for the ride up to the drone thing, Angus, of course. Yeah.

Jon Oxer: And so I had lunch with him a couple of days ago and we were talking about the ESP8266 and now the ESP32. And so the issue is that ESP8266 has become so popular and people have built it into so many things that it becomes the go-to thing that you reach for when you want cheap Wi-Fi. So now the ESP32 is coming out, which has many advantages over it. Yeah. Yeah, more memory.

Chris Gammell: It has Bluetooth, all that stuff. Yeah. Yeah.

Jon Oxer: More radios in there. But I was just saying to Angus that my reflex when I was designing this board was go for an ESP8266 just because that's stuck in my head, even though I could just go for an ESP32 instead. And I possibly should just for, I don't know why. Yeah, like future-proofing almost. Yeah, future-proofing. It's the newer chip. Why go for the old one when the new one is here? Right.

Chris Gammell: That's an interesting point too because there's people who've done that before. I mean, I think this is a pretty low impact kind of experiment, right? But people have done that before and they've said, I'm going to move to the next generation and then the next generation just never takes off, right? Because the previous generation gets cheaper or there's just too much engagement with the older generation type thing. It has happened before in the past and it's an interesting question on whether you should.

Jon Oxer: Yeah. Well, for my particular little personal project, it really doesn't matter. Right.

Chris Gammell: Yes. Low impact, yes. Yeah.

Jon Oxer: It's just that I became aware while I was talking to him that I'd made this engineering decision without actually stopping to think about it. I'd made it totally out of reflex. It was just, oh, that's what I have done before. So I'll stick that footprint on the board and we're done.

Chris Gammell: Yeah. I mean, it depends though too, right? You're making prototypes, right? I think that's almost always okay for prototypes. Maybe it's different with processors because that has a much bigger implication in terms of like, okay, you can't have more memory in the future. There's a higher switching cost. But like if I build a prototype and I put an LM317 on there, there's not really that much implication other than like dropout voltage, right? And I can always go and redo that later. You know what I mean? Like unless I really mess something up with that first iteration, I don't know. I think prototyping, it's not a huge deal. It's like it's when you take it. If you then didn't do anything, you go, okay, this is going to production now. And you don't. That's when you say, I'm not going to take a look at this again. That's when you can really get into trouble.

Jon Oxer: Yeah. And that really comes down to the point at which you start doing optimization. So there's an old software saying that premature optimization is the root of all evil. Because a lot of people start to think about optimizing things as they are designing it, which is not necessarily when you know what needs to be optimized.

Chris Gammell: Exactly. Exactly. So, yeah. Right. And most of the time you're just trying to prove like, I mean, you might realize that you don't even like the motor in the original device. I mean, obviously you've already been using it. But, you know, similar kind of things where, yeah, you don't know what you're actually, most of the time when you're prototyping, you don't actually know what you're even looking for. Right. But, yeah. It's the point of a prototype. Yeah. So, yeah. I don't know. I've still been trying to do that better this year. This year has been my year of getting faster prototyping and half a year in, I'm still not much faster. But, yeah. I'm trying, man.

Jon Oxer: Yeah. Apart from that. Yeah. Yeah. Apart from that, I'm working on a few different things. Probably not anything that's particularly interesting. But one of the little boards that I've also just been working on is, once again, ESP8266 based. It's a little board that I can use to hack into an existing remote control for something. And this is solving the problem of you've got an appliance of some kind. It comes with a remote control, you know, like a garage door opener or whatever.

Chris Gammell: Yeah. Right. Right.

Jon Oxer: And there are several different ways of solving the problem of linking that into something else. And it's so you can emulate the protocol and have your own transmitter or whatever. And it sounds really nasty and hacky. But I found that one of the best ways to get around this problem is rip out one of the factory remote controls and just wire across the back of the buttons and trigger it and send the signal.

Jon Oxer: And I've actually done that so many times now with different devices that I have like a semi-standardized little Arduino shield layout. Where I have outputs that switch the buttons on the remote control. But it's always very labor intensive to make it because there are lots of little jumper wires and bits and pieces. Yeah.

