Ungovernable Misfits

Share on social media:

Welcome to the The Confab, the term derives from "confidential talk", which was commonly used in the Prohibition Era for meetings and conversations that took place in the smoky, rule-breaking speakeasies of that time.

The informal, privacy focused and clandestine nature of Ungovernable Misfits lends itself to these discussions.  So, grab a seat and a stiff drink from the concealed bar, listen and revel in the conversation.

On this episode, Matt Hill from Start 9 drops-in to speak with Max

Show Discussion

I. Networking Challenges and Limitations in the Current Start OS

- Matt explains the issues with the current networking architecture in Start OS, including the reliance on Tor and limitations around VPN connectivity.
- He outlines how the team identified these problems and decided a complete rewrite of the networking stack was necessary.

II. Designing a Flexible and Extensible Networking Solution

- Matt describes the team's approach to redesigning the networking capabilities in the new version of Start OS.
- The focus is on providing users with more choice and control over how their services are exposed, whether via Tor, VPN, clearnet, etc.

III. The Start 9 Origin Story

- Matt discusses how the Start 9 team came together, originally with the goal of making it easier to run Bitcoin and Lightning nodes.
- The team realized the broader problem they wanted to solve was making self-hosting and decentralized computing accessible to regular people.

IV. The Start 9 Router - Extending Sovereignty to the Network Layer

- Matt announces the upcoming Start 9 router product, which aims to give users more control and security over their home networking.
- The router will integrate tightly with Start OS and allow for easy remote management of networking configurations.

V. Reflections on the Journey and the Path Ahead

- Matt reflects on the progress of Start 9, the challenges faced, and the rewarding moments of building something he believes is important.
- He expresses excitement about the increased awareness and support for Start 9's mission.


IMPORTANT LINKS

•     https://freesamourai.com
•     https://p2prights.org/donate.html
•     https://start9.com

SHOW SPONSORS

FOUNDATION (https://foundation.xyz/ungovernable)

Foundation builds Bitcoin-centric tools that empower you to reclaim your digital sovereignty.

As a sovereign computing company, Foundation is the antithesis of today’s tech conglomerates. Returning to cypherpunk principles, they build open source technology that “can’t be evil,”

Thank you Foundation Devices for sponsoring the show.

Use code: Ungovernable for $10 off of your purchase


Thanks for listening you Ungovernable Misfits, we appreciate your continued support and hope you enjoy the shows.

You can support this episode using your time, talent or treasure.

TIME:

- create fountain clips for the show
- create a meetup
- help boost the signal on social media

TALENT:

- create ungovernable misfit inspired art, animation or music
- design or implement some software that can make the podcast better
- use whatever talents you have to make a contribution to the show!

TREASURE:

- BOOST IT on the Podcasting 2.0 apps (https://podcastapps.com)
- STREAM SATS
- DONATE via Paynym @ https://paynym.is/+maxbuybit
- DONATE via Monero

83H2AcSXfJv69K7SqPkBAcTVsEXXB84ppXviMmhHydyoXJxEyfavH3qRg3X6GufTRfYW4LF233w8bPbbDfqYeJdLC8vBv1D

- BUY SOME CLOTHING @ https://ungovernablemisfits.com/store/

(00:00:00) INTRO

(00:01:11) BOOSTS

(00:03:55) THANK YOU FOUNDATION DEVICES

(00:05:12) The Start 9 Rewrite

(00:22:44) It's All About Remote Connectivity

(00:27:44) Your IP Address is Showing...

(00:35:28) A Family of Start 9's

(00:41:12) A Server in Every Home

(00:52:24) Being Better At Backups

(00:56:20) What Came First: The Server or the Bitcoin?

(01:08:16) It Starts @ SALT

(01:21:18) Future Plans: Start 9 Router

(01:29:48) Go Spin Up a Node!

[00:00:04] Unknown:

Bitcoin   is close to becoming worthless.  

[00:00:16] Unknown:

Now what's the Bitcoin?  

[00:00:19] Unknown:

Bitcoin's like rat poison. Yeah.   Oh. The greatest scam in history.   Let's get it.  

[00:00:27] Unknown:

Bitcoin will go to fucking 0. 0. Yeah.   Welcome back.   Today's episode is with Matt Hill from Start 9.   I thought it was about time to get him on the show.   I've been running one of these start nines on an old laptop for   probably about 6 months now. I've been really impressed, and I wanted to dive into a little bit more detail   about   how it started   and where they're going.   It was really good to chat to Matt, and we'll be getting something else booked in very soon.   I wanna say thank you to everyone who boosted the last show with Max Tannehill.   It's really great to know you're enjoying the content.  

I love reading the messages, and, of course, I love the tasty   gonna jump into the boost now, but I also wanna say thank you to Chad Farrow, who has been streaming a significant amount of sats to the last few episodes.   I appreciate you, mate.   Now I'm gonna read the top 10 boosts   from last episode.   User   85445985,   a 101,000   sats.   Max t is absolutely right. Samurai left behind an incredible community with a clear view of where the line in the sand is drawn.   Lincoln Park rules.   Never heard of him. I've been selfishly waiting for this one for ages. Max Tannehill is criminally underfollowed on Twitter.   Maybe it's that pesky shadowban at work,   but he always delivers with insightful takes rooted in both rationality   and his wealth of experience in the cryptocurrency   space.  

Hopefully, this episode can impact some people that haven't heard Max before.   Late stage Huddl,   33,000   sats.   Donkey work?   Well, that's a first for me. Was concerned to Google that.   John,   21,000 sats.   Rooke Goldberg machine, drink.   And if you're out of drink, have fun staying poor whiskey on the ungovernable misfit store. Thanks for the shill, mate.   8.   Max, got them sick Singapore p to p hookups. Jealous as fuck.   Bubba,   bored? How can that be? Could we that, like me, you think the mission of Bitcoin has been compromised?   Mass adoption by limp dick faggots crying for   government approval,   hence permission.  

I ain't bored. I came for Bitcoin. I might leave because of Bitcoiners.   Expatriotic.   Oh,   got your coffee cup there?   Mister mister,   dear listener,   run your own node. This is very important.   Much more important than liking my post with more than 10 sats.   Please do not   like this boost   unless you have slash want your own node.   Message ends.   And finally, FOMO Medtronic   with no message.   Right. That was the top ten. Thank you to everyone who's boosted,   everyone who supported the show, sent in messages,   shared with friends and family, and everything else you're all doing. I appreciate it.   Before we start the show, I wanna say a quick thank you to Foundation   Devices.  

They make my favorite hardware.   They have an incredible team.   And if you aren't already using one of these things, you've got a question what the fuck you're doing in life.   They're beautifully designed.   They're fully open source.   They have the best team in the Bitcoin world, including our good friend, Bitcoin q and a.   Any   fuckwit can use these.   I've proven that. I use it literally   every day at the moment.   You can use it with their companion app, which is incredibly slick, makes labeling   very easy.   If you care about keeping your Bitcoin secure, you care about open source, and you don't wanna be using some horrible piece of shit with a disgusting UX,   check out Foundation Devices   at foundation   devices   dotxyzed.  

And if you want some money off,   you can use the code ungovernable.   If anyone has any questions, you can reach out to me. I'll answer in any way I can. But I work with these guys because I love what they do, and I think you will do too.   Enjoy the show.  

[00:05:12] Unknown:

Hey, Matt. Hey. How's it going? Good. Thanks. Welcome to the show. Yeah. Thanks for having me.   So this is not video. Is that correct? Yeah. Not video. Okay. We try and keep our privacy.  

[00:05:25] Unknown:

All good. Great. Yeah. Well, thanks for joining me. I thought it was about time I get you on because   me and q and a and everyone else who's involved in,   this show have been talking about start 9 a lot recently.   We do a monthly show covering all the updates on things, and this is what we're running most of the time for most of the things that we do.   Big step up from what I've had previously just in terms   of usability.   It's been bulletproof   ever since I started running it. So,  

[00:05:59] Unknown:

yeah, congrats, mate. It's fucking cool. Thank you. It's really funny you say that too because   how do I put this? We have such high standards for what we're building that,   because we we want it to work. Right? Like, long term and in a big way. We we won't we were serious about the idea   of many, many people, you know, maybe not everyone,   but most people or most families at least having a server in the future.   In order for that to happen,   it needs to be, as you mentioned, bulletproof, and it needs to be really simple as well, and it needs to be very extensible. It needs to scale to accommodate   all the various different   applications that somebody might want to run now and in the future.  

And so while what we currently have is 0 351 is our latest version of Start OS, and we are officially still in beta very much so,   we made the determination about a year ago   that we got the architecture wrong of the current version that you are using and and praising, which I appreciate.   We determined that it wasn't the proper   foundation   that would enable us to achieve our long term goals, and so we set out to rewrite it for the 4th time.   And   and that rewrite, is now nearly complete. So I I I I smile because we do get a lot of,   positive feedback on the current product.  

