Episode 788 transcript

FLOSS-788

Jonathan: This is Floss Weekly, episode 788, recorded Wednesday, June 19th. Matrix! It's Git, but for communications.

Hey, this week Simon joins me, and we talk with Josh and Matthew from the Matrix Foundation and the company Element. It's the Federated Uncentralized Communication Platform and some of the challenges they're facing right now between funding and some EU and UK laws that really threaten to put a damper on things.

But there's a plan. You don't want to miss it. So stay tuned. Hey, everybody. It is time for Floss Weekly. That's the show about free, libre, and open source software. I'm your host, Jonathan Bennett. Today, we've got a really a great and interesting show. Let's go ahead and bring on our co host. We've got Simon.

Simon the man. The, the open source guy. The, one of the, one of the stewards of the open source definition, which is kind of important these days, isn't it?

Simon: Yes, it, more and more so. There's a lot of pressure. Yeah. On the definition of open source, but particularly because of what's happened in AI and you know, there's a whole show to have about the the there is a word I could use to describe what's going on there But I don't think it's allowed on this show, but it involves cluster Oh,

Jonathan: yes.

Yeah, we try to avoid that one See my kids watch the show and I would prefer to not have to tell them to not say things that they hear on the show That's just a preference thing.

Simon: Yeah, so they so the open source definition is, you know, it's You It it's, it's now venerable and we're OSI is in the process of creating a definition of what open source AI means.

And that is extremely lively that discussion. I'm

Jonathan: sure it is extremely lively. That's yeah. Well, today we've got we've got something. Probably the opposite problem. It's, it's some people that are trying very hard to do something the right way, the open source way. We're going to talk with the folks behind matrix and Simon, you're familiar.

In fact, Simon is the one that said, Hey, when we set up the back chat for floss weekly, make sure it's something that we can integrate with matrix. Cause that's where I've landed. And then I found out in the, in the past 30 minutes that, you know, where all the bodies are buried, because you've worked with at least one of the guys we're going to interview.

Simon: Yeah, so I've worked with Josh at OSI and then I've known Matthew for a while as well. And we, we team up against the European union occasionally on European legislation. So I've been using Matrix for really quite a long time now. And I have, I actually have two Matrix IDs because there's a lovely community member that they have called Beeper that is making an integrated chat client that uses Matrix as its infrastructure.

And if you're a member of Pidgin, it's kind of the. The the, the modern equivalent of the, of, of the pigeon instant messenger. So I've got a lot of, a great deal of matrix in my life, both element and beeper.

Jonathan: Wasn't,

Simon: isn't Beeper

Jonathan: the one that tried to go toe to toe with Apple?

Simon: Yeah, yeah, they did and they're still, they still are.

They, they made an iMessage bridge that makes you show up as an, as a, an iPhone user to other iPhone users. And that's still available. So, you know, the iPhone, I can talk to iMessage people. I just have to open an iMessage account.

Jonathan: Oh, interesting. I, I thought I heard that that went away. Let's ask them.

Let's ask the experts about it. Yeah, yeah,

Simon: yeah.

Jonathan: Let's not waste any more time. Actually. Let's go ahead and bring them on. We've got Josh and Matthew, both of you. Welcome to the show.

Matthew: Hello. Hi, thank you for having us on.

Jonathan: Yeah, excellent. I am, I am super excited to talk about matrix and element and all of these things.

So let's, let's start with. Who does what? Let's try to keep it, our two, two guests, try to keep them straight. Let's start with Matthew and kind of give us the rundown of like, where are you coming to this conversation from? How, what, what part of this do you represent?

Matthew: Sure. So I guess I'm Matthew and I guess I came up with the idea of Matrix about 11 years ago now.

I've always been enthusiastic about telecommunications and open source, and eventually got the opportunity to smash the two together to try to come up with a new communication protocol that could be the missing communication layer of the open web. It would literally be a real time communication fabric, a bit like Tim Berners Lee originally hoped for on the web.

That there would be a kind of read write semantic, and it would be real time, and you can publish real time information, like instant messages, or VoIP, or files, etc. And it would have to be end to end encrypted, because it would be decentralized, and it would smear messages around the place. So that was what I came up with, with my co founder, Amandine, who isn't with us today.

But, I guess I've spent the last 11 years of my life trying to make that happen. And I guess nowadays my day job is Element, which is a company that we founded as the, well, the guys who created Matrix founded basically as a way to try to keep the lights on and fund us to work on Matrix.

And my position there is CEO, CTO. And I'm also a guardian of the Matrix Foundation. And I'm also project lead. For Matrix, so the kind, not really a BDFL, but more the guy who started it in the first place and keeps an eye on trying to keep it roughly on track on a technical side of things. So that's me.

Jonathan: Alright, Josh, same question, how do you fit into the Matrix slash Element conversation?

Josh: So I fit only into the matrix conversation. I am the managing director of the matrix. org foundation, a foundation that was started back in 2018 and really is the steward of the protocol. And really the, the, the body for open governance of the protocol.

And Basically, I'm tasked with operationalizing the foundation and supporting the broader ecosystem of which Element is a major player but only one.

Jonathan: Okay, I want to go, I want to go to Matthew then next because I want to understand more about the, the matrix protocol. What you described actually sounds similar to the Fediverse, but it's another approach to kind of solve the same problem.

Matthew: It's similar. I mean, ActivityPub, I think, postdates Matrix by a couple of years. ActivityPub focuses, I guess, on multiplayer RSS, for want of a better description. So, it's about micro blogging. It's not really real time. It's not decentralized in the, I, in the way, So Matrix is very much about relatively low latency, instant messaging, or, I don't know, live location sharing, or setting up a voice or video conference in real time.

So the whole thing is meant to feel like WhatsApp, or iMessage, or Signal, or Telegram, or Slack, or Discord, or Teams, or whatever your favorite real time communication tool is. Whereas ActivityPub in the 30 verse is very much, obviously, trying to compete with Twitter. And micro blogging systems. And the two can overlap.

People have built micro blogging on top of Matrix. We did it, in fact, ourselves, as a thing called Cerulean, which was a proposal for BlueSky that we showed to Jack and Parag at Twitter. They chose to build their own thing, rather than building on Matrix, sadly. But you can do micro blogging there, just as if you were feeling better.

Slightly masochistic, you could try to do instant messaging over the Fediverse too. So, I see them as kind of sister technologies. We've got good links with the ActivityPub team. I've spoken a lot to Gargron at Mastodon over the years. They used the end to end encryption for Matrix at one point as an attempt to do encrypted DMs in Mastodon.

