POD256 | Bitcoin Mining, Freedom Tech, and Awesome Tangents

Share on social media:

In this episode, we go deep on two fronts: protecting open-source projects from trademark hijacking and advancing real-world hash rate heating. We share the ongoing battle to oppose fraudulent USPTO filings on the BitAxe mark, why “TM” vs. registered matters, and how we’re navigating opposition, Madrid Protocol options, and the broader goal of keeping open hardware open without enabling scammers. We then switch to practical engineering: Tyler walks us through immersion mining powering radiant floor heat, dynamic performance scaling, control loops with Home Assistant, thermostats and dry coolers, and why tight software control beats expensive hardware band-aids. We unpack LibreBoard and Mujina plans, APIs, Stratum v1/v2 quirks, Intel vs. Bitmain chip behaviors, and how PyASIC/ASIC-RS standardize miner control. We also touch on FreeCAD pains, open-source CAD needs, educational content plans, and a wild idea: launching a BitAxe to low Earth orbit for space-mining experiments. The throughline: building a sustainable, open-source mining ecosystem where entrepreneurs can profit while dismantling proprietary roadblocks, especially for heat reuse at home and in buildings.

Resources we discussed or referenced include: USPTO trademark process and oppositions, Madrid Protocol for international marks, Home Assistant integrations with open thermostats/APIs, LibreBoard and Mujina firmware architecture, BrainsOS and DPS/ATM concepts, PyASIC and ASIC-RS (standardizing miner APIs), FreeCAD/KiCad vs. proprietary CAD, and Dyson Labs’ BitAxe-in-space concept. We wrap with shout-outs to community hashers supporting 256 Foundation and an invitation to contribute, test, and build on these open platforms.

[00:00:06] Unknown:

Welcome to pod two fifty six, episode 91,   October   22,   2PM Nashville time. We're live   with Scott and Tyler. How are you guys doing?  

[00:00:20] Unknown:

Fantastic.  

[00:00:21] Unknown:

Doing awesome.  

[00:00:23] Unknown:

Scott, you're just saying,   the Chinese are being dicks or what?  

[00:00:29] Unknown:

Yeah. Yeah. Yeah.   There are   two   different,   Chinese   people,   entities, I don't really know, that are trying to   register   the BidEx trademark in The United States. That's right.   Which is   messed up. It is also illegal,   like, fully illegal. So   whoever did that, you know, they're they're trying to register the trademark. So for those who don't know, you can   you can have a trademark. You just put TM on your logo and just   trademark it. And the bid acts has been trademarked for over two years.   But what it hasn't been is registered. So that's where you go to The US,   patent and trademark office, and you pay them, and they list the trademark   on their website.  

So not really a fan of bureaucracy   and things like that.   So, I didn't I didn't register it. And so they're trying to do it.   And it kinda it initially comes off as just like, well, no one else had did it so had done it. So, like, you know, they can do it. But that's not the case.   When you   register or or apply to register a trademark   with the US Patent and Trademark Office, you have to sign,   not just the lawyer who's doing it for you, but the actual person   attempting to register trademark has to sign something that says, you know, I declare under penalty of perjury that this is my trademark.   And they actually have, like, you know, potential fines in prison,   laid out,   if you   are lying.  

So these people lied.   And,   yeah, it's going through the it's going through the you know, their applications are going through the   USPTO PTO   process. And,   now I've got a lawyer,   and we're attempting to,   oppose this formally, which also,   you know, requires paying fees and things like that. But   it's just totally lame.  

[00:02:37] Unknown:

So   do I got questions. Did they get it approved? Like   or is it still pending?  

[00:02:43] Unknown:

It's still pending.   You know, the the bureaucracy moves very slowly. It takes, like Yeah.   Eight, ten months for them to   approve it. And as far as I understand, you know, if there's no   formal opposition, you know, the the USPTO, it's not like a patent,   where they'll actually, like, investigate.   The trademark is just like, alright. You know? We don't know anything about it. We don't do attempt to do any research. Like, they'll just approve it if it doesn't get opposed.   So   Wow. I have,   you know, with   a legal team, have formally opposed it now.   So that's   that's good.  

We have also applied for a   trademark. You know, I have applied for a trademark through the lawyers.   One of the things that's kind of a bummer is there's no sort of, like, public such thing as, like, public domain trademarks. Right? You can't just be like,   this is a registered trademark   of no one.  

[00:03:41] Unknown:

Right. Because I would like to be the copyrights.  

[00:03:44] Unknown:

That's what I like because I don't wanna be the center of this.   Right. I don't want people have to trust that I will not try and,   do, like, trademark violation takedowns and things like that. But you don't want scammers to be the center of it? Right. Yeah. Because want is someone else to be able to go to Amazon and take down sellers,  

[00:04:05] Unknown:

for, you know, violating their trademark. So that's why they have to do this now. Or they could even the risk that I see is they could go after you and say,   you,   the instigator of this entire project, doesn't have   the legal claim to have that logo on your website or anywhere.  

[00:04:23] Unknown:

Yeah. Yeah. Or, like,   I mean, that you know, they there's, like, legal proceedings like that, right, where they could come after me for trademark violation. That would be pretty bold. But Mhmm. One of the sort of just   I don't know. Just little bitch moves you can do is, like,   go to GitHub and be, like, take down this GitHub repository because it's violating our trademark. Right?   And those are the kind of things where GitHub just is like, oh, yeah. Comply. And then   it's up to you to sort of deal with the fallout after, that's happened to  

[00:04:52] Unknown:

various open source projects before. So that would suck. We don't want that to happen. And, like, dude, anyone who's ever received an attorney letter, like, it can cost tens of thousands of dollars to even form a reply   depending on the situation. You know? It's   and you can there doesn't have to be a actual case filed or anything. Like, anybody can hire an attorney at any time and send you a letter for any reason.   And then, like, what do you do?  

[00:05:20] Unknown:

Yeah. Yeah. And I guess, like, registering the trademark doesn't prevent that either. Right? I mean, people who have lawyers,   can just   I don't think there's any penalty for sending a totally fraudulent, like, cease and desist. You know? You can just   do it and cause havoc, like frivolous lawsuit.  

[00:05:36] Unknown:

Yeah. Look at Craig Wright.  

[00:05:37] Unknown:

Yes. Yeah. Exactly.   Exactly.  

[00:05:41] Unknown:

Does a individual have to own it, or can an organization or an entity own the trademark?  

[00:05:45] Unknown:

The registration entity can own the trademark, for sure. So I don't have to have my name attached to it necessarily.   Yeah.   So dealing with that, I don't I don't like I don't really like this stuff. I like engineering.   Yeah. But  

[00:06:01] Unknown:

I can't wait to see how all the FUDsters on Twitter are gonna twist this one around.   And,  

[00:06:08] Unknown:

It's it's already started. Right? You know? Really? I think, technically, you're supposed to, like, have the little TM mark, you know, next to like, if my shirt you know? You imagine, like, a little t m mark. Right. Right. Right. It's like, I just wanna do that. R with a circle around it. Yeah. That's right. That's when you're registered trademark. Yeah.   So dumb.   I hate seeing companies that make, like, these rad stickers for their product. Right? And then there's this little, like, r on there. You're like, that just looks stupid.  

[00:06:34] Unknown:

Right. Yeah. It takes away from   the icon.   Mhmm.  

[00:06:39] Unknown:

It really does. So   we'll get it sorted out. It's just an annoying thing. We'll get it sorted out. And then, you know, there's the rest of the world,   too. So,   bid ex trademarks have been   have been approved in China. I mean,   I don't think that really matters. But   What does that mean?   You know, China has their own trademark system. I mean, as a country, they don't I see. They don't seem to give a shit about intellectual property or anything like that, so they can  

[00:07:09] Unknown:

figure that out themselves. So in China, someone already owns the trademark on your  

[00:07:14] Unknown:

logo. Yeah.   Yeah. Really cool, That's what it ended up. In another in in other countries, I think it's been approved in Germany,   to someone who is actually a friendly organization. Someone had the wherewithal to do that there, and I'm in contact with them. And so that that's fine. I think The UK as well, someone got one. I don't   think I know who that is.  

[00:07:36] Unknown:

Wow.   But So there's a right way to go about it where you have trustworthy folks kind of representing it in different jurisdictions?  

