Episode 813 transcript

FLOSS-813

Jonathan: Hey folks, this week, Simon Phipps and Aaron Newcomb joined me and we talk about an interesting smattering of topics. We start with talking about legacy computing and retro hardware and how open source still makes that work. And then end up with politics. It's actually a lot of fun and you don't want to miss it.

So stay tuned. This is Floss Weekly episode 813 recorded Tuesday, December the 10th. Turn off the internet. It's time for Floss Weekly. That's the show about free Libre and open source software. I'm your host, Jonathan Bennett, and sometimes we talk about a little bit of hardware here. And today is interesting because I was absolutely certain that we had a guest schedule to talk about the open source AI definition, and I sent an email yesterday, just checking up to make sure that everything was good.

And I got sent back, you never confirmed that we were going to be here. So we don't have it on our calendars. And I went, Oh no, yeah, that's right. That's one. So. Thankfully we have a couple of willing volunteers. Simon and Aaron jumped in sort of at the last moment and they're here and we're going to talk about a couple of things.

First off. Welcome guys. Thank you so much for being here.

Aaron: Absolutely. It's all your fault, Jonathan. It is.

Jonathan: It's all my fault. So this, this was basically the crew that was supposed to be here. We were just supposed to have one more, sort of another expert in the field of open source AI to talk about, for one thing, what that even means.

And we may get into that a little bit here towards the end of the conversation. But I've got, I've got here two experts in retrocomputing. For two very different reasons. ,

Aaron: well, I don't know, is, I don't know that they're that different. I mean experts because we lived through it. Is that what you're getting at?

Because we're so old 1, 1,

Jonathan: 1 being an expert because he lived through it and the other being an expert because you've got the YouTube channel on it. Yes. That was sort of what I was getting at. Okay. Yeah. Okay.

Simon: You know, I have to say that I, I, the, one of the things that makes you feel old is going to the Computer History Museum and walking around and going.

Hey, that's my computer.

Aaron: Yeah, I worked

Simon: on that. Yes. And I, I, my, my SWH 68, 000 is indeed in the computer history museum. So you know, that's retro computing. Yeah. We, we call that a live systems production. That's cool. Yeah,

Aaron: I get asked, I get asked a lot by the younger crowd, like, you know, the question always starts with back in your day.

I was like, all right, were Atari joysticks really like this when you, yeah. Okay. Yeah, I get it now.

Jonathan: I'm old. Yeah. Yeah. I'm starting to feel that too. I had somebody the other day. We've got some really cool retro computers. You should come and take them, take a look at it. And you know, I'm expecting a certain era of machines and I get in there and I'm like, yeah, here, it's this Windows 95 machine.

I'm like, okay,

Aaron: That's not right. I try not to be too. I try not to be too. Yeah, because there is starting to be That nostalgia for things even out of the 2000s Now, you know and it's like well, yeah, it was 20 years ago. So I I get it. I it's not my thing, but I understand, you know So that's what I usually tell people.

It's about a 20 year window, right, where people start to feel nostalgic. Until then, it's just old crap, you know, and then all of a sudden it's like, oh, I remember how long that took to do.

Simon: Yeah. Yeah. So one of the curious consequences of always assuming that Microsoft was going to audit me I've, I've kept all of the original media for, I have it, I have it for Windows 3.

I have Windows for workgroups, I have Word version 1, I have all the original media and original packaging all down in the office, so I'm waiting for a museum to call, or I need an auction house to give me a call.

Jonathan: So one of the, one of the things that I have seen both on Aaron's channel on RetroHackShack and some other people's channels is that we're starting to see like open source, both software and hardware get applied to some of these old retro computing things.

And like just a couple of things that come to mind. I've seen, I've seen people take like Commodore 64s. I am the age that I consider a Commodore 64 to be retro. Simon, that's probably. That's probably just old to you, right? ?

Simon: Yeah. I, so I never, I never had a Commodore 64. I, I, I worked on Sinclair qls.

Mm-hmm. They were, they were my, my thing. 'cause they had built in networking and everything. But yeah. That's good. You know, I've got friends who've, who've got those on a shelf somewhere.

Aaron: Yeah. And the, the QL was the one with that little tiny tape drive, right? Yeah. With

Simon: the micro drive. Oh yeah, yeah.

Aaron: The micro drive.

Yeah. Yeah.

Simon: But I, so I actually used to run a network of them for an office. And because the QL had had built in networking in the hardware. So you just ran a cable from QL to QL, and they could all share each other's drives and resources. And so I added a hard drive on, and so that we effectively had a hard drive in the cluster.

And you add a printer on, and everyone can access that. It was actually really great computing. It was much easier than anything I ever encountered subsequently on PCs.

Aaron: And I think that, I think that one thing that lends itself to open source in retrocomputing is there was always an element of, even though they didn't call it open source back then, but certainly hacking right on these things.

And those became projects that were published in a magazine that anybody could pick up and do. So there, there is some, for me anyway, there is some root, some kernel that was going on back in the seventies and early eighties. That was akin to open source to an open source project where someone would come up with something It would get published in bite or it would get published in popular electronics or something like that and maybe they got paid for their article, but the point was they were sharing that information Yeah with the community and anybody could pick that up and do whatever they want with it if you go back all the way to The TV typewriter, for example, and different people will say, Oh, the TV typewriter, I think like Waz, for example, said the TV typewriter wasn't an influence in the Apple one, but it was, there was some, some sort of DNA there because what was used in the Apple one, which led to the Apple two, which led to.

Apple, as we know it today used all the same elements that were in Don Lancaster's TV typewriter. There was some memory there. There was a character generator. There was a keyboard. There was the way that it interacted with the rest of the circuitry was there. And that was published in a, in a, you know, one of those early magazines.

So I really do think that there's some open source DNA in there somewhere to the whole. Computing back when it was actually happening in the seventies and eighties.

Jonathan: Well, so something, something important on that, on that note, something important to keep in mind is that when, when, when RMS came along and came up with free software, he wasn't inventing the idea of let's make software freely available.

Like he was, he was in fact saying, this is the way it was. And this is the way it should still be and then they came up with the licensing hack that is the gpl And it's the same thing with open source when when the open source definition first was was kind of developed They were not they were not inventing open source.

They were looking at this is this is what people have been doing This is what's worked and let's try to codify that. So yeah It's, it is a much longer tradition than, you know, just the, even the free software foundation or the, the, even just the OSI, it has existed longer than that. It's just, those are, those are the groups that we sort of have to thank for, for taking all of this nebulous idea of let's share the code, let's share the information and sort of distilling it down into, let's use this as a definition for that.

