FLOSS Weekly Episode 688 Transcript

FLOSS Weekly Episode 688 Transcript

July 7, 2022

Please be advised this transcript is AI-generated and may not be word for word. Time codes refer to the approximate times in the ad-supported version of the show.

Doc Searls (00:00:00):
This is FLOSS weekly, I'm Doc Searls. And this week, Jonathan Bennett and I are talking with Matthew Hodgson of matrix.org. If you haven't heard about it yet, you will. Matrix is like probably we think we are sold to think by this show as incredible that it is going to be the way that we start communicating by chat and in groups openly, or at least using open an open protocol. It's a protocol, but so much can be built on it. It has enormous promise and it's moving fast. It's probably behind a lot of things you're already seeing without knowing it, and it's open-source, and it's very cool. And that is coming up. Next podcasts you love

Speaker 2 (00:00:47):
From people you trust. This is TWiT

Doc Searls (00:00:53):
This is FLOSS Weekly episode 688 recorded Wednesday, July 6th, 2022. Matrix.Org. This episode of FLOSS Weekly is brought to you by new Relic. That next 9:00 PM call is just waiting to happen. Get new Relic before it does, and you can get access to the whole new Relic platform and 100 gigabytes of data free per month. Forever. No credit card is required. Sign up at new relic.com/floss. Good evening. Good morning. Good. Whenever it is wherever you are. I am Doc Searls and this is FLOSS Weekly on which I am joined this week by Jonathan Bennet himself, who will, for those not visually impaired appear on the screen, he already has. There he is. Hey

Johnathan Bennett (00:01:41):
Doc,

Doc Searls (00:01:42):
You look so much better than you did when we, when you're doing show prep and you're turning around backwards and trying to get your mic working. <Laugh>

Johnathan Bennett (00:01:49):
It's always something

Doc Searls (00:01:50):
There's, there's always, Mike.

Johnathan Bennett (00:01:52):
My setup here is a little too complicated, I think. And so there's always something that wants to go wrong. I have a, I have the new HP dev one came in, and I'm thinking I'm just gonna redesign my whole setup, set the laptop on the desk in front of me, plug the camera into that, plug the mic into that and just go that way. And hopefully that'll take care of some of the complexity

Doc Searls (00:02:13):
<Laugh>. Did it, is this a plan or is this something you just did?

Johnathan Bennett (00:02:17):
It's a plan.

Doc Searls (00:02:19):
It's a plan. Okay. We'll see. We'll see. Next week, how that works. <Laugh>

Johnathan Bennett (00:02:22):
That, that will take more than 30 seconds of work hustling before the show.

Doc Searls (00:02:26):
Yeah, yeah, yeah. In my case, it just came back from a trip where I was, you know, and now I'm back in my own. You might call it a studio. It's actually a basement <laugh>, but stuff here hasn't changed as in the fact that I put, put the laptop into it and it, and it seems to be okay, except that zoom keeps wanting to jump to other microphones and other cameras. So, and that's apparently a zoom thing on the laptop thing. Anyhow. So our guest this morning is Matthew Hodgson of, of matrix. And are, are you familiar with with, with, with this the topic of the person?

Johnathan Bennett (00:03:02):
So

Doc Searls (00:03:03):
I, I am we have 'em on hold right now. <Laugh> right

Johnathan Bennett (00:03:06):
Ahead. We chat a little bit before now. I'm I am real excited about the entire idea of the fed averse, which I'm pretty sure that matrix.org is part of, and that's one of the things that, that they're trying to tie into and make work. And I, I think that is just the coolest idea and, and the best ideas from like an open-source viewpoint for replacing the mess that we have ourselves in with Twitter and Facebook and LinkedIn and every other social media site, that's almost entirely closed source software and almost entirely controlled from the top. And the, the, the key though is, you know, getting this to a point to where the people that want to use it can use it without having to jump through a whole bunch of hoops and kind of getting that critical mass to, to make it something that's self sustaining. So I'm, I'm super interested to hear his thoughts on, you know, kind of the future of matrix, the fed verse and social media as a whole.

Doc Searls (00:03:59):
Yeah. I'd thought about the fed verse. I, my, one of my interests with it has to do with having been involved with Jabber and X PP two decades ago, and having a great deal of hope for it that didn't entirely pan out. So and we pushed it real hard with Linux journal and and it's there, but it ain't what it was or what it wanted to be. So that's, and we wonder where these things go. So interested in talking about that and because we're up to a little bit of a late start, I don't wanna delay, especially since we don't have it, we don't start with head this morning, which gives us just a little bit more time. So our guest this morning is Matthew Hodgson. He is the project lead for matrix.org. It's an open-source project, responsible for the matrix open standard for secure, decentralized communication.

Doc Searls (00:04:45):
And perhaps we're lucky the fed his day job is CEO and CTO at element the company founded by the matrix core team to fund development and provide commercial services for the matrix industry. So, and Matthew's coming in from somewhere in the UK, I assume since <laugh>, he should appear on screen in a moment. There he is. And I, I know he is rushing over from the house of Lords where he was dealing with a topic we'll talk about later, I suppose, but I think you changed before you got <laugh>. You got your, are you wearing exactly what you wore when you went over here?

Matthew Hodgson (00:05:20):
I'm literally wearing what I wore as a round table in the house of Lords about 20 minutes ago. And so I apologize that I'm looking slightly flustered and sweaty having run across town to find somewhere with good internet connectivity. But thank you so much for having me onFLOSS Weekly. It's a pleasure to be here.

Doc Searls (00:05:38):
Oh, it's, it's great having you here. And and I, so I take it, therefore you were in London, we always wanna check it where people are. You know, I didn't mention that <laugh>, that Jonathan's in Oklahoma and I'm in, in Bloomington Indiana, and this being the internet, it doesn't none of that matters except that we have to check in with it. So tell us a little bit about matrix and where it come from, came from and what you're trying to do with it and in kind of the past and future context of it. And, and if you feel like a throw in the fed of risk, is Jonathan just brought that up?

Matthew Hodgson (00:06:10):
Sure. I mean it's interesting to correlate matrix with the fed a so matrix as a project and began in about 2013, when we came up with the idea of trying to build a communication layer for the web, and it's heavily inspired by X and PPP and other open communication standards of the past and before matrix, we spent about 15 years building open standard-based communication solutions using lots of set, awful, lots of IRC, quite a lot of X and PPP. And a bit like the sort of realization you mentioned earlier that 20 years ago, this looked as if it was a really exciting open future that we were heading into the reality didn't quite happen. And instead communication obviously has been fragmented onto zoom and slack and discord and Twitter and all these other centralized, proprietary, closed silos. And we thought that was a bit of a shame because we've basically been denied the opportunity of the open web in the context of communication.