[00:07:45] Unknown:

Yeah. I think so. I think that's that's right.   There there's a process. I think it's called, like, the Madrid   Treaty Accord, whatever,   by which if you have a trademark registered in The US, there's, like, some sort of streamlined   application to do it in, like, the EU and other places.  

[00:08:03] Unknown:

It's just more money to the lawyers, but we'll get it figured out. Well, now that you're appealing and fighting it, will they do their research and take a look at both applications and it's pretty clear who actually   is the behind this project?  

[00:08:17] Unknown:

Yeah. Yeah. I mean, I have a a very good body of evidence to the fact that I've owned the trademark, not the registered trademark, but the actual trademark for years at this point.   So I think we have a very good case to oppose it.   The lawyers were,   I think, they're right. You know, they said that the first avenue is to contact   the people who are making these registrations and get them to withdraw it,   either their lawyers. They so that the Chinese people attempting to do it have US based lawyers that filed these Oh my gosh. Negotiations.   So contact them. You know, let them know maybe that what they're doing is illegal.  

Get them to withdraw it.   We'll see how that goes.   You know, if it does get approved, there is there's a lot of,   articles out there about the USPTO just   straight up canceling,   registrations that were fraudulent. They were like, just in August,   there was they caught some   Chinese group that was, like, mass registering trademarks   and canceled, like, 25,000   trademark registrations just, like, in batch. Like, no. These are all, like, invalid. So Well, you have to show use too. Right?   Yeah. You have to show use in commerce.   You have to show it, like,   for that specific product and that Yeah. You've been using it, in commerce for a while, all of which   the actual project has been doing.  

Boo. Boo.   Boo.   I was   yeah. I feel bad. I mean, people told me about this early on. I was like, ah, no. Fuck the system. Like, I don't have time for that. I really should have listened to them.  

[00:10:00] Unknown:

That would be Yeah. This is the unfortunate, like,   analogy to hardware and software development of the the financial advisers that have to say this is not financial advice. They have to say all that bullshit whenever they talk. And it's like, well, we have to deal with this, unfortunately.  

[00:10:15] Unknown:

Yeah.  

[00:10:16] Unknown:

That's right. That's how you know your project went viral.  

[00:10:20] Unknown:

Yeah.  

[00:10:21] Unknown:

Kudos to you, Scott. Yeah. Yeah. I get it. The other option is just to have your project for being totally obscure and no one cares anyways.   It would be a lot simpler that way. That's how it started out. It was like, I don't know. Like, what's the point? Right.   You know, paying to get the trademark registered.  

[00:10:41] Unknown:

Crazy.  

[00:10:43] Unknown:

Do things that are MIT license   go through this process, or is it a nonissue because   it's available for everyone?  

[00:10:53] Unknown:

So the   these these licenses,   whether they're MIT or otherwise, are copyright,   sort of So unrelated.   Yeah. They're copyright instruments.   And copyleft is sort of using that copyright mechanism   to,   make things open source.   Mhmm. And the license applies in the copyright. And   I'm not exactly sure I know the distinction, but   I found out that trademark does not fall under  

[00:11:22] Unknown:

that. Yeah. Gotcha. If I recall correctly, there's something   even in the MIT license,   to the effect of,   this license doesn't give you   the right to use the logos or name   of the   licensor.   This is just   for the software itself. Gotcha.  

[00:11:49] Unknown:

Yeah. I was so the,   Debian, the Linux distribution, it's been around for a long time. And they they have an interesting take in that their name and logo   is trademarked by whatever group it is, and they actually don't allow anyone to use it. Like, that little,   devil's   logo   that they have is, like, you you can't just use that. So they maintain very strict control,   I guess, over their logo and name.   It's only for their use. Now it's open source software.   Anyone can fork it, but they have to call it something else. Mhmm. Presumably, they have the lawyers to back that up. I think that's what really matters.  

[00:12:28] Unknown:

Yeah. Unlike the bid acts where   anyone   is free to just trample all over the license and there's, like, not really any recourse.  

[00:12:38] Unknown:

Yeah. And again, I mean, there are rules here governing this. Like, you can't just register someone else's trademark, but that didn't seem to stop them. So Right.  

[00:12:49] Unknown:

Well, I hope it doesn't rack up too much in expenses. And Yeah.   I hope you, at the very least, get the protection so that someone else can't come along and tell you to stop using your own logo.  

[00:13:04] Unknown:

Yeah. Yeah. That's the goal here. It was a very interesting conversation with the lawyers, you know, because they're they're a trademark law firm.   But, you know, I'm like,   I wanna trademark this so that everyone can use it,   is is a little bit of a a strange paradigm, especially in hardware. You know, there's there's a little bit of precedent in software   or there's a lot of precedent in software or open source software. But in hardware, this is like,   you know, uncharted territory, I think. Yeah. We were talking about this the other day. Like, what   other, you know,   copyleft open source hardware projects are there?  

I think Arduino is one, but there there isn't a lot. Right.   I still remain fully committed to it. I think it's it's absolutely worth doing. You know, this is this is   protecting people's   freedom, right, to use, modify, understand, and distribute   these tools,   which   are, you know, very important. This is Bitcoin. This matters.   So it's definitely worth  

[00:14:08] Unknown:

doing. Does someone hold the trademark on the Bitcoin logo?   That's a good question.   Oh, that's a really good question.  

[00:14:17] Unknown:

I don't know.   You could go and do, like, a uspto.gov  

[00:14:23] Unknown:

search. Someone probably snagged it.  

[00:14:26] Unknown:

Yeah. I'm sure somebody did. Probably Craig Wright. Yeah. Alright.   Tyler, you're in an interesting place.  

[00:14:34] Unknown:

I'm in the basement.  

[00:14:36] Unknown:

What's going on?  

[00:14:38] Unknown:

Well,   behind me is our radiant floor plumbing setup, and littered around me are a bunch of different   mining systems.   We were swapping out   behind me here is an m 64 that's from our friend Neil.   Mhmm. And that was the water cooled miner   powering the radiant heat. And now I'm testing out this   fog hashing c two, which holds two   s 19 or s 21 series miners  

[00:15:06] Unknown:

Nice.  

[00:15:07] Unknown:

Which is cool. It's I got a little messy with immersion oil. I don't love that, but I do love the fact that these miners are, like, 240   a piece.   So it's a really cheap way to get someone   200 Terahash.   The system's actually going to a guy   in the Rocky Mountains who's got all electric heat, and so that's a perfect example of someone who Right. Doesn't need a $10,000 miner. He literally is paying for zero hash rate heat right now. His bill's gonna stay the same. He's just gonna get sats. Yeah. So it's fun. We're testing it all out here.   And, really, the big challenge I get asked us a lot is, you know, not getting these things running or getting the miners. It's the control.  

And so marrying that to   his plumbing loop, his boiler,   his thermostats on the wall. Right.   That's why we're bullish about  

[00:16:00] Unknown:

all the two five six stuff. That's for sure. Yeah. What what do you have to do to those Antminers to get them immersion friendly? Did you change the firmware?  

[00:16:10] Unknown:

Yes. So So they're running brains OS right now,   and we have the fans off.   They're just dropping.   We're playing around with the dynamic performance scaling,   which,   I guess, is a mode. LuxOS has a competitor called ATM, automated thermal management,   where it dials the mining power up and down, kinda walks it down, walks it back up more conservatively   based on the actual temperature, the chip temps, the temps in the machine itself.   And so   that's our fail safe, I guess, because if that oil all of a sudden stopped flowing or something,   the machines would walk themselves down, eventually shut off so they don't burn themselves out.  

But I've also seen guys use that for their control. I have a friend Cody up in Idaho where,   you know, if it's warmer upstairs and the radiant heat is on, less of that heat is escaping the floor into the environment,   then the fluid coming back to the miner is going to be hotter because the building took less of it.   And so the miner itself should walk itself down. So I think if you dial those parameters in, you can effectively have a good radiant system that instead of having a thermostat on the wall control when it kicks on and off,   it's dialed in to just actually balance itself to how much heat you're using.  

But I think that takes some time to actually get those settings dialed in. That's where we want   a way to communicate the mining power back to an external temperature sensor. Right? Mhmm. And this is where we're bridging that gap through Home Assistant right now just on a little pie at a bunch of Zigbee temperature sensors or a Wi Fi thermostat with public documentation, public API   to marry the two together right now. But I do wanna test the the DPS version.  