Simon: Yeah, I mean, it all comes from a real long time ago. I've been working with somebody on a short book recently, and you'll likely love it, you know, because he's written a book about why open source is capitalist, and he started out by talking about how Stallman invented free software, and then he looked back at history and discovered that Stallman was really quite late to the game.

Yeah.

the folks over it to Berkeley University were there ahead of him. Were it's just, they, they had a different way of expressing their expectation. Their expectation was that this was before the IBM consent decree. And so it was before. The productization of software. And so there was no need for a GPL because there was no value to software.

People just gave it away. And but the bill joy was there creating BSD and then going on and creating the world's first open source software company, which was some microsystems and people look, they, they, that's so far back that people assume that it all happened in the nineties, you know,

Jonathan: It's interesting. You mentioned that book, you know, Steve Levy's book, Hacker's Heroes of the Computer Revolution. That was sort of the introduction for, at least for me. And I think for a lot of us to some of this history and he makes the exact same point, but with Stallman and I think it was the, the, the hacker club at MIT where they were playing with the PDP machines and Stallman came again.

He was late to that game. He was towards the end of it and he saw things. Changing away from the, let's just all share this code so that we can all make it better to again, this idea of productizing it and the free software foundation and the GPL itself was sort of a response to that. And so, yeah, it's super interesting to see the, you can sort of trace that line down through history to where we are today.

Aaron: Yeah, so the, so the, so the nuts and bolts have been there for a long time. But now it's kind of come full circle where, you know, people are creating these, these cool hardware and software projects that can then bring these retro computers and give them more functionality or make it easier to do things for people that, that maybe aren't you know, that don't want to do it on their own.

And there's some great examples of that. Of course you know, the one that comes first to my mind is RGB. Yeah. To HDMI which is an open source project. And I sell some of those boards on my site, shameless plug. But you can also, it's an open source project. So you can go get, I had a guy this morning that emailed me saying, are there DIY versions of this thing that you sell that I can just go do on my own?

And of course the answer is yes, you can. But yeah, this project allows you to take an old. Sinclair QL, for example, or TRS 80 or whatever, you know, computer you, you want to use that had some of these older protocols and allows you to hook them up to your modern HDMI. Monitor. So you take the output, convert it to HTMI and, you know, cause a lot of these old monitors aren't available or are very expensive.

I mean, try finding a IBM 5154 EGA monitor these days, it's going to cost you three or 400 bucks and hopefully it's working. It probably isn't, or probably won't be for long.

Jonathan: Or it's dim, or, yeah.

Aaron: Yeah, and not everybody can fix those. They're awfully hard to work on. So yeah. So that's a great example that, that first comes to mind is, is that project because it works with so many retro computers and is so accessible for people.

But there are also a lot of projects that are coming up right now around the Raspberry Pi Pico. So, and of course there's tons of things going on, but, but especially in the retro community, people are taking the Raspberry Pi Pico and making it do things inside retro computers as well. So there's two projects there that come to mind.

One is the PicoMem. Which adds memory for older IBM computers with a ISA slot. You can put this in there and the Raspberry Pi Pico will emulate the memory and give you, you know, two, four megs even. Can you believe it? Of memory for your old IBM computer, right? And what people found, I think, partly based on that product and based on other products, once they had the ISA bus figured out, is that they could do all sorts of other things.

So there's another project out there called the PicoGus and that project has gotten a lot of momentum lately, which actually takes the Raspberry Pi Pico and emulates a Gravis ultrasound. That's what it is. I was trying to think, what does G U S stand for? Gravis ultrasound, again, incredibly hard to find these days.

Not very many people had them cause they were expensive when they came out. But you know, this is an a sound card that was that worked with a lot of games. You added it onto your Sound Blaster and it did incredible, incredible things, but they're super hard to find these days, but it'll emulate a Gravis ultrasound.

Or a sound blaster or a a Roland system or a Tandy three voice system. It'll do all of that just based on this little, I don't know what they cost these days, 4, 5, a little open source or open hardware microcontroller. That's, that's cheap and easy to use. So there really is a lot of, a lot of great open source projects out there.

And I, I, I think it speaks to the legacy of open source. The fact that open source is established so much now that when people create these projects, the software that they create. Of course, it's going to be open. Of course, I'm going to give it a GPL license or whatever. It's not like a question where you have to convince people, Oh, can you please, like, just make that open source?

No, it's just like by default, right? They're choosing to make this open source and it's for the betterment of the community.

Jonathan: Yeah, every once in a while I come across a project where it's like somebody obviously, like they'll put it on GitHub and everything. They obviously intend for people to just grab it and use it, but they won't put a license on it.

Aaron: Yeah,

Jonathan: like you you realize that you're all rights reserved by default, right? Please pick one of the open source licenses I don't care which one you pick just one of the OSI approved licenses and just to add the file like here I'll even make the pull request for you. Just accept the pull request put a file on there.

That way we can use it Yeah,

Aaron: yeah. Yeah for people that don't know I think github does a pretty good job though of pushing people into choosing a license when you start a project Which is good. They try to at least Yeah Yeah. Yeah. Now , so there's a lot of there's a, there's a lot of great projects. One last project I'll, I'll mention is one that I just found and I'm super excited about it.

It's called The HID Man. Hid Man. Mm-hmm . And you might know HID from USB HID devices. You know that that's what makes USB so ubiquitous and easy to use. You can kind of like plug. A USB mouse into anything and it just works. And so there's a guy, he goes by the name of Rastiri on GitHub, but he he also has a YouTube channel and he's been working on this device that takes.

USB devices and converts them into either PS2 compatible or AT or even XT compatible protocols so that you can take, if you don't have one of these old keyboards or whatever, or you just want to use, like in my case, I have a fancy new mechanical keyboard that sounds wonderful. I love to type on it. But I want to use that on my retro devices.

But it's a USB keyboard. You can now take that plug it into this device and then plug from that device into your computer and use, you know, a mouse or, or a keyboard on your, your retro computer. And it's just a little tiny device that does a wonderful job. He made it open source. And I talked to him about it.

I'm going to be carrying the hardware portion of that on my, the hardware is all open source as well, but I'm going to be going ahead and making some of those and putting them on my shop as well. Because. I think it's just a, it's just a great, it does such a great job. One of the things I love about it is the, in order to do the configuration, because he has in, in the software, he has control over the PS2 keyboard.