Matthew Hodgson (00:07:12):
There, isn't a kind of crazy open, vibrant wild west where anybody can pitch up and start building communication solutions. Instead it's kind of almost like a dystopia where the web never happened. A bit like a world where everyone's still on AOL, then cohesive and at and T net. And yet they're not actually inter-operating at a IP layer, we have the same problem, but for communication. And so we looked at where X and PPP hadn't taken off and why CE hadn't taken off and why IRC had remained this very geeky elitist thing beloved by nerds four nerds, and wanted to know if we could do better and basically add an HTTP based communication protocol to the web that would become the universal open standard way to communicate. So that's the history. Basically, we got fed up with zap and X and PPP, and wanted to build something on top of HTTP to see whether we could be the missing link and replace the phone network and replace email as the general-purpose way to communicate.

Matthew Hodgson (00:08:13):
And it's kind of similar to the <inaudible> and the activity pub ideas, which emerged a few years later targeting more micro republishing and micro-blogging and Twitter style use cases. And I kind of say that matrix and the fed averse are sort of parallel initiatives. Some people say that matrix is part of averse, but then again, nobody's really defined what fed averse means. It's like cyberspace or I don't know, even the metaverse heaven forbid. So if you could say that we're part of the fed very, we're certainly a decentralized open communication system, like activity pub and light apps like master do.

Johnathan Bennett (00:08:51):
So I I've gotta ask, did you just say why sip didn't take off? Is, is that what I heard? <Laugh>

Matthew Hodgson (00:08:58):
Yeah, absolutely.

Johnathan Bennett (00:08:59):
So, so as a phone system guy, and way back, who has played a lot with asterisk and done some commercial stuff that uses sip, what are you talking about? <Laugh>

Matthew Hodgson (00:09:10):
Well, I'm talking about is where is the sip Yari on your business card? And if I go and phone your no phone number on your business card, what does that even mean? Do you have DNS set up such that I can resolve a sip ROI for your E one sixty four number to call you on sip? I mean, sure. Sip is used under the hood in the PBX, and it's used for the kind of VoIP backhaul in your big telcos, but it's not a consumer product. There's not a sip software client that Joe public uses day in, day out now unto the hood. Whatsapp is doing something very similar to sip, but it sure isn't exposed to the user and it isn't part of a global network. So we see it as a failure, frankly, that there isn't a big public Federation of sip folks out there on the internet who are just talking, calling each other seamlessly instead it's all these very private closed off things and inside PBXs and telcos.

Johnathan Bennett (00:10:05):
Yeah. So sip sip has only really had a life chained to the pot system, the plain on telephone system and where you see it used the most is, you know, connecting point a to point B somebody will have a sip telephone on their desk, but then that goes to a server somewhere, which then goes back to the telephone system. And yeah, I I've seen a couple of these and I, I can't remember the name of it. There was one of 'em I got real excited about until I realized that nobody used it it was part of asterisk and it was one of these deals where it was like, it was going to auto-discover and, and route things directly over the internet when you tried to call somebody. And I, I, I really got excited about it. And then I figured out that there were like three people using it. And it wasn't nearly the cool thing that I thought it was. So for a certain, yeah, for a certain goal for a certain end-use case, what you're trying to accomplish, yes. Sip never took off. And so how is matrix trying to be different? What, what is kind of what's the, the path to get from here to there from matrix?

Matthew Hodgson (00:11:05):
So I guess the things we did differently on matrix was to first of all, use HTTP and Jason as the simplest possible most stupid transport for sending messages. So in matrix, if I wanna send you a message, I do an HTTP put with a blob of Jason to my server, and my server will go and replicate that out to any other servers, participating in the conversation again, via HTTP pop. And then if you want to receive a message, you just hit the slash sync CRO, and it will just sit there until there's a message to receive. And then it will return with a message. So I could literally write a functioning matrix client right now in about three lines of bash, a wild loop to go and grab stuff from the keyboard and post it into Kohl and another wild loop to grab stuff from slash sync and echo it to the terminal.

Matthew Hodgson (00:11:55):
And that's it. So really, really, really try to basically build an open communication protocol using the web standards that developers these days, just default to know where if, if it's what Twilio uses for their API, if it's what Twitter uses for its API, why wouldn't you use that as an open standard, as a kind of flavor of how to do things? And then the other big changes we made is that everything in matrix happens inside what we call a room, and it's not literally a chat room, but it's basically a virtual space with a bunch of people in it. So you don't have one-to-one conversations. So it's the polar opposite on se where onset, as you said, it's BA basically modeling an hundred 50 year old phone call from the plain old telephone system where Joe blogs phones up Ann or Barb, whoever it might happen to be, but it's always a one-to-one call in matrix.

Matthew Hodgson (00:12:45):
It's always a group conversation. And the way we think of one to one communication is just a room that happens to have two people in it. And this is fundamentally different to say X and PPP or IRC or anything that came before where you always had this sort of weird duality that the whole protocol gets implemented twice one for one to one communication, and then a totally separate thing for group communication. Whereas in matrix, just everything is group. Another thing we did differently is to have end-to-end encryption baked in as a concept from the beginning, it took us a few years to implement it. But because we have this architecture where the conversation gets replicated between the servers participating. And it's not just the messages, it's actually the history of the chat conversation gets replicated. You are basically smearing that data over as many servers as participating in a room.

Matthew Hodgson (00:13:40):
And if you take a big room on matrix, something like matrix HQ, the first one we ever created, it's got about 40,000 people lurking in it. And it spreads about 5,000 different servers. So that room, as it happens in public. So we don't bother encrypting it, but it's quite possible to have an encrypted room with hundreds of servers. And you do not want to smear that data unencrypted over all of them, if it's private messages. So we had to have end to end encryption from the outset. Now, I guess from a dev perspective, the way to think of matrix is that it's very, very similar to get. And if you think of X and PPP being more like subversion, where you have a sort of single centralized ma server in X and PPP terms or subversion repository inversion terms, and then get came along and decentralized everything.

Matthew Hodgson (00:14:26):
And everybody had that horrible sort of brain tweak moment where they had to start understanding that now everything is decentralized. And why is it get clone? What do you mean? I got an entire copy of the repository or my computer know what's the main repository, oh, wait, there isn't one. You have that whole sort of paradigm shift. It's identical for matrix that there is no same single central server at all. That everybody who brings those server to a conversation equally owns that history and those messages, and you are kind of subversively decentralizing it, you couldn't even talk to somebody on a different server without sharing ownership of that conversation with them. So I think those are the main differentiators over say, sip or X and PPP for that matter.

Johnathan Bennett (00:15:08):
Sure. So boy, there's a bunch of different directions that I want to go. Let's, let's dig into something though, that that thought that I had at the very beginning of the show about the fed averse, which as you say, is not defined quite as well as maybe it should be. So if I have a Mato on handle, can I bring that ma on handle to matrix and say, because of the background magic of the fed averse, I can prove that this me on Masteron is this me on matrix? Is that something that's wired up?