And there are bridges like Kazama, the link. Fediverse and Matrix together, and indeed with things like the Bridgy thing on BlueSky, I think you can go, now go all the way around the houses from Matrix to ActivityPub to AppProto and BlueSky if you so desire.

Jonathan: That's great. So Matrix does have the same kind of a federated approach though, right?

Like, so I can, I can host my own Matrix server and then open up a chat with, and in fact, I think this is actually what I'm doing. I'm not hosting it myself, but I'm using somebody's. But you, you can make the two Matrix servers, Talk to each other, right?

Matthew: Yeah, absolutely. So it's federated, but it's more than that.

It is decentralized. And the conversations get replicated between the servers. And this is the unique thing that nobody else does. I guess Blue Sky is somewhat close to it in some respects. But the whole idea is that it's like Git. The, the model is the sort of subversion to Git is the same as Blue Sky.

IRC or XMPP to Matrix. So in Git, everybody has a copy of their source code repository, and if you work on it, you basically have your own independent clone that you can do whatever you like with, and you push and pull commits with other people. And Matrix is basically identical. The dirty secret of Matrix is that it is effectively Git, but for communication.

And so, I might have my server, and if I invite Josh and his server in, and perhaps Simon on his server, then every time I send a message, it gets pushed to the repositories effectively of the chat rooms on the other servers. And when they communicate back, they will push to their copy, which in turn, you know, fans out.

pushes to my copy and the other servers as well. So in the biggest matrix rooms like MatrixHQ, I think we've got about 20, 000 servers with about 60, 000 participants. So every time somebody sends a message, for better or worse, it has about 20, 000 HTTP heads to all of the other servers in the room in order to fan it out.

But the lovely thing is that there cannot be a single point of control or a single point of failure. Because nobody owns the conversation. It doesn't exist on a single server. That's it. And kind of it's multilateral communication intrinsically, so that the very act of me talking to somebody else shares ownership of that conversation with them.

And it turns out this is novel, and it's actually academically novel on the access control side of things, because we were the first people to figure out how you can ban people and kick them and give them permission to change the room name and all that sort of thing. In this sort of system, where you don't have a centralised authority, because normally you have like an access control server somewhere, or you have a focal point which determines the rules of who can do what.

Whereas in Matrix, every server goes and executes its own access control algorithm, to make sure that if I speak in a room, Am I in that room? Am I allowed to be in that room? Have I been kicked? Have I been muted? And all that sort of thing. So, that's the real interesting idea. The matrix isn't really a messaging protocol so much as a way to synchronize real time chatroom history of whatever flavor around in real time.

It's more like a pub sub database. Multiplayer CouchDB or multiplayer Redis with OpenFederation if you think of it in those terms.

Jonathan: You, you say it's like Git for communication, does it literally use something like a Merkle tree? Is it, is it sort of that, that blockchain y thing? I know people, people get mad at me when I call, call Git blockchain, but I, I stand by it.

Matthew: It is a chain of blocks, it is a Merkle tree. And basically every time you send a message or anything happens in a room we call them events rather than messages But a message could be an instant message, but it could be a join or a part a I know topic change In fact, it's arbitrary key value data. So it could be a I know floss weekly dot Defcon Defcon level 5 or whatever, whatever metadata you want put into there.

And each one gets signed by the server into a Merkle tree, so that you get a transcript consistency for that conversation. As well as these rules about how you merge together the Merkle trees, just like it has its horrible Octo merge algorithm for going and actually merging together branches.

Matrix has an equivalent merge resolution algorithm called state resolution. that allows different copies of the room on different servers to converge on the same result.

Jonathan: I, I was, I was actually going to ask you about this. How do you handle the split brain situation? Because on a, on a very different project, I've been, I've been noodling for a while about this idea of trying to do guaranteed delivery messages on a very lossy Low bit rate medium over over radio waves.

And like 90 percent of it, I can work out, but this idea of, you know, if the two networks, they move away and then when you bring them back together, how do you, how do you figure out which is the, you know, the canonical. Message tree and which you know, how do you deal with the ones that are not on that?

That's a difficult problem And I actually I need to go look at what you guys came up with because I might be able to Borrow some of your ideas I mean

Matthew: that is literally the core of matrix and it also does support some low bandwidth Environments there is mse 3079MSE being matrix spec change, the mechanism by which the spec evolves.

And that is all about using CoAP rather than HTTP, as an IoT transport for these horrible lossy links. But in terms of how the actual merge resolution works, basically, every time a server sends a message, it includes a proof that says, Matthew is allowed to talk in this room, because he had got invited.

And Everybody in the room, the servers, execute that proof and check that it, you know, that it stacks up. And if it does, then they accept the message and I'm allowed to do whatever that operation is, and if it doesn't, they know that I'm lying and I will get ostracized and the event will not get agreed to.

So, this was done as a very pragmatic engineering solution to the problem and then we started getting emails from the University of Karlsruhe about or Coulter Institute of Technology, technically, where there is a distributed systems networking group where they got excited because they started analysing it and discovered that it was a unique, special snowflake way of doing decentralised access control.

And if you go and Google for, KIT, DSM, and MATRIX, you'll find all sorts of papers and interesting presentations on the, on why this is academically interesting. But the, the end result is that you basically replicate the room across the world.

Simon: Yeah, so that, that, that, that, yeah, that all, that all sounds standards ish.

And it sounds like the Matrix Foundation could end up being a standards body. Is, is that something in your future, Josh?

Josh: Well, you know, it's funny you should say that. Because after spending six years at OSI, which is itself a bit of a standards body for for licenses. I am now at the Matrix Foundation, which is also a bit of a standards body.

Of course, not officially recognized as one. So I believe that's an open question as to how we proceed with that but you know, pre show we're talking about the Eclipse Foundation and the work they're doing to be recognized as such and I would not be surprised if we, we go down a similar path ourselves either individually or in partnership.

Simon: So how is Matrix Foundation set up? I remember talking with Matthew about it two years ago in Brussels. And he was talking about wanting to fork Element so that the community stuff was in a separate organization, and the corporate stuff that paid everyone's salary was in its own organization.

Seems that's what's happened now. So where does Matrix Foundation fit in?

Josh: Right. So this is again having spent time at OSI and being deeply opinionated about the course of open source projects. I was really excited and encouraged when I was speaking to Matthew and Emma Dean and the rest of the team before I accepted this role about the way that they created the foundation.

They handed over the assets like the trademark for Matrix and the Okay. Copyrights to the extent that they had them in the specification And all of those things live within the matrix. org foundation at this point and the spec core team, which is the the team that more or less manages the matrix spec change process That is all within the bounds of the foundation.