And then we look at each other and we go, well, we just trashed it. We literally we rewrote the entire thing. Nothing survived. We rewrote it from first principles   in order to fix   all the problems that we see with it, maybe not on a, you know, daily   basis that somebody would notice, but in our imagining of where it's going. Right? Because I know where we're going. You might not. A lot of people might not, and the current OS will not take us there.   So we rewrote it, and it's almost ready. It's in alpha testing, actually, the new rewrite.   Very exciting.  



[00:07:59] Unknown:

What was it that   you weren't happy with   even though I'm singing the praises? I am   one of the most technically inept people in the Bitcoin world,   which is which is why   q and a helps and holds my hand, and lots of smart people jump on the show to help me through things. But,   in my mind, at least, when I compare this, I think I've run built every single node package out there. And,   usually, it's me, a bottle of whiskey, a weekend,   a lot of swearing,   and calling people and trying to sort things out,   followed by finally getting it sorted. And then   every few months, everything failing and me throwing my toys out the pram. That's been my experience for the last   6 or so years.  

For me, it's fucking amazing because   I built this thing in and set it all up on a laptop in, like, I don't know, couple of hours.   No dramas, no swearing, no stress, and it's been rock solid and stable   running not just my Bitcoin stuff, my lightning stuff, everything else. So for me, it's great, but why is it not great for you or not good enough for you? Yeah. That's so great to hear. You know? I I love the feedback.  

[00:09:17] Unknown:

Well   so, unfortunately, it gets a little technical,   but I can,   I can keep that part thin and talk about the meaning?   Right? Sort of Mhmm. At a at a higher level. So let's let's start there. And if you have deeper questions, we can dive in.   So first and foremost,   our networking stack   was too rigid.   And I bet if you had a complaint, which you have not stated one yet, but I bet if you had one, it would probably   be the lack of networking options in terms of connecting to the server while you're away from home, which currently is Tor only. Right? That is how you connect to your server while away from it   is using   the various   Tor dot onion URLs that every service interface and the operating system itself receive.  

And Tor is slow   and unreliable.   It was a great place for us to start because it   had NAT punching built in, meaning the user does not need to set up any kind of configurations on their router. They can just plug this server in, go to the other side of the world, and access it securely   and privately   using the Tor URLs   without any additional setup or or,   trade offs. It's just great.   The trade off of that   is the slow,   unreliable nature of Tor in general. Mhmm. And so   we've known for a long time that this was not going to, as I mentioned, scale to everyone. Right? Not everyone is gonna open up the Tor browser   and put up with spinners. Not everyone's going to run Tor on their laptop or run Orbot on their iOS or Android device.  

This is like niche stuff.   This is techie, geeky stuff.   Even though it's achievable, it's still not   what we're going for.   And so as we talked through   how to add more networking options, we realized that we had made a mistake   in the, current architecture of Start OS,   which is that we thought it was really neat,   clever,   you know, to   host every service interface   on a unique dot local URL,   just like every service interface gets a unique dot onion URL, right, which is a public key dot onion. We said, well, why don't we just reuse that same public key and spin up a dot local URL for that service? That way if somebody's home, they can use dot locals. And when they're away, they can use dot onions. And this, like, was very simple. It was a very simple paradigm.   And it was very clever because in order to, do these unique dot locals, we had to use a technology called Avahi,   and more specifically, it's Avahi aliasing. What that means is that it's utilizing mDNS,   multicast DNS, on the local area network,   and every service   is actually   the main dot local. Right? Your server has a main dot local. That's a adjective, noun, right, word word dot local.  

Well, all unique dot locals that you see for all your services are actually just aliases   of that primary dot local.   And then under the hood, we are resolving DNS. So we're saying, oh, okay. They came in from this pubkey dot local URL.   That's gonna map to this internal service interface.   And essentially, we're running our own DNS server in StartOS   and using this aliasing   to achieve it. And we were proud of ourselves, and we thought we were really clever. What we didn't realize is that Avahi is a ancient   pile of trash that is no longer maintained.  

And so just buggy. Right? It's an open source project that is barely maintained. We certainly do not have the resources to go in there and take over the project, and so we found out kinda too late that there's bugs associated with this, and we had to hack around them. So it works pretty well on StartOS,   but that's because we had to build up a lot of hacks to make it work,   to accommodate the shortcomings of of Ahi.   And secondly, for Windows users, anyone with a Windows laptop,   this doesn't work out of the box. They have to install something called Bonjour Print Services   onto their machine Uh-huh. In order to make this work, which is annoying and kind of a letdown when these people discover that, you know, they have to do even more configuration on their client device. It sounds shit as well, doesn't it? Yeah. It's like, why are you selling why are you selling print   services? Make to, you know, just to reach my URL. It's it's weird.  

And it is. It's introducing an another yet another, you know, third party piece of software. I mean, it's all open source stuff, but still, it's like you don't wanna be cluttering what should be a very simple experience with a bunch of extra dependencies.   And the biggest problem   is not everything I just mentioned. The biggest problem is actually that by doing it this way,   services on the LAN so, you know,   I wanna clarify one term. I keep saying the term service interfaces,   and I want everyone to know what I mean when I say that. When you install Bitcoin,   you are installing a service,   but Bitcoin has multiple interfaces   through which you can access   Bitcoin. 1 is the RPC interface. This is what the client wallets use. There's the p two p interface. This is what other nodes on the network use.  

And then you have a z m q interface,   which is what the dependence of Bitcoin use, like LND.   And in theory, you could imagine a user interface for Bitcoin   where you could visualize statistics about your node, its peers, and stuff like that. We don't have one, but you could imagine a future where somebody slaps a user interface onto the Bitcoin node just to see what's going on with your node in a very visual way.   And so when I say service interfaces, that's what I'm talking about. I'm talking about all the various ways that a service exposes itself   to,   either the the Internet or a local net to be used by you or another service or a peer.  

And so each of those gets its own dot local. Right? Every single service interface. It's not like Bitcoin has an address. Bitcoin has 3 dot onion URLs on start OS, each Mhmm. Being the different interfaces that I just mentioned.   By using the dot locals on the LAN, these service interfaces did not get their own ports, which is a much more common approach to networking. Right? You would expect an IP address   or a dot local address   with a port after it, a unique port that would map to that service interface, We didn't do that because we had this clever little dot local scheme.   Again, the mistake here was the oversight of, well, what happens when you want to use a VPN   to connect to your home   to access your Bitcoin RPC interface?  

Well, it turns out, and we didn't know this at the time either,   that VPNs do not support Avahi   and dot local aliasing.   So now if you were to go get your you know, turn on your VPN, whichever, you know, one you   use, and plug in a blah blah blah dot local URL, it's just gonna throw up. It's not gonna work. And so we effectively   disabled inadvertently VPN access to start a lot services by doing this Avahi aliasing.   That's long winded enough. Mhmm. You can understand why this is a problem.   When we set out to fix it, we began thinking about all the other ways that we had made mistakes. We were like, okay. Well, you know, clearly we don't we're not perfect and we get things wrong, so   let's really think this through. Like, let's   let's reevaluate   the way we do networking   in general.  

And what we found was that we had created a relatively   rigid   system.   Every service   gets a dot onion URL.   Every service gets a dot local   URL. Sorry. Every service interface   gets one of these URLs.   So we said, okay. We're gonna get rid of the dot locals. We're gonna switch to ports.   So now every service will get a dot onion, and every service interface will get a a unique port.   But what happens if someone wants to use I2P to connect to their service interface?   Or what if they want to use some peer to peer technology like hole punch or something like that to connect?   Or what if something new comes out? Right? They wanna use Nostr   to, you know, send a special,   note   that, you know, indicates   speaking to a a service on your land. What if they want to use a VPN? What if they want to use Clearnet? They wanna   host their blog on a Clearnet domain.  

We realized that we we   not only we weren't recovering   all the currently available options, but we definitely were not building a system that could accommodate all future   unknown   forms of networking as well. And so we wanted to build a much more general   configurable,   flexible   networking   stack   where, for instance, a user could   not have a Tor URL if they don't want 1. Right? Like, why do we force users to have a dot onion URL for all their service interfaces? We shouldn't. That should be user choice.   In fact, when you install a service for the first time, this is the conclusion we ultimately came to,   it shouldn't have any addresses   because we cannot pre assume what you are going to want.  

Take Nextcloud for instance,   or here's a good one, Synapse. You're familiar with Synapse and the matrix protocol and network? It's not something I use, but, yeah, I know what it is. So Synapse, like   like some federated networks,   your server URL   is part of your network ID and username. Right? So you don't just it's not like if we're using signal or Telegram and you say, oh, here's my handle. It's, you know, at matthill or something. Mhmm.   When it comes to Matrix, you have to say, oh, here's my handle. It's   [email protected].  

Right?   It's it's it's more like an email address. It's more like because this the the server the location of the server   matters in these federated networks.   And so you only get one.   You can't change it. Right? It's it's not like   Bitcoin RPC   where the Bitcoin RPC   could be hosted on 5 different   network interfaces. Like,   I could put my Bitcoin RPC   on 2 different dot onion URLs,   a unique port on the LAN, a dot local URL.   I could put it on a clear net domain, bitcoin.mathill.dev,   and the wallets would respect all of them equally. Right? I could plug in my dot onion to a wallet that supports Tor.   I could plug in my,   IP and port   to a wallet with a VPN enabled on the device,   or I could plug in my Clearnet domain to a wallet and not need to do any kind of client side network configurations.  