Chris Gammell: And then you could have like the jumper wire might break. And that's kind of a point. That's right. Right.

Speaker ?: Yeah.

Jon Oxer: So it's got to the point where when I'm faced with doing this, I sort of sigh and go, oh, no, I've got to sit there with jumper wires and make this up. And it's going to be a bit of a pain. So I've got to the point where I said, stuff this. I'm going to design a board for this purpose. So I've designed a little circuit board that fits into a standard size plastic project box. It's got the ESP8266 and a little switch mode power supply down one end. And about half of the circuit board is a blank space. So what you can do is take your little remote control and basically stick it on the PCB, bridge across a couple of wires, and you're done. So, yeah. I don't know if anyone else will care about that. No, no. That's good, actually. That's not a particular problem.

Chris Gammell: And so you've never – do you ever do the thing where you try and like capture the – like do like a replay, basically? Oh, yeah, definitely. So you capture the IR and then replay the IR? Yeah. How do you deal with that?

Jon Oxer: And radio as well. So I've done quite a bit of that. And in fact, I've got several boards, several products that Freetronic sells specifically for that purpose. Oh, really? Which ones? There is a 433 megahertz receiver shield, which is – you basically plug it into an Arduino and then you get a raw waveform coming out that – so if you're – there are a couple of frequencies that are really common just because they are in the unlicensed band. There's 433 megahertz and a couple of others, 315 megahertz. It's country dependent. So what I've found in Australia at least, just about everything transmits at 433 megahertz. Things like, you know, once again, garage door openers, but, you know, power board, remote controls and whatever. So this is – assuming they're radio, they're not IR. And so I've done quite a bit of work. And in fact, I did a conference talk about this. And there's a video of it up online about hacking – basically reverse engineering communications protocols for IoT devices. And so I've used it in the past for things like intercepting data coming from wireless weather stations.

Chris Gammell: Oh, cool.

Jon Oxer: And from power monitoring systems. So, you know, those things where you clip a current clamp around a wire and then inside your house you have a little display that says you are using, you know, 2.3 kilowatts or whatever. They use the similar thing. There's a little transmitter where the sensor is located that's measuring the power usage. It's just blindly transmitting on 433 megahertz. And so I've got some code that uses our receiver shield. It just intercepts that signal and then you can log how much power you're using. So I've done lots of those sorts of projects in the past where I look at the protocol and I often do that with something like an amateur receiver. So I got my amateur radio license a little while ago and I've got like a handheld transceiver. So you tune that to the right frequency and you can get a raw waveform coming out of it and open it up on your computer and you can actually see the shape of the data. It's actually one of those things that's quite cool. It's very crude, but it works. So you can, you receive the signal on your radio and you get that, you know, the classic modem screech buzz coming out of it, which is the data going through. And to you, it just sounds like a screech. But if you record it and then open it in an audio player on your computer that shows the shape of the waveform, all of a sudden you can see it's binary data.

Chris Gammell: Yeah, it's a down sample to like 40 kilohertz or something like that. That's right.

Jon Oxer: So you can sit there and decode it by hand if you want to. And then pass. Yeah. But sometimes that's necessary the first time. Like if you don't know what the actual structure of the message is. Right. So you need to figure out, you know, what's the preamble? What's the actual payload data? How does the checksum work?

Chris Gammell: So are you saying you're actually, you are reverse engineering? You're not just, you're not just doing a replay.

Jon Oxer: Exactly. Yeah.

Chris Gammell: Okay. Do most of the products that you do this to, do you find that they have some kind of protection against a replay attack?

Jon Oxer: Nope.

Chris Gammell: Yeah. I wouldn't think so. I mean like, yeah.

Jon Oxer: It's, it is ridiculous. One of the, in fact, this was part of the point of the conference talk that I did. I demonstrated, I started ordering things online because.

Chris Gammell: You just want to find one that doesn't have this? Yeah.

Jon Oxer: And I started becoming even more ridiculous. Like I ordered wireless infrared motion detectors. So the sort of thing that you use in a security system. I ordered a few different types from different manufacturers and they just blast out data. So the scary thing is that you can sit outside someone's house and read the data from their motion detectors. Yeah. And there's, there's no protection. It's crazy.