So backing up a little bit the matrix. org foundation. We are a Community interest company founded in the United Kingdom that is more or less a not for profit, right? There is no beneficial owner, no private owner no private gain that comes out of that. We are committed to keep all of our assets dedicated to fulfilling our mission and that mission is, of course, stewarding the specification doing advocacy as we see fit on matters like Encryption and privacy and security and human rights.

Because those are all highly related to, you know, secure communications protocol. So the role of the foundation is. If nothing else to look after the standard and to convene people around the standard. But our aim is as you see elsewhere in, in FOSS we see foundations that are able to, you know, look after the core technology, but also be a vehicle for nurturing the broader ecosystem.

Right. So with the Matrix. org foundation. For instance, we are hosting our first ever matrix conference later this year and bringing together policy makers and engineers and product managers and everybody who has a stake in the work that we're doing. So. I don't know. You could also use a bit of a gardening metaphor.

You know, we're here tending a garden as as the Matrix of Art Foundation.

Simon: So that it's, I'm kind of interested in having picked a UK CIC as the foundation. You know, from two perspectives. Two perspectives. First of all, I spent like five years advocating that open source organizations ought to be CICs.

So in my view, it's the perfect legal vehicle for an open source foundation. How's that going? You know, I know I was looking on Companies House just now, and I, I see you've had to stick the CIC on the end of the name.

How's it going? Being a C3 is very well understood to Americans like you, whereas being a CIC is almost unknown to Brits like me, so what's it like? Is it working well?

Josh: You know, I have to say it has not presented a significant barrier to us. It definitely has taken a little extra doing in terms of understanding what our constraints are and how we need to operate.

We are not tax advantaged. So, you know, that's one of the perks that comes with a 5. 1c3 or a recognized charity in another country. But it's been working out for us. Now, that said, we, we see like the Eclipse Foundation has rea Rehead quartered itself in Brussels. And we see foundations discussing having entities in a number of jurisdictions, right?

Whether that be the U. S. or the E. U. or the U. K. So And there are a number of reasons for that, right? You might be able to be a recipient of funding based on your jurisdiction. But for us, the United Kingdom has been a perfectly acceptable home and the CIC has worked out all right, I'd say.

Simon: Right, well, I'm happy and in fact somewhat relieved to hear that's the case.

So I had a look at the corporate structures around matrix and element. There's quite a lot going on. I noticed that you seem to have the, so being a CIC that you can't have any beneficial owners, so that's good. Over on the other side, I see, Matthew, you've got a holding company whose name you don't use in trading.

And then you seem to have three subsidiaries in Germany, France, and America. You know, how is that working out for funding the matrix work? Because that's very, you know, you're a critical part of the open source infrastructure for the future. And if you fail to fund this activity sustainably, we are all in trouble.

So how is it going? Well,

Matthew: that's a big question. So yeah, on the element side, first of all, it's completely disconnected from the foundation. Myself and Amandine are on the boards of both Element and the Foundation, but that's as far as the relationship goes and we're deliberately in the minority of both boards so that we can't go and do crazy things with either organisation.

In terms of Element being key to funding Matrix, historically, absolutely. So the, the origin of Matrix is that there are about 12 of us who were working at Amdocs, a telecom vendor when we came up with the idea of Matrix. And Amdocs actually funded the first three years of Matrix dev as a crazy R& D project in the hope that perhaps this thing could replace the phone network or replace email and be the next generation communication system that they could benefit from.

Eventually though, they got fed up in funding the whole thing and we somewhat abruptly left and suddenly had to make payroll for 12 people. Now, in the ideal world, I think we would have just grown, we would have set up the foundation and we would have asked for donations and slowly bootstrapped our way into being a Mozilla style non profit vehicle going and building out mission driven things for Matrix.

But We didn't have any money and any revenue. We literally pooled our redundancy package from Amdocs together to buy ourselves a couple of months in order to set up shop. And luckily the folks at Status, who do decentralised communication on top of Ethereum went and invested and put I think four and a half million dollars behind the bar to get us started.

So, on one hand, that was great because we were able to set up shop as Element and make payroll and actually continue our work. On the other hand, it set us on the path to basically a VC investment model where we, on the Element side, raised in total, I think, 55 million so far, of funding. As element in order to first of all build out matrix and then obviously try to do something that would make money as a return to our investors.

So before we set up the foundation in 2018, you ended up in the strange model where basically 100 percent of the matrix work was being done by us as element. And we were basically taking investor money and building Apache licensed open source software with it for the greater glory of the world to try to terraform this new industry.

Okay. Now, that obviously has worked relatively well. I would say that Matrix itself is wildly successful. We've got 152 million people addressable on the network. We have people like NATO and the United Nations and all of France, most of Germany, Sweden, Ukraine, you name it. The different countries and organizations are enthusiastically using Matrix.

And we've always had this really delicate balancing act between Do we just give away everything as liberally licensed open source and get as many people using it as possible and pray that there will be a way for Element to do an automatic on top of WordPress and basically sell value added things, whether it's support or SaaS hosting or whatever?

Or, or what the correct balance is. And so that has been the eternal funding challenge. These days Element is at a point where we are not profitable yet. Hopefully we will be in the near future, and we've been killing ourselves over the last couple of years trying to get to that point. About 30 40 percent of our effort still goes into core matrix development, effectively donating the work to the foundation, hopefully to the benefit of the foundation, and then the rest of it is spent doing commercial work.

element work. So I feel, I think of it a bit like Red Hat or Intel or somebody who has a whole bunch of people who are frantically committing code to the Linux kernel or whatever your favorite bit of the Linux stack is. But they also have the commercial business going on where they sell Red Hat enterprise Linux.

And so for element, the equivalent is this thing called element server suite, which is a commercial distribution of matrix, very similar to realm, but for matrix land.

Simon: So do you have anyone else in the ecosystem who's Contributing significantly to the core code. You know, I, I mentioned in the, in the, in the rundown Beeper existed.

And there's Eric over there who's doing that, I think a fine job building a, a you know, replacement for libpurple. Do you, is that meaningfully contributing to Matrix or are you still really carrying the world on your shoulders? I'll let Josh

Matthew: answer on behalf of the ecosystem, but from my perspective it's you've got to think of Matrix in two levels.

There is the kind of core technology, the stuff in github. com slash matrix dash org, like the spec. and the encryption, and lots of SDKs and things, which, without which Matrix would never have existed, and continue to be a kind of fundamental building block. And of that, about 90 percent is still written by Element.