Matthew Hodgson (00:15:36):
No, not natively. So the fed averse, as you describe it, there is using a protocol called activity pub which has a bunch of clients and services implemented for it, like master Don, like canoe, social, like PLE and many others. And all of those instances are talking the same protocol and they can directly talk to one another. So somebody's handle like I am@matthewatmasterdondotmatrix.org in the fed averse on activity pub will work within that matrix. On the other hand is a different namespace and a different protocol. So you would need to bridge between the two it's like, I don't know, SMTP and N NTP is two different similar looking protocols, but they're actually not able to talk each one another. And unless you've got a bridge of some kind, so on matrix, I'm at Matthew co on matrix.org, although we pronounce it@matthewonmatrix.org and you know, it's a similar setup, it's a username added domain or on a domain.

Matthew Hodgson (00:16:36):
But we deliberately have a slightly different syntax to make sure that people don't confuse it Eva for email addresses or for activity pub identifies. So as I say, it's a kind of parallel world where everything is optimized for real time chat and these big, large virtual rooms with persistent history that are shared between them with encryption. Whereas activity pub is much more focused on the Twitter style, micro blocking semantics, and it doesn't do the replication trick and it doesn't have the encryption there. Although more recently Masteron actually took matrixes encryption and used it to do end-to-end encryption for DMS and mastered on specifically.

Johnathan Bennett (00:17:14):
Very cool. So, oh goodness. I had three questions and they now have all slid it out of my brain. Oh, sorry. Oh, we have a question from the chatroom. That's all. It's what I get for trying to type and talk at the same time question from the chat room. Is this blockchain <laugh>

Matthew Hodgson (00:17:29):
Oh God, no, <laugh> no, not so. I mean, well, I guess it's as much a blockchain as get is. I mean, if you look at how get works, you've got a whole bunch of commences in a branch and each of those is a kind of block of data. And each one has pointed to the previous block and includes the hash of the previous block within it. And if by that metric, it's a blockchain. The matrix is a blockchain too, because it has the same behavior that every time I send a message, I'm basically building up a direct to data, click graph data structure that is replicated between these devices of these servers, I should say. And each message I send goes and points back to the previous message in the room. So you could keep on, you know, following these links in a local tree, as it's called back to the, the original event, what blockchain people would call a Genesis block and what we call a end room dot create event, because it's where the room was created.

Matthew Hodgson (00:18:28):
But the, we don't have any double-spend or proof of work or proof of stake or anything like that because it's not a distributed ledger. It's not a cryptocurrency. It's not designed to replicate a ledger of who transferred, what value to who. And in fact, in matrix, it's a feature that you can net split indefinitely. So on the blockchain system, you expect to close a block every 10 minutes or whatever it might be. And at that point, you basically seal the history of the world into this one global logical centralized, or be physically decentralized data structure. Whereas on matrix, we just don't have that. If you are on a matrix server on a raspberry pie in your car, and you're chatting to people on, on, in your car in it, and you drive out of connectivity, then everybody inside the car can keep chatting away.

Matthew Hodgson (00:19:18):
Even if there's no internet connectivity. And you could do that for six months, if you like building up huge, beautiful chunks of chat room history that only a year people on that local server can see. And eventually when you get back into connectivity, it would link up with the rest of the network and it would replicate those messages through and everybody would continue on their way. Basically having merged the room back with the history that happened elsewhere when you weren't in the room. And we're actually working at the moment on presenting that as a thread, so that physically you would have a sidebar showing the thread of conversation that was happening was you were outta the room or outta the network. And you, you wouldn't have that in the blockchain world because it would try to create a single global snapshot of what happened every 10 minutes or is matrix you're allowed to diverge indefinitely.

Johnathan Bennett (00:20:06):
Oh, that is, that is a really fascinating use case. I, I am in, as we mentioned, I'm in Oklahoma, which means I'm not in one of the largest cities in the world. I'm, you know, we, we have this joke that says that let's see, how does it go? I don't live in the middle of nowhere, but it's close to here. And there's, there's literally a gas station and a house that is an unincorporated town called nowhere, Oklahoma. So we, we kind of care about this idea out here of off the grid communications and making it work. And that what you just described is just about the most compelling solution for that. I think I've ever heard. That is really fascinating. Okay. So I I've gotta ask and I, I don't mean to be insulted, but this is just what comes to mind. It sounds like matrix is a smarter version of email. It's like email 2.0 which that could be taken as an insult. And I don't mean it to be, but here's the question. What do you do about spam? Has spam become a problem yet?

Matthew Hodgson (00:21:03):
Oh, yes. Unfortunately it has particularly in the last couple of months. There's been a real uptick in spam to the extent that some communities have moved off matrix onto discourse, very depressingly just because of some targeted brig, abusive spam attacks. And we've always known that this is gonna be one of the more obnoxious risks for matrix, because it's a big decentralized communication network. There are about a hundred thousand servers that we're aware of on it. And about 65 million users. So it's not quite as big as email. Well, the way Paul said, but you know, it's getting in the right direction, roughly doubling or tripling in size every year. And, you know, if it's a military grades, crypto com system that anybody can run on their own terms and has no single points of control or censorship whatsoever, it kind of begs the question as to how abuse is gonna manifest itself.

Matthew Hodgson (00:21:57):
So an element which is the company we set up to basically keep the lights on and keep us employed to work full-time on matrix. We've got, I think, five, six people in a dedicated trust and safety team whose fun job it is, is to define anti spam and moderation and anti-abuse tooling. And right now we're actually going through a real pivot from looking after our own servers versus building tools that everybody else can run to look after their own ones as well. In terms of whoever metrics is like email 2.0, I'd actually say it's a lot more like Usenet 2.0, it's much more like N NTP. And if you remember how the old Usenet newsgroups worked, it's pretty similar. They were replicating messages in real-time, over a big decentralized network. And also, they had a pretty big abuse problem in the later years, poor old Usenet just filled up with all sorts of dodgy stuff, rather than the high quality conversations of peers gone by.

Matthew Hodgson (00:22:57):
And we don't wanna see that happen to matrix. So what we are doing is basically building the concept of morally relative reputation systems into the matrix as a first class citizen. So this is pretty opinionated. Imagine if the email or N NTP or even HTTP had the concept of reputation built-in now, before anybody goes nuts and thinks that we're building a kind of Chinese social credit, absolute reputation system. We're not, we're literally doing the opposite where instead we empower everybody to have their own personal view of what content they want to see and don't want to see on matrix. And anybody is welcome to publish what we call reputation, gray lists of stuff that they think is of interest. So for instance, I might subscribe to a NSF w reputation list and say, I'm gonna use this to block any NSFW content from ever appearing on my matrix client, or I could subscribe to it and I can turn it up and say, actually, I, I don't mind seeing NSFW stuff is the kind of Google equivalent of turning off safe search or whatever, but with safe searching Google, there is only one reputation list that you can subscribe to from Google.