[00:17:50] Unknown:

How are you are you monitoring the,   dielectric   fluid temperature?  

[00:17:56] Unknown:

The tank does that   as does   the dry cooler here. So I guess   this is kind of like a a janky way to do it because it's another   it's another hardware expensive piece of hardware. That's the most expensive part. It's got all the copper in it,   where if the miner to prevent the miners from overheating, we've we've set a temperature on that dry cooler. If the fluid is x, so this would be after the floor   before returning to the miner.   That fan will just bleed off excess heat to not cook the miners,   which when you're trying to heat is a waste. In a in a traditional mining environment, that's what you want. You wanna keep your miners cool, keep your hash rate up, but we don't wanna waste that heat and just bleed it into the basement.   Where I would use something like that, and that's where we've got some extra zones on our manifold back here, We're just gonna have one mounted outside, and   then depending on which valve we open or which pump we turn on, that's when we can, like, monetize our solar if we don't need the heat. So the control, there's a handful of ways to do it.  

I I do think the best way to do it is with software, like a temperature sensor or thermostat talking to the miner itself.   And that's where we need Mujina.   We need LibreBoard   to make that easier without all this stuff in the middle and relying on manufacturers,   ASIC manufacturers to either release   documentation, APIs,   jailbreak with aftermarket firmwares. Mhmm. It's a lot of steps.  

[00:19:15] Unknown:

So   Go ahead.   The   are you still using the so you have those   immersed hash boards in the tank.   Are you still using the air temperature sensors on there as that, like, immediate fail safe?  

[00:19:33] Unknown:

That's the temperature sensor that is   commanding   the dynamic performance scaling, the power of the miner.   There's a temperature sensor in the immersion tank itself and throughout the plumbing to the dry cooler   on the inlet and outlet of the dry cooler, and that's what triggers the PID loop to change the dry cooler fan speed   or the pump in the immersion tank. And are those systems connected at all, or is that just a closed loop on the There's isolated loops. The miners don't know they're in immersion   aside from me telling them in brains OS.   Immersion tank doesn't know what kind of miner or what power I put in there.  

It's it's very janky.  

[00:20:10] Unknown:

It seems like it must be   like, it seems like it'd be tough to sort of tune that whole system. Yeah. When you when you, you know, you have PID running, but it's just on this closed, you know, single component   of the system.  

[00:20:23] Unknown:

Right.   And that's why I've seen almost every immersion setup in a home or business case, like, a Hashrate heating scenario.   They always have these dry coolers as, like, a   a separate loop fail safe to bleed the heat, which is a waste. I mean, that's an expensive component for someone to buy if it's heat that's not going to what they want it to. Mhmm.  

[00:20:43] Unknown:

Yeah. I mean, I guess you wanna have those those fail safes. Right? Somewhere   Yeah. You want a really tight loop that's like, if this gets too hot, like, just   shut it off or shunt the heat out somewhere.   It's nice to have those fail safes, but when you're tuning the whole system for optimum   efficiency and and performance,   you know, having those   those all at least somewhat connected seems like it would be helpful.  

[00:21:07] Unknown:

And the other way we could do it, I agree, is you could have a temperature sensor upstairs. This is how normal boilers work, is you just kill the thing. Like, it goes to a 220   volt Wi Fi relay switch, and it tells the miners to it just kills power. And I think that shocks the miners. I'm sure there's some benefit to being an immersion if we were to do that because you kinda have this big heavy thermal oil jacket. It can't really drastically change temperature very quick like a hydrominer, but   I don't wanna do that either. I mean, they're not designed to run bang bang like regular heating setups.  

[00:21:41] Unknown:

Right. Right.  

[00:21:43] Unknown:

And if you That's a cool problem to try and solve. I hope that we have this platform. People can Totally. You know, these control It's it's it's why the water heater hasn't been solved. I had a call this week where a guy was like, is there any kind of turnkey solution yet for a water heater? Just a 50 gallon. I got one out of the shot here.   And the answer is no because it's a closed loop.   Your water heater just heats up   to the temperature, and then it stops, and you're not always using that hot water. Right. And it's kind of like charging an electric car where   your charging rate has to drop off as you fill the battery or I like the analogy people use where you're trying to fill, like, a stadium,   and you can pump a lot of people in. They find their seats quick, but as it gets more full, it's harder to find your seats. It's slower to fill a stadium. That's how heat transfer works too. So as the water tank gets hotter,   you need to walk the miner down in wattage   because it gets ever so harder to dump more heat, to cram more heat in that tank. And so that's where I think it shouldn't even be those are insulated too. I think one, two, maybe a couple more Ember one boards   and a slick control setup with, LieberBoard and Magina firmware,   you will be able to build the hot water heater.  



[00:22:53] Unknown:

That's awesome.  

[00:22:55] Unknown:

That's so cool.  

[00:22:57] Unknown:

Someone should work on that.  

[00:22:59] Unknown:

Alright. Yeah. Because if if the   I mean, how hot would the water get? Like, what? A 120   degrees? One twenty. Yeah. And if if you got a 120 degree coming out of your faucet, that's, like, that's pretty decent.  

[00:23:13] Unknown:

Yeah. You don't want it much more than that. Yeah.   Because I shout out to my friend Travis. He he put an m 64,   one of those,   to his hot water. It's like a 160 degree water of 5,000 watt miner.  

[00:23:28] Unknown:

Amazing. And then and then when the water jacket's full in his hot water tank, he has to have radiator set up  

[00:23:32] Unknown:

jacket's full in his hot water tank, he has to have radiators set up throughout the house as   backup heat dumps.   Right. And then walk that yeah. It's it's really   not ideal.  

[00:23:45] Unknown:

So when you when you're doing this,   are you running, like, your shower hot water, like, through the water block in that miner?  

[00:23:54] Unknown:

Travis did, and then,   he started   eroding some metal. Okay.   Yeah. Aluminum,   steel stainless steel, copper with brace plate heat exchangers, they don't like to,   live together with   water, which has oxygen in it, so it can cause corrosion.   We separate the loops.   Like, when we had that hydrominer for the floor here oh, I just heard someone flush the toilet above me. That's the benefit of sitting in the basement with the plumbing above me.   We have the loop separated. So there's, like, the minor cooling loop, and then there's the building heating loop separated with the heat exchanger. And I think that's just better for us.   Heat   exchanger. And I think that's just better for isolation, for service, for repair, hot swaps.  

I don't want the the floor water going through the or your domestic tap water flowing through the water block. Right. Yeah. I mean, at least the the floor water is a closed loop. Right? But the True. Domestic hot water, that's just, like, whatever you get from the street or your well or whatever. Yeah. It's from the city. It's like I mean, there's there's a benefit. It comes in pretty much 50 degrees   year round.   Cool your miner a lot.  

[00:25:02] Unknown:

I used to watch, This Old House all the time, and they had a   a segment where they, like, just cut open a, domestic hot water heater.  

[00:25:11] Unknown:

Oh, boy.  

[00:25:13] Unknown:

Oh my god. Like,   it's traumatizing.   They're gross. Disgusting, wasn't it? The calcium disgusting. You're like, I don't even know if I wanna shower anymore in hot water because that was just filthy.   Yeah.  

[00:25:25] Unknown:

Yeah. Mine at home doesn't even have a it doesn't even hold enough hot water anymore to fill a bathtub, so I'm pretty sure it's gunked up.  

[00:25:34] Unknown:

Yeah. Of course, you've been sticking to the the very strict schedule of flushing your  

[00:25:38] Unknown:

hot water tank. Yeah. What is it? Every quarter you're supposed to do it? Yeah. Of course. Everyone does that.   No one does that. Are are those valves remote controlled, or do you have to manually open and close them? They will be remote controlled. Right now, they're cannot not connected to a relay or anything, but we're gonna get,   a controller. That's actually something Dylan's working on is   we wanna be able to control those valves with home assistant because that's where   the minors are also controlled, and that's where the smart thermostat can also be controlled. So we're treating that as our central hub in this   wheel spoke of all the devices in the building.  