Let's say you're using, when you want to change the configuration, you open up any text editor. Maybe it's just edit in DOS, right? And you push a button on the device and it prints out in the editor, what the configuration is. So instead of having to open a file or go in and change something, you know, it just prints out the configuration and it says, what do you want to change?

And you hit one. And of course he can interpret that he knows you hit one. So then he says, okay, I'll change that configuration to this. You know, this is like a brilliant in my mind is something that would have been done back in the early days. And. He's using this to actually allow people to change the configuration on this device.

And I was like, that's, that's cool. That's cool. And it's brilliant. And I love it. It's

Jonathan: kind of like using a teletype machine almost. Yeah, exactly.

Aaron: Exactly. That's fine. So

Jonathan: one of the things that fascinates me about this, and I've got the same bug and it's this idea of, of so someone, someone might ask.

Why don't you just emulate? Right? Like, so if you, if you want to play with a Commodore 64, well, there's Commodore 64 emulators. Why bother with the real hardware? Why bother with any of these adapters? You know, why would you use a, a machine that actually has an AT keyboard or an XT keyboard plug? Why not just emulate it all?

And, I don't know that I know the exact answer, but there's just something special about running the real hardware. And, and, Why, why is that?

Aaron: Yeah, there is something about, about running on actual hardware. It's, it's a great test bed. I use emulators to test stuff all the time. Testing code. If you're writing code, for example, for some of these old things, it's, it's a lot easier to pull out an emulator and just do it that way.

But even what I find interesting is even the younger crowd, right? Kids to me, anybody younger than 20 still has that same response to the actual hardware, you know, and you would think they would be like, I'm just going to run this on my steam deck or on my phone or something, but no, when you. When you give them the actual hardware, it like something clicks in their brain.

It's like, Oh, it's tactile. I can use it. I see how that worked. I understand like how the cartridges work and how the joysticks works and how that all came together. In, in a special way, I think because all of those systems at the time were, you know, there wasn't a lot of standards back then, you mentioned the Commodore 64, you know, that was as different from an IBM PC as it could be.

Right. But. you know, learning how the systems work and how the hard drives work. And you had to kind of be your own expert on some of those things. I think there is just an appeal to people that like to learn about old systems of having the physical cartridge, the physical hardware to play with, seeing how the keyboard felt in those days, how it was either bushy or clicky, or, you know, there's just something, something about that tactile response to those systems.

I mean, Simon, you mentioned those, the micro drive, right? Yep. For the Sinclair QL or the keyboard on the Sinclair QL, I know was quite a bit different. Of course, it wasn't the the, the old plasticky one of the original Sinclair, you know, ZX Spectrum, but

Simon: the membrane keyboard, only one keyboard. And so with the key, with the, the space on the top of the case with the keys attached to the, the the, the printed circuit board that was underneath, I mean, the whole thing was a single, was a single board computer.

And that's obviously a desirable thing because I, you know, Raspberry Pi are coming out with the Pi 500 at the moment, which is just the same. And it looks just the same. Just like what I remember my Sinclair QL looking. I mean, it's white, but it's very similar kind of experience and they expose all the ports on the back.

The GPIO pins are all out there. I think that one of the things that makes it compelling. Knowing that you can

yeah,

I think that the the difficulty with with lots of stuff, you know I've got it. I've got a device here that I picked up from a well known Scandinavian furniture shop that's for controlling my Sonos speakers And and it's it's all sealed And I can't do anything to it.

And if I take it apart, what's inside is sealed and the box that it's controlling over there is sealed and I can't mess with it. And all this stuff is stuff you can, you can get your hands on and not because anyone who made it intended you to do that, but because the people who wanted to stop you doing it are all dead now.

And there is that sense of, I've got this. This, you know, ancestral artifact here and I, and I can, I can touch it and I can, and nobody is compelling me to do things in a particular way and there's problems to solve, but I'm not going to find myself getting a cease and desist letter from Sony for doing it.

And, and I think that part of that is what's compelling, and I think it's, that's the same thing that's compelling about Raspberry Pi. It, it is that you can, you know, it's got all the GPI pins are, are on the back. You, you know, you can take it apart. They sell the whole thing as a chip you can put on your own circuit board, and people do.

And I think that's, that's always been, you know, the things that have always driven me in computing are the fact that I can and also the idea that I can do things here that make something happen over there. Those have been the two things that have always driven me in computing is action at a distance and being in control of my own destiny with my hardware.

This when when people try and take those things away from me, that's when I feel diminished

Jonathan: This is why simon and I get along so well We we have those two very core core to our being things in common All right. So simon, do you find yourself or maybe you have this? You mentioned the the Sinclair. Do you, do you have some of those still kicking around that you, you boot?

Is that something that even appeals to you?

Simon: Yeah, so I do have a box up in the attic that's got a couple of Sinclair QLs and I've got, I do have a, a, I do have some other Sinclair hardware up there and we have a dinner guest who comes every weekend for, for Sunday lunch who keeps a compute library.

He's, so he's got his old systems and we have our eight inch floppy drives and things, but I found with the QLs that the QLs are very

mechanical

and their mechanicalness is also very what we would call in this country, Heath Robinson. I forget what your term is for it over there. It, it, it, you know, it, it does involve rubber bands and, and it does involve little thin pieces of cassette tape in a, in a little plastic box.

And they've all stopped working and getting them back to being, working again is gonna be really quiet. A challenge. Yeah. And and I know I could, I've got a degree in electronic engineering. I could get them working again.

Yeah.

But there's other things that are more interesting to me at the moment.

Sure. So. They stay up in the loft, really.

Jonathan: And that's something, you know, kind of coming at things from the, like, even the Hackaday perspective. That's something I find fascinating about some of these retro machines, you know, in the right era. The parts inside are just the right size that you can actually get in and work on.

And so they're, they're super useful for young people. Teenagers, people in their 20s that are like working, maybe working to get their electronics degree or just got it Or they're in high school and they find all this stuff really fascinating You can actually get inside of them and work on them you can actually take these old machines and fix them and We're still thankfully in a kind of at a time period in history where if someone has a you know a not working Again, I keep coming back to the Commodore 64.

It's the best selling computer in history, so I guess that's why. But like, you can still find not working Commodore 64s on eBay and places like that without having to spend a fortune on them. And so it is, it's really fascinating for, for all people of all ages, but specifically young people. You can grab them and really get into the hardware, and it's a good, it's a good learning platform, I think, among, among other things.

Aaron: Oh, it's a tremendous learning platform, especially since in the olden days people would actually have schematics for these things as well, right? So if you get a bug in your, under your whatever, and you want to actually go in and fix that thing, or you want to learn why it's not working, you can pull out a schematic usually.