Matthew Hodgson (00:24:09):
And you've gotta trust them to be filtering it out perfectly. Whereas in matrix, we're chosen to make it completely subjective. People can publish whatever reputation lists about whatever thing they like. And it could be the world's best analytics content. It could be the world's worst BSD content, and it's up to them whether they want to dial the volume up or down on it, you could also use these to decide what content is allowed into your communities. So if you are a community admin and you want to run a non S F w content the community, then you could subscribe to one and apply that. And if you're running a server and you want to block particular content from ever coming near your personal server, that likewise you could subscribe to it and say, Hey, you know, I don't want to have illegal content at all on my server because I don't wanna go to jail.

Matthew Hodgson (00:24:58):
If people discover that some people have gone and started using it to communicate about terrible stuff. So this is pretty ambitious. Nobody, as far as I know, has got this to work before everybody else goes for either a, basically an absolute view, like on spam assassin, it has a bunch of how code drills about what is spam. And it learns a little bit from your personal activity, but it's not like you can then share that with other people or subscribe to it. And you look at DNS blacklists for email. And again, it's very it's pretty corrupt. Honestly, it's a sort of centralized system, almost a, a shakedown where you have to pay to get your IP address taken off the block list and this sort of thing. Whereas on matrix, we want it to be a much healthier, much more reverse, sort of open ecosystem where people can make filtering choices based on whatever metric they like.

Matthew Hodgson (00:25:52):
So that's how we're trying to solve it. And we're in the middle of building that stuff out right now. It's actually live with Illa and FOS D and a few other big matrix servers. Ober is a pretty blunt granularity where we're just sharing band lists at the moment, rather than the sort of shades of gray level. But we have experiments going on where you can actually visualize the different reputation lists and see that particular content is being filtered from your view, because you've subscribed to this and it might say, Hey, by the way, you are, we are showing you this thing. 98% of the world has filtered it out for this reason, but we're showing it to you anyway, by the way, you know, that's where it sits in the context of things. And we even visualize this as maps where you can do cluster analysis on the various different communities out there.

Matthew Hodgson (00:26:42):
And you can imagine a big sort of map of the world made outta these virtual continents. And you can move a pen around that map to see, you know, what viewpoint it is. And you can increase the radius around the pen to, until it covers the entire world. And then you have all the filter taken off, or you can drag it over to one particular community and then zoom in and view the content from that community. So that's the sort of thing we're doing, but frankly, we're a bit behind the curve on it, given the level of abuse we've seen recently, and we're having a better of battle stations to up our game and particularly provide that tooling, the wider community ASAP from within element, rather than today, you have to run a separate bot for it, at which point, know 90% of people run away because they don't want to run a bot to mitigate spam. So instead we need to really hoist it up and build it into the core app that people use.

Doc Searls (00:27:36):
I have a boy thinking about markers of success. I, I suppose getting spam is, is one of them. You know, suddenly you have an attractor there for bad acting, but I have some other questions about that. But first I need to let people know that this episode of pH weekly is brought to you by new Relic. If you're a software engineer, who's, you've been there, it's 9:00 PM. You're finally unwinding from work your phone buzzes with an alert that something's broken and your mind's already racing it. What could be wrong? Is it the back end of the front end? Is it global? Is it the server? Is it a network? Is it a cloud provider? Do we have slow running queries? Did introduce a bug in my last deploy. Now the whole team scrambling from tool to tool and messaging person after person to fix the issue.

Doc Searls (00:28:19):
According to a new Relic report, only half of all organizations are implementing observability for their networks and systems. The report showed how maintaining network observability continues to be an issue for companies around the world that would happen if you get new Relic, new Relic combines 16 different monitoring products that you'd normally buy separately. So engineering teams can see across their entire software stack. In one place, you get application monitoring, that's APM, unified monitoring for your apps and microservices Kubernetes and pixie instant Kubernetes observability with pixie distributed tracing, see all your traces without management headaches. So you can find and fix issues, fast network performance monitoring, stop guessing where performance issues start and ditch data silos for a system wide correlated view, and much more, more importantly, you can pinpoint issues down to the line of code. So you know exactly why the problem happened and can resolve it quickly.

Doc Searls (00:29:23):
That's why the dev and ops teams at door dash GitHub, epic games, and more than 14,000 other companies use new Relic to debug and improve their software. Whether you run a cloud in ATO startup or a Fortune 500 company, it takes just five minutes to set up new in your environment. So that next 9:00 PM call is just waiting to happen. Get new Relic before it does, and you can get access to the whole new Relic platform and 100 gigabytes of data free per month forever. No credit card required. Sign up at new relic.com/flos that's N E w R E L I c.com/flos new relic.com/floss. Okay, so, so Matthew I've, I've done a little bit of homework here and, you know, I'm looking at, or have looked at him. I'm not looking at him right now. You started really pushing this thing in 2015, it's been seven years.

Doc Searls (00:30:21):
And you just gave some numbers about how, how big it is and, and all that. And yet, I, I don't think matrix is a household term yet. So I'm wondering, I mean, obviously you're big enough to attract spam that's that's so that's a sign of a sign of success cuz excess chases success in certain ways. So I'm wondering what, what you're looking forward to there. What are the markers of that? I mean, you, you have a team there at the company and it's moving along, but I'm so mindful of what happened with X and PPP that I, I personally feel a little burned about that. So though I was outside the kind of the crater, but it's still it's there and there's some bitterness around it too, you know, and some of that has to do, I'm sort of jumping on a possible answer you're with having it, having been adopted by Google and others that then dropped it. So, or partly dropped it. So I'm wondering if a sign of success might be, Hey, the bigs say, wait, we like this, we like this a lot. We're gonna use it now, what, you know, if they decide they don't like it.

Matthew Hodgson (00:31:30):
Yeah. I mean the history of X and BP is obviously an interesting one to study to understand things which went wrong. I would say it was both the fact that Google hugged it to death, frankly, by kind of betting the farm on it and then very publicly dismissing it and also not necessarily pushing it forward that much in the public domain, as well as the fact that the original Jabber team who wrote it, a Jabber rink got acquired by Cisco quite early on in the life cycle of the whole thing. And it sort of went a bit off the boil after that point too. So what we've trying to do with matrix to avoid the same fate is, first of all, to never, ever pursue one single big Dyna not dinosaur, what am I saying, whale? Shall we say like Google?