And so, typically,   you can get analog ones that just talk to thermostats for different zones. I want   radiant floor heat on that side of the building, not on this side or vice versa.   But we want a way to bring that into software. A lot of this heating industry stuff is all just done through   analog   leads.  

[00:26:36] Unknown:

I was at a restaurant   over the weekend, and they had the same Honeywell   smart thermostat that you have.  

[00:26:44] Unknown:

Yeah. We actually,   I we learned the hard way. We're just buying thermostats left and right. That one   same thing with Ecobee, unfortunately. They used to have public open APIs, and now they make it where you have to request a developer account, which cannot be the solution if people wanna just buy this and get it in their house. So we've we've been waiting, like, two weeks to get approved.   I just bought one from a company called Vester   that does have a public API.   And there's something we learned too. Not all of them through the API endpoints let you just turn on the furnace fan, the circulator fan through the API.   Obviously, on the thermostat,   there's usually a heat or an auto heat and cool and then just a blower circulator button.  

But the reason we want that is that's how we're defining the control for the the HVAC upstairs, which is   I want the minor heat,   but I want the furnace circulator fan to distribute it because the minor fans   don't have enough pressure to force it through all of the ductwork. And if we did get fans that could do that, that's when they get loud. Right? I just want the minor to, like, leak heat out the back of it,   and then the circulator fan sucks it and pushes it around the building.  

[00:27:53] Unknown:

We need to be,   we need to be jailbreaking these thermostats.   I know. They're just running Linux. They're tiny little embedded Linux devices.   I think some people are gonna start working on this because   Google famously,   end of life to all the,   well, all the older,   Nest thermostats.  

[00:28:13] Unknown:

Yeah.  

[00:28:15] Unknown:

Let's make an open source thermostat.  

[00:28:17] Unknown:

It's not that complicated of a device. I think I think LibreBoard will be it. I wanna build LibreBoard where it actually just swaps into the wall,   and you have your red, green, white, yellow   thermostat leads. Right. And so   LibreBoard can analog manually control the furnace blower fan. Right? And there's no modifications. You don't have to splice wires, get into the drywall. You don't have to hack something to your furnace. And then via software, it talks to the miner. So then all you have to do is strap a miner to your furnace, which is easy. You cut a hole in the duct,   get a LibriBoard   thermostat,   swap it onto the wall, and, bam, you just married, and now you have, like, hybrid hash rate heat in your house.  



[00:28:58] Unknown:

What are those,  

[00:29:00] Unknown:

the green, red, and white wires? What are those like, what protocol are those communicating over? It's just, like, analog signals. Right? I think it's just analog signals. We have one triggers the flame, one triggers the fan, one triggers the air conditioner pump. Yeah.  

[00:29:15] Unknown:

They're they're,   Yeah. You could totally do that with the Libra board. It's like a standard or a group of standards, but, you know, if you have those wires popping out of your wall in your house, they are almost always dry contacts. So one of them will be 24 volts   AC, and then the this thermostat really just   connects   the 24 volt AC to whichever one it wants to turn on. They're like   really simple devices.   In fact, one of the hardest things to do with these thermostats, and I think Nest was kind of the first one to to solve this, is that   you need to you know, because there's these new thermostats are smart devices. Right? They connect to Wi Fi. They have a screen. So they need power.  

But there isn't necessarily power coming through those wires.   If if you've ever tried to hook up an ecobee   to what you you know, say you had a dumb, like, mercury switch or just a dumb thermostat on the wall and you wanna get an ecobee on there or a nest.   The the challenge is   getting power to the device. And sometimes it means going down the basement and, like, putting a little adapter   onto where those wires those thermostat wires come out so that you can send power   to the Nest thermostat on the wall. And they're   they're like   it's such a hack. I think the Nest has a battery in it because there   there is power coming sometimes, like, when the,  

[00:30:37] Unknown:

like trickle charges?  

[00:30:38] Unknown:

So it trickle charges a little battery in there so that when your heater isn't on, it's still connected to Wi Fi. But,   if you're if you're gone for a long time, like, you'll definitely come back to your nest being, like, you know, needs to be charged or, like, battery low.   It's,   it's a real hack. But  

[00:30:57] Unknown:

Yeah. They must do a lot of work to make it ultra low energy consumption too.   Definitely.   Would,   would would LibriBoard   need more power than, say, a Nest if I if someone were to make a a LibriBoard thermostat?  

[00:31:14] Unknown:

Yeah. For sure. For sure. Okay. It's it's not like a low power device.   Right. And and Raspberry Pis are not   really good   low power devices anyways.   They're just not optimized for that. But I think with the Libra board,   you know,   that's gonna be close to the miner.   And the miner obviously needs to be powered all the time. So you can get your power there. That's true. You do need it close to the miner.  

[00:31:40] Unknown:

And then Especially if it's air cooled because the Libra board will be the fan controller too.  

[00:31:47] Unknown:

So I guess you need a a separate little device that   uses those red, green, white, yellow   signals   to your thermostat but is also talking to the Libre board.  

[00:32:00] Unknown:

Yeah. Talking to the miner. Repurpose the same wires. Because the the problem that the Nest and the Ecobee have is that they're designed so you can hook it up to any old furnace.   But if we if we sort of control both sides of the program here, you could   you could provide   power, 24 volts AC or whatever it is, to your thermostat all the time.   I know that when you I did this. When you get an ecobee,   it comes with this little box.   And then it's like, you know, if you're in one of these   situations where you I forgot the colors, but it's like, if you don't have the white wire coming out of your wall, you need to be able to get your furnace and clip on this little box, and it will then provide constant power to the ecobee on the wall.   But if that little you know, if   if the furnace side of it is a is a Lieber board, then we could just build that in.  



[00:32:50] Unknown:

Gotcha. Well, so then maybe the thermostat device   also runs Home Assistant,   if that's the way we were to connect these things to the miners.  

[00:32:59] Unknown:

The furnace. I don't know. Yeah. Certainly. Yeah.  

[00:33:03] Unknown:

We'll hack it together somehow.  

[00:33:05] Unknown:

Yeah. We'll hack it together. And or we won't hack it together. Right?   We won't make a platform and people who know how to do this kind of thing can contribute stuff. And,  

[00:33:16] Unknown:

that's how we win. We should have the Libre board design all finalized,   maybe as early as this week.   Wow.   Schnitzel's   traveling right now.   So to be determined, might be next week. But,   this will just be the   the first one for the prototyping phase, and then we'll need to build the prototypes and validate them. But, yeah, he's getting really close.   There's like   he stumbled across something kinda interesting. And, Scott, I think you probably helped him with this. So you could probably help him or help explain it better than I could, but it's like the impedance matching on the traces.   Like, instead of having to, like, calculate   exactly how much copper   and, like, what diameter and length each trace is, the   PCB   manufacturing company can   automate that process if you just give them the parameters.  

Because if you think about, like,   the signals going   across those traces, right, if the impedance is different than, like, the sensor or whatever is gonna be getting a different signal than you want it to be getting   if the impedance doesn't match.  

[00:34:30] Unknown:

It it it matters most for the high speed stuff.   USB,   in the case of the Libre board, I think it it matters most for the, the PCIe connection,   because that's gonna you know, we need to support   full on PCIe for   whatever people wanna attach. So that could be,   SSD or or some sort of, like, a NVMe storage,   which would require   high speed data signals. And high speed data signals need to be impedance matched because they're they're like a differential. So the they don't just   have a single wire where, like, high voltage is a one and low voltage is a zero. It's actually,   they send, like, the opposite signals,   and then the receiver   will   take the difference of those,   which is nice as it cancels out noise,   but it needs to   it needs to be   matched so that you can   recreate the the signal on the receiver side.  

You do need to length match the traces. So, you know, there's these two this, like, differential pair. The traces on the PCB need to be the same length.   But what you can do and this is what I was talking with Schnitzel about is   you can then assuming they're length matched, you can then call out to the PCB manufacturer, like, these are my,   impedance controlled traces.   And because the PCB manufacturer will know   the,   characteristics of the board they're manufacturing, they can adjust the width   of those traces so that they have a set impedance, like 120 ohms,   I think is usual.  

And there's there's a fairly standard process to do that. It costs a little bit more to get the boards made like that, but but not much.  