Or a service manual, you know, I mean, to Simon's point about things being sealed up and not people don't give you any information about what it is or how it works anymore, but they're used to and give you a full service manual with a theory of operations and a schematic and you could go in and you could learn to your point, Jonathan, about how it works.

You could learn about electronics or any of those things just by, you know, breaking open your home computer and getting in there with a screwdriver and a soldering iron. So it is a tremendous learning experience just because those things were available at one point in history.

Jonathan: I've got a couple of devices, old devices that You can tell like you pull you pull it apart and you look at the pcb And you can tell someone wrote this pcb out by hand Right, like there was a time where that was that was the way things were done and then you know, you would get a you You would replicate it, but it would the first thing it would be someone would literally draw it out by hand And on the other extreme of that today, you know you talk about You get five layer PCBs, you get eight layer PCBs made to be able to fit all the things into a tiny little space.

And so, you know, on on one hand, we get enormously more powerful devices. And and so, you know, we're we're not luddites. We appreciate the the enormously more powerful devices. But on the other hand, that means that it's so much more difficult to work on, to understand, and The, the old stuff, the old stuff is pretty neat, because, again, you can, you can see inside of it.

You can understand how it works on the inside. That, maybe maybe not by accident, is an interesting segue into something else that is enormously powerful that is very difficult to see inside of.

So we were going to talk about open source AI and the open source AI definition, and we've got half of our expert crew here. And so

Simon: You got the skeptic department here.

Jonathan: Well, that's interesting because I find myself being sort of skeptical about a lot of these things too. So maybe, maybe Aaron will hold down the, the AI enthusiast side of the conversation for us.

Sure.

Simon: So we, we, we, we've worked on the open source AI definition at OSI. And I was the staff skeptic. I stayed out of the process completely because somebody had to do the other stuff that wasn't AI, so I got on with it. With you know, looking at the legislative schedule of the European Union and influencing that for two years and, and I kept on throwing rocks at the AI stuff, you know, I, I would make rude comments like, isn't it strange how all those people who were into crypto last year into AI this year?

No. And but nonetheless, you know, I've, I've got a reasonable amount of respect for where we've got to. We do have open source AI definition. It exists for excellent reasons and it's the subject of a program that you'll be doing in January, I think.

Jonathan: Looking forward. I am very much looking forward to that.

You guys, you guys have gotten some pushback on that though. People don't like your definition.

Simon: Well, actually, when you, when you look at what people are saying about it, you find that Generally speaking, free and open source software people analyze the problem the same way. You look at what the, the, the way that the Software Freedom Conservancy has analyzed the problem.

You look at the way the Free Software Foundation has analyzed the problem. And they've, they, you know, we look at it the same way. The, the, we, we recognize that it's a, A system that involves software and data that it involves life, life cycle phases. We, we, we assume that for something to be open or free, it has to include all the source code to the software under an approved license, that it must include all the code, all the, the, the parameters that you need to configure the system, and that it must include all the data that you need to populate the model.

And, and like, like all good high detail movements what we're arguing about is one of the tiny details, which is, well, how much of the data do you actually need to trade and train the model? Yeah. And the answer to that question is quite complicated. The answer is that you don't actually need all the data to train the model.

The problem you have is that you don't know which of the data you do need to train the model. And so you give the model all the data, and it It then trains itself with the bits you needed, and if you could magically know which bits it needed, or if you could artificially synthesize those bits so that the model could be primed, that would be great, but we keep on being told by AI experts, and it's important to talk to AI experts as well as free software experts about this subject.

We keep on being told the data doesn't really matter. You can use any old data because the model will end up roughly the same. And also that they, the other thing that they say, which is really head scratching, is they say that if they run the training process twice, they get a different, a different model at the end with even using the same data.

And so the things that free software people want is they want the, The full corresponding source, you know, they want the, the, the form normally used for making modifications to use the software language about it. And you talk to the AI experts and they say, well, there isn't really one of those. And so all the arguments we're having have been up have been from free software and open source experts trying to force fit the free software concepts.

Yeah. into the world of A. I. And the answer is you can't do it. And we then disagree with each other about about the compromises that have to be made to try and do it. Now, I think the A. I. Definition we've got is is perfectly adequate for the next step, which is stopping meta taking the market over by claiming that their llama is open source, even though every any fool can see that it's not.

That's the big presenting problem. The big presenting problem isn't, you know, is it okay to train a model on transient data that you can't give to people? The real presenting problem is we've got these completely closed AIs that you can't see anything inside that people are saying are open AI.

I mean, it's even in the name of the company.

Or they're saying is open source AI. They say about LLAMA and neither of them is in any way. Well, LLAMA is actually, you know, less closed actually. But the real reason we need the definition now is to, to make sure that we've got time to have this conversation, because if we hadn't made this definition now, the conversation would be as moot as the conversation about you know, about cloud is the free software and open source community completely blue cloud.

Yeah, we made no ethical directive statements that helped. And consequently, the cloud industry has gone whoosh off in that direction. It's completely out of our reach. And the same thing was going to happen with AI. We were going to have an argument for five years about, you know, whether we should call it AI or machine learning.

And meanwhile, Meta was going to completely control the market and tax everybody who innovated. So you know, i'm i'm glad we've got the definition. I don't necessarily agree with all of it but I'm glad

Jonathan: we've got it. And it's, I guess, it's important to point out that like, you guys are not writing the licenses.

You're writing the things that must be in those licenses, like the minimum baseline requirements for them to be considered an open source license, right?

Simon: Yeah, we don't, so ISI has never written an open source license. What, what we've, what we have in the open source definition is a you've heard people say, you know, I don't know, I don't know what the, I don't know what one of those is, but I know it when I see it.

Well, the open source definition is the list of things to check to make sure that it's one of those. And that's AI definition is as well. It's a, it's a list of characteristics that let you know, Whether your freedoms are going to be respect respected that whether you're going to be able to take that AI and do the thing that you wanted to do rather than the thing that somebody else wants you to do.

And I, I think the definition we've got is pretty good for that. I think it's a little bit over generous with what it allows in terms of availability of data. But I think it's got it absolutely right on the source code. I think it's got it absolutely right on the model weights. I think it's got it, Absolutely right.

All the way down to this continuum of how much code do we need. And no one has been able to say how much code you need. And so there's some people say, well, so to be on the safe side, we want all the code. That means we'll never have an open source medical AI. Because, you know, medical AI is never going to ship with all the original data.

Aaron: Right.