And Bitcoin doesn't care. It's like if you have a if, you know, your house is located somewhere and you're gonna give somebody   the directions to your house,   You could arrive at that house many different ways. I could tell you to take the highway,   and then go left, left, right, or I could tell you to take the back roads and go right, right, right, right.   You know? And it's like there's many different ways to reach my house,   but my house doesn't care how you got there. My house has no knowledge of these routes. That's how you can think of Bitcoin's RPC.   Bitcoin RPC is like, hey, if I receive a signal, I don't care where it came from. I don't even need to know where it came from. As long as it reached me, I must assume that it is valid traffic.  

Sure, I might check authentication, make sure the password is correct before I give you a response, but I don't care how you got here. And that's these different addresses.   But when you talk about something like Synapse,   Matrix,   it does care.   Right? There is a very specific   route that you must take   to get to the server.   And so if we, as a company, start 9 and as a development team,   decide   sort of a priori before the user even   expresses their their desire or intent, if we just assign them a dot onion URL for their synapse server,   well then we have condemned them to only being able to use their synapse server on Tor,   which isolates them from 99%   of the world's population who is not gonna be using Synapse on Tor. And so what we realized was that   the networking experience we want is for somebody to download and install a service   and then choose   how they want that service to be exposed   to these various   networks. Either the public Internet,   a private Internet,   VPN access, the Tor network,   I2P network. They get to choose. It's user choice.  

And so that's what we designed is we redesigned the entire networking stack to be infinitely extensible,   totally customizable,   putting the user in control of how the service is exposed to the world,   and in a way that is intuitive   and with, you know, same defaults. Like if you don't know what you're doing or we can   sort of present you with what we think   you want to do and then you can just hit okay.   So that's one major rework of Start OS that's coming   is this total overhaul of the entire networking stack for the purpose of extensibility   and flexibility and user control.  



[00:22:44] Unknown:

It makes total sense. I mean, what if you're saying a server in every home,   for me, at least, it's fine, like, doing everything over Tor. Even though it seems stupid because I'm, like, sat here next to my node, and then I'm connecting over Tor. It like, it's feels a bit silly, but it doesn't really matter because nothing that I'm doing is that time sensitive. It's like, okay. I wanna send a payment. You know? I'm connecting to it over Sparrow. I'm sending a payment.   If Tor's being a bit buggy or glitchy, it's not the end of the world. But if people are uploading photos or keeping their chats on there or doing any of the other stuff that you might wanna do with the server,   I can imagine tool would be a major issue.  

A lot of friends of mine are using things like Tailscale   to connect to their nodes. And from what you've just said, I'd imagine   you couldn't do that   as it currently stands.  

[00:23:36] Unknown:

That's correct. Yeah. That's one of the big   and really only,   to be honest,   complaints, the right recurring complaints that we get is this remote connectivity   limitation,   but the new architecture   not only allows for tailscale, but wait till you see what we got coming.   We have networking options   that not only do what Tailscale does, but do it without a trusted third party, which Tailscale is.   And then secondly, we have clear net capabilities.   Start OS itself will have WireGuard capabilities so that you can   not only, connect to your server   privately   using a VPN,   but you could also   host services on your server   on a clear net IP or domain   without exposing your home's IP address.  

That's just There's gonna be these tiers   of   choices with trade offs, and half of our battle is documentation, of course, and making sure that we don't   set defaults that have unexpected trade offs, but really messaging to the user, okay, by default, your service can't be reached, period.   It's useless.   And if you're a super you know,   if if privacy,   censorship resistance,   anonymity are most important to you, well, you should just toggle, you know, the d dot onion on and use that and deal and the trade off there is this user experience.   But the next best thing, you know, if it's for you accessing your server, would be to,   use this VPN strategy, and here's a guide for how to do that. And all of it, again, is built into StartOS,   might require you to do some configurations   on a client device,   might even require you to spin up a, a VPS somewhere that has a static public IP address for the sake of reverse tunneling,   for lack of a better term, such that when somebody visits your public   IP address, they're actually reaching   your home server, but without knowing the IP address of your home Mhmm. Which is a clever thing that you can do with WireGuard. And it's effectively what Tailscale does, except Tailscale is a VLAN   setup. It's not hosted on the public Internet. It's only for your private access.  

But we will have Tailscale on the marketplace.   I'm aware of somebody packaging it, 4036.   We will have this really cool thing called Static Wire available on the marketplace. Static Wire is a way to lease   public   IPv 4 addresses   over Lightning Network   anonymously   for varying periods of time. So for instance, you could you would install static wire,   select how long you want an IPV 4 address for, say a month.   An invoice will be presented. You'll click pay because LND will be a dependency of this service, so you won't need to actually open a wallet or anything.   You'll just fund it right from the LND node running on your server.  

And then this service will provision you an I p v 4 address   such that if anyone visits that IPv4 address in a particular port   or visits your domain that you have now, you know, mapped to that IP address,   then anyone in the world would be able to access   the services that you choose to expose. So if you wanna host your blog from your home or something like that, you can do so   without exposing your home IP   using an IP address   that you rented anonymously   using Bitcoin over the lightning network, and this is   super drop dead easy user friendly to do.  

Now you are introducing this third party, this business that is renting you this IP address for a period of time,   but they don't know much about you at all. They will know your home IP address and that's effectively   what they will know. They will not be able to see,   introspect the traffic. It's all encrypted going through   their server and but it is. It's a third party, and so we will have to message that. We're gonna say, hey, you know, you're gonna rent this IP address from you know, we are not doing this. This is somebody else who spun this up and built this technology,   and they're gonna know, you know, your home IP address. And that's it. Other than that, they're they're just a a node, basically. You know, they're a relay. They function as a relay.  



[00:27:45] Unknown:

Is there a way around that? And the reason I'm wondering is when I think about using Lightning,   I only ever spend   from my own lightning stuff. Like, I won't receive because of privacy concerns.   I don't wanna leak any information, so I'll only   fund my Lightning channel. I run c lightning on my start line. It's I'm pretty happy with it. It's pretty good. And then I'll fund with post mix, and then I'll only spend.   But someone was saying to me recently,   there's a way around this where you can get an invoice paid, and then it does a couple of hops from that point to where you are.   Is there any trickery that could be done   with Start 9 where I could effectively run, like, a separate wallet or separate service where I could receive rather than having to do this weird, like, Rupert   machine situation that I'm currently doing?  



[00:28:39] Unknown:

You know, I don't know   because I'm not familiar with the strategy that you just   relayed to me about, you know, some multi hop thing. I I don't know if that's unique to   CLN   or Lightning or if it's I just don't know. I'm not familiar with that. But I will say that   being private,   like   so there's a lot of chatter about IP addresses in the world. Right? And for good reason, but also   it's important I think to   address reality   for real. Right? To not have some sort of fantasy about   how the Internet currently works and how imprivate   it really is.  

It is   very, very difficult   to not expose your home IP address or   your phone's IP address or whatever you're doing. Right? Using a VPN   is   inadequate   depending on your threat model, who you're trying to   be private from. Okay? So for instance, let's say I'm using my   my   computer to   at home to do internet things, okay, my IP address is being blasted   all over the place. I mean, people don't necessarily realize this. They seem   surprised when they find out there was, like, this kind of uproar around Umbrel a while back where,   you know, Umbrel is pinging their server,   every I don't know what it was, like, 30 seconds to check for updates,   and that this request   to Umbrel's servers   was being made using a standard   Internet request. Right? It's just like, you know, one my my computer or my server pings your server and says, are there updates available? This is, like, very, very standard normal stuff.  

And the the controversy was, well, Umbrel knows your IP address.   And they seemed a little sort of taken aback by this, and they're like, well, well, of course. Like   like, how you know what I mean? Like, everyone knows your IP address. Like, every single website you visit, this it's made using the same Internet protocol,   and your IP address is being blasted all over the world 247365.   Like, your IP address is not a private thing. It's a public thing.   To keep it private requires   extraordinary   efforts.   And as I mentioned, using just using a VPN is inadequate to a degree   because, one, it must always be running. If that thing is off for even a second, you've just leaked your IP all over the place without knowing it. Yeah. Number 2   is instead of blasting your IP address to 20 different   websites that you're visiting,   what you're doing is you're blasting your IP address to 1 party, namely the VPN provider,   and telling them the 20 different sites that you visited.  

So you're actually consolidating   your Internet activity   into the hands of a single  

[00:31:35] Unknown:

party. Yeah. You're putting all your trust in one company.  