[00:36:14] Unknown:

Yeah. Yeah. Is this the stuff that's Like, we're not.   Yeah. For sure. Is this the kind of   thought that's required to make   something like the Libre board   energy efficient too, just circling back to   what we were talking about with the Nests too? Because I had a my roommate in grad school, he he tried to make his own thermostat, and he tried to make it,   like, Bluetooth low energy and not run off a battery. And   through his circuit board design,   I don't know what was off of his calculations. He thought it would last, like, three months, and it lasted, like, five minutes. And so,   we ended up just having to get a little five volt, like, wall plug for it.  



[00:36:53] Unknown:

Well, low energy electronics   is,   the name of the game is sleeping.   Okay. If you have a battery powered electronic device, the   the you need to put it into a low power state and it needs to remain there for as long as possible.   With with RF electronics like Bluetooth,   there's no   you have to use a certain amount of power to get that   that RF signal,   that you needed from the antenna   or receipt. And so there's no fool in that. Like, you just have to do that. That's just physics. But   the art isn't that   how long can I stay asleep for in between Or pulling frequencies and stuff? Doing things. And then   making sure that that's that sleep current is as low as possible,   so that when you're asleep, you're really in the, like, microamps   or or microwatts   range. There there's a bunch of tricks there and things   you can do. But, yeah, that's the name of the game is is, you know, making sure your microcontroller can go into a low power state and stay there for as long as possible.  

Bluetooth is kinda crazy in that there's, like, specifications as to how often you need to transmit.   Mhmm. And that's defined by, like, the iPhone. Right? The iPhone will   will disconnect your Bluetooth device if it doesn't wake up every, like,   hundred and fifty milliseconds   and check   Gotcha.  

[00:38:18] Unknown:

A lot of So not an optimization   target for   EberOne?  

[00:38:24] Unknown:

No. No. EberOne's Yeah. A 100 watts. Like, low power mining is not a thing. Oh, sorry. The Lieber board. Yeah. Yeah. Yeah. For for Lieber board, the Lieber board is designed to be lined powered lined powered. Right? Like Yep. It needs to be plugged in. You need to provide your 12 to 17 volts   at   potentially a 100 watts continuously. And I I don't think it makes sense to make that low power. But if you were   trying to make some sort of accessory,   that would need to be low power. And I think we can get away with using a lot of off the shelf stuff that has already done that optimization there for low power.  

[00:39:01] Unknown:

Man, that's so cool.  

[00:39:05] Unknown:

They they always have these, like, oh, you know, it takes a little coin cell battery and it lasts ten years.  

[00:39:10] Unknown:

Yeah.  

[00:39:12] Unknown:

That's,   like, just means you're asleep for longer.  

[00:39:16] Unknown:

What about my Casio?   That last ten years.  

[00:39:20] Unknown:

Yeah. I mean, it they have a a very low power crystal oscillator in there, to keep time. And then those LCDs   are very,   low power. Right? They only   take,   it only takes power to change it.   It's   Right. Almost nothing to, like, you know, keep it,  

[00:39:41] Unknown:

showing the same thing. So is it like an e ink display then maybe?  

[00:39:45] Unknown:

E ink's a similar idea. LCD Gotcha.   I think that they do require a little because, you know, if you took the battery out of your watch or it goes dead, like, the LCD will go off.   But it takes a very, very Gotcha.   Just trace amount of current to, like, keep the LCD   segment,   switched on.  

[00:40:06] Unknown:

When someone wants to download KiCad,   they   bought a number one because they wanna support the project   like many of us did.   I know that you can put an initial order, right, Eco,   for some of the first   v zero zero?  

[00:40:21] Unknown:

Yeah. I mean, just to be clear, I haven't taken money from anybody yet. Right. So Right. I mean, nobody's, like, purchased any. But But people wanna get their hands on. Yeah. We've got a list of people who've expressed interest when they're ready to go. How do they   and developers mostly because,   you know, we still gotta figure out the heat sink, the fans,   power supply,   and, you know, the stuff that we're gonna recommend as reference material. So  

[00:40:52] Unknown:

Yeah. I didn't mean to put you on the spot there. I'm one of those guys. I can't wait to get my hands on one. I would love to start to look under the hood,   get the the PCB and KiCad and start playing around. Where does one go to start   learning the fundamentals?   The   for the number one and the two five six project platforms, is it as simple as the documentation with these repos when they're all live? Is there any   existing knowledge that we're expected to have to a certain level?   You guys have any thoughts on that stuff?  

[00:41:22] Unknown:

Well, like, KiCad, I would just, like, go to the either your system package manager on your computer to get it or, like, I've got Linux, for example. So KiCad was in the package manager,   and I just installed it from there. Or you can install it from the website.   Kicad.org.   Yep. And then,   really, it's like   what's been because I had never used KiCAD before, so I'm in that same boat. Sure. And it's it's been a lot of YouTube videos and then asking people like Scott who have good experience with it.   And then, like, as far as the, like, two fifty six projects go, so   the I think the GitHub repo will be the, like, first   source of   valuable information   Mhmm. For each of them because that's where you're gonna, like, get the   the files or   yeah. Like, that's where you would get the design files for the hardware or that's where you're gonna, like, clone all the files to compile the software.  

So,   I also have,   URLs and websites for each of the projects.   And my plan is to   make those websites   have, like, all encompassing information   on them about the project. So that should capture everything that's in the GitHub repo, plus I'm gonna do like a step by step guide for each one of these things on, like, how to get started using it.   So, like, for example,   we're, like, just on the cusp of   doing the   hydro pull release. In fact, I think jungle.ly   technically did do the first hydro pull release on GitHub.   And we were going through it this morning, and I was stepping through the whole process with him and sharing my screen.  

And we found a couple of bugs that we need to fix, So we're gonna do that.   But, yeah. If anyone's like familiar with my technical guides, like imagine a very detailed step by step guide   on the website for the project where you can get all the information you need on how to use it.  

[00:43:36] Unknown:

This is so cool.   Like,   seriously, support the two fifty six foundation. Right? This is where all the money goes to is, like, making this stuff, but then also   all the rest of these things. Like, how do you use it? How do you get involved? How do you contribute? How do you modify it? Like, how do you run it? What to do when it goes wrong? Like, there's a lot of work   that needs to be done there in creating that documentation and stuff. And,   yeah, I think   two fifty six   Foundation,   is very committed to doing that, but we could certainly use your support.  

[00:44:08] Unknown:

Yeah. And,   along that education component, so I decided I'm gonna launch   a new video series   called assembling freedom.   Oh.   And it's I'm gonna, like, do   YouTube like, I I haven't decided if I'm gonna put them on YouTube or yet or not. I just I really fucking hate YouTube's, like, trigger happy censorship.  

[00:44:32] Unknown:

They might just ban you. Yeah.  

[00:44:36] Unknown:

And we have bitcoin,   tv.com,   which is like a peer tube video server maintained by Bitcoin Park.   So,   that's where, like, I put all my videos anyways, and I'd like to just use that. The only problem is it doesn't really get any traffic   like YouTube does.   So,   I don't know. I haven't decided   exactly how I'm gonna distribute it. But the point is I'm gonna do, like, videos on   everything I'm doing here in the shop in regards to, like, building the electronics.   So, like, I'll be doing, like, screen shares and showing   what I'm doing in KiCad, like, getting set up for a project,   how I'm, like, checking everything and ordering   it to get the PCBs made and, like, getting the components on order.  

And then, like, what it looks like when you receive your orders and you're unboxing them and, like, getting everything prepped on the pick and place machine and   the solder paste and all that. So,  

[00:45:37] Unknown:

stay tuned for more. That's so cool. That, like That'll be fun.   You know, this stuff is complicated,   but we need more people doing it, figuring out how to do it, you know, getting started down this rabbit hole, more manufacturing.  

[00:45:51] Unknown:

100%, dude. Like, if there's really gonna be, like, millions of devices in homes   all across the country and the globe, like, there's gonna need to be a lot more manufacturers.   So  

[00:46:03] Unknown:

Yeah. And more people will put out educational content on how they went about it, tricks they learned,   and maybe tutorials for tweaks they made, everything.  

[00:46:13] Unknown:

Yeah.   I hope so.   It's it's a superpower.   Like,   you know, understanding   just even a a part of this this whole process that, you know, you can build stuff. You can design stuff. You can, like, change it to do what you need.  