Chris Gammell: Yeah. Well, when we had Sammy on, he told us about a lot of this stuff. And I remember Sammy told us about that. I think Joe mentioned that a little bit. Joe, we just said Joe Fitzpatrick on a couple of weeks ago. And like, yeah, maybe, maybe you have a future as a security researcher too, you know, like, I don't know, man. Yeah. It's, it's scary, you know? Yes.

Jon Oxer: One of the things that I like to do is take things that were never intended to be connected to the internet and connect them.

Chris Gammell: Yeah. Yeah. That's great.

Jon Oxer: One of my little hobbies. So, which, which all falls into that home automation thing as well. When you take an appliance.

Chris Gammell: Like when you, when you, when you like, uh, have it transmit on the internet, do you have it say anything snarky? Like, hey, I'm a stupid fridge.

Jon Oxer: I should do that.

Chris Gammell: Yeah. Yeah. Yeah.

Jon Oxer: Not everything needs a Twitter account.

Chris Gammell: That's right. Yeah. This stupid fridge is at 32C. It's broken. It's on fire. All your food is bad. Yeah. Right. Don't eat it. Yeah. That's cool. No, no, that's good too. I mean, like, uh, and I finally, I found the, um, the, uh, the shield you were talking about too. So. Oh, right. It is, it is still for sale on your site, right? I mean, like the receiver shield. Yeah. Yeah. That's cool. Hmm.

Jon Oxer: And that's kind of, um, I actually did another project, uh, outside of Freetronics. I did it for a super house and I made a little board, which is an add-on for a raspberry pie. It's essentially an Arduino with a 433 megahertz transmitter and a receiver connected to the Arduino, but it's in a form factor that lets a raspberry pipe plug into it. And then it communicates over the serial interface. So what happens is that the, you know, the Arduino can receive a, like a random 433 megahertz signal, and it can also transmit whatever signal you want it to. And it passes that information back to the pie. Yeah. Also on 433. Okay. So what it means is that I can, um, have it do things like, uh, and then I run open hub on the raspberry pie, which is an open source home automation, um, management, like a, uh, a hub, I suppose you would call it. Yeah. Where you can put in rules and get it to do things. So what I've done is set it up so that the, so that open hub can say, I want to send a code that controls this blind. And then that blind opens when it sends a signal. So, um, yeah, made quite a few of those. Now I've just been doing those on the side and, um, that's cool.

Chris Gammell: Yeah.

Jon Oxer: There are probably, I think about 30 houses now that have those installed around Melbourne.

Chris Gammell: What is the, what is the range on those, uh, on those 433 receiver transmitters?

Jon Oxer: Um, it varies a lot depending on the quality of the module itself, but they're, they typically will easily do 30 meters. So a hundred feet.

Chris Gammell: Oh, that's not bad. Yeah. Yeah. Like through walls and stuff too.

Jon Oxer: Yeah. Yeah. That's right. So normally if you put one of these in a house, it'll cover the house. Unless it happens to be big.

Chris Gammell: And you can probably do like repeaters and stuff like that. Yeah.

Jon Oxer: That's what I've done. Um, yeah. So for a number of houses that I've installed these in, they, they've been rather crazy properties to say the least.

Chris Gammell: Like, um, so these are, uh, maybe higher end clients. Is that what you're trying to say? Yes, that's right. Wait, wait, wait. Is the housing market in Australia a little bit, uh, the high end? That's weird. I didn't notice that when I was there.

Jon Oxer: No, well, not the ones that I normally see, but, um, yeah, for some customers that I've set up home automation systems for there. Yeah. You need a segue or something to get around the house.

Chris Gammell: Yeah, exactly.

Jon Oxer: Right.

Chris Gammell: No, that's what I expected. I mean, like I meant that it is like high end. Yes. High end customers, I assume have big houses, right? Yeah. Yeah. That's right. Yeah.

Jon Oxer: So in those situations, you sometimes have to set up multiple transmitters and have them all, um, linked together. So it sends it from the transmitter that is nearest the destination.