[00:31:38] Unknown:

Well, correct. Yeah. So if you're a state actor and you're interested in somebody   and you know that they are a privacy advocate using a VPN service, you're almost like, oh, thank god. Because you just have to go to that one person, that one VPN provider and be like, tell me everything about this person. Everything. And they know everything.   Whereas, if you don't use a VPN,   then they would have to go to every single place that you visit. They'd have to go to Facebook. They'd have to go to Google. They'd have to go they'd have to go to all these different places, which they can do, but it's just a lot more work to create this profile of your Internet activity. So using a VPN   is a good strategy   if you are trying to obfuscate your IP address   from the various places that you go on the Internet.  

But it is a bad strategy   if you are trying to obfuscate your Internet activity   from a state actor.   So it really just depends on your threat model.  

[00:32:32] Unknown:

Theoretically,   they can go to anyone   and find out who you are. I suppose the only way around that and and like you said, this gets into the weeds a bit. It's like   you could   pay for a VPN service   via lightning. You could do things to obfuscate who you are, and you could run a router level VPN, and then that connect to a second VPN. You can do these things, which   I do, and   I've probably made mistakes. So, like, I'm not I'm not doing it thinking like, oh, fucking hell. The government isn't gonna know who I am because I'm really boring. I do these things because I'm just trying my best, and I'm trying especially being involved in Bitcoin is, like, you don't wanna, like, be talking about this stuff because everyone thinks you're a weirdo and, like, you know, it's it's just, like, best practice. But I totally get what you mean, and I think people   overstate what they're capable of or   think that they're kind of like some   secret squirrel spy when actually it would be relatively easy to find out who they are. But anything that makes   it more simple   for more people in my eyes is a good thing, especially around Bitcoin. You know, there's a lot of loud mouths on Bitcoin   Twitter saying things they shouldn't probably be saying. And if people are even close to being correct about where this thing goes in the future,   you know, they could be in some danger.  



[00:33:57] Unknown:

Yes.   And there are strategies as you mentioned that, you know, are effective using multiple VPNs   in different jurisdictions,   meaning each VPN you're using is under the purview of a different state authority, is a very effective way to maintain your privacy. Because what it requires is cooperation   between the 2 jurisdictions.   So that is a very, very effective way to not   have anyone know   what you're doing on the Internet, because neither VPN provider has that picture. They would have to talk to each other to get it. That is definitely a way. Tor is also   a good way, except   Tor is only really   provably effective   if you are visiting hidden service URLs,   dot onion URLs. If you are using Tor to visit clearneturls.com.net,   there's a very good argument to be made that that is not effective, that the exit nodes necessary to complete that transaction are compromised.  

Yeah. It would make sense that they would be. Yeah. As far as we know though, so long as you are using Tor to do Tor and you're not leaving the Tor network that there's no exit node involved that it should be safe unless you're unlucky and got   a series   of relay nodes that are all owned and operated by the same bad actor, which could also be the case. But anyway,   so   should I go back into a little bit more of what the new OS fixes? I mean, we talked about networking, why it's important, how it relates to privacy.  

[00:35:27] Unknown:

Yeah. Definitely.   Before we do, I just have one more sort of nerdy question on this stuff before Yeah. For sure. Jump into that. What are your thoughts on   running   multiple   start nines   for different things?   For example,   would you have any concerns   with someone? Let's just say they play with all the toys that you guys offer. They download every single option, and they've got all the Bitcoin stuff. And they're storing their photos on there, and they're doing their chats on there, and they're doing everything else.   Is there any concern that you'd have by having,   effectively all your eggs in one basket, or is the new architecture built in such a way that it partitions things out and it wouldn't be a worry,   if that makes  

[00:36:14] Unknown:

sense. It makes perfect sense and is something we get asked a lot. Our very strong   and and official stance on this   is that you should partition services amongst different servers   based on   how critical   their failure would be to your life. How devastating   an attack or failure would be. So for instance, the naive approach to this and what a lot of people say, and I've refuted this on multiple occasions,   is run your Bitcoin stuff on one server and your other stuff on another. I think that is a very arbitrary   line to draw. The reason people draw it is because Bitcoin stuff, what they really mean is   financial,   right, private keys. They're talking about stuff you don't want to lose on 1 server   and stuff you wouldn't mind losing on another server, and that is a much more   accurate   way to phrase it, which means it doesn't need to be Bitcoin and non Bitcoin. It's just if you have something that's really, really important let's say you're running   a massive   powerful lightning routing node. Okay?  

There shouldn't be anything else on that server. There should be Bitcoin   and either CLN or LND. That's it. Don't install anything else. Maybe you could have something like RTL   or ThunderHub installed so that you could manage that node, right, and do the things you need to do, but you want the bare minimum number of services possible.   The reason for this is because   every single service   introduces   another entry point to your server.   It's like putting another door in your house, you have to secure that door.   And with services,   updates are always a danger.  

So if you have 20 services   on your server   and any one of them   gets compromised at any time   and you update to it, then you could lose the whole system.   Updates are very   risky.   You should do them carefully. We recommend waiting, oftentimes   weeks or even months after a new service has been released   before updating to let other people update first and let the community   test and and maybe discover any bugs. And the more important   the things on your server are, the more diligent you should be about   what software gets installed onto that server, including updates from Start OS. Right? If Start 9 puts out a new update to Start OS,   you should think hard before   clicking the update button.  

It is open source code.   It is   increasingly   being used by more and more people around the world, which is you know, a lot of people argue that just because something's open source, it's secure. I find this to be a ridiculous   argument.   I know tons of open source projects out there. In fact, probably most of them   that are giant piles of insecure crap.   What matters equally to being open source is is the code being looked at? Are people auditing it? Are people running it, testing it, using it? Is it widely used and audited? If so, then open source is really, really important. If nobody's using your stuff, then open source doesn't even really matter, to be honest. Open source only matters if your stuff is popular, which is part of the reason why we weren't open source to begin with. We just were like, well, that's kind of a LARP.   I mean, sure. You   know? Like, could we have been? Yeah. Was it necessary to   have a noncommercial license and stuff on our code for a couple years there? Probably not. But it also was this sort of benign,   doesn't really matter in the early days thing because nobody was using our stuff. Right? Like, you're trusting us completely. Everyone who's installing our OS is completely trusting us even though the source code is published. Even though the source code is completely open source now. You're still trusting us. Very few people on Earth have the ability   to audit StartOS,   And I guarantee you, none of them have to date. So there is an enormous amount of trust   being put into Start 9, the development team right now when it comes to running StartOS,   and we don't want that. Right? I I hate being in that position. I can't wait to get more, you know, recognition and acclaim. I can't wait till competitors   start forking our project to compete with us because   it hardens the entire   technology   and makes it more secure. Bitcoin is secure because it's popular,   because people fork it.  

The less attack surface you have on a given server the better. So   I have multiple servers, and one of them is running the absolute most mission critical things in my life, and I barely touch it and I am slow to update it. I have another one that has some kind of, you know, we'll call it mid level stuff where I create backups, but if it's not storing private keys, it's not storing indispensable   data.   And then I have like a, you know, a test server where I install everything immediately the second it hits the beta registry. And I test and I play and I blow it up and I re flash it and all the rest. So,   you know,   it's a good strategy to have a few servers,  

[00:41:09] Unknown:

in those tiers, I think. Cool. Alright. That's enough of the nerdy stuff. So getting back to having a server in everyone's home,   like, when you think about that and you work towards that, what's the kind   of vision you have   and how far away do you think we are from that?   Because it still to me seems   pretty niche. Like, if I spoke to normal friends in normal life about, yeah, just, you know, run everything on your own server. Forget about Icloud. Forget about all this stuff. Most of them just go, oh, fuck off, mate. Don't be such an idiot.  

[00:41:43] Unknown:

Yeah. That's that is the the state of the world right now.   Tell you what, 10 years ago,   you'd have no interest.   None. I mean, if if start 9 was 10 years ago, like Urbit was 20 years ago, you know, Urbit was a server OS   with the same essential goal as what we have today, and they or they came around in   2002   or 3 or something like that. Now, one, they didn't do it right, in our opinion. Urbit built an entirely new paradigm where application developers would have to come in and build apps for Urbit,   which was just never gonna happen. That's a, you know, chicken and egg problem.   Whereas Start OS   doesn't require people to build   to write new apps for Start OS. It accommodates the apps that already exist in the world. Now they have to be packaged for Start OS, but that is a very different thing than just rewriting   new apps. But they were also just too early, way too early. There was no need. The reason that nobody runs a server   is because   there   all your needs are being met in this   third party   custodial   model of cloud computing.  

And to date,   the consequences of that model   have not been apparent enough, costly enough   to warrant change. People don't change   unless they are in pain.   This is sort of true in any endeavor.   And so   the pain just isn't there. You know? For the most part, people are still just using these third party   SaaS products and cloud services   smoothly and without   consequence,   at least   visible consequence.   And so expecting people to take seriously   the idea of, you know, making a big change is a big change. It's not like   just   some one hour thing.  