Simon: And that's, that's the AI I want most is an open source medical AI because I can never get an appointment with a doctor. So, you know, I want to be able to ask my computer what that rash is down there. Why I can't bend this arm. And yeah, here's a,

Aaron: here's a picture. What, what, what is this? Yeah, yeah, yeah, exactly.

Simon: Do you recognize this rash?

Aaron: Yeah. Yeah. Here's my symptoms. Here's a picture, diagnose it for me. And there's already been examples of that happening, right. Where people have used AI to get a diagnosis that, that many doctors couldn't figure out. And then when they see it, the answer that they got back, like, Oh, of course you put these things together and it means you have this, this thing.

But, you know, AI did it in. A couple of seconds.

Simon: Thank God it got it right. I mean, the big problem with all these AI models. So, you know, you're getting into my skeptic self now.

Aaron: You

Simon: know, the generative AI models, you have to remember that their purpose is to persuade you that they have responded to your prompt.

Their purpose is not to tell you the truth.

Yeah.

And sometimes in persuading you, They accidentally tell the truth, and that's really handy. But the people who should be interpreting the output are the people who don't need the AI. The AI is there to help them get to an answer sooner, not to help them get to an answer without an expert.

Aaron: And,

Simon: I hear so much of this discussion being about, oh, you know, the AI hallucinated. The AI made an error. No, it didn't make an error. It was trying to persuade you that it was right. It has no idea what right actually is. Yeah. What it does for a living is persuade you. And, and it did a great job because you were persuaded.

And and the fact that you're persuaded about something that was completely false is beside the point because it's not here to tell you the truth.

Jonathan: Yeah. Yeah. You've got even, you've got even another, like, angle, another point to dig into with this, and people call it trusted AI, which sometimes boils down to we always want the AI to give you the answer that we've deemed appropriate, and there's a whole, like, Boy, that's a whole ball of wax in and of itself, right?

But then on the other side, you have things like one of the car manufacturers here in the United States, you know, Ford or Chevy, one of those guys, they put put an AI chat bot on their website and someone convinced it to sell them a car for a dollar. And apparently the courts upheld it and said you put the AI out there as your official spokesman So you you sold him a car for a dollar have fun with that.

Yeah

Aaron: That's funny.

Jonathan: I I've I've made the point repeatedly that I am I am just I am waiting for the next big thing So simon mentioned this with the idea of all the cryptocurrency guys are now into AI I'm waiting for that crowd to move off to the next big thing. Whatever it may be Maybe it'll be back to cryptocurrency or whatever like the the AI bubble will kind of pop And then we'll actually be able to see what it'll be useful for as a tool, once people stop trying to fit it into everything.

I'm looking forward to that.

Aaron: Yeah. That'd be nice. But again, that speaks to why this type of definition is so important right now, right? And even regulation. I'm not a big fan of regulation. But if you're going to be looking at putting regulation in, now's the time to do it. Because as fast as cloud moved, AI is moving a hundred, a thousand times faster, maybe, maybe even faster.

So if you don't do it now. It's, it's over, right? It's gonna be very hard to put the genie back in the bottle. Yeah,

Simon: I'm, I'm curiously a fan of that. I, you know, regulate early, regulate often is the, is the, the watchword. I think that what happened in Europe is we saw the, the absolute need to have privacy regulation on the internet.

Which was which became that became necessary like 15 years ago and GDPR was way late to the To the to the scene and as a result the scene that it was way late to was already extremely important economically, and so they compromised and they, they got it, they got it wrong. It's, it's an awful piece of legislation with that, that, you know, if anything, strengthens the arm of the people that it was trying to control.

And that's why I quite like the AI act in Europe, you know, it's not, it isn't perfect, but it's come out early enough that you can change it. And it can actually address the real problems rather than coming out in 10 years time when it, when it, when I is already a trillion dollar industry and you can't possibly regulate it because everybody will simply remove your political funding, regulate, regulate early and then you Change it every time you find it's wrong and that way you end up with a regulation that fits.

So I'm, I'm a big fan of regulate early, regulate often. I think that's the way to go.

Jonathan: You know, when you first said that I bristled and I'm like, ah, regulation, it's terrible. We shouldn't be doing it. And then you actually explain what you mean. It's like, okay, that's a reasonable approach.

Aaron: As long as there's a mechanism to fix it, right?

As long as it's not.

Jonathan: Yes.

Aaron: You know, a U. S. Amendment to the Constitution, which never happens anymore, right? But as long as there's a mechanism in there where we can say, go get, you know, go, go fix this thing and make it better, then it's fine. What is that? I'm not familiar with the European. What is it called?

The act or the act?

Simon: Yes. Yeah. So it's it's it's what it's it's It's been quite widely criticized by, by tech utopians. But it, it puts in place some regulations that will stop harms happening and they may be over conservative, but I think that it's being held reasonably lightly in the hands of the regulators.

And when you look at. In, in Europe regulations, you find they get changed, they get amended quite often. And so you look at a regulation like GDPR, you find it's, it's getting modified many times a year. By other political instruments. So by, by delegated powers from other pieces of legislation or from new pieces of legislation that modify the early legislation.

And the problem with GDPR is there's, there's all that awful money associated with spying on us and regulating it harms the people who are getting rich out of all of your and my private information. And so I, you know, what was needed was to stop them from creating the advertising surveillance industry in the first place, rather than to try and regulate it after it existed, because you know, when you try and regulate something evil, you have something evil and regulated.

Jonathan: And the other, the other side of that that we see here in the States from time to time is when you add regulation onto an industry, it is a minor burden. For the established players, but is a nearly insurmountable burden for anyone coming along trying to do anything new And so, you know when when you hear people in in my country decrying regulation, that's sort of the thing that they're getting at and again, it's one of those where when We try not to get political on here on the show, but I think we're going to, we're going to get into it just a little bit today by the very nature of the same things we're talking about.

So when you hear like a conservative talking about, we got to get rid of regulation, that's what they're getting at. We've got some of these regulations that are just because there's so much of it and because they're written to the benefit of the established players, it, it can be a problem. And again, it's one of those deals where, If you, if you just, oh, let's see, how should I put this?

If you just come at it from the kind of the partisan viewpoint, well, you have an an A versus B and, and, you know, the other side is wrong. But if, if you have something to where you can kind of cut across that and I think open source is an interesting tool for doing this and, and listen to some of the points that are being made, there, there are some things that we have in common here that people on the very far left and the very far right, if you actually have a, like a A well intentioned, good faith conversation.

There are some points that you can come and actually agree about. And that's sort of a useful thing.