But then on top, you have loads of things, like Beeper, which have been written by different people in the community. So Element obviously has its own app sitting on top. But there are literally hundreds of Matrix clients, and servers, and bridges, and bots, and carrier pigeons, and God knows what else, which people have built on top of the underlying technology.

So Beeper is interesting in that they have contributed a bit back to Synapse, which is the Python server. I think I've seen, like, tens of PRs over the years which is pretty good. They've also contributed some MSCs, so metric spec changes to the spec itself. Although on the client side, I don't think they ever contributed anything back.

So they went and forked the iOS and Android code bases which we'd written that element, improved them a lot, and kept them proprietary, and despite asking a few times that they might possibly upstream them, they never did. So, it's always a bit of a cure it sack. What do you think, Josh? Ha ha

Josh: ha!

Yeah, So, like a big part of the role of the foundation is to to build up the matrix ecosystem into something that has a diverse and sustainable contributor base. I think it's fair to say that We have a very active ecosystem. There's a lot of great work going on. But we also we are still far too reliant on element.

And I think that's, that's a fair characterization both in terms of you know, Contribution upstream to core libraries as well as support for the foundation itself. And on the one hand, I'm very grateful to element for the generosity. Also, not surprised because, of course, it's run by the founders of matrix.

But it's. It's a dangerous position for any ecosystem to be in when there is one organization that is doing so much of the upstream development. And so, you know, we do have you know, to Matthew's point, Beeper, for instance, has done great work in creating a ton of bridges that really make the matrix experience much more powerful and interesting for people.

I love that I run my own home server and I use that home server and a bunch of bridges so that I'm recreating Yield, Pigeon, or Meebo experience but all FOSS which is pretty great. Then we, we also have I believe it's the etk. cc folks who have created some great software that packages matrix and bridges and other sort of related libraries to make it easier to run for home server operators like me, who are good enough sysadmins to be dangerous, but not to be employed as a sysadmin.

So there, we, you know, I'm, I'm very encouraged by the number of contributors and organizations that are contributing in the ecosystem. But it is, we have a lot of work to do for sure, to to get more upstream development.

Matthew: Hopefully the good news is the element isn't a evil corporate upstream trying to screw up the community or whatever, because frankly, element success is completely contingent on matrix of success, it would be sabotage if we did something.

to make we as Element did something to make Matrix less successful. And so there is that guiding principle throughout, and has basically been enshrined in the Element, sort of, core values since day one, that the company exists in order to realise the Matrix ecosystem, and then hopefully provide a flagship killer app on top of it.

Noticing that many of the previous technologies b abysmally failed due to a lack of a killer app. Like, how was Activity Pub doing before Mastodon came along? It was a disaster and I was, well, what was the killer app for SIP X and ppp, IRC? You know, they, they never had the kind of obvious go-to thing, whereas nowadays, the better or worse, well, if you ask what the killer app for email was, depending on the year.

It's gonna be Hotmail, Outlook Express, Gmail, Lotus Notes, I would argue. Profs. Link. Yeah, or Pine. CompuServe.

Jonathan: Yeah.

Matthew: But basically, it had the kind of, you would be able to ask somebody and they would have a strong opinion as to the killer app for email is this, and email itself is the killer app for the internet.

So, Element was trying to do that and is trying to do that for Matrix.

Jonathan: So I want to jump in and we've got a question from the chat room and I think it's going to be a great segue to talk about something important. So MashedPotato asks, Are matrix elements aligned with signal etc on their position To pull out of territories if end to end encryption is outlawed or if they're required to have back doors And this is a reference to I believe the the the way the governments are saying it now is chat control Which I don't know a whole lot about But apparently it's a thing.

It's become a thing in the last few days. And so I don't know whether Josh or Matt wants to take this first. I have a feeling that you're both going to want to comment on it. So let's go yeah, let's, let's, let's, let's kick it off. Take it away.

Josh: I would be happy to jump in on this to get us started.

So this is this is proposed legislation. That's been in the mix for at least a couple of years now. And toward the end of 2023, it seemed like like maybe, maybe we were going to be all right. But it has come back up in 2024 and they, One of the sticking points has been basically their way of undermining end to end encryption.

I think their, their new turn of phrase for this is upload moderation, which in their view is somehow not undermining end to end encryption, but it's just mental gymnastics. It's the same thing by another name and I think so we're, we're monitoring the situation you know, we have our, our, our foundation's DPO has been paying A great deal of attention to, to this and how it impacts matrix is different than how it impacts other communications technologies.

So that's one thing I want to signal and I want to unpack that a little bit, but suffice it to say we should be fighting tooth and nail against legislation like this because it makes everybody less safe. It's everybody less secure whether or not you're based in the European Union and no matter what technology that you are using.

So We're very concerned about chat control. One of the things that is a little different in Matrix, and this is something that Matthew spoke to earlier is the fact that it is not just decentralized it is, excuse me, it's not just federated, it's also decentralized. Or, you know, it is both of those things.

And so when I think about you know, If something like chat control were to get passed, and let's, let's assume for a moment that it doesn't get hollowed out or blocked by the courts, right? And, and there's reason to think that it might. Let's say that it makes it to an implementation and enforcement stage.

What does that look like for us? And the way that I think about that is how would that legislation interact with client developers, server developers the specification itself, and then home server operators, right? Because all of these things are different moving pieces and are you know, sometimes in different jurisdictions are developed by different people who are subject to different laws.

And so. When I think about what the implications of this passing might be for instance, The foundation, again, its core role is to steward the specification and and something like this goes completely counter to the mission of the foundation. And so I don't see any world in which we would see the specification change to accommodate the magical thinking of legislators.

So let me, let me start with that. But then also all of the software is open source. Right, you know most of the popular clients in matrix And indeed also the popular servers, they're all open source, you can scrutinize these things. And if for instance you see that a developer of one of these things, because they're based in a certain jurisdiction, is starting to make changes that would be concerning, fork it.

Right? Fork it. And, and, and don't, don't adopt those changes. Now, of course, you have to be mindful of not only the technology that you're running, but the technology that the people you're communicating with are running. So, in an ideal world, this stuff doesn't pass. This stuff doesn't get enforced. But I just want to highlight that the way that Meredith Whitaker and Signal have very rightly Taken a principal stand and say, we would rather pull out of the market than then undermine our technology and people's security.

You know, that's the right thing to do. And we would take a page out of that book, but also what that looks like in the world of matrix is different because we are structured very differently.