[00:46:29] Unknown:

Mhmm.   But if you don't follow the license   and you you build something   and modify it and produce it and start selling it and then close source your changes,   we will make fun of you on Twitter. Mhmm.  

[00:46:44] Unknown:

Yep.   Yep.   Yeah. We we need we need to get, like, these skills in the hands of,   you know, more people who are doing cool stuff and and get it out of the hands of just these, like,   Chinese sweatshops, basically.   Right. Because they don't hear shit.  

[00:47:02] Unknown:

That's such a good point. I think that and it it makes sense. Like, not everyone's as in the weeds as you guys are or I try to be.   But, like, the BitHacks   is   a physical representation of what's fucked in the mining industry. But people just think think seem to think it's a fun little toy. You know?   And it's like, no. This is our token   of the mission we're on.   Right. And same with the Ember One platform and all the other projects as well.  

[00:47:30] Unknown:

Yeah. Yeah. It's   it's frustrating when you see,   the chatter online about   the BitAxes'   low hash rate or the high cost per terahash   Yeah. And or the fact that, like, you have to disassemble   a perfectly good miner to make BitAxes with the bit main chips. And   it's frustrating to be confronted with those arguments because,   like, the the point is that   Mhmm.   This is the first open source miner and, like, unnecessary   first step that has to be taken in order to get to   a dismantled proprietary mining empire   where there is widely available   low cost per terahash,   high efficiency   Bitcoin miners that are open source.  

And we'll get there, but, like, you know, it just   it's   the shit has has to be reverse engineered   initially because it's all closed source and proprietary. There's no other way to do it.  

[00:48:41] Unknown:

Yeah. It's it's a really backward system that we have now. I I can't believe that people stand for it. I mean, I guess that was their only option   for so long. But, you know, we're gonna change that, and then it'll just be   forgotten that we had this ridiculous   setup where people couldn't understand how it worked and had to, like, do these   gnarly hacks just to get their miner to operate the way they want.  

[00:49:05] Unknown:

Right. And that's the beauty of OSMU and BitX and all these projects is   everyone can play a small part in dismantling the proprietary mining empire. This isn't, like, just wait for it to happen. The tools are being built so that you can make it happen. Right. So get involved.  

[00:49:21] Unknown:

And yeah. And they're, like, being licensed so that you have full unrestricted access to them   so that you can inspect and modify and even distribute and profit off of them if you want. We encourage it.  

[00:49:35] Unknown:

Just follow the license. Don't be a dick about it. And that's a that's a really important point, I think, that deserves   driving home is that, like,   this is not like, being involved in this is not just some, like,   you know, altruistic   endeavor that, you know, is a vow of poverty. Like, no. No. You can make money doing this. You can make real money doing this, like building this stuff, selling this hardware,   developing on top of this software for your,   company's platform.   You can make real money. I mean, there's we're early. There are already profitable companies working in this in this ecosystem.   So,   yeah, don't don't get the right the wrong idea that   if you do this, you just it has to be given away for free or anything like that. Right? Like,   it's free as in freedom,   but it doesn't have to be, you know, free as in beer. Like, you can definitely make money doing this. Please do.  



[00:50:32] Unknown:

Yeah. I think a lot of there's a common misconception   that open source just means nonprofit.  

[00:50:38] Unknown:

Right. Yeah.  

[00:50:39] Unknown:

And it's it's not true at all.  

[00:50:46] Unknown:

Yeah. Anyways Obviously, Two fifty six Foundation is a nonprofit,   which is the right structure, I think, to be at sort of the the base level of this.   Mhmm.   But,   you know,   please   please go ahead. Feel free and, you   build a profitable company  

[00:51:05] Unknown:

on top of this. It can be done. Yeah. I mean, that's the whole point   is the   tech is out there   free and open source for   entrepreneurs   to build businesses on top of. So, like, like, Exergy is a great example. Like, if you can leverage   Mujina and Liber Board in such a way to   manipulate minors and turn them into home and office heating systems,   That's a, like, brilliant   business model   being built   on a foundation of open source.  

[00:51:40] Unknown:

Well, thank you. And it's not just an example of, like, you can make money doing this.   This is going to be   the primary   base load of hash rate and businesses   and system builders and people selling Bitcoin miners and creating all these,   you know, ingenious business   ideas because   it literally costs more   to wait around,   negotiate with Bitmain, try to reverse engineer it, to build whatever you wanna do   versus when those layers, those are commoditized and they're done for you,   any company   in the right mind will start there. And so   I think I don't know how long it's gonna take, but   we're gonna look back and laugh at   how Bitcoin mining,   the rigs looked, how the firmware   worked, how you talk to them, how you coordinated with them. Right. And in the future, the vast majority of Hashrate on the network,   we hope, is built on open source software.  



[00:52:40] Unknown:

Satoshi,   you know, famously posted on on the forums back in the day that, you know, Bitcoin miners could be used for heat reuse. Like, we've known about this for a long time. It did. Yeah. Why hasn't it happened at scale yet? And I think it's it's it's entirely because of the proprietary nature of miners   and the fact that all the hardware out there   is built   for, you know, on grid use in data centers.   Yeah. And you can't   you can't   build a real business with real hardware,   hardware, products   by hacking someone else's design. Like, that just isn't a go to market strategy. You need to have a real system that you can build on if you're building   products that you're going to sell to people,   you know, and they're gonna rely on it for their heat.  

So, you know, it's it's an obvious idea that has been going around for a long time that you can use Hashrate for heating. It makes so much sense. It's a huge potential market as you've dug into, Tyler, but it's just like we need to have   a way to develop real products to satisfy this market.  

[00:53:45] Unknown:

Yeah. That's why I,   I'm not trying to make enemies out there, but   I don't think I have too many yet. But  

[00:53:53] Unknown:

I I don't get I I hang out with us long enough. You'll you'll get a ton of enemies.  

[00:53:59] Unknown:

Dude, I have absolutely zero interest, and I actually just think it's lame and fake and stupid.   The,   like, the district heating. Sorry. I'll make some enemies right now. The, the people that say I heat 11,000 homes in Finland with Bitcoin. Oh, what are you working on? You can just do it now. Just use it now, and it's like, that is a 99%   uptime mega mining operation. You run a shitload of what's minor products. You don't care that the firmware sucks. You don't care that they're not open source. It's like, it's it's the opposite of this mission that we're talking about right now. That's why I get so excited about   hash rate heating, not like Bitcoin mining operation heat reuse because   we need all the stuff from the two five six foundation, all the heat punks out there. We we need these building blocks. And so to just say, well, why are you spending all this time working on this now? It gets me so jazzed up because I'm like, what you're   doing, you're part of the problem.  



[00:54:50] Unknown:

It's interesting. I didn't even, like, think about that. Yeah.  

[00:54:55] Unknown:

The industrial heat reuse is part of the problem.  

[00:54:58] Unknown:

Yeah. Who who does that in Finland? Isn't it Jared Melerund? Or there's a couple. Like, Mara has a site.  

[00:55:06] Unknown:

Shout out to my friend. Don't wanna make an enemy. Kent at Sasmizing mining has a Norway site. You You know, this is, like, how it's done. Some of them are hosting facilities. Hash Labs out there has this to name drop a few.   But they're just traditional mining operations where they don't feel the pain, I guess, is how I'd put it. So they just got they just need need, like,   a massive amount of heat, and then they're, like, directing it. Yeah. They're preheating setups. So they, like, preheat. It literally looks like a mining operation. They have operators. It's 99% uptime. They just run the the cold water returns from the city   or from the college campus or from Disneyland or whatever. These are all, like, district heating setups or some industrial plant, and you preheat that water with the miners, and then   they'll use some other heating fuel to get the temperature they need.  



[00:55:51] Unknown:

Interesting.  

[00:55:52] Unknown:

Yeah. I think   that they are also running into roadblocks trying to set this up because it's not the normal model of Bitcoin mining.  