Chris Gammell: Mm-hmm. That's cool. And they'll just like pass messages on like, so it's like, do you have endpoint 35 near you? No, I don't pass it on to the next one. Do you have endpoint 39 five near you? No, I don't pass it on the next one. Like that kind of idea.

Jon Oxer: Um, the ones I've set up, I don't have it learning that what I do is configure it so that I have a list of end points and it says, you know, these end points are associated with this transmitter, these ones with that transmitter. So it just-

Chris Gammell: Oh, so there's like a lookup table somewhere. Yeah.

Jon Oxer: Yeah, that's right. I see. It's, um, I normally configure that in OpenHAB. So it's just a list of all of the devices around the house, which could be lights, blinds, you know, heating, whatever, door locks.

Chris Gammell: Hmm. Do you, uh, do you commit the cardinal sin of IoT in that, uh, you don't have, do you still have switch, localized switch control or is it all, all by wire or, well, wireless, I suppose?

Jon Oxer: Um, I-

Chris Gammell: Like if they hit a finger switch on the wall, does it still control the light or does it? Yes. Or does it completely replace the switch?

Jon Oxer: No. Um, I really don't like it when it's hard for people to come into a house and use it. And I am guilty of the sin of, um, of making my house harder for guests to use. And, um, so in my- It's a security measure, folks. Yes. Um, there was a, um, there was a project a few years ago that I wasn't directly involved in, but one of my friends worked on it. And the brief from the homeowner was that they didn't want a single switch in the entire house. Oh God. So his objective was that everything would be automatic. So you walk into a room and the lights turn on. And, um.

Chris Gammell: I just imagined myself in the dark waving my arms wildly being like, what is going on here? Yeah.

Jon Oxer: Um, and, uh, so that was what the customer wanted and the, um, the engineers argued against it and said, no, no, you really need some local control as well.

Chris Gammell: You may have to sell your house at some point, sir. Yeah.

Jon Oxer: And then it needs to be totally rewired. So.

Chris Gammell: Yeah. Right.

Jon Oxer: Um, so I've seen that done and, uh, I don't like that approach. I do like the approach where things happen, uh, automatically. So things should just behave the way they should, but you also want some sense of control. And if you want to walk into a room and turn a light on, you should be able to do that without pulling out your mobile phone.

Chris Gammell: And, uh, this was also without, without a touchpad, I think that's another thing. Yeah. Yeah.

Jon Oxer: Yeah. I think touchpads should be a supplementary thing where you want extra functionality, not just, um, replacing a, um, a switch with an entire touch screen.

Chris Gammell: If you want to change the color palette of your room, fine. Right. If you want to turn the switch on, if you want to turn the light on, I'm sorry, the switch, we are, we are at that point. It's a final control. Right. Exactly. Yeah. Until, until I'm talking to the computer for everything. And even then I probably don't want to turn a switch on like that. Yeah.

Jon Oxer: That's right.

Chris Gammell: Yeah.

Jon Oxer: Yeah. So in my house, what I have are, um, light switches in all of the normal places where you would expect them. But instead of the normal rocker switch, they're a push button. And when you push it, it sends a command to the home automation system. So it knows, oh, they've pressed this button and I need to turn this light on. So basically what it does is, um, it puts software control into the middle of it and you're not directly controlling it because power isn't going through the switch. Switch is just like a control input. But it also means that if you turn a light on with your phone, you can then walk over to the wall and push the button and the light will turn off again. It's, um, it's just another way of controlling the same thing.

Chris Gammell: I don't like that one, but it's your house. So it doesn't matter.

Jon Oxer: So you don't like the fact that there is the, the software link in the middle.

Chris Gammell: Exactly. Yeah. That's, and that's like, uh, that's like a soft power off button, right? Like, and like those. Yeah. Cause then like what happens when this, what happens when the software locks up? Yeah. Yeah. That is the problem. I still want to, I want the button to work even if the software is dead. Yeah. And that doesn't, that doesn't fit that. Yes, that's true.

Jon Oxer: Yeah.