Simon: Yeah. It's a, it's a, it's a turning fine line to walk. So, you know, before, before we, before we, you pressed record, we were talking a little bit about that. Yes. And What we found at OSI over the last 25 years is it's really important that we focus on the, on our swim lane,

you

know, and our swim lane is open source licensing.

And the reason that swim lane is so important is because every side. And there is, there is at least three sides to this conversation. Every, every side recognizes the need for a copyright anchored license that gives you the rights to exercise your freedoms. And. So as long as we focus on that, everyone's happy with that because, you know, there are some people who are, you know, they're very keen to have that happen so that their freedom to make sure everybody can use the software under copyright on the copyleft terms is protected, there's other people who are very keen that there's The license makes it available to everybody without restriction.

There's other people who are very keen that the license protects their, their rights in a way that is, is, is, is lightly touching the software, like Mozilla license does. And So that means that within the community of people who are wanting this to happen, we can have people who are, you know, over at the libertarian front or over the, the, the social good front or at the, Over at the intensely personal.

Leave me alone front. And and they can all be served and we don't have to argue about the things that we disagree about. And that's where we're going to run on the rocks with AI is because already in the conversation there is all of this is being woven together with ethical purpose and acceptable use policies.

And if there's one thing I've learned over the last decade, it's that what is an acceptable use to one person is an invasion of freedom to another person.

Jonathan: Yep. Or, or, One of the,

Aaron: one of the,

Jonathan: oh, go ahead, Jonathan. Or, so, or, it falls into their definition of, of evil. Like, that's not a thing that everybody agrees on.

And there've been some, there've been some brouhaha's in various open source projects, like the X, Y, or Z military uses some of this source code. And so, and to, to one group, that's the U. S. military uses it and therefore it's evil. And to another group, well, the Russian military uses it and therefore it's evil.

And I don't know that I could say that either of those. are like illegitimate points of view. They don't work well together when you try to add the, except the idea of acceptable use into the conversation, which I think, I think that's why it was, it was brilliant that when the open source definition was first founded, it was explicitly stated that we are not going to include.

Definitions of even no, no acceptable use policy because even back then they realized, and they could see it even in culture and, and, and all of that, that nobody's, we are not going to get on the same page on this.

Simon: That doesn't mean there's no place for them. So so I think it's absolutely fine for a project community to say that we are here to achieve these objectives.

Jonathan: Sure.

Simon: And the, and we're going to make sure that there's no one in our community that doesn't want to achieve these objectives. And when you find that excludes you, you fork, because you're free to do that, because the software license makes you free to fork. And then if there's a critical mass of people who share the other view, you get two pieces of software that are serving.

But you, that, that is a, the, that is a role for the governance of the community, not for the license of the software. very much. And so that's, that's the line that I, I think I'm still trying to work out how we draw on AI. Yeah. Because we're being encouraged by legislators to fold acceptable use policies into the core licensing.

And that isn't gonna work. Right. That's That's going to result in the fragmentation of the community down to individual company sizes. And that's going to result in there not being open source AI and thus not the, we won't get the social good of the network effect of massive collaboration.

Aaron: Yeah, I think there is a, I'm hoping this is true and maybe you guys can corroborate this or not, but I think that what you were alluding to there is this idea, I always worry, right, about open source, that it's going to be somehow influenced by, The conspiracy theorist, right?

It's going to be given a bad name. Just the reputation of open source could be affected or something. But I remember to your point earlier, one of the most interesting times I ever had at Sun was when we were down at Fizzle in Puerto Alegre. And I was giving a talk, so I was in the speaker room, and I'm sitting there, and there's two guys talking, right, in the corner.

And one of them is Richard Stallman. The other one is Peter Sund. I don't know if you guys know who Peter Sund is. I have a

Simon: photograph of that conversation. Do you? That's awesome. I do, yes.

Aaron: Wasn't that crazy? That was like That was reasonably

Simon: wild. You know, Michael Tiemann was also involved in that conversation.

Aaron: Yes! Like, how do these people get on the same page, right? Like, you would think they would be diametrically opposed or in some way, you know, how does the guy that created, or one of the guys that created Pirate Bay, sit down and have this philosophical conversation with Richard Stallman? But they were having that conversation, and I think it was, you know, my point is, can open source be more of a unifying factor?

And of course, this is again, why we need to get there with AI. But can it be the leverage that we need between all of these different political groups with different opinions to say, look, That's the nature of it. If you don't like it, fork it and do your own thing and see if other people like it. You don't have to have this conversation like this is, we need to change this or it's the end of the world because you can just go change it.

Jonathan: Yeah. So I'm, I'm, I'm involved in a, to, to kind of tie into this, I'm involved in a project, it's mischastic. And we, we have kind of a, because of what it is, it's off grid radio. We kind of have this like higher than usual population that are I do not, I do not want to insult them. So how shall I put this?

Well, they're, I guess, prepper community, right? We could, we could just say that. And we were talking again before the show, like I, I have some level of very much empathy towards that. If nor the reason, then I live in Tornado Alley. And so I definitely get the idea of let's try to be prepared. Let's try to prep.

And Then, of course, you have some people that just go way off into the deep end. And I, I've made the, I've made the point here before also that you, you have sometimes these people that, like, they will have reasonable points to make, but it's so wrapped up in the, sort of, the tinfoil hat language. So, there was a There was a deal where we, we founded a corporation to, to try to, you know, do some things around this, this particular open source project.

And we had a guy that goes, Oh, well, that's it. They're getting the Rothschild money now. And, you know, it was just sort of, it was off, it was off the rails. But when I actually read what he was saying, the, the point that he was trying to make was essentially I'm worried that now that there's a corporation, there is a leverage point for, you know, the, the government or whoever to be able to come in and say, we would like back doors in this encryption as well.

And like, I don't know if you've been watching the news, but that is not a that is not a tinfoil hat theory. That is a thing that government officials are actively trying to do. That's not a conspiracy theory, that's being done out in the open. And so like, yes, he was off the rails and he was wearing the tinfoil hat in the way that he presented it.

But there was a legitimate kernel of concern that you may not agree with but is a legitimate point to make And so this is something I spent a lot of time thinking about like for one thing How have we come to the point to where people just cannot make their legitimate? Concerns known without getting into this ridiculous language.

And then how do we how do we help? How do we help somebody get back from that point? It's all the internet's fault, you know yeah Yeah, that's probably true social media

Simon: you know, seriously, I think it is because I think that the The ability to compromise involves having a small number of opinions that you can think through rather than having a constant echo chamber reinforcing your first, your, your, your first surmise.