Totally de googling yourself is a major undertaking.   And that's just a broad term for you know, becoming a digital sovereign, becoming   opting out of 3rd party clouds and trusted third parties and intermediaries altogether.   Doing that fully   is it's almost impossible. Right? Like, I still don't have that, and I build this stuff, and I've been doing it for years, and I still have in my life the use of some third party stuff like Start9's own community channel. Our primary community channel is on Telegram, and that's because that's where our customers are. That's where our community is, And, you know,   to a degree, we could just say, no. You can't be you we're not gonna interact with you unless you use our server. And it's like, okay. Well, we're we're, you know, we're making our job harder at that point. So Yeah. It's it it can be intimidating,   and so nobody's even gonna think about it. They're just gonna be like, oh, there's no way I'm doing that. Like, you're crazy. I'm not doing that. So part of our messaging to help people with this   is to not pressure themselves to doing it all.  

You don't have to get a server   and then install everything on it and use it for every part of your life. And then next thing you know, you know, 2 days later, you're done with third parties and you're totally free. That's just it's an it's an   it's not a realistic expectation   and people get frustrated and then they give up. And so what I message is I say, Look, just start. Okay? We're called start 9, not finish 9. You don't have to do it all. Just   get a server,   select one thing,   usually passwords. Right? Like, take back your passwords. That's a really great place to start. Get your passwords off of a 3rd party cloud. Get them onto your own server.   It's easy. It works really well. Right? The Vault Warden self hosted password manager and Bitwarden client applications are just brilliantly designed. They work flawlessly.  

And so it's like a it's a win. You know? It's a quick, easy, and powerful win. And then just let that sit, like, bask in that success of yours for, you know, a month, 2 months, 6 months. I don't care. Just   start. And what you'll find is that in time,   there will be this   urge to keep going when you have a long free weekend or something. Okay. I'm gonna take back my photos.   Okay. I'm gonna take back, you know, my my   blog hosting site. I'm sick of being charged $50 a month by by, you know, AWS or Bluehost. I'm gonna host my blog. So you can just sort of piecemeal take these things on without pressure,   and I think it's a much   more realistic   approach, sort of like getting fit, you know, like going on a diet or something. It's like you don't just go cold turkey. You have to be realistic with yourself and wean   off of the drug   or habit and form new ones over time. And I think based on my perspective here, like our position in the world and market,   that the demand   for self hosting   is growing and growing fast. I no longer believe that we are too early.  

I believe that we are   slightly   too early or, like, slightly before   we'll call it   large scale demand,   but we are definitely past the   total   geek, hobbyist,   niche   phase. In terms of interest, the people actually doing it are still the geeks, hobbyists, early adopters,   but we are seeing interest now from very   just   normal people. They are increasingly becoming aware   of the   problems,   abuses,   corruption, and consequences   of broadly speaking cloud computing.   What it means   to have third parties intermediating   and custodying your entire digital existence.   It is starting to become a mainstream   awareness.  

I think we're just in time, to be honest.   And   I think that   the product is nowhere near ready   for these   larger markets.   When I say nowhere near ready,   I'm being a bit hyperbolic.   We are closer than people think. As you mentioned, it works.   It worked pretty well, and   you don't self define as a super techie person.  

[00:47:48] Unknown:

Far from that. And you were able to do it. Yeah. Great. It was simple enough. But and and you made a good point, like, it's becoming more   understood by more people that there are issues and   just simple things like, you know, what is a cloud? Like, literally, I think people thought it was something in the sky a few years ago, and people are starting to understand that their data is being used and abused.   And like you said, it's niche now,   and there are still things   even, like, for me   that I won't quite touch yet. Like,   I won't quite touch yet my family photos.  

To me, that's more precious than Bitcoin. If I lose my pictures of my kids, I'll never forgive myself, and I'm too much of a retard to take that on right now and be comfortable with it. But these are the things where as soon as it becomes   more comfortable and I'm like,   okay. I've not only got my back up, but I've got a backup of a backup. And then I've got that backed up again, and this is how it works, and I understand it fully. Then I'd be like, well, I don't wanna be trusting,   you know, any other service keeping these. I wanna make sure I have it myself. And so,   yeah, I don't think it's a 1000000 miles away, and it's just gonna be probably the more you know, like, obviously, you'd host my Bitcoin stuff because I'm like, I only trust myself to do that. But step by step,   there'll be, more and more stuff that even someone like me will wanna take off the table and get away from these third party services.  



[00:49:16] Unknown:

Yeah. You know, we,   we my my wife and,   and I   just did that.   You know, I've been I've been building this for   years and,   only 2 weeks ago did she finally move all of   our, you know, family photos. We have a couple of children and, we moved them over to our I set up a new server specifically for this purpose Mhmm. With a, you know, big old big old SSD on it, so that we could do videos and stuff too and not have to worry about space. So I set it up and, and we moved it. We we downloaded all of our stuff from Google Photos, which is what we had been using previously to do albums for friends and family and all that stuff, you know, just pretty standard. And we we downloaded it all, and then we uploaded it all to our next cloud server over the LAN, which so it was pretty fast. You know? This didn't take that long, and we had quite a few. And then once it was up, I immediately made a backup   that, you know, is not stored in the same room as the server. Yeah.  

And then we, you know, we went to she went to work. She started creating new albums. I mean, it's an endeavor. It was like this thing we had been wanting to do for a long time   and, finally did, but now I have it set up such that if I go anywhere in the world and I take out my phone and I take a picture,   it automatically gets encrypted on my   phone and sent over Tor to the server in our home where it gets added to my inbox of my photo album, the default place that they get uploaded to. Mhmm. I can then open up my Nextcloud   UI   from   there using my dot onion, or when I get home, I can open it up using the dot local   and add that photo to an album if I want that is shared with her so that it automatically shows up on her phone and computer no matter where she is in the world. We can drop photos into shared albums for friends and family and provide them links, and so we're we're sort of back. You know? We got it all   transferred up and running, and it is super smooth. It's basically identical   to the   Google Photos experience. Mhmm. Especially since all the uploads are happening in the background, you don't even notice the Tor part.  

So it just works, and it works really well. And we have a backup,   and, you know, I will make   additional backups every so often. But one of the big features that's coming   to StartOS, not in this next version,   because this next version is largely   gigantic   refactor for the sake of reliability,   performance, extensibility,   etcetera, but in the immediate, release after this one as a new feature,   we will be adding automated   remote encrypted backups as a feature. So you will be able to, instead of plugging in a drive and clicking backup,   you'll be able to schedule   backups as many as you want, as many times a day as you want, whatever,   to pretty much any target out there. You'll be able to use things like Dropbox or Google Drive where you could dump your encrypted backups on their servers. You could use just an FTP server sitting somewhere. You'll be able to use other Start OS servers. So if you have multiple servers in different states, they can automatically back each other up to each other 3 times a day. Oh, that's clever.  

Yeah. So that's coming. You know, we've already   built a decent amount of that, so it's it's well in the works. And in order to accommodate that because,   one of the big changes we made in 0 3 6 that's coming soon too is that, backups sort of had a problem.   Backups are a challenging   feature in general because   so much can go wrong during the process, and you need to be able to make sure that   it ultimately works. Backups are   mission critical. They cannot fail.   You want them to be encrypted, and then you're dealing with different file systems. So backing up to an external hard drive that you have plugged in is different than backing up to Google Drive or Dropbox or to another starter OS server. These are all different interfaces, and and these systems have opinions,   depending on what file system they're running. We had a problem in StartOS. We tried to,   StartOS   currently uses something called duplicity.  

Duplicity is a, you know, a Linux   based   backup   application   that is very opinionated about how it does backups,   but does ultimately work. The problem with duplicity is that it does   incremental backups. Meaning,   if you were to create a backup of Nextcloud   and then you were to go in   and delete files that you no longer   need and then create another backup,   the backup would still contain the deleted files.   It only adds new data. So backups just get bigger and bigger and bigger over time.   Any new file you put into Nextcloud   will get backed up, but any file you delete from Nextcloud   will not get removed   from your backup, and so they just grow. And that's unexpected because if you delete something from your Nextcloud and then you create a backup, you don't expect it to be there if you were to ever recover from that backup. That is unexpected   behavior in our opinion,   but it's how duplicity works. And so we tried to switch to this backup strategy using   a technology called rsync.  

Rsync does differential backups, so it would   add anything new and delete anything that was deleted,   effectively   creating the smallest possible size backup and a backup that contains only what you would expect it to contain.   And it's faster.   It's really efficient.   Well, the problem is that Rsync,   depending on the files that are being encrypted so if you combine rsync with encryption,   then oftentimes it will produce file names, and there's no way to avoid this,   that are too long   for certain file systems like ext 4. So if you were to try to back up Nextcloud   to your MacBook,   the backup may contain file names that are so long   that the Mac will just tell you to screw off and the backup will fail. There's no real way around this, no simple way anyway.  

So what we did was we just kept duplicity. We were like, okay. Well, I guess we can't switch to rsync even though it's better because it won't always work. It'll just fail for some people sometimes.   And so what we did for 036   what we, I say we, what my brilliant cofounder and CTO did was he wrote a new file system module, a fuse module for Linux. He   just   cooked it up and wrote it, and,   it uses rsync under the hood to do the efficient differential backups,   but it normalizes   the not only the file names, but the entire   format   by basically overlaying rsync with a new file system he wrote called start FS. So   there's now this thing called start FS.  