Chris Gammell: Yeah.

Jon Oxer: So I have, um, yeah, that's a compromise that I've made.

Chris Gammell: Yeah. Well, you know, yeah, it's okay. You, you know, you know the trade-offs more than more than most. So, yeah. Yes. And also, you know how to fix the software if it does lockups. That's the other thing. I'm like, yeah. Right.

Jon Oxer: Yeah. That has been a bit frustrating recently. There have been some, there's been some weird memory leak issues in open hab, I think. Maybe in Java.

Chris Gammell: And you, you posted about this. This was one of your, one of your super house episodes. Wasn't it about, about going off to like a US based servers on Amazon or something?

Jon Oxer: Yes, that's right.

Chris Gammell: Yeah.

Jon Oxer: So, um, this was, this isn't my system. And in fact, uh, this was some other system that I saw where controlling it required use of an external service. Right. So, you might have a light switch in your house and a light that it's controlling. But when you press that switch, it's sending a signal off to some server halfway around the world. And then it's sending a command back to the light saying, hey, you need to turn on now. Right. And that sort of thing just seems ridiculous to me. Like if, right. If you.

Chris Gammell: Because even a best, like right now we're, we're in a pretty direct connection to California direct, like we're both down in California. You still have a 200 millisecond drop or delay. Yeah. Right. I mean, like it's just, that's a best case scenario.

Jon Oxer: That's right. But even more than that, I've got a, um, a very unreliable internet connection here. It's just because of the, um, the circumstances of where I'm located and the cables in the street, that sort of thing. So my router drops out regularly. It can happen a couple of times a day sometimes. And if, um, my internet connection dropping out means I can't turn on the lights, then that really sucks.

Chris Gammell: Right. Exactly. Yeah.

Jon Oxer: My general philosophy with home automation is that the home should never be dependent on external services. That's one of the rules. If you can't control everything locally while the house is offline, then you're in a bad place.

Chris Gammell: Yep. Yep. I totally agree. And, uh, I think even if you don't know, if you don't know where your data is going, right? I mean, like, I mean, how, yeah, it's like your data is going off into the cloud. Like, okay, that's fine. But that also means someone might be able to read when you're home or not. And like there's security stuff with that too.

Jon Oxer: And it's just, yeah, definitely. Yeah. It's just a whole, it's leaking a whole lot of information that just doesn't need to be leaked.

Chris Gammell: Exactly. And it's like, uh, I don't know. I, I, obviously I've been scared into thinking about security more than I ever had before. Uh, and I'm sure that a lot of my practices are not great, but, um, yeah, I don't know. Like that kind of, that kind of stuff. Like, okay. I'm going to be, I'm going to be a little honest here, John. I don't really get a lot of this automation stuff. Like, I don't think much of it's needed, but yeah, I know you're doing it for fun too, right? Like you, you're just interested in this stuff.

Jon Oxer: Yeah. So I think there are certain justifications for it and there are a certain, what are they? I'd love to hear them. Okay. So, um, a lot of people use justifications that I don't think are valid, like power saving by having lights turn off automatically. I don't think there are better ways to do that. But, um, one of the really interesting ones is assisted living. So either the elderly or people who are confined to a wheelchair. Um, I've seen quite a few people who have benefited tremendously from some kind of remote control or automation. And, um, that could be lighting. Um, something as simple as that. Um, it could even go as far as actuators on doors so that someone who is confined to an electric wheelchair and only has control using say a joystick of the wheelchair. Right. And like reach and stuff like that too. They can't necessarily reach a light switch, which is high on the wall. Or, um, you might come up to a door that opens towards you and how do you get through it if you're in a wheelchair? So there are lots of things like that. And, um, so I think one of the.

Speaker ?: Okay.

Chris Gammell: So assistive, assistive makes a lot of sense. You're right. And, and in those cases too, like those are usually like the monitoring is actually kind of a feature. Like, I mean, yes, there's still security concerns, but it's like, you're, you're hoping to watch out for someone, right. Or something like that. Yeah. A geriatric or handicap or whatever. Right.