And I think that we've, the, the internet has produced this this dynamic for human beings that results in an inability to, to to, to empathize with. People from, people that you only know virtually. And I, I, I have no idea how we fix it. You know, the most obvious fix is to turn off the internet.

Bring

Aaron: back modems. Modems were underrated.

Jonathan: Media types.

Aaron: Not from a hardware perspective, but from a social fabric perspective. Yeah.

Simon: But I can't see that working out very well in the current environment. So I, I've no idea what we do, but and then the social media has then put that that effect on steroids and allowed people to create divisively boundaried communities for profit.

Yeah. And and you know, so what do you do about that? Well the best, very best thing you can do about that, if you're, if I'm honest with you is regulate it. Right.

Aaron: Right.

Simon: And and I don't see a great deal of that effectively happening just at the moment. So it's

Aaron: especially hard here. Cause we already have regulation that says you can't sue the social media companies, right?

So, you know, that's the thing that's like, I go back and forth on that one. I was like, well, I like that regulate. Cause I like technology and I don't want them to be sued. And I don't want to lose my. Whatever the latest social media thing is. Right. But at the same time, it's like, well, yeah, but the outcome of that could be this.

Other bad thing which we see happening right now, which is they can't be held responsible for what people say On their on their platforms.

Jonathan: Yeah, that is very much a two sided sword So I I fall into the camp of and I I acknowledge this is not working all that Well at the moment, but I very much fall into the camp of rather than solving it with regulation Let's solve it with open source technology This is why i've been trying to push people to mastodon because it it it cuts out all of these problems You can run it yourself.

You can, you can do, you know, whatever you want to with it because you can host it yourself. The problem really with it is two problems. One, there's not quite the critical mass that enough people are on it. Everybody is on, everybody's still on Twitter and then everybody's moving to blue sky and then you have a few people on threads.

But the other problem with that is that people just don't want to host their own mastodon server. Like the majority of people don't want to. And that. That sort of dries up a lot of that Advantage to it. So it's a it is it's a it's a tricky problem So simon, I see that I see the wheels turn and the smoke pouring out.

Yeah

Simon: so, you know that does concern me because I last I checked the the little islands of of isolationism Were running on mastodon Yeah, you know truth social is

mastodon

for example And I actually don't think that it with that that solves the problem because the problem is not a software problem.

The problem is a social problem and that social problem is being exacerbated by the technology and applying more technology to the social problem just gives you a bigger social problem. And I, so I. I don't really think that's the answer. I tell you what does happen though, when you get lots of open source software forming the basis of the economy like it does now, like whatever it is, I forget what the most recent research was, it's between 80 and 95 percent of all software systems are open source.

Yeah. And the what does happen is it's driving sufficient economic value to need regulating. And so that's what's happened over here in Europe. The European Union has decided to, to the Cyber Resilience Act is applying product liability to the whole product, including the software portion of it.

And so manufacturers are now liable for what their software does as well as what their hardware does. And that's going to change everything in Europe. I can't see you being able to get away without doing the same thing in the U. S., honestly. But Europe is a big enough block that it can, it can freely do that and not need everyone else to come along too.

And then we're going to see more of this. We're going to see more of people saying, hang on a minute. What is it that's driving this trillion dollar economy, economic force here? Shouldn't we be taxing it? Shouldn't we be regulating it? Shouldn't, shouldn't we, or, or, or alternatively, shouldn't, shouldn't we give, be giving it pork out of our barrel?

And those things are going to happen because it's big enough now to be visible from space. And as a consequence, the regulators are coming. And there may be regulators that, that pork barrel things that they took to, to, to help their electors. There may be regulators who want to stop people having fun.

There may be all sorts of different regulators, but now that open source software can be seen from space, the regulators are coming for it and you can't stop it.

Jonathan: And so there there is an interesting little quirk here and it's it's a difference between the united states and sort of the rest of the world And this is this is one of the things that was one of the latest kernel dust ups really had me livid in the united states we have the first amendment that that like enshrines freedom of speech as As one of the core principles of our entire existence And not to say that other countries don't have freedom of speech, but it's just it's so central To to the u.

s. Government to what the u. s. Government is supposed to be about and then you have some court rulings over here that basically say that code is a form of speech And so, there is a, there is a very interesting blocker to regulating, particularly open source code written by individuals, because that code is an expression of speech, and the freedom to be able to do, you know, to say what you believe is so bedrock to the United States.

It, it, it's just, it's interesting. I, I, I don't know that I have anything more to say at the moment other than that, but it's, it's really interesting what that's going to do when you try to get into regulation and all the other things.

Simon: I don't think that's going to cause too many problems for regulating it, because, you know, you regulate banks and people talk in banks.

And so you're regulating a modem and the software in the modem is what makes the modem do its things. I think you can regulate the behavior of the system. And when you recognize that the, that you may well have freedom of speech to create whatever system you want, you have freedom of speech to say whatever you want in the U.

S. And it's only when you start acting on it that you get, that you get restrained by the state. And that's, that's all this regulation is doing. It's, it isn't going to infringe anybody's freedom to write the software that they want. What it is going to do is remove their ability to be able to avoid the consequences of having written that software.

Jonathan: So isn't, isn't this though one of the things that, that you particularly had to fight for in the European Union and the laws there, the idea that the liability for the systems would be reflected back to the individual open source developers?

Simon: Well, well, in actual fact, the, the European Commissioner had done a pretty good job of understanding that, that, that there should not be mm-hmm

Liability imposed on people who were not benefiting from the availability on the market. What they hadn't understood was that the language that they used accidentally affected the people who were in the supply chain.

Mm-hmm .

Because they've, they, they'd had a view of the supply chain that, that there was always somebody placing the software on the market and that there was always a company and they were always an entity that was able to accept liability and they hadn't really considered some of the way that they had phrased things meant that You know, a guy in a barn in Omaha maintaining an open source library might well discover that he is a party to a liability claim.

Yeah, and that was what we had to fix. We had to to make sure that that their original intent of the legislation, which was to make parties placing software on the market in Europe liable for the consequences of doing so, which is a very fair proposition in my view, and we need to make sure that that intent Was the, was both real and bounded to the people who had actually placed it on the market and not the people whose work they had exploited to do so.

And, and we succeeded at that. We've, we largely succeeded at that. There's a couple of corner cases still. And so it's the, I, I believe the CRA is in pretty good shape in that regard.

Jonathan: And I would, I would, I would actually agree with you that I think similar legislation is coming to the U. S. You know, eventually, even, even if it's not like a federal legislation, probably what we'll see here is California will introduce it first.