Matthew Hodgson (00:32:21):
At a time we would always go and pursue multiple ones and try to make sure that there are sort of multiple 800 pound gorillas in the room to try to avoid the ecosystem becoming lopsided. And we've had a few instances already where we've basically had to tell individual ones from G M to no, it's very kind that they're interested in matrix, but frankly it's not the right time for them to jump on board and make some massive billion user platform start speaking matrix because we were still in beta or if it was just them who did it then suddenly matrix would be seen as that public API that so and so used. And then they would basically control it. And if they turned it off tomorrow, we would see the same mess that happened X and PP, but you are right that we are not a household name yet, even though we're seven years in.

Matthew Hodgson (00:33:13):
If we were a normal like software startup, this would be a disaster like seven years and you haven't taken over the world. I mean, you should have IPOed by now or whatever, but the reality is that we are building something way bigger than a typical startup. I mean, the, the mission on matrix is to try to build it to be a trillion-dollar industry that replaces the phone network and email and use that and all these other things. And the element should be a major player in that space, but you know, have 10% market share of that industry, leaving 90% market share for everybody else to jump on board. And there are lots, lots of companies now who are building out really interesting things on top of matrix, but the reason it's not mainstream yet is two things. First of all, our focus has ended up bizarrely being towards big enterprise and public sector staff.

Matthew Hodgson (00:34:05):
So we are one of our first big breakthroughs was in 2018 when the entirety of the French state adopted matrix as its official communication platform. So that's 17 ministers and five and a half million public sector employees who were told that, by the way, if you're gonna communicate about stuff for work, please stop using WhatsApp and telegram, particularly telegrams. It's not encrypted and it's definitely not French. And instead, please, can you use the matrix deployment that we're providing for you? And that was a real sort of step change in matrix because frankly it meant that we had enough revenue to build a sensible company on top of it. And after France wanted it, then Germany wanted it too. The UK is also deploying it bits sort of the us are also starting to roll it out, lots of Scandinavia looking at it.

Matthew Hodgson (00:34:52):
And we found ourselves very unexpectedly going down a little bit of rabbit hole, making damn sure that our paying customers are able to run these huge public sector deployments successfully, but that sort of pulls in a slightly different direction to say building a signal style app. And I'm sort of jealous of the signal guys who have a diametrically opposite approach to all of their stuff in that they build centralized closed communication networks, but at least they are just building a messaging app for mainstream normal users. And they're not having to worry about complicated single sign on, or I'd know, how do you handle audit capability in an into end encryption world if you're doing some regulated business or whatever else it might happen to be. And so we're in a slightly weird situation where the day job is very much looking after these guys whilst slowly but surely improving the quality of the underlying apps until eventually we hope that the app quality itself will be as good as the discords or the slacks or, you know, the WhatsApp and as a normal mainstream user, it suddenly becomes a lot more compelling to use it.

Matthew Hodgson (00:36:03):
But right now, if you flag somebody down on the street, unless they work in communication tech, or perhaps if they work in open source or government stuff, they're not gonna have heard of us, but we keep no moving things forwards and we are seeing more and more organizations like Illa and red hat and Wikipedia and these sort of outfits moving over to matrix KDE and no as well. Rocket chat, one of our competitors element are editors that element have also adopted matrix as their way to communicate in a decentralized fashion, which is really positive to see other players in the open-source communication space, adopting the underlying protocol. Even if they're now doing their own thing with it, but as the years go on, we are hoping that we will improve the quality and the usability of the apps that normal people will start just using it a bit, like people have started to use signal on a day to day basis, even if they're not cipher punks or open-source evangelists.

Matthew Hodgson (00:37:08):
And it's kind of reassuring to see that signal was able to bootstrap itself out from being a, a pretty hardcore techy thing to become mainstream. And just like, I don't know, GitLab has done the same and WordPress has done the same. I'm pretty confident that we'll manage it too. But it's hard to put a good user interface, particularly on open source stuff, particularly on a open standard with a big open community, all of who want to paint the app a different color and have different features. And we have all the people who want to replace the composer in element web with them or EAX or both. And we need to sort of push back on that sometimes and say, Hey, let's just simplify it and do simple, beautiful apps that normal people can understand whilst also keeping the lights on by selling variations of it to governments and big enterprises.

Johnathan Bennett (00:37:59):
Goodness. So I have three different questions I wanna cover. Let's start with the first, the simple one. I'm sure there is a command-line interface. You, in fact, you already said you could put it together with about three lines of of, of bash code so that we can run matrix from within VM and EAX right.

Matthew Hodgson (00:38:15):
<Laugh> yep. There's a really good EAX client. It's quite mature. It's about five years old now called, oh God, I'm gonna get it wrong. It's eman dot El. So it's a anagram of element <laugh>, but dot El because it's, EAX li and it's actually really good. It sort of supports feature parity with element and has all the exciting bits of the matrix back in there, but implemented as a mode in EAX and there are also lots of other really nice text user interface ones. There's one built on WeChat, which is, you know, a bit like IR SSI but no really, really good implementation first in Python and now a rewrite happening predictably in rust. And then there's also a native matrix, one called GOM, which has been written in go Lang and is a completely standalone stack written by really talented member of the community called tiller, which again has a pretty comprehensive feature set with end-to-end encryption and reactions and edits and typing notifications and all that good stuff, despite it being a command-line client.

Johnathan Bennett (00:39:23):
All right. And then the other question I have to ask, you mentioned wanting matrix to eventually replace the public switch telephone network. And I have to ask, does that mean that there is more to matrix than just sending text?

Matthew Hodgson (00:39:34):
Oh yeah, definitely. And we're quite serious about replacing the PST with matrix. I mean, matrix itself, if nothing else, was a rebound against many years we spent working on the phone network and getting more and more frustrated with the fact that the pots haven't really changed in the 150 years. So our matrix, you are synchronizing arbitrary blobs of data, and the simplest one would be a text message, but it's basically anything you can put inside Jason that is smaller than 65 kilobits, and you type your messages with a Javas style reverse DNS namespace. So for instance an instant message is M room message. I an instant message inside a room inside the special Mt. Namespace. But if you wanted to do a VO call to somebody, you would send an Mt call dot invite event, which is kind of similar to a sip invite except we just extended it to also support group calls natively.

Matthew Hodgson (00:40:31):
So you can go and build out zoom style, decent or zoom, zoom style, video, and voice conferencing, but keep it decentralized and end to encrypted. And we have a app called element call called element.io, which looks and smells quite a lot like zoom or JY, but is powered entirely by matrix. So you don't need any other servers to use it. If you already have a matrix server hanging around somewhere, you can just create a room and start firing all of your Mt. Call, invite events, back and forth, and people will M call to answer them. And we trackle ice candidates to rapidly set up the call using web RTC as the actual VoIP media stack. And it gives you full voice and video communication over matrix multiway. And another thing we just added is the ability to collaborate spatially over matrix.