Jonathan: Do you, do you foresee a possibility, so let's just think like worst case scenario where the, the Matrix Foundation would then have to move out of the UK, just as a result?

Josh: Well, that's interesting because I don't quite know how this is going to impact the UK specifically, right? Being in, in this post Brexit world, I don't think it's not clear to me what, what, what would our obligations would be.

Matthew: I can take it on the UK side. Let

Jonathan: me, let me make sure I have one thing clear because I, I pay attention to US politics, not nearly as much to EU and European politics.

So the chat control legislation, that is a piece of EU legislation? Or an effort. Okay. So not, not directly applicable to the UK. Interesting. Good to know. I did not actually know that.

Simon: And it's, it's important also to know that that legislation is, is not primarily a piece of technology legislation. That, that piece of legislation is primarily a child protection measure.

And so the legislators who are involved in it are not people who understand that open source even exists, let alone how software works. And so you know, that, that piece of legislation itself is something of a special case in the overall landscape. But we do in the UK have another piece of legislation called UK Online Safety Bill which is just as as ill advised.

And unfortunately, because of the way that that legislation works in the UK, it's harder for us to access and influence than the European legislation is. Because the European Union is actually extremely open to discussion with outsiders, whereas the UK government, not so much. But Matthew, I should be asking you to say these things.

No, no, not at all. I

Matthew: mean, honestly, last year was a really terrible year for matrix and element and by very many different metrics. And one of them was the UK online safety bill. Not least because historically I think we've depended a lot on our European friends to wave the flag to prevent Well to protect human rights and avoid mass surveillance kicking in post brexit We can't suddenly turn to the french and germans and say hey guys, can you please Confirm that it's actually a really bad idea to have mass surveillance, so instead I find myself doing it and Spent basically a year turning up on the main tv and radio programs trying to explain to everybody That a backdoor that is there to protect the children Will be abused by attackers to attack the children as well as everybody else at the same time And meanwhile the people actually doing illegal things will just continue to use actually secure systems rather than the ones You where the technology has been deliberately weakened in order to scan what everybody is saying.

And it didn't work because the punchline is that the online safety bill is now the online safety act and it got passed by government So the worst case scenario that we're worrying about with chat control on tomorrow I guess thursday when they vote on it has already been and gone for the uk and yeah, it failed and so I feel massively burnt and a bit stupid for going and Investing the time to try to influence the process there and it is terrifying to see the eu suddenly You Sprouting off in the same direction particularly in the dead period immediately after the four yearly election cycle which happened last week and It's all in a bit of disruption and it kind of feels like it's suddenly come out of nowhere.

There's something To try to rush through when everybody isn't looking there is one good piece of news though on this which is canada somewhat unexpectedly very vociferously said they were not going to scan end to end encrypted messages Because they had been listening to the debate in the uk around the online safety bill and entirely agreed with the argument That this would be catastrophic to online privacy and would create a surveillance state and they didn't want to do that So even if I failed in the UK, apparently Canada was listening and took the hint, but the EU, meanwhile, has gone in a different direction.

So it's scary times. This is a dark timeline that we are in. Yeah,

Simon: I don't think you actually failed, Matthew, because, you know, when the Online Safety Act was passed was passed. Ofcom, which is our regulatory agency in the UK, was given a fairly open ended remit to go work out, to interpret how the the law was going to be enacted.

And that means that we've got a significant delay, and we've also got an agency which is going to report back to government that they've discovered it's act, unfortunately, impossible. Which of course if you're a computer scientist, you know, it's impossible to to, to having, you know, the, the very phrase encrypted back backdoored encryption contains within itself a totology, you know, it's either encrypted or it's backdoored.

It's, it can't be both at the same time. So you know, there, there's a good sign there and I'm not. completely desperately miserable about chat control in Europe either yet. All that's happening is the European council is doing part of the forwarding that the ongoing process in pushing it forward and the new parliament is going to have to stand up and Do something with chat control.

And fortunately, the right wing did not take control of the European parliament in the elections. That what did happen was all the people who understand these issues in the green and pirate parties got decimated, but all the centrists are still there. So we do have a job on our hands to go and educate this new wave of centrists about how a, a functioning democracy needs to recognize the, the statistical or the proportionate relevance of different aspects of protection.

And they need to have people explain this to them. So I think you did a great job, Matthew, honestly, and I'm, I'm really hoping you're going to, you know, so Matthew did a memorable presentation in the digital markets act presentation in Brussels where we heard from all of the big companies. Platforms about how impossible it was to do interoperability and instead of standing up and arguing, Matthew just said, I'm going to do a demonstration of making what, what, what did you have interoperating?

I think it was Facebook and Google or something, didn't you? Yeah, it was WhatsApp and Google chats. What? Yeah. So, so you'd heard the spokes, the spokes lawyers for those two companies saying how impossible this was and how it would take decades of research. Whatever. And then Matthew said, well, I just knocked up a quick demo of how it's possible.

And I'm going to send a message from, from Meta to Google and I'm going to do it live on live on the screen in the meeting room. I thought that was utterly memorable and I want you back doing the same thing because I think that makes a real difference. It breathes a. Breath of fresh air. It's totally alternative to all the boring people in suits saying, think of the children and you know, we need you back.

So don't give up, please, because we need you. And honestly, this is the role, not for you as a company, but it's for the matrix, the matrix foundation needs to get in there. So Josh, do you have any plans to go and engage in, in particularly in Brussels? To go and join in with those of us. You know, there's a little posse of us from a, from NL Net Labs and from Eclipse and from some other organizations.

And we are going out there every week and going and talking to these people and explaining which end is up and why the thing has to be plugged in. And we need, we, we really need people to come. Are, are, are, are you coming, are you gonna send someone.

Josh: So I am, I'm pleased to say that Denise Almeida, our data protection officer is already very engaged in these instances, responding to the comment periods of Ofcom and other regulators.

And the foundation we recently joined as a supporter of Open Forum Europe and are participating in a number of Eclipse working groups so that we are a party to these things and can do some advocacy. Transparently. The foundation is still an early stage nonprofit and and funding is a, is a thing that I lose sleep over regularly.

And so we do the best that we can on matters of advocacy with the resources we have we would like to be doing a whole lot more. And that is is ultimately going to be in contingent on on us being able to, to rally the support that we need to, to fund that.

Simon: I was also doing

Matthew: my stuff as Matrix by the way for this rather than Element.

So I was unashamedly wearing my Matrix t shirt and boxes and socks and everything else.