[00:56:00] Unknown:

That's true. Yeah. It's not the on grid operation. I'm sure they've had some challenges they've had to overcome to give them credit where credit's due, but it's   it's worth pointing out that it's different than running the miner dependent on your heat load. I'm not accommodating the Bitcoin miner. The Bitcoin miner's accommodating me. Right? That's an important distinction.   Yeah. Thanks for my rant. I'm preaching.   TED talk.   Hey, Scott. I had a question for you. Yeah.   So what is the name of the   firmware that the BitX runs? Remind me.   ESP miner. ESP miner. And so when   Mujina has been under development by Ryan   And through some conversation I was having with him, it sounded like ESP miner   and Magina   both had to run on it. Or how do you hand that off? And   can you talk to me a little bit about this? Because it it didn't make sense in my head.  



[00:56:59] Unknown:

Sure.   Well, ESP miner is called ESP miner because it was designed to run on a a low power,   low performance,   embedded system, a microcontroller.   Right? This thing runs. In The ESP. Yeah. An ESP 32.   Yeah. It's a cheap Wi Fi microcontroller.   It's it's so small,   and cheap that it can't run Linux.   There's just no way to run Linux on it. So it runs this code, ESP miner. It's bare metal. Bare metal means there's no OS running.   Gotcha. It just runs the code directly on the CPU.   And this is not a CPU like is in your computer or even in   on the Libre board. It's it's a it's a super small, super cheap processor.  

And   there's a bunch of things you have to do to accommodate   that,   that processor. So ESP miner,   you know, does all those things,   so they can run on this, like, you know, it it's it's a CPU, but it has,   like,   eight megabytes of RAM total,   and, like,   I think 16 megabytes at the max of, like, code space. Right? It's a very underpowered system. So there's a lot of sort of tricks you have to do with a code to get it to run on a system like that.   That is much different than Magina. Magina is   we call we call it a firmware, but it it's an application.   Like,   you could download it and build it and run it on your desktop computer. It's an application.  

Now there's there's   some supporting   code like Linux and stuff that also goes onto that same system so that this application can be run.   But   it would not be possible to run Magina on the ESP 32 that's in the BitX.   It's just totally different.   And so I think Magina   is more targeted towards these larger systems. When you start getting more hash rate,   when you,   you know, more hash boards, when you wanna interface with,   other network devices easily,   like,   Home Assistant or   or whatever. It it gets it gets to be very difficult to do that with,   with ESP miner or any embedded platform.  

So  

[00:59:21] Unknown:

And and Ryan had to   test Mujina, like, on a miner and have an ASIC chip to communicate with   initially.   So when he started building this, he used,   special   firmware,   I guess, for lack of better term, called BitX Raw.   Right? Yes.   And he had that loaded on a BitX, and that gave him the ability to, like, start   building Magina and, like, seeing what it was like to   communicate with the,   the ASIC chip and,   what else is on the BitX? Like, a temperature sensor,   the fan controller.  

[01:00:01] Unknown:

So is BitX Raw   talking back to his computer running Magina, feeding it data about the chip so that he can build Magina.  

[01:00:12] Unknown:

Yeah. Yeah. BitX Raw is a it's an ingenious little bit of code. It replaces ESP miner. So ESP miner is no longer running on the BitX. Okay. I think that was my question then. Yeah. You load BIDX raw on there. Shout out to k one, a community member who who developed this. But BIDX raw just   your BIDX won't function as a BIDX anymore. It it basically passes the data Gotcha. From the ASIC and the other accessories on the board directly through to your computer over USB.   So now you can run   your mining   firmware, really an application on your PC   and treat   the,   Bitaxe  

[01:00:48] Unknown:

like it's a a hash board. Yeah. Like a dumb hash board. Gotcha. Like the Amber one. And the ESP 32 is no longer the tiny control board on there.  

[01:00:57] Unknown:

Mhmm. Yeah. It's just it's just, like, it's hardly being utilized at all. It's just passing commands through directly.   Cool.   Ryan had some   initial,   difficulties using it because, like,   you know, if you have the ESP the ESP raw firmware on your BitX and you turn it on, like, it's not going to do anything. And in fact, I think it overheated immediately   because you have to, like, control the fan. Yeah. Because the fan didn't turn on. Yeah. So it Yeah. It's a it's a it's a development tool.   It's really neat. It's a really good idea.   But, yeah, it's   a it actually is is it makes your bid acts operate very similar to the way the ember one is going to operate.  



[01:01:38] Unknown:

Gotcha.  

[01:01:40] Unknown:

Yeah. Because the ember one doesn't have any   fan controller   or control   mechanism on it. Like, all the   the everything it's gonna do, it's gonna get told to do from the Libre board. With Magina. Yep.  

[01:01:55] Unknown:

Gotcha.  

[01:01:56] Unknown:

Yeah. And he was I was talking to Ryan earlier today, and he was, like,   found this rabbit hole where   if you're hashing with more than, like, 260   or 280   Terahash   on Stratum v two,   then   the,   there there's not enough.   You you can't divvy up the work enough   to that many   different chips   without, like, duplicating work   because it just it, like with that much hash rate, like, all the bits you can the chips can roll. Like, it rolls through it too fast before you can get more work to it.   So   he was   it was just a he's finding all sorts of interesting things out about mining, and it it just kinda makes us wonder, like, well,   like, have what are the other firmwares do like, this is clearly an issue, and, like, what are other firmwares doing about this problem? Are they addressing it at all? Or is, like,   is, like, the entire Bitcoin network at, like, one Zeta hash right now   unnecessarily   because   x percentage of that is being wasted due to some sort of weird limitation,   where, like, you just can't roll bits fast enough. But So this might limit how many  

[01:03:20] Unknown:

how much how many hash boards you can have per control board. Am I understanding that correctly? Like, that specific   bug?  

[01:03:27] Unknown:

Yeah. But it if it's on Stratham v two.  

[01:03:31] Unknown:

I mean, yeah, there's miners out there with more than that. So  

[01:03:34] Unknown:

Yeah. And it's,   it it it can be fixed, like, a number of different ways that are it's pretty easy to fix, but it's just something that   Ryan was trying to decide. Like, is this something I, like,   try to address right now, or do I just focus on, like, getting   Mujina shipped   for, like, the smaller 100 watt devices   that we're actually gonna be running it with.   So  

[01:04:01] Unknown:

just Yeah. He sent me a message about this. I need to get back to him. I you know, my under I I think Stratum v two has a mode where you can essentially   the pool can send,   work   directly to the miner.   The way it works with Stratum v one is you kinda get this, like,   mining template, but it has a lot of holes in it.   And that gives   the mining firmware that's on the miner,   this optionality is to sort of it can change, you know, little   fields in this template.   And that actually gives it a lot of,   space to hash, a lot of bits that can be rolled   to hash,   like, a a ton. I think   most pools will give you, like, eight bytes of, extra nonce to   to roll. And so that gives you plenty of   of work that it can iterate over, and it can do it very quickly. But,   yeah, I think I heard about this JADM v two mode where   that the the pool does the sort of extra nonce two rolling,   and sends it to the miner. And, yeah, I think that could be a a bandwidth like a network bandwidth,   bottleneck there. But   they gotta have a way around this because, I mean,   200   you know, we already have miners that are well   over 200, 300 terahash. So Right. Yeah. You gotta have those extra nonce bits in there. He he was also saying the,  

[01:05:27] Unknown:

like, with the bit main chips,   the firmware will just broadcast the job,   and then, like, the chips will intelligently   decide   how the work gets divvied up so that they're not, like, overlapping and duplicating work. Right?   But with the Intel chips,   the the chips don't intelligently divvy up the work. So the firmware has to make sure   that the chips   are only getting   the appropriate work so that they're not overlapping   the work that other chips are doing.   And it's not only the chip that it needs to send that job to, but it has to send,   like, the work to, like, each of the individual, like, little engines running in the chip.  

So it's gotta broadcast out, like, for, like, a 12 ASIC design   for, like, an ember one, for example. The firmware will have to send out, like, 3,000 different,   work orders   for each of the little engines and all of the chips to work on   because the chips don't do it like Bitmain does. That's  

[01:06:33] Unknown:

yeah. That is madness.   I'm a little bit worried. I haven't fully dug into how this amber one or how the the Intel b z m two protocol works.   I can say that I have two different miners with hundreds of b z m two chips in it, and they work.   So we do   have multiple examples of how this can be done.   Yeah.   So it does seem possible,   but it yeah. It's gonna require some finagling to   to figure out how to do it. I know that the Intel chips   don't necessarily use the same broadcast mechanism that bit main chips do. They have this very complicated thing called time division multiplexing   where, like, each chip has, like, this window,   and it's like you you broadcast it during the window and it's like this chip gets it, and then the next window, the next chip gets it, and then they send back their responses with the same   time division multiplexing strategy.  