Matthew Hodgson (00:41:25):
So there is a project to do text editing with CR DTS. In fact, there are several projects now to do text editing a bit like ether pad or Google docs over matrix, but the one that's closest to my heart is doing 3d spatial collaboration. So going and collaborating in virtual worlds effectively using glTF as a open standard format for the 3d assets, which get uploaded into the matrix room. And then you use the matrix VoIP in order to do spatial VO to communicate within that space. And you run around interacting with the world, doing whatever you want with it as a blank canvas importing new assets, adding in user generated content, and basically extending the chat rooms into a full physical virtual environment for whatever sort of S spatial collaboration it could be. And it could be gameplay.

Matthew Hodgson (00:42:18):
It could be GIS data. It could be, I don't know, a Google earth style environment and it's brand new. We haven't, it's very much alpha. We haven't properly launched it yet, but it sits out there@thirdroom.io. And we're pretty excited about it as being a very open, equitable alternative to the distinctly closed or crypto backed visions of the metaverse, which we see flying around out there. So that's just how flexible matrix is. We also have a thing called element drive, which is Alon of Google drive built on top of it as another example of using it for something that really isn't chat.

Johnathan Bennett (00:42:57):
So it sounds like you can build pretty much any just about any service that you want to on top of matrix. So I'll just throw out a simplified example say I wanted to build a chess match service all about, you know, people playing chess against each other over the internet. You could build that on top of matrix fairly easily. It sounds like.

Matthew Hodgson (00:43:18):
Yep, absolutely. And I think I saw somebody had done that. And what you do is to use what we call extensible events, where you would have your domain specific blob of Jason, and it could be something like, I know com dot chess dot move as the name of the event. And if you own chess.com, then you could define that event type. And you could say to the world, Hey, it's gonna have these fields inside adjacent object to describe which piece you're moving from, what location to what location, but you can also define the fallback that to a normal text message formatted in HTML or just plain old text. So that if I was on a client, they didn't understand how the chess move should be interpreted. It would show it as a plain old text message. But if I am on my client or if I have what we call a widget, which is an extensible, I embeddable app embedded within that client, which does understand that message type, then it could throw up a proper 2d or 3d chessboard and show the move with all the rich semantics of that, which you would like.

Matthew Hodgson (00:44:19):
So for instance, in the metaverse third room stuff, you can enter those rooms via a normal chatline. And if it has voice capability, you can talk to the people audioly in the room. If it doesn't, you can just chat to them. And if it has the 3d bells and whistles, then it will actually rent to the room, or it will at least be able to show you snapshots of the 3d state within that room. So we had this extensibility to allow you to go as domain-specific and as richly as you like whilst also falling back to something that you can just join from any old, random come blind client. And it will either render the text equivalent of what's going on or worst case. It will give you a link which will then open in a browser, a sort of more dedicated viewer so that you can see more the, the richer stuff of what's going on.

Doc Searls (00:45:06):
So in our own back channels here, there are questions about bridges, and you actually have a page on bridges, which I'm sure an can bring up on screen in a second and interoperating with all these, all of these different things. And I'm wondering the question I have about it is if you can bridge to all these other things that are probably in some ways more brand familiar than, than is matrix, does matrix get lost in that? So there are really two questions there. How, how does the bridging work and how would we even know what's happening I suppose, is another thing, and that may not be a problem. And is that a, and is that a risk that you just become kind of a, not so much a backend, but invisible to, to end-users who we would like to, you know, we'd like to have branding happen here too, right?

Matthew Hodgson (00:45:55):
Honestly, it's a result if people use matrix to bridge together different platforms, and they're not even aware they're using matrix, that's fine by us because it's just opening up those silos, those ward gardens and giving the users the openness and the ability to communicate with other platforms or to build or use other systems on top of those communities. And people use 'em a lot. I mean, we see bridging as the first plus citizen in the matrix. In fact, it's the reason why metrics is called matrix because it is a matrix and much different things communicate to one another in a kind of X by Y or N BYM sort of set of things talking to each other. It's then again, it's always gonna be a bit of a compromise because there's always gonna be some feature that I know discord has or slack has that the other guy doesn't, and it might be something that isn't natively in matrix, although, as we said, matrix is incredibly extensible.

Matthew Hodgson (00:46:50):
So if we've done our job, right, it should be expressible in matrix, but it doesn't mean that, I dunno if I'm talking from WhatsApp to telegram and I know telegram sends me some funky phone that just doesn't exist in WhatsApp yet, like some sticker or whatever. What are we going to do to bridge it over and to answer your questions about how they work. It's a mix of open APIs that we connect through to. So for instance, slack has an open API that we integrate against and it gives a pretty rich experience allowing you to send and receive messages and threads and reactions and edits and typing notifications and all that good stuff, such that the people on the slack side will not realize that they're actually talking to somebody on matrix or perhaps somebody on another system in turn, which is bridged into matrix.

Matthew Hodgson (00:47:38):
And likewise, the people on matrix just won't care that they're talking to people on slack. And in terms of the risk of matrix getting lost in the noise when this happens often what we see is that the balance kind of tilt over time. The very first slack bridge we ever did was for the decentralized web summit, which is a great event that the internet archive has arranged since 2016. And we just launched the slack bridge that year and they set up on slack and we said, guys, you can't run a decentralized web summit and communicate via slack. The most centralized com system there is please can we push it to matrix? And they said, yeah, sure, whatever. And we started off with like a handful of people on the matrix side using it and everybody else on slack. But then as the years have gone by more and more people have just kind of absentmindedly almost automatically moved over to the matrix side.

Matthew Hodgson (00:48:28):
And nowadays the balance is like 95%. There, likewise, we see this when people bridge to discord or IRC, that sure there's gonna be some people who just blindly use discord or IRC and are happy with it. But people actually kind of like matrix when they start playing with it and where they realize just how open and fun it is and they move over to it. And you see that the balance and conversation often does shift to native matrix. But if it doesn't, that's cool too. The whole point is to make it optional on what platforms people use to communicate. And if some people wanna stick out on IRC, that's fine. And if some people just really like discord or dunno better, then that's fine too. And that is why the Flo weekly discord should be bridged to matrix

Doc Searls (00:49:11):
<Laugh>. Well, well, we'll hand that question over to the people who are responsible for it. We do like to open, we do like to support open source as much as we can, whatever it happens to be. A, a couple of thoughts. One, one is just as it occurs to me, you've got a protocol. Obviously, if you have a protocol, then that is inherently bridging in lots of ways and lots of protocols get used in ways nobody ever imagined when they came out. I mean, HDTP, for example, was just a way for high energy physicists to look at each other's documents inside a land <laugh>, you know, and it turned out to be the Trojan horse in which the entire, what we now call the web ran without the phone companies, knowing what they were saying yes. To, by allowing it to run on T C P I P.