Jon Oxer: Yes. Yeah, that's right. So I think. Okay. So that's one. You got one. You got one. I got one justification. So I'm planning ahead for my own dotage. One day I'm going to be stumbling around, unable to hit light switches and the system will accommodate me.

Chris Gammell: I hope you've updated your system by that point.

Jon Oxer: So I think that is a very clear, it is kind of niche, but I think it is a very clear application for this sort of thing. I think in a more general sense, there are benefits to it that it comes down to the old thing about, okay, so you've got a TV with a remote control. Do you get up off the couch to go and change the channels or do you sit there and press the button? And sure.

Chris Gammell: No, I change it from my phone. Come on, man.

Jon Oxer: Well, sure. You could get up and change the channels on the TV by pressing the, you know, the TV itself, but we've become accustomed to the convenience of it. And there are little convenience things that you can do with home automation. Things like grouping of actions together. One that I did-

Chris Gammell: Oh, so like real life scripting almost.

Jon Oxer: Yeah, exactly. So one- That's interesting. One script that I did fairly recently in my own house, and I'm going to do an episode about this in the future, is a leaving home script. And so what you do is hit a button on your phone or a button on the wall just inside the front door, like it's on the light switch. And it turns off every light in the house, closes the blinds, closes the electric windows, sets the door locks to locked. It can change the, like the HVAC settings so that you're no longer hitting a house that you're not in. So it can do a whole lot of different things like that just by pressing one button. And it avoids the problem of, oh, we're going out now. So walk around the house and, you know, make sure all the windows are closed and lock the back door.

Chris Gammell: I think I'm going to have to give you a minus one on this one though, too. I'm not sure that this, that's not like, I mean, like you're saying it's more for an insurance against whatever, but-

Jon Oxer: Yeah, well, it's like central locking for your car as well. It's the same sort of thing, but it's central locking for your house. So walk out the door, press a button and, you know, the house is locked and every light is turned off.

Chris Gammell: I bought $20,000 worth of electronics to make sure I didn't have to do the thing that would have taken me three minutes. Yes. Yes. You're right. Okay. I'll give you a quarter of a point for that one. A little bit of a point. Okay. You've got one and a quarter of points right now. Yeah.

Jon Oxer: Okay. So I'll have one final attempt at it. This is a justification for the electric blinds that we have around the house. So we have a conservatory, which is a glass walled room with glass, like a glass cathedral ceiling. And it curves around the end. So we have 12 blinds around the end of this room. And when they're open, basically the whole room is glass and you can see out in all directions. But opening and closing 12 blinds is a pain. And we would just leave them permanently open or permanently closed if they weren't motorized. So just being able to press a button as you walk past and they all open means that we open them when we wouldn't otherwise.

Chris Gammell: I don't think I can give you more than a quarter point on that one. I am coming across so badly, aren't I?

Jon Oxer: I know. This is like... I'm digging this hole and I just keep digging.

Chris Gammell: Yeah, man. I mean, it's like, okay, this is like high-end stuff. And like, I get it, right? Like, this is... But I think you just proved my point. Like, none of this stuff is necessary. Like, truly assistive stuff... I agree. Sure. Yeah. Like, that actually does make sense. But like a lot... And I think that's a lot of the things that like people are like, oh, why don't we have more connected stuff? I think that there's not really needs for it. But there will be some stuff that pops out, right? People have said, oh, you don't need a garage door opener at some point as well, too. So...

Jon Oxer: Yes, that's right.

Chris Gammell: But it's... I think probably more than anything, it'll become a normalized thing, right? It's like, oh, of course you're blinds or why would you turn the thing yourself? Yeah. You just have that automatic thing.

Jon Oxer: Yeah. It's not necessary in the same way as a TV remote control is not necessary and a garage door remote control is not necessary. Right.

Chris Gammell: Right. It's a further first-worlding of a first-world problem. Yes. Yeah. Okay. But I mean, but at the same time, I think the thing that's interesting about that, too, though, is that probably in the same way, right? Like the people that had the first garage door openers were probably very high-end clients. And then eventually, it just got... It dropped and dropped and dropped and dropped. Yeah. And eventually, it became normal. Yep. Yep. Yeah. Same thing with like Bluetooth and car, right? That's like almost ubiquitous these days. Yes. Only the highest-end cars had that at the beginning. Yeah.