Simon: You know, I totally want you in the discussions about the Cyber Resilience Act and Open Source Software Security. Talking realistically about that. You know, your stuff on the Digital Markets Act was great. You know, come and join us in the Cyber Resilience Act meetings as well.

Digital Markets Act looked to me like it was going to be a gift to Matrix, of, you know, forcing the big platforms to open up and tolerate your existence. Is that how it's worked out, or has the reality been a bit different? I'm

Matthew: also a little bit jaded on this one , and that we hoped indeed that it would be a gift to us.

And we basically used Matrix as the reference to show how you could do end-to-end encrypted, interoperable communication between the big players. And we actually did a full integration with WhatsApp using their DMA APIs basically helping them develop the. Open APIs that they are obligated to expose to the world for DMA.

And as we speak, I'm meant to be writing a blog post to announce that and get feedback from the wider world on it. Although I'm about three weeks late, much to the irritation of WhatsApp, as well as everybody else who works on it. I promise I'll put it out soon. The catch though, the reason I'm jaded, is first of all, it only applies to people in the EU.

And WhatsApp are within their rights to say look, you're only allowed to use these APIs if you are also in the EU. At which point one would have to start IP geolocating matrix users to check that they're really in the EU. Which is not something that we want to do. But then Apple managed to wriggle out of it.

And then Microsoft and Google weren't really even in scope in the first place. So the only organization who currently is considered a DMA gatekeeper for messaging purposes is WhatsApp and Facebook. So Meta. So, whilst it's been great working with them, and honestly it has been. I saw that Will Gaffgut, the CEO at WhatsApp, was retweeting all of our chat control, propaganda yesterday and agreeing that, you know, you know, you know, it's a serious situation if WhatsApp and Matrix are doing kind of joint PR against legislation in the EU.

But on the other hand, they are the only people who are currently in scope for this. So either there's a way that Apple manages to un get, obligated to get involved, which seems quite hard because they've done a pretty interesting judo move where on one hand, They've added some really funky post quantum encryption to iMessage that makes it incredibly hard to interoperate with anybody.

And then at the same time, they've started speaking RCFs, the horrible Green bubble thing that Google have on Android. And so they can simultaneously say, Oh look, we interoperate anyway via RCS. Plus, obviously we would never be able to interoperate encryption that nobody understands. And that I'm not quite sure how we dig away from that.

So if nothing else, we might be able to at least do matrix to WhatsApp if you're physically in the EU.

Jonathan: There's about, there's about three different directions. I want to go with this all at once. On the, on the political side, We talked a little bit about left wing versus right wing. And of course, my perspective on that is entirely U.

S. centric. So I understand that it's a different conversation. I would just say this, regardless of which side someone is on of that political divide, I think if you actually explain to people what this means, like the idea of backdooring encryption, it means that Facebook or WhatsApp, whoever is doing it, they can read your private messages.

And as an extension, that means that the government can read your private messages. I think Basically, everyone understands that that's a problem. Okay. It's, it's just well, as we say, it's kind of a matter of how it's packaged by the different groups and the people that are in charge of them. Anyway, that's all I will say because I do, I do not want to dive any further into the politics here.

I, that is what I call a cup of coffee conversation. If I was with you in person over a cup of coffee, I would be glad to talk more about it, but not here on the show. In person over a cup of, I know I have my coffee too. But you've mentioned And this is something that intrigues me with encryption Apple is post quantum.

And so I, I guess the setup question here is, is Matrix considered a quantum? Post quantum secure or just as soon as someone gets an actually working quantum computer are all of the matrix conversations going to be a decryptable.

Matthew: So excellent question. We started working on it last year and in fact, one of our lead crypto and rust developers went on sabbatical for a couple of months, came back from it with a great big push request or pull request.

For the dosimates, which is, end to end encryption implementation for Matrix on the Matrix Foundation side, which adds PQXDH, which is the amazing five letter acronym from the SIGNL team that describes Diffie Hellman key exchange with post quantum protection. And it's what SIGNL launched, I think, about a year ago now, and they kindly documented it in the public domain, and it's what we have Dutifully gone and cloned and implemented in Rust.

Just like we did our original encryption as a port of the signal double ratchet over to matrix. Now we've done the post quantum double ratchet over to matrix too. Now it isn't live yet. We're hoping to finish it off in around September. We also potentially have a really exciting set of announcements around that, which I can't talk about yet, but watch this space for September, in terms of basically trying to position Matrix as a playground for post quantum.

Because it's not really clear what the best architecture is. We've got them in CloneSignal here because they're smart people and we trust them and we think they know what they're doing. But, for instance, there was a Chinese paper a few months ago proposing an attack on lattice based encryption, like in Kyber and NTRU.

Kyber being the, Primitives used for key encapsulation in PQXDH, and the only known solution to that, which is post quantum resilience, is this thing called Makilisi, or Makilise keys, with the catch that the keys are about two megabytes large. So, imagine putting that in your ssh. Authorized keys.

So, I think it would be really interesting to see what happens if you use Matrix's modularity to go and experiment with different post quantum or indeed normal encryption implementations. So today we already have double ratchet as well as MLS. The IETF RFC 9420 Implementation of Group Encryption.

Both of those can in turn be extended with post quantum primitives like PQXDH and Kyber. But then perhaps somebody might turn up and also experiment with Matalese on top of that. And perhaps it turns out, if the lattice based stuff never worked in the first place, everybody who was using the experimental 2MB keys gets the last laugh, even if it takes 5 minutes to set up a conversation, because it has to copy hundreds of megabytes of data around the place.

So it's looking kind of exciting and the apple stuff is weird because they just did their own thing. Yeah, and they did a Kind of a second order of post quantum key exchange on top of the normal stuff Which they call pq3 to make it a bit more different.

Jonathan: Shall we say? Yeah, well, I mean that's kind of apple's calling card these days.

They've got to do their own thing so obviously like i'm gonna put my You My amateur cryptographer hat on here for a second like the the threat of quantum computing to encryption We've got to take it seriously, but like what is what is your gut feeling? Is this really a thing or our quantum computer is going to be forever five or ten years off?

Matthew: I think it is definitely a real concern In terms the time frame who knows There is definitely some silicon snake oil going on where people are pushing this as an urgent thing where everybody has to invest and buy post quantum things now, now, now, now, now. But on the flip side, it's just a matter of time.

As in, any, the reality is that any competent intelligence agency is just going to be sitting there storing everybody's traffic onto disk. putting it in a big mountain somewhere and whether it's two years, 20 years, 200 years. At some point, somebody is going to find a way to brute force using whatever technology, the current encryption, and replay it all and see what everybody was saying.