Doc Searls (00:49:59):
And I, I kind of wish that you, I consulted BT for about six years. I kind of wish that I'd known about matrix at that time. I dunno if they would've welcomed it. But but it, but it is interesting that to think of, of, of other uses one that comes to me and, and I'm not sure whether this is, and I'm sure it's a legitimate use, but one of the things I especially hate it when I go to some companies customer support site and they have a little chat thing and it's always awful and it's always run you sense by some third party that some big proprietary thing that sold them a bill of goods and they're, and they're running it in that I'm wondering if, if there's a, either a business for element or a, a business at all in, in doing that with matrix there's matrix already behind some of those. And I don't even know it.

Matthew Hodgson (00:50:51):
So I mean, this always answers if it's a planted question <laugh>, but it isn't your audience at all. But literally tomorrow as stage of July the seventh, we will be launching chatterbox, which is a matrix powered chat box to go into your website, built using hydrogen, which is our lightweight open source. I mean, honestly, this wasn't a plum, right? <Laugh> you had no idea

Doc Searls (00:51:15):
It was out of the blue <laugh> and now there it is very good. So what's it called again,

Matthew Hodgson (00:51:20):
Obvious it's called chatterbox. And I think the press release is going out tomorrow. It's a completely open-source, of course. And it's built on this very lightweight web SDK that we've created called hydrogen and hydrogen is an entire really glossy, fully end-to-end encrypted matrix client within 70 K of HTML, CSS, and JavaScript, and a bit of web assembly for the crypto. So it's tiny, tiny, tiny thing smaller than your typical image that you can embed into your site with complete visibility as to what it is because it's all open source. And then it gives you this Intercom style customer support chat experience sitting. There is a little icon in the bottom, right? That you click and up comes the matrix client. And obviously if you wanted to then connect a full fact matrix client to it and integrate it, perhaps on the customer support side with a great big element thing, juggling thousands of conversations in different groupings and spaces, then you can do that too. But yeah, we have the same fourth and then fourth GE I wish that somebody would do customer support on top of matrix. Well, if no, one's gonna do that, we might as well do a element and then provide pay hosting for people who can't be bothered to run their own matrix server. So that's basically the business model on it.

Johnathan Bennett (00:52:37):
All right. I've gotta, I've gotta jump in and ask, you mentioned this idea of bridging the TWI discord with matrix and I'm all for that. The problem is the twit discord is whitelisted. It's tied into one of the one of the ways that twit is supported. I'm curious, does matrix work with a whitelisted discord? Is there a way to make that work, even though, you know, matrix is intentionally decentralized and kind of opened to everybody.

Matthew Hodgson (00:53:04):
Yeah, yeah, absolutely. So we have access control in matrix, and in fact, academically, it's the big novelty of matrix. We didn't realize it at the time, but in a matrix, you need a way to be able to limit what users can do, and what operations you want to have invite rooms. You want to have whitelisted ones. You want to have groups which are bridged through to another limited-access system, like a discord. And you want to make sure that only people with the right token are able to get in for your sort of use case. And the way we do that in matrix, even in a super open decentralized environment where nobody trusts anybody else. So it's kind of busy and time full tolerant is that you have to prove everything you do. So when I send a message between service and matrix, I say, Hey, I'm sending a message in the room.

Matthew Hodgson (00:53:50):
And the reason I can send that message is because I got invited by doc and he created this at this point, or it could be I'm gonna kick Bob from the room and I'm allowed to get Bob because on admin, because I got given admin privileges by Charlie at some point, blah, blah, blah, blah, blah. And if you do that consistently, every time you send something and everybody executes the same proof and checks, whether it matches their map of the world and accepts that operation, if it matches their view of the world. And if youd lie and say, I'm going to take over the world and kick everybody because I've got admin, but I'm not gonna prove why I got admin. Then people will just ignore you like a crazy person and not go and let you vandalize the room. So it's that technology that we use, which turns out to be a novel, a bunch of German researchers ER, Institute of technology, distributed systems, network group started writing exciting academic papers about how we had cracked the problem of decentralized access controls without finality, I, without a blockchain.

Matthew Hodgson (00:54:53):
And apparently, it's a interesting academic result, but basically it allows us to bridge through to your discord with its whitelisting, whilst upholding the white list. Now the discord bridge of the box doesn't do that. We would have to figure out a way to actually let the users on matrix, prove that they have, you know, put money behind the bar or whatever the metric is that they need to be allowed into the discord. But conceptually it's absolutely fine. And we do it a lot for ILC and teams and slack already. I just don't think that the discord bridge has native awareness of it yet, but no, it wouldn't be hard to add.

Johnathan Bennett (00:55:32):
Okay. I I've gotta ask two things real quick. First off. How does someone get started with matrix? Either connecting to one of the existing servers or I wanna build my own little server here. That'll be an island. If you know, we, we lose connection to the internet. What does that, what does that process look like?

Matthew Hodgson (00:55:49):
So for getting started, just if you wanna hop on, on a public server, go to element.io and hit the get started button, and it will fly you through to a web or a mobile app, like app element.io, which is the web app. And you can then just jump straight in. So you can see get started up there at the top, right. And that will tell you either go download the app or go and launch element web at the top and, and get started thing. And that will default to the matrix.org server, which is the first one we created. And it's got about 10 million people on it who just use it as a quick try before you sort of commit side of things. But what we often see is that people try like it and then say, oh, well I want to run it myself.

Matthew Hodgson (00:56:31):
And to run it yourself. Best bet is to properly run synapse, which is our Python based server implementation, which you can grab off GitHub and you can grab it from the sauce on GitHub. You can go grab it from pep. You could grab it from Debian packages. You could get it as a Docker composed image. And honestly the best way to get going is probably to Google or search using your favorite open source search engine. The tutorial that I did at the beginning of the pandemic of how to run your own matrix server and jet C conferencing system, because we historically have integrated jet Z for our conferencing before we added the native matrix voice and video conferencing. And it's a terminate YouTube video of me basically speed running, doing a digital lotion droplet from first principles going and setting up DNS, setting up in apps, using the M packages, putting the certificates in the right place and pointing element at it. And Hey Presto, you've got yourself a matrix client that could be running on a VPS, or it could run on a raspberry pie, or it could run on your laptop or wherever you want to run it.

Johnathan Bennett (00:57:40):
All right. You mentioned an app. I assume there is a muggle friendly app, one that you know, grandma and grandpa could use

Matthew Hodgson (00:57:48):
Element aspires to be that app where it is debatable how far we've gone on the friends and family side of things. But, you know, it's an open protocol with loads of really nice apps out there. Fluffy chat is one that is really nice from a kind of messenger use case written entirely independently by a community based out of Germany and flatter. So you get kind of nice cross-platform flatter style semantics built on a Dar SDK element itself, I think is quite good. But then again, I wrote large chunks of it. So I'm probably biased <laugh> but it does cater a little bit more for the sort of slack or discord style use case rather than the WhatsApp one. And we're building a new app called element X at the moment that is starting off much more focused. And in fact, it's so new, it isn't on that page which is much more focused on the kind of messenger use cases. And it's built on a rust test, DEC that we've created and the UI is written on iOS and swift UI, and it looks and feels very similar to iMessage or some other sort of mainstream messaging app, but that will be launched in the next couple of months.