Maybe. Because that's, that's the way things tend to go in this country, but

Simon: It can be you know, I, I also, I don't think you should underestimate the will of certain political actors to to cut the tech sector down to size. This, this is true? And I think you might well see some regulation that that has effects that are not unlike the CRA, but are there to make sure that the Googles and Facebooks and Microsofts of this world are not the the political forces that they have been for the last 15 years.

Jonathan: Isn't it, isn't it interesting how different groups and different political forces act in surprising ways sometimes? Yes. Well, I, so, I think something that we would all agree on with that is we hope they don't mess it up.

Simon: Well, yes. And, you know, that was, that was where, you know, I'm allowed to make political statements about the UK, aren't I?

That was where we, we really screwed up with, with Brexit, was we did something that was not Not realistically reversible still at the moment in the U. S. everything that's happening is still you know, conceivably reversible and it's not obvious that there's anything that's going to be different about that in this political, this, this upcoming political presidency.

And I think what you've got to watch out for is people erecting no return barriers. And you know, the place that that happens is actually in, in, in much more subtle places like the appointment of judges than the adjustment of electoral laws. I think that's where you want to be watching very carefully to make sure that your freedom isn't being abridged by anybody of whatever color.

Jonathan: Yeah, yeah, that's a, I think that's a reasonable statement to make. Let's see, we are, we are just about at an hour. Is there anything else that you guys want to touch on that we haven't gotten to? We, we talked about retrocomputing and politics.

Simon: I mean, you know, we, we, we could talk about sex or religion. We've never had shows on those. No, I think, I think we have avoided that on

Jonathan: purpose. I think I just try

Aaron: to imagine how that conversation would go. I can't even like.

Jonathan: So I've got, I've got somebody that follows me on, I think on Mastodon that will occasionally like and make comments and it's like an app to be able to get to some religious text.

So yes, there are, there are ways that that, that circle can be squared and that we could pull it in, but I, nope, not interested. We

Simon: can all, we've all seen the Saint Ignatius photographs.

I will say that so over here in Europe if you're going to be in Europe at the end of January. FOSDEM is happening. FOSDEM is coming, folks. You need to get yourselves over to Brussels for the first weekend in February, where more open source and free software developers than you've ever seen in a single place will all be trying to squeeze into too small a gap.

And we'll all be geeking out and probably drinking rather too much beer and pizza. And that is a completely free event. So FOSDEM you can attend completely free of charge. All you have to do is get yourself there. I think it's probably the highlight of the event year for me. The event that I, that I enjoy most from a health and welfare position is the SFS con that happened in Italy in November, but the one for simply seeing everybody and making the most progress ready for the new year that's got to be forced them.

So come and come and come, come on over. It's really easy to fly over to Europe these days.

Aaron: And when is that event again? I'm sorry I missed it.

Simon: That's the first weekend in February. So that's February the 1st and

Aaron: 2nd.

Simon: And there are actually quite a lot of adjacent events now. So they, it's now being called EU Tech Week.

And there are events as early as 29th of January. And there's events carrying on as late as 4th of February. So you can quite easily fill your calendar up there. And the best way to get to Brussels if you're flying from America is to fly to Schiphol, to Amsterdam airport and take the train.

There's a high speed train from Schiphol to Brussels. So you don't have to work out how to fly yourself to Brussels. And you get to try a really good high speed train.

Jonathan: Yeah, it's a lot of fun. Insider tips. Yeah. Aaron, anything you want to cover?

Aaron: Well, just the usual, you know, go check out my YouTube channels on Retro computing, vintage computers, all that kind of good stuff.

I just released an episode this morning covering the controllers that Atari put out the CX 40 plus and the CX 78 plus, which are reproduction Atari controllers but they are wireless. So, and, and they have dongles nine pin dongles, so you can actually connect them to your original hardware and get a wireless experience with a, with a controller that looks ju identical to the original, but it's wireless.

So, but then you miss out, they're pretty cool.

Jonathan: Then you miss out on the experience of playing your game and trying to dodge something and ripping your entire console off the shelf, ripping the controller out.

Aaron: Yeah, exactly. I did find a bug inadvertently actually, because, you know, me being me, the first thing I did was like, okay, well, let's really put this to the test.

So I grabbed my, my original heavy sixer, right? The first system that Atari made, you can't get any older than that and plugged it into there. And when I turned on the game, all of a sudden I was hearing all of these, all this new music that wasn't in the original game. And there's something actually in the wireless Controller that doesn't interact correctly with the heavy sixer and the light sixer It's only when you get to the one the four switch model that that problem goes away So there's something that was connected differently in those original ataris that causes the wireless.

Noise essentially that's going on in the circuit to bleed through Into the system and so it was really weird. I was playing with my son. We're like, I don't remember there being a soundtrack to joust on the atari 2600 And then we disconnected the dongle and it went away. I was like, oh So I wrote atari and said yeah, I think you need to be testing this stuff with the even older equipment than you're using but anyway, fun episode, check it out.

And there's a bunch of other Atari 10 things you didn't know about the Atari 2600. If you like Atari stuff, go check out my channel, RetroHackShack on YouTube.

Jonathan: Aaron, that's, that's not a bug. That's a feature. It, it introduces generative music to games. They tried to tell me that. Yeah, they were joking.

Aaron: They were joking.

It was all tongue in cheek, but yeah.

Jonathan: They

Aaron: said, Oh, I think I'd love to tell you that we planned it that way. But I think I'll forward this to our engineers.

Jonathan: All right. Anything else you wanted to plug Simon, you didn't get a, is there, is there a radio

Simon: I

Jonathan: just,

Simon: Put a new post up on, if people go to blog.

opensource. org, they can see my my article today all about the role that open standards are playing in the new regulatory environment and some unexpected consequences of it. I would love Focom Mastodon to, To to follow the OSI blog and anyone that finds any of the outrageous things that I'm saying interesting can follow me on Mastodon.

I'm webmink at mesh dot cloud. To follow along, say hi. I even reply sometimes.

Jonathan: All right. Very good. Thank you guys both for being here. I sure appreciate it.

Aaron: Absolutely. Anytime.

Jonathan: Yep. All right. So you can find my work of course, at Hackaday. We appreciate Hackaday being the home for Floss Weekly these days.

We've also got the Untitled Linux Show over at Twit still, and you can you can find my security column talking about sort of the other side of the coin there at Hackaday on every Friday morning and have a lot of fun with that. We appreciate everybody that watch both live and on the download, and we will see you next week on Floss Weekly.