I'm speaking in vague terms about it because I, quite frankly,   have not talked to him much about it. This is something he went off and cooked up about 3 weeks ago   and showed up and said, well, here's what I did to solve the backup problem. And so,   anyway, you could ask him, and he could tell you every little detail. But but it's it's cool because now backups on start OS are gonna be faster, leaner,   and more reliable. And that's all in 036 already, and we're it's currently in testing.  

[00:56:18] Unknown:

So Very nice.   When you and your cofounder   started this,   did you start it from   a Bitcoin perspective of we wanna create a node,   or did you start it with we wanna create a server? Because, obviously, I found out about star 9, and I was like, okay. It's a node competitor. That's what I thought as a node competitor. And I was like, I'm already running stuff, which I'm sort of happy enough with. Like, it's, you know, whatever. I'll, you know, maybe try at some point. But then when I started to see that you were doing more and more, these things were becoming more capable and I could run more stuff on this, it's sort of like my ears started to prick up a little bit. Was that always by design, or was it more of, like, an evolution?  

[00:57:03] Unknown:

It was it began as a Bitcoin endeavor, Bitcoin and lightning node endeavor. But I also wouldn't call it an evolution   towards the   reimagining of the Internet through a mass proliferation of personal servers that are accessible to normal people and running, you know, replacing cloud services. We went there in about 3 days. We   we, you know,   it really it it was lightning fast. You know, we as a team, there were 4 cofounders of of Start 9. There are only 2 of us currently,   but 4 originally.   And we all came over from a prior company where I was the CTO, and we had a big fall I had a big fallout with the then CEO.   And   I left.  

And, you know, I had built the engineering team at that company. Wonderful people, great engineers. And so we kind of all I kinda you know, I don't wanna say took them with me because, you know, everyone made their own choice, but Mhmm.   It was a mass exodus,   from that company, and   we didn't know what we were gonna do. We just knew that   we were the best people any of us had ever worked with and that we could do something special if we just, you know, rolled the dice and took the leap. So That's cool. There was a few months where we were doing, like, you know, contract   coding for a few different companies just to keep, food on the table and it was easy stuff and it paid okay.  

We were just kind of passively doing this contract work   while we   hung out at my at the common area of my then apartment building   every day, which had a pool and everything. So we were basically just kinda brainstorming about what we wanted to do, and we had a few ideas. And, ultimately, this is when,   2019   when, you know, it was hashtag   reckless for the lightning network.   And,   so being who we were,   Keegan McClelland, who is now doing protocol development over at Lightning Labs, by the way,   he set up his LND node and he wanted to play. So he was like, alright. Go get it up. You know? Go go set up your lnd node, and we'll we'll test this thing out. We'll put a little Bitcoin on it. We'll play.   And, I sat down to do that.  

And about   10 minutes later,   I called him back, and I was like, this there's no way. I was like, come on, man. Like, of course, I can do this. You know? I I'm a I'm a developer. I can figure it out. I was like, but it's gonna take me all weekend.   You know? I I was looking at the guide for setting up an LND node back in 2019, and it was like I mean, I'm an application   developer. Alright? Meaning, I   live like 4 layers up in the development stack. I build   websites. I build apps. Yes. I can write back end code as well. I'm proficient in a few languages and   I'm pretty good. I can crush out code,   but I'm not formally educated in computer science. I never was at the lower end of the stack in terms of, you know, setting up servers and configuring   Docker images and, like,   you know, I I came in I'm a self taught   high stack programmer. I I use JavaScript and TypeScript and   Python, you know, these higher level languages.  

So anyway, I say that just because   I said, okay. I'm a fairly technical person,   and I can definitely get through this. But if this is what it's gonna take to run a lightning node, nobody's ever gonna run a lightning node. So   what's the problem here? You know, I called and I said, you know more than I do about, you know, low level   DevOps and sysadmin and computer science. Like, what's the problem? Why is this so hard? Why can't this be easier?   And at the time,   Casa,   you know, Jamieson,   over there at Casa had the   the plug and play Bitcoin node. Yeah. I remember.  

And they had just announced that they were gonna   add l and d to it. Were gonna add Lightning. And I was like, okay. Well, here's somebody who's doing it. They're they're they're trying to make it easy for people to run a Bitcoin node in a Lightning node without having to go through this command line experience,   and, you know, all the different ways you can screw that up.   So   we rather than you know, our our first thought was not to go build something ourselves. It was to   call Jamieson and Nick   and, like,   see if they wanted help. Like, hey. This is an interesting problem. We think it's an important problem.  

So they   agreed to me. We went out for dinner in,   Denver. It was me, Keegan, Jamieson, and Nick. Jamieson was the CTO as he still is, and Nick was their then product manager, and he's now the CEO.   And we started asking them questions   about their product,   like, you know,   you're coming out with a lightning node. It's great. You know, what strategies are you using? What technologies are you using?   Do you have plans to generalize this? Because by then we were already talking about this. We were like, well, do you plan to   put SMTP servers on there? What about file servers, messaging servers? Like, you know   because by then, as I as I mentioned, we had already realized that if you could solve the problem   of making it easy   to not only install and run   a Bitcoin and Lightning node, which, by the way, is what most of the other kind of node products   solve,   is their goal is to get you up and running   with a Bitcoin node as quickly, seamlessly, and beautifully as possible. Right? It's like, okay, you go from 0 to node in 5 seconds. That's their goal. But we knew that if self hosting, node operatorship,   and all the other things   that someone could do   was ever going to be a viable,   scalable,   like, realistic   paradigm in the world,   that getting people up and running was not actually the hard problem. That's an easy problem to solve. It's   allowing it to work reliably   even in disaster scenarios. Like,   servers are very difficult   to own and operate.  

They're very easy, in fact, to get set up. But what happens when the Internet goes out? What happens when   the drive fails? What happens when a backup doesn't succeed? You know, what are all the fail safes?   And can a normal person,   when confronted with these inevitable   challenges,   do it themselves? Or do they have to, you know, immediately   go get their super technical friend to help them again? Because if that's the case, then you haven't solved the problem.   You've not solved the self hosting problem. What you've done is you've tricked people   into thinking   that you've solved the problem. You've tricked them into thinking that they can do it, but then at the first sign of ad ad adversity,   they're back to where they were, which is but now it's worse because they've put their stuff on the servers. Now   not only can't they do it, but they have to do it because they've become dependent on it to a degree.  

And so we were like, okay. Solving this for real is a serious, serious undertaking. We're   trying to codify the entire Linux sysadmin experience   into a graphical interface that a normal person can use, and that one that works reliably   despite Mhmm. The inevitable   challenges and and and failures that will take place.   And we brought this up to Jamieson and Nick. We were like, are you solving that problem?   Is that what you guys want to do? Or are you running a Bitcoin node?   And not only did they avoid   answering us directly,   they showed almost no interest   in their Bitcoin and lightning   product, the node product. They they were like it was almost like   they just didn't wanna talk about themselves at dinner. They they didn't wanna talk about their product. They didn't seem interested in our excitement about what their product could become someday.  

And we left a little baffled as to, like, what had just happened. Because we were   like, well, we have a very talented team of, like, 5 people just sitting around right now waiting to dig our teeth into something, and we think you're solving a really cool and important problem. Like, do you want help? Are you hiring? Should we build something? Is it open source? Should we fork it? Should we blah, blah, blah? And they were just, like, dodging, dodging, dodging, basically. Mhmm. And we found out why a few weeks later, which is that they discontinued the product.   You know, they didn't wanna tell us at the time because it wasn't public information. I don't think they wanted to leak it,   but they   they didn't they weren't interested   in their own product and solving the problem that we were interested in solving.  

And they pivoted the whole company. Their CEO,   resigned,   and they pivoted the company to the, you know, multi sig product that they have today. And it's cool. It's cool product. You know? Like, do you. That's what they wanted to do, and they're doing it well and kudos to them. But they abandoned   completely the thing that we were interested in. And so we said, okay. Well, let's pick it up. You know? We didn't base any of our stuff off of their stuff. As soon as we looked at it, we knew that it was,   we'll call it a a a very you know, I don't wanna use a I don't I'm not using this in a derogatory sense. It was a naive,   very naive approach to solving the self hosting problem. In other words, what they had done   was they had created   we'll call it a   shortcut   to getting up and running with a Bitcoin node by creating an extremely   opinionated,   rigid,   hard   coded   product. It's like, Oh, you plug this in and Bitcoin, poof.  

And that means it is massively inflexible. There's   service?   We just went to the whiteboard and said, Okay. Well, how do we build   a server OS   that   one can do everything   that   a current Linux,   OS can do,   but do it in such a way   that a normal person can competently administer it through thick and thin.   And that's the essence of the problem that we set out to solve from day 1.   I shouldn't say from day 1, because again from day 1 it was just me wanting to run my lightning node   easier. But after even a few days of talking, we realized that by solving that problem   in a general way, we would actually be solving one of the fundamental   problems   in all of computing,   which is   decentralized   computing infrastructure, like   a decentralized Internet.  