Jon Oxer: Yeah. That didn't take long at all.

Chris Gammell: I don't know. I understand your skepticism. I live in a 500-square-foot apartment, so I don't really... Yeah. It's not much. Was that 10 square meters? Something like that?

Jon Oxer: Yeah.

Chris Gammell: No. No, 50 square meters.

Jon Oxer: Bigger than that. Yeah.

Chris Gammell: Yeah. 50 square meters. Yeah. I forgot. I thought it was 50 or 10 to the... 10 to 1. You know, whatever. Yeah. So, yeah. Well, I'm glad you're working on the problems regardless. Well, quote-unquote problem. I'm glad you're having fun. That's what I'm trying to say. Yes. That's nice.

Jon Oxer: Yeah. And I think that's probably one of the biggest adjustments that I need to make now with my change in my life circumstances is a lot of these things. I've been doing this because I wanted to and for the fun of it. And now I need to justify doing things because I need the income. Right. Money. Money's a thing.

Chris Gammell: Yeah. Yeah. Welcome. Welcome. The fact that you did it the other way so long is actually the most impressive thing. So, I mean, it's unfortunate how it all worked out. I do feel free on that stuff. That sucks. Yeah. Thanks. But, yeah. I think that you are entering a period of rapid growth and interesting problems to solve. And I'm sure you're going to be fine.

Jon Oxer: Yes. There will be many new opportunities. That's right. Great.

Chris Gammell: So, if people have an opportunity, how do they get in touch with you?

Jon Oxer: They can email John at Oxer, which is O-X-E-R.com.au. It's probably the best way. Okay. I am way behind on email. Only 30,000 or 40,000 unread, but still, give it a try. Yeah. Right. Twitter too, right? Yeah. I'm on Twitter at John Oxer. Yeah. And those are probably the best ways.

Chris Gammell: Yeah. Usually, Twitter is the one where it's like, if you really need to, you can keep pinging that person. Yeah. That's right.

Jon Oxer: Yeah.

Chris Gammell: Here's the real question. Do you get Twitter notifications on your phone?

Jon Oxer: I do. Yes. Then, eventually, you're going to get a hold of them. If you ping me on Twitter. Yeah.

Chris Gammell: People ping me on Twitter. I'm going to see it eventually. Yeah.

Jon Oxer: And if you want to see some of the totally unnecessary projects that are totally frivolous, check out superhouse.tv.

Chris Gammell: That's right. Right. Which also has an associated YouTube channel.

Jon Oxer: Yes. That's right. So, I put everything up on YouTube and then just embed it in the superhouse site. So, one of the things I try to do with that is not just make videos, but also make associated documentation and guides.

Chris Gammell: Right. Yeah.

Jon Oxer: So, it's a project page that has the video embedded in it. And then, it will also have things like links to the source code and everything else that's referenced in the video.

Chris Gammell: Yeah. That's great. That's great. Cool, man. Well, I'm going to definitely check. I want to check out this 433 megahertz radio thing. It's got me thinking about some things about my apartment, too, that I might, yeah, who knows? Maybe I'll have some connectedness around here.

Jon Oxer: Yes. You might go to the dark side.

Chris Gammell: Yeah. The one I was thinking about is like, okay, here's a first world problem for you. You go out running, right? Yes. And you don't want to carry a key in case you lose it, right? So, you ring up to the house, you ring up to the apartment, it has some kind of buzz, having like an auto answer, and then like somehow buzzing the person in, right? Like, so like having like, I'm like actuator sitting there pressing the buttons and listening to my voice or something like that. I don't know. So, that's my idea for our home automation thing. Okay. So, we'll see. Yeah. Could be an interesting project. Yeah. Cool, man. Well, we'll see you online, and we'll talk to you soon, I'm sure. Absolutely. Thanks for being on the show. Thanks, Chris. All right.

Chris Gammell: Bye.

Speaker ?: Bye. Bye.