So the sooner that you can get additional defense in depth against that, using wacky new encryption techniques, the better. And I honestly wish that we had the funding to have worked on this earlier, rather than relying on somebody going on sabbatical and doing it in their spare time, unpaid.

Jonathan: Yeah, yeah, the only the only real danger that comes to my mind when I think through this is that that concept that They're wacky and new encryption techniques And anytime you're talking about encryption and you're talking about new you have a a Potential problem, right?

Like it's there's something very useful very very valuable and using encryption that has been around for 10 or 20 or 30 years and people still not figured out how to break. But anyway, that's That is all the time that that really I have to dive into the the the cryptography thing, which utterly fascinates me I do have one very quick thing, which

Matthew: is that the trick is to do both You do hybrid.

You take the existing stuff, which you know works, and you put the wacky stuff on top, and then you hedge

Jonathan: your bets. That way if the wacky stuff turns out to be broken, you still have the existing stuff. Yeah, that's, no, that's clever. That's great. Okay, so I want to ask about the Matrix Conference. Is there actually going to be a proper Matrix Conference coming up sometime soon?

Josh: There sure is. We will be gathering at Mitosis Labs in Berlin September 19th through 22nd of this year. And for folks who are frequently on the conference circuit, that is Immediately following the Open Source Summit Europe. So that same week. So for anybody who's making the trip, it'll be a little easier to tech.

Matrix conferences on. So we are expecting to have a an unconference on the first day, two days of pop proper conferencing. We've got the call for papers is open presently. We're seeking sponsors, of course. And then on the Sunday, we will be having a contributor sprints and community meetings.

So the aim is to as we've been partnering with the matrix community summit. Team to create a space for the engineers, project managers all sorts of contributors in matrix, as well as policy makers and people who care about things like data sovereignty create a space for everybody together.

Jonathan: Very cool. I, I know there's a bunch of questions that you guys had listed in the rundown and we have not gotten to about half of them, I think. Let's say, pick one topic that we did not get to, that you would consider the most important, you want to let somebody know about. What, what did we not ask you about that we should have?

Matthew: Matrix 2. 0! Okay, what's up with 2. 0? So this is basically Matrix that outperforms the centralised normal apps. Like, historically, Matrix, first of all, was alpha, and it barely worked at all. Then it was beta, and it worked most of the time, but it was pretty slow, and the UX was terrible. Then we got to 1. 0, which was about five years ago, and it was starting to get properly usable.

The UX still wasn't great, but critically it was slow. Matrix 2. 0, we've just been doing nothing but optimization work, both for the usability, UX, as well as the performance. And it's things like instant login, instant sync, instant launch, invisible end to end encryption, so you don't have to mess around verifying devices and things, because when you logged in you did it automatically, like, Signal or whatsapp does and basically getting the ux polished to the point where we are out competing our dear friends at meta or indeed signal So to me, this is what the last 10 odd years have been building up to and I Should be going live at matrix conference on september the 20th, assuming that we Meet our deadline and I cannot wait to actually finally say guys It's here.

Please install your favorite matrix client, which supports matrix 2. 0, and you will see that it is as good as it was cracked up to be 10 years ago. And I'm so sorry that it took us 10 years to get here, at which point I assume the heavens will open and I'll ascend in a blaze of glory and head to Valhalla or something.

Yes. And that's all still interoperable, is it Matthew? Yep, it is. So it's backwards compatible with normal matrix, but it's basically new APIs. That are designed to not suck in terms of performance and usability.

Simon: Right, because I'm sitting here I've got a box full of very old computers on the floor down there.

That I'm busy, I'm busy reconditioning and, and dealing with. And one of the things you discover is that that you can't install anything new. And the, the, the, typically the reason is because the certificate, the, the, the crypto certificates have expired and there's no way of getting the new ones because of a version level problem.

And so, you know, one of the things I, I really look for in something like Matrix is to make sure that as you make progress you're not leaving all the, the old junk that people like me run. Behind so, so I'm very pleased to hear you saying it's interoperable.

Matthew: We haven't broken backwards compatibility yet.

It's been 11 years. I, it's a bit contentious. Some people on the team are pretty upset that we jumped through hoops to maintain compatibility with day one matrix, but I see it like the web. Like the Space Jam 1996 website that still runs fine on today's browser despite being 30 odd years old. And we want to do the same thing for Matrix, but on the other hand, encryption needs to move onwards, and so if we turn on post quantum, because we have it, And you're on some old client that doesn't support it.

I think we rely on the open source community to then patch your Windows 3. 1 client or whatever it is. So they can actually support it. If it's any consolation, one of my pet projects is to get my Dragon 32, a TRS 80 color computed free clone for our American viewers to talk matrix. And I actually had it up and running last weekend and I was very pleased that the very first computer I ever owned 40 odd years ago was able to talk through to Matrix.

So I'm gonna, that's my baseline compatibility to maintain. That's

Jonathan: very reassuring. Matt may have just answered the question I was about to ask. I like to ask people, especially with projects like this, What's the weirdest thing that someone has done with your project? What's the strangest, most surprising thing someone has done with Matrix?

And the answer may just be running it on a TRS 80 clone.

Matthew: Probably isn't. Some of the most creative things are definitely not appropriate for the show, unfortunately. But suffice it to say, there are some IoT use cases which surprised me, particularly as the German healthcare industry, I think, ended up hiring the people responsible for these IoT devices to do very serious work turning a slightly blind eye to their previous expertise.

But we'll see. So we did media for matrix, which is kind of fun. So you can jam over matrix, a musician. We've built an entire metaverse system called third room, thirdroom. io on top of matrix, which is still better than any other online virtual world. Infrastructure, frustratingly. What other weird ones have we seen?

There's a brand new project called Posca, which I saw yesterday that looks really fun. And it's written in Occamul, a well known language. And it is both message boards, and chat, and microblogging, and TikTok clone, all on top of Matrix. And it looks really exciting, and somebody going full crazy. Let's see how far we can push this thing.

Jonathan: Yep, excellent, excellent. Alright final two questions for each of you. We'll actually start with Josh. And I want to know, what is your favorite scripting language and text editor? Ha ha

Josh: ha, okay. Favorite scripting language? Alright, well, I got my start with with ASP and spent like a decade working in PHP.

So, PHP is, is my old standby but these days when I have to I have to do any coding, I tend to go to Python. Text editor, you know what? Nano has never let me down.

Jonathan: Ah, yes! I do, I do, I'm a Nano fan. And I think it's Well, for me, it's because I got my start in QBasic. In, in Microsoft's Basic. And the editor for QBasic is, is, Well, when I come to Nano, it just, it feels, it feels like home because that's where I got started, so.