The only way to achieve that   is by people running servers. That's it. There is no other way. Anyone who presents a, well, we can decentralize the Internet or decentralize   information,   and data and all that using software, they're just they're they're wrong. You can't. You   you have to have the bricks in the homes. Like, otherwise,   everything is running on one machine and whoever has access to that machine has control over it, power over it. They can, at minimum, turn it off, but usually they have so much more than that. And so   we set out to solve the big problem, what we viewed to be the big problem   in   computing,   which is   to   resist   and reverse   the current trend towards centralization,   which   is in our opinion   a dire, if not existential threat to   a free society in the future.  



[01:08:16] Unknown:

Was there anyone in that   crew of 5, I think you said, who left the previous company? Was there anyone there who was like,   guys, what the fuck are you doing with this Bitcoin thing?   What are you thinking? This thing's magic Internet money is not gonna catch on,   or was everyone   already involved for a decent length of time and that you're all just like, yeah. Yeah. I mean, this is obvious. Bitcoin is the future. We're obviously gonna do this.  

[01:08:47] Unknown:

The latter. We all came from   a company that was doing Bitcoin already.   Okay. You're already involved. Yeah. That's how we met. So, you know, the previous company was Salt, Salt Lending. Was it some sort of yeah. Yeah. Yeah. I remember that. Yeah. Yeah. It was a BlockFi   Unchained competitor,   same as, well, I think, Celsius. There was a few. You know, it was the crypto lenders. And this was in,   early 2017,   I joined Salt, and I was their 1st developer hire. I was the 1st person at Salt who knew how to code,   but I did not know much more than that. Like I said, I'm an application developer, but I'm also an entrepreneur and have a long history of management   prior to that. And so Sean, the CEO of SALT at the time,   brought me on to,   really build out the engineering department,   both from a product   architecture and design all the way to hiring   more product and engineers   to, to build it. So I was acting CTO.  

That was not my title at the time, but it was what he hired me to do.   And my first hire I was so lucky   in retrospect. This is just a life changing   moment. My first hire was Keegan McClelland.   He was in Denver, had just left Amazon where he was coding.   And again, he is a lifetime hacker, totally formally educated in computer science, like, you know, brilliant, brilliant   young man.   And, I hired him,   and he really complemented all the things I didn't know. And then together,   he helped me vet   more and more engineers.   Together we designed   SALT's entire   engineering stack, software stack and product.  

And to this day, I will tell you that what we built at SALT is very solid. Salt has all sorts of other problems, and it's ultimately why we left. There was all sorts of   crazy,   not wholesome stuff taking place at that company, and I fought tooth and nail to   legitimize   their endeavors. And, ultimately, they were most interested in,   protecting the,   the non security nature of their shit coin   and not going to jail   and basically doubling down on all their attempts to,   you know, just enrich themselves. And I I won't go into details, but, I was there to build a lending product. I wanted to provide fiat loans backed by Bitcoin. That was why I was there, and,   that is not why the cofounders of that company and the CEO were there. So   but but, Yeah. Well, this was too early 2017   too. This was you know, even at the time, I was, you know, I was I was buying some of these coins to try to try to catch pumps and stuff like that. Like, this was before   I had not been around for long enough to know the dangers of this. Like, it's a very common common Bitcoin or journey of kinda getting into Bitcoin, realizing, oh my god. There's so many other things. Bitcoin's just the beginning, and then you   hopefully quickly realize that these are all just affinity scams Mhmm. And   come back.  

But, this was my 2017. It was like, oh, boy. You know, cryptomania.   And, Salt was prominent. I mean, boy, I I felt very lucky to be at SALT at the time. It was a prominent brand.   They had done a huge raise   selling SALT tokens, which were supposed to get you discounted lending rates and all sorts of other perks when you use their platform and get loans and, you know, it was a utility  

[01:12:18] Unknown:

token. I remember it well because the person who got me involved with Bitcoin is a   massive shit coiner,   and he was really into salt and all these types of things. And I remember him rapping on nonstop to me about this in that sort of time. So, yeah, I remember it well. Yeah. Well, it didn't work out so well.  

[01:12:39] Unknown:

But   but at the time, it was it was a lot of energy.   You know, I had the resources to hire anyone I needed. I had the resources to, you know,   really anything to build this product. And   so I had a ton of freedom, a ton of creative freedom and power,   and I built this wonderful engineering team. Really just found 15 people by the time I left. And   of those 15 were Keegan McClelland,   his younger brother, Aidan McClelland, which was funny because when Keegan suggested that we hire his younger brother,   I believe Keegan was   22   at the time. He may have been 21.  

He had already been out of college for 3 years and working for Amazon.   And he goes, we should hire my younger brother. He's   19   or maybe yeah. I think he was just about to turn 20.   And he had already been out of college for 2 years, and he had been working at Symantec Security out in Palo Alto.   And I was like, your younger brother who's graduated college   at 17 or 18 is working in California. And Keegan goes, he goes, yeah. You think I'm a good developer? Wait till you see my brother. He goes, he's better than me. He's a better coder. And I was like, dude, you're like a prodigy. What do you mean better than you?   So, anyway, we brought Aidan out from California, and,   here comes this 19 year old. And from day 1, he just I mean,   boy, what a monster Aidan is.  

So we brought him on, and then, the last of the what would eventually become the founders of Start 9 was, Aaron Aaron Greenspan. He   He was actually contracting for Start 9 through another dev agency.   And from the minute I started working with this guy, I was like, Oh, we have got to   like, get you here for real. So I went to his dev agency and I was like, you know, what do what do we have to pay you to, like, buy out Aaron's contract? You know?   So we did. We we bought him, and he wanted this, of course. So he came in and, got him at Salt, and he was brilliant too. And slowly over the next 2 years, you know, we we learned a lot. We learned a lot about Bitcoin. We learned a lot about shitcoins. We learned a lot about finance,   you know, building lending products and all that that takes. And, I mean, at the time, Salt was trying to   to   offer loans backed by all sorts of crypto. Right? First, it was Litecoin, then it was Ethereum, then it was Doge, then it was Ripple. I mean, they were trying to basically accept any form of crypto as collateral to provide a fiat loan.  

And we, as a team, and this is a kudos to Salt even at   the time, we refused   to do a,   virtual custodial model. Like, we were going to put every last   coin that somebody gave us, whether it was Bitcoin or Dogecoin, didn't matter.   We were gonna put it into a multisig address   where they could see it. Like, we were not gonna rehypothecate.   We're not gonna do fractional reserve. Like, your coins that you put into custody with Salt are sitting in a window that you can look at, and it's multisig, and we can't even move it unless we get, like, lawyers involved. Like like, we have a key, the lawyers have a key, and you have a key. And so it was this really kind of, you know, almost high integrity approach to doing crypto backed lending, except that we were backing the loans with everything under the sun, which is extremely risky because these things are massively volatile, right, and have no liquidity. And so   we learned a lot about these chains. I mean, when when we were told   by the CEO,   again, I still was not even CTO until 2019, but when we were told to integrate   XRP   Ripple   into the platform such that people could   put up   XRP and get a fiat loan.  

Well, we had to go figure out how to run a Ripple node   and how to do Ripple transactions and how to store Ripple securely using multisig or some variation thereof, like using Shamir secret sharing or something.   Mhmm. And it was a research project. It was like, okay. Well, how do we accept XRP as a collateral asset securely?   And in the process,   we learned   firsthand   that the entire network was a scam. We were like Yeah. We were watching the transactions.   Right? We had a transaction. We had a XRP node up. We had a transaction monitor. We're categorizing them. And what we realized is that everything was fake. All the transactions on the XRP network were faked. They were all, like,   pumping.  

It was very clear that somebody or a group of people   were faking   the activity on the network to make it look like it was being used, but it wasn't actually being used.  

[01:17:19] Unknown:

It's just Brad Garlinghouse and his missus just sat on separate laptops Yeah. Pinging each other. Yeah. They just had fucking bots set up that were faking  

[01:17:27] Unknown:

network volume.   And we were like, oh, this is bad.   And it was, like, impossible to use. Right? It took up, like, terabytes of space, and it was clunky, and it crashed. We were like, this is   horrible software on a fraudulent network.   We, like, went back to to the executive team and we're like, guys, you don't wanna go anywhere near this thing.   And they were like, no. We do because we have people that have we have customers that have   1,000,000 and 1,000,000 and 1,000,000 of dollars of XRP,   and they want to use it to take out loans. So, like, we're gonna do it. Figure out how to do it. And that was the beginning   of the, like,   the end, so to speak, where there were now these ideological   and very practical too, just limitations that we were unwilling to compromise on as a team   and slowly,   just became   it became increasingly apparent that we were going to have to leave. But during that time, Salt founders were infighting quite a bit because   the original founders of Salt were a real estate guy,   a restaurant manager,   a Okay. Creative designer,   and a lawyer, a bankruptcy lawyer. Those were the founders of Salt. That was their prior careers and skill sets. And they had stumbled onto the shitcoin scene and basically hit it at exactly the right time, and next thing you knew, these 4 people who didn't know each other before this, they had all met at, like, you know, a couple meetups and bars,   suddenly had, after the Bitcoin pump of 2017,   had, like, $70,000,000   in the bank,   amongst 4 founders with equal power who didn't know each other.  