Johnathan Bennett (00:58:58):
All right. We, we are at the end of our show and there is something that we absolutely have to ask you about, and that is you hustled over here to do this from someplace pretty interesting. You wanna plug the meeting that you had just just an hour ago?

Matthew Hodgson (00:59:12):
Sure. so I was at the house of Lords, which is one of the legislative bodies of the UK government at a round table discussion about the online safety bill, which is a piece of legislation that the government is trying to put through. That basically puts a huge onus on tech companies in the UK to frankly, censor the content that goes through them in order to protect their users from online harms, which is a very, very broad definition of anything that ranges from insulting people to encouraging eating disorders, to terrorism, to child sexual abuse, or many, many, many other things. And the idea that there should be some obligation for tech companies to know, be responsible for moderating their platforms, I think is probably a legitimate one, but the implementation is terrifying and horrific in that, for instance, they are as of this morning suggesting that all content has to be scanned for a child abuse and terrorism content, even if it's end to end encrypted.

Matthew Hodgson (01:00:23):
So the home office literally amended the bill about eight hours ago to add this as a stipulation that end to end encryption would not be what had to be undermined in order to scan for content like this, which is the worst that the UK has ever gone down the path of trying to make encryption effectively illegal. So this coincided with an opportunity to meet with members of parliament, to tell them firsthand what a terrible idea it was. And so myself and the few people leading similar projects, including made safe on the decentralized storage side of things turned up. And we tried to explain to the politicians that all they are doing is making it very difficult for the good guys to run communication services because they are fundamentally undermining it. And meanwhile, the bad guys will still continue to run their own communication services and we'll ignore the stuff that says you've got to censor everything.

Matthew Hodgson (01:01:22):
So you literally end up harming the, the, the good actors and undermining privacy for everybody while not actually doing anything to solve the underlying problem that they're trying to fix. And we may or may not have got through because the politicians obviously love the idea of a quick and easy way to say, don't worry, we're making the internet a better place and we're saving the children. And, you know, we are being crack active and passing laws. I'm hoping that I, I literally compared it with the time when politicians tried to redefine pie as 3.0, because it's easier to think about 3.0 than some other horrible number. And I think the point landed, but who knows

Doc Searls (01:02:03):
<Laugh> unless what IM says, it's not 3.0. I thought it already was. I've been rounding it all this time, making a spiral. <Laugh> oh my gosh. We are so out of time just two final questions, we always ask everybody. And then and we'll head out. One is, what is your favorite text editor and scripting language?

Matthew Hodgson (01:02:25):
Oh, well, I mean, that's easy, I'm afraid it's original VI, which is how I learn to edit text on an old Silicon graphics, workstation and Iris 1400 running. Even before I system five Unix with BSD enhancements from the 68,000 processor, which my dad brought home from his workplace. And I just love VI to this day and it's totally muscle memory in, and whilst we're on the subject of bad, bad habits, I'm afraid that my favorite scripting language continues to be PO for much the same reason that I have PO in my veins. It is literally I, if I was just gonna write a massive script without thinking and having to engage my bread tool, it will come out as a whole bunch of Pearl. So it's never gonna go. It's a sunk cost fallacy that I'm never gonna escape.

Doc Searls (01:03:12):
That's fantastic though. That's good. So, Matthew, this has been fantastic. It's been a way too quick hour. You're clearly moving forward. We're, we're gonna have to have you back and to reflect on additional success and and, and thanks for being with us. It's been great.

Matthew Hodgson (01:03:29):
Well, thank you so much for having me. It's been lots of fun.

Johnathan Bennett (01:03:32):
Hm.

Doc Searls (01:03:34):
So Jonathan, that was good.

Johnathan Bennett (01:03:36):
Oh, it was excellent. Still is. I, I kind of feel bad. I, I, I sort of monopolized the conversation there and asked all the questions. We'll have to, we say this

Doc Searls (01:03:44):
Everybody. No,

Johnathan Bennett (01:03:45):
That great. You're we say this about everybody, but we absolutely will have to have him back in six months cuz they, you know, they're launching multiple things right now. And you know, a lot of the things we asked about are things that they're launching tomorrow or in a couple of months. And, and that'll be great. I am trying to, I am right now trying to convince a friend of mine that, Hey, you know, that old messaging platform we were using that we're not really using anymore. We totally need to replace it with matrix. And then I'm, I'm also kind of thinking in the back of my mind, how do I do an off-grid matrix server here at the house that I can then use ham radio to connect to for the, the, the times when, you know, a tornado comes through and there's no internet for anybody this, this definitely has gotten my my gray matter percolating, very interesting stuff.

Doc Searls (01:04:37):
So I, I, this is the point at which I usually plug next week and we, we have lots of really great guests in the queue, but all I see there right now is new Craig bot maintainers. So that, that has to stay mysterious as to the person and exactly what we're talking about there. So, but there are a bunch of them lined up. They just have they're in a crowd, not a queue <laugh> so put it that way. <Laugh>, they're sort of crowd they're, they're crowding the desk, but we don't know who's on first. And also it is also time to do our plugs, even though we're almost outta time. So give us yours there, Jonathan.

Johnathan Bennett (01:05:11):
All right. Well, it's the same too that I have every week, but they're super important. The first one is over@hackaday.com. Make sure and watch for my articles. Particularly the security Roundup goes live every Friday morning. That one's a good one to keep track of, particularly if you're running your own services people get a heads up about things that are broken and yeah, it's important to keep on top of, and then the second one is club TWI. We've got the untitled Linux show, a bunch of other things you need to be on club TWI, come join our discord. I'll see if I can talk to powers that be into getting us a matrix bridge. We'll see how that goes, but get on club TWI and go subscribe to the untitled Lennox show. We have a lot of fun over there. That's it. Thanks doc.

Doc Searls (01:05:51):
Thank you. And we are, I, I think we just got incredibly well sold on matrix <laugh> so <laugh>, I dunno we can how we can resist that. So, so, so, so thanks everybody. We will see you next week.

Speaker 5 (01:06:06):
The world is changing rapidly so rapidly. In fact that it's hard to keep up. That's why Mica Sergeant and I, Jason Howell, talk with the people, Mac and breaking the tech news on tech news weekly. Every Thursday, they know these stories better than anyone. So why not get them to talk about it in their own words, subscribe to tech news weekly, and you won't miss a beat every thursday@twit.tv.