Alright, Matthew, same two questions. Favorite text editor and scripting language?

Matthew: Oh god, it's tricky. So, on the text editor side, it's pretty easy. I have to say, I'm afraid I'm a Vi. Person, not a VIM person, but a VI person. And the reason is that even if my first computer was this TRS 80 clone, my second computer was the Silicon Graphics workstation.

In fact, the 149th machine that SGI ever made. Beta hardware back in 1984 at the expense of showing my age. And this thing was amazing. It's like half a height rack. It had 24 bit color. It had 32 megs of RAM. It had a 70 megabyte hard disk. And it had UNIX. System enhancements that turned into IREX eventually about three years later, and it had a 68, 000 processor running at like 8, 8 hertz, and that was my first proper computer, and that is why I'm such a geek, because I spent my entire childhood learning C and Iris GL and network programming and all sorts on this thing, and it was 100 percent done with Vi, because Vi and Edlin were the only editors that you had on this thing, and I, I've never looked back.

Now, on the scripting side, this is controversial, but I have to fess up, people who know me well know that this is my darkest secret, that I'm a massive Perl addict, and I used to run the world's biggest Lord of the Rings website, the one ring dot net, that was written entirely in mod Perl, and statically compiled down and run, it's served by Tux as a Linux program.

HTTP, Linux kernel HTTP server, and so I write huge amounts of Perl for that, and I still think and dream in Perl, unfortunately. Nowadays I would probably use Python or Node or Rust or something, but if you ask me to do a quick one liner to look through a ton of logs for something, I will produce the most unreadable Perl straight off the top of my head that you have ever imagined.

Jonathan: When, when Randall Schwartz checks in and listens to this episode, he will get a kick out of that. Alright. Guys, thank you so much for being here. We've had people in the chat room already tell us we need to come back and do a part two of this cause there are things that we did not get to. So we will be in touch about that.

Bring you back in probably a couple of months or six months. We'll see. We'll see when, when it works out. But both definitely have you guys back on and chat with you again about what's going on with elements and matrix and maybe a little EU and UK politics because that's just kind of all tied up into it.

But thank you so much for being here. Awesome, thanks for having us on.

Matthew: Pleasure to be here.

Jonathan: Yep, awesome. Alright, Mr.

Simon: Simon, what do you think? Well, you know, I have a lot of time and respect for what Matthew and Amandine have been able to achieve with Matrix. I think it's a great project that they have been good stewards of.

And the direction that it's headed, you know, the way that it's, you know, Become adopted so universally in the open and free software movement the way it's been adopted so widely in so many government contexts, I think is a testament to both the technical work and the also the stewardship of the brand and the, and the code.

So I'm, I'm a massive fan and there was nothing in the last hour that. Suggested to me that I've made an error in coming to that conclusion

Jonathan: about you.

Simon: What do you think?

Jonathan: No, I enjoy it. I, I very much like the, well, obviously it's, it's, it's federated. And I love that. That means that there's no, you know, there's no one server runner that can, that can say, you know, for whatever reason, I don't like this person and therefore they don't get to use any of the matrix services.

And then. The fact that it's decentralized, you can run your own server. I think it just, it ticks all of the boxes of what you want a communications protocol to be there's, there's no way there's no way to use the technology against anybody. And I think that's important. That that's that's important because let's see.

How do I want to put this? Even the times when you think you have a good reason to use a technology against somebody It's still a bad idea. And so the fact that you have that built into Your project is is super important there's no, you know There's no government that anywhere around the world can put their thumb on the scale because you could just fork it and run your own server So I, I love that aspect of it.

And then I think it's, it's really fascinating. And this is something I wish we could get, could have gotten into more. It's an open source project that they're actually making money with. And, you know, we didn't ask him a whole lot about the details about that. Maybe, maybe I should say we're using the term making money a little bit loosely because they're still on the path to profitability, but like they've got income and they're, they're, there's actual income from it.

And that in and of itself is a big hurdle for some open source projects Turn the corner to even having income associated with it without, you know, having to metaphorically speaking, sell their soul to be able to do it. And it seems like, it seems like matrix and element have avoided having to sell their soul to have any income.

Simon: Yeah. So, you know, I forgot to ask the question about who the backers were and that can, that can be in the next episode. And I think that, that question about, you know, looking for, you know, a way to sustainability models of projects that have become sustainable. I'm hoping matrix makes it, you know, Matthew makes me feel warm and contented that they're heading in the right direction.

They, they, they, you know, they're operating loss the last few years has been significant. But then so has the investment that's getting them to profitability. So I I'm very hopeful that we're going to see matrix under undergirded with a a profit making ecosystem that is going to make sure that, you know, just like Thunderbird has gone on forever.

Because there are so many people who want to contribute to making it succeed. I'd like to see matrix be in a position where whatever happens, there are people there to both fund it and to work on it. And, and I, you know, I think the separation of the matrix foundation out was, it was, it was a good move.

And I think that the the the funding for element is a good move. The big missing ingredient, as Josh said, is more than one party making a profit from matrix so that matrix can have a sustainable funding, whatever happens to the, the commercial counterparts. And I think that's the key missing. And I was going to say missing element.

That's, that's, that's. That's the key missing component at the moment is, is, is Oh no, we've lost

Jonathan: Simon. With such a, with such a serious look on his face. Making, making such an important point. Well, I was going to let him do his plugs. I know he is webmink. I believe you can follow him at webmink. com.

You can find him on GitHub. I believe also under webmink. And he's got a he's got a GitHub sponsorship there that you might want to check out as well. So hopefully, hopefully Simon will be satisfied with the things that I have plugged for him. If not, I'll get him to record a little, a little a little bit afterwards.

All right. I think that is it for the show. Thank you everyone for watching and listening. We, we did an experiment this week. We tried something. We shot out straight over YouTube. And that seems to have worked okay. And boy, that simplified things for the setup. Hopefully it will make things simple for the edit afterwards as well.

We will be back next week with a couple of developers from Armbian. That is the Debian port for arm, which. Gets used in a lot of places. You may not realize it, but it gets used in a lot of places. So we will be talking with them next week. If you want to follow me, of course, you've got my work over on hackaday.

You've got the security column goes live every Friday, and then, of course, there's the untitled linux show still over at twit. We appreciate twit over there, keeping that going. And we appreciate Hackaday stepping forward and keeping Floss Weekly going. We will see you next week at the same time, same channel here on Floss Weekly until then.