And so all hell broke loose, man. They were they it was like one assassination after another. Like, the CEO got ousted, then another CEO got ousted. The board turned over 6 times. The I mean, it was like it was like a war zone for the Bitcoin. Everyone was just trying to to secure the control over the over the money. And, now I don't wanna be overly aggressive. There's a couple of these people that were were   well intended,   and just found themselves in a very,   difficult situation.   And there were a couple who were not so well intended.   But, anyway,   through the the chaos   and the, you know, the executions, I'll call them, like, people getting wiped out left and right, I just ended up running the company. I ended up as the CTO.  

There were 80 people at the company,   and I was running the operation on a day to day basis. I basically was was the owner operator of Salt or I shouldn't say owner, but the operator of Salt.   There was a Mhmm. Token CEO who lived in Mauritius on the other side of the world who really had no idea about anything. He was the bankruptcy lawyer. He the bankruptcy lawyer ended up in the CEO seat. No should be no shock there,   But he had no idea what the company was doing or how it did it, so I was running it out of Denver. But like I said, he was insistent   on   on this, you know, endless legitimization of the SALT token and accepting shitcoins and all sorts of rehypothecation   that he wanted to get into. Oh, did you know we could take all the borrower's collateral and go shove it on exchanges and gamble with it type of stuff? And I'm like, oh my god. You are you are not gonna win. So   so, yeah, I left, and,   Keegan and Aiden and Aaron,   and then shortly thereafter,   Lucy and Drew and Bluejay, we   all left and became start 9.  

Ultimately, 7 engineers from SALT were at one time or another at start 9. There's currently   only 3 of those 7 left,   including Aaron and Keegan who both moved on,   in a very,   we'll call it mutually agreed upon way.   And Keegan is   tearing it up over at Lightning Labs right now. He's doing great things over there with Ola Lou on LND.  

[01:21:19] Unknown:

Well, it sounds like quite a ride.   Yeah. That was quite more information than you asked for, but there you go. No. No. No. It's good. I'm glad although it sounds extremely stressful, I'm glad it happened that way and you met the people that you did because   as I said to you before, like, I love running start 9. I'm really happy with it. It's exciting for me to see all this stuff happening and groups of people who care about Bitcoin making it easier for the non technicals like me. And   more and more people who listen to the show have started reaching out and joining and jumping on because we're talking about it a lot. And it's been cool to get to know you a little bit, and I'm very excited to see what you guys keep doing.  

[01:22:00] Unknown:

Yeah. I I appreciate the interest,   and the support and the compliments, really. It's,   it's so rewarding to, you   know, be doing something that you think is important,   and difficult.   You know what I mean? I know it's difficult.   I think it's important. I know it's difficult.   And to bust your ass on it for coming up on 5 years now   and,   really kind of finally. I mean, we've always had people that   recognized and supported us. But, really, in the last few months,   we've broken through some kind of,   inflection point of awareness,   And, it's been really, really rewarding to have not just people, but good people, like, people that I respect,   you know, the right kind of   partners   and fighters, I guess,   approach me and just be like, hey. You know, like, good job.  

Like, we're you know, appreciate what you're doing. Keep going at it. It's it's super rewarding. You know? So thanks.  

[01:23:00] Unknown:

My pleasure. Yeah. Keep doing what you're doing. We'll keep talking about you, and, I look forward to all these updates and bits and pieces. But, yeah, it's been cool to get to know you, and, appreciate you coming on the show, mate. Yeah. For sure. You know, it might be worthwhile to  

[01:23:15] Unknown:

get another one on the books because   we didn't get into it at all, but part of what we were talking about earlier   around the networking   and just privacy and security and remote access and all that   is such a   complex and difficult problem   that we realized we couldn't solve it all in StartOS.   That ultimately there are greater forces at work here,   you know, like your ISPs for instance. They can see quite a bit. But the next logical step for us   was to take   not take, but give users greater control over their networking,   including   stuff that could not be done directly through StartOS,   like the rest of their home and stuff like that. So   at Bitcoin 2024,   last week,   announced very quietly to only people who came by the booth, and we demonstrated   our new router product.  

So Start 9   is coming out with a router.   We expect it to be ready in the first half of next year. I know that's a broad time frame, but I prefer to be broad when it comes to software.   Yeah. Yeah. First half of next year, we will be releasing the router. We are currently accepting   donations and contributions   that will also get you a router, you know, from the first batch. So we're crowdfunding the router efforts.   Every last piece of software that we are writing for this router is MIT open source, so we're we're writing it and giving it away. That's why we're running a crowd fund for it too. Nice.   And, it is going to be   100%   open source all the way down through firmware, so it will run our custom   flavor of OpenWRT,   which we're calling start wrt.  

It runs coreboot firmware,   meaning the firmware on the device has no closed source drivers for any of the hardware,   and we are disabling Intel's management engine in firmware.   It is a   device,   but it's also an open source router OS that people will be able to flash   to their existing,   in many cases, to their existing routers.   But also if they just wanna grab a mini PC that, you know, accommodates a couple antenna,   then they could   just go get that and flash our router OS start WRT on that as well.   And this sort of high level claim and, you know, we shouldn't dig into this now because I think our time is up, but the high level claim of the router   is that it is   just as powerful as the most powerful router you've ever encountered. It has all the features that any power user could ever want in a router, but it is completely   accessible and usable,   not in full, right, because if it's an incredibly, you know, esoteric, rarely used power feature, we're not gonna put it front and center in our GUI.  

But our custom GUI for OpenWRT,   we think, strikes the right balance,   the best balance because we evaluated a lot of routers as we set about to do this. And it's why we didn't because as we evaluated routers, we realized that they all suck. They're all very complicated and difficult to understand and use.   And   so we set out to build a router that was   struck the proper balance between power,   like, features,   and,   accessibility, right, in intuitiveness,   usability by a user,   while also allowing those who do need that extra little bit of power that 99.99   percent of people won't need, can still   do via a secondary   advanced GUI   or via the command line if they feel competent there. So it's sort of a device all the way for a normal person who just wants a very secure,   open source, easy to use router   all the way to somebody who is a, you know, advanced networking engineer that needs the full power of OpenWRT.  

Our router will accommodate both of those users   equally well. And the reason we did this   is not just because we felt like building a router. It's because   Star OS   cannot,   on its own,   solve networking problems that exist outside of it. Right? Inherently, that's just an inherent truth. It is the router the server   is within a   larger network always. No matter where the server is placed, it is always on a LAN. There's always a WAN to the server's LAN.   And so what we realized is that to solve the sovereign computing problem   doesn't just require us to design a new kind of server,   it also requires us to design a new kind of router because the router is the   creator of,   guardian of and traffic controller of the network.  

And so   we had to enter that space eventually. And now was the time because we're blowing the door open on all these advanced networking options. So the start 9 router   will   plug and play integrate with the start 9 server. Meaning, if you have,   if you have an open WRT router at all, any open WRT router, it doesn't need to be our flavor of it, then your start 9 server   will be able to connect to it. You'll simply go into your   router settings   and paste in the public key of your server.   And from then on, your server   can   effectively take over the router. It can manage the router   remotely, which means you'll be able to go into your server and say very simple things like, I want my Bitcoin RPC to be on bitcoinrpc.matthill.dev,   and click save.  

And the server will then prompt you and say, Start OS would like to make the following changes to your router. It's gonna forward this port. It's gonna create this firewall rule. It's gonna list the things that it wants to do in order to comply with your command,   and you'll click okay. And then it will change those settings in your router, and you'll be hosting on ClearNet   in the, you know, safest way possible,   without needing to go into your router settings and do all these advanced configurations. So it for us, was a very logical next step and product. And from a business perspective, it's also,   you know, a way for us to to keep bringing new products to market and and hopefully funding ourselves so that we're not, you know, struggling for for resources and   donations.  

You know, we are a business.   We are a good business, but you know, doing good things, but we we do need more products and more customers. So the router is a great way for us to do that. It solves it's it's win win win across the board.  

[01:29:49] Unknown:

Very exciting.   Let's get something booked in.   Until then, it's been good to chat to you. I appreciate you and what the team are doing. So thanks again. You as well. Thanks.   I really hope you enjoyed that. It was a pleasure speaking to Matt, and it's been a pleasure running Start 9.   If you haven't already checked them out, it's worth having a look at. If anyone has any questions, you can reach out and ask me, and they have a very useful Telegram chat group.   Thanks again to everyone who's been supporting the show. I really do appreciate it.   Every time you share this with friends or family,   make clips,   send in sats, send in messages, and all the other things that people are doing. It really does make a difference.  

If you haven't already checked out what we're doing with our clothing,   articles, and everything else, go to   Ungovernable   Misfits   dot com.   Catch you on the next one.