It's   seems like a weird way to do it.   I think that   had Intel continued making ASICs, they probably wouldn't stick with this,   fairly obscure   method.  

[01:07:40] Unknown:

My initial concern was if Ryan was if Ryan correctly interpreted   that that's how the Intel chips are working,   that my concern was that it would shift more of a burden onto the CPU   where Mujina is running.   Because, like, the chips you know, instead of sending, like, one broadcast job, now it's sending out, like, 3,000   little jobs.   So I don't know. It'll just be interesting to see how   different chips affect the overall system differently.  

[01:08:12] Unknown:

Yeah. I think that's right. I think that's right. Ryan has a big job,   to make this modular enough that we can support these,   sort of fundamentally different architectures for how the ASIC works.   It might it'll require some trial and error. I think we can get it to work.   The the miners that I have that,   use these Intel chips,   one of them is actually over USB also. So   they're they're doing it that way. The other one uses sort of   more traditional,   asynchronous serial,   but it does operate at a very high baud rate. And that's one of the troubles we've had with integrating these b z m two chips is they default to,   five megabaud, which is just an insanely fast,   asynchronous serial link.  

And I I'm assuming they do that because of the increased bandwidth that they need to directly address each chip individually.  

[01:09:07] Unknown:

Wow.  

[01:09:11] Unknown:

Speaking of   modularizing things,   that makes me think of,   what is Brett working on now? Not PI ASIC, but the Rust implementation on the two five six recall. S. ASIC RS. Yeah. What's the what's the story behind that and why he's working on that? Because,   you know, that makes me think in the future too when people want to put Hashrate into   all sorts of fun businesses and homes and buildings and stuff,   It would be great if there was a standard way to talk to them that aggregated all the APIs. Either the API was a standard from different   firmwares and manufacturers   or someone just compiles it altogether. And you send a request, and it figures out how to talk to it. Yeah. So, like,  

[01:09:56] Unknown:

the Stratum   protocol,   it's it's   if you can even call it that, it there's no, like, actual spec for Stratum you want, if I understand correctly.   It's just, like, developers that have been   digging through, like, old forum posts and, like, finding relevant   relevant bits of information   and then looking at,   like, work that, like, doctor Khan Kollivas   had done   and, like, figuring out how to,   you know, communicate using Stratum v one.   And   because there's no spec, like, that means that, like, the clients can, like,   package the information   in, like, different orders.  

They can, like   you you never know how that information is gonna show up to the server. And likewise, you never know, like, how a server is gonna, like, send things out.   And it's just, like,   without having that   standard, it's,   almost impossible to be able to, like, communicate effectively with so many different kinds of miners out there. So what Brett did originally with PyASIC,   if I understand correctly exactly what it's doing,   it basically just takes all the information from the various miners, the clients,   and standardizes it so that it's in, like, a nice, neat, organized format. And   it makes it easier to communicate with them and see what they're doing and check,   you know, different metrics on each of the miners. And so, like, you can imagine   if you use that to power something like a nice   dashboard,   like, you'd be able to have, like, a range of miners and be able to see temperature settings and, you know, hash rate and best difficulty and all the information that's coming in from those clients.  



[01:11:48] Unknown:

Yeah. There's an important distinction there. Right? Like, stratum is how the miner,   gets work to actually hash and submits its responses back to the Bitcoin   pool and and network.   And then the the API, which is what PieAsync and ASIC RS deals with, is kind of everything else. Right? The control of the miner, like start, stop, adjust your hash rate, you know, are what's the temperature,   change the pool that you're mining to,   all those kinds of things.   In most cases, the API is   just sort of an encapsulation   of how the dashboard works. Like, you think of everything you can do on the dashboard,   is done   for the miner is done over the API.  

In fact, on Mid Axe, we actually use the API. The dashboard, which is called Axe OS on the Mid Axe, communicates with the miner over its API.   So you can make a dashboard somewhere else,   or, you know,   get full control of your miner over that API. But then it's still   doing   Stratum   to do the actual Bitcoin mining part of things.   With with Magina,   you know, we have the opportunity   to create our own API.   I think it'll benefit from being similar to what's out there.   But the problem with what's out there is that every manufacturer essentially has their own   API. It's all Exactly. Different, and it's poorly documented. And so that's what   has done is collected all of those   proprietary APIs and tried to, you know, get them all under   one roof so that   it's easier to deal with different   manufacturers, miners. We have opportunity with Magina to make our own API.  

Of course,   in this case, it will be very well documented, and it'll be extensible so that if there's something you   really gotta have, right, you could add it yourself, or you could   submit it back to the main project and get it rolled into the mainline firmware.  

[01:13:50] Unknown:

This makes me think of   are you guys familiar with, like, the CAN network?   Mhmm.   Yeah. It's like   there's hundreds of automobile manufacturers out there, but all the gadgets, everything communicates to the ECU over CAN. And then you've got your OBD two OBD two port, and   any code reader or scanner can talk to any type of car. I mean,   this goes back to what we were talking at the beginning is   unfortunately, like, standard bodies   are helpful in that scenario, but   you can also go about it like like Brett is, and I hope it becomes the the status quo, which is just where the community is and where the development is   and where the people are working on implementing all the new APIs as they come out, and it's sort of the source of truth to communicate with these things.  



[01:14:38] Unknown:

Yeah. And so he's got,   PyAsync   for anyone   who's good with Python, but then he's also,   turned it into a Rust project, and that's   asic RS.   And so he he wanted those projects,   like,   in a place where they could be   kinda under, like,   a a open source organization   umbrella.   So he   asked if he could put both of those projects under the two fifty six foundation   organization on GitHub. So they both live there.   And then we've since,   added,   Py   Asic to the,   Py   py. I think I'm saying that right. It's like PyPI. P y p I,   the PyPI repositories.   And then,   for ASIC RS, we added that to crates.io.  

So you can just, like, add those   repos into your package managers on your systems and just, like, when you do your system updates,   any updates that are   pushed out to those repos, you'll you'll pull into your system.  

[01:15:54] Unknown:

Amazing.   This is such the win with open source development. Right? It's like now those,   you know, those libraries like PyASIC and Asic RS are just available   on your system. And so you don't have to spend all this time reverse engineering this. You just, like, you know,   APT   install,   PyASIC,   and now you've got that library there, and you can just start building your application on top of that, whether it's a mining management tool or,   you know, a home heating type system when you need to have a overall controller for everything.   This stuff is is pretty neat. And I think we're already seeing it. You know, we're seeing a lot more development in this space   because of things like that.  



[01:16:35] Unknown:

Mhmm.  

[01:16:37] Unknown:

Yeah. Speaking of Python,   one of my friends reached out to me here from the space and said that he vibe coded a Python script that runs on a Raspberry Pi   that is in real time. I shot you a message about this, Scott. We gotta dig into it more. I need to see it. But it's every thirty seconds, it's adjusting the frequency and voltage and fan speed of his bid acts,   which I didn't know you could do without a reboot, but he's claiming that his shares have doubled.   So every thirty seconds, the miner is   readjusting itself, I guess, to the environment, which I could see is helpful.   Not everything's like a static environment. Even our our bid acts upstairs,   overheats every day when the sun sets, and it's, like, perfectly in the window.  

But what is the plausibility of this? I need to actually take a look at it. It needs to bring it over here. But can you do that without can you adjust those parameters without rebooting it?  

[01:17:33] Unknown:

Yes. Technically,   yes. You can.   What I was telling you earlier is,   adjusting the the fan speed and the,   ASIC voltage is is easy.   Adjusting   the ASIC hash frequency, which is sort of how fast it hashes,   is a little more difficult.   Their chips are a little bit they're a little bit sensitive. Right? You can't just, like, massively stepwise change   the hash frequency. They'll   crash,   more or less. So you kinda have to ramp it up.   There has been some effort to   make dynamic hash rate,   or hash frequency changing work in the BitX.   I don't think that's made it into,   ESP miner yet. I could be wrong. I know that,   Wonklu, one of the,   actually, the the core maintainer for ESP miner,   had it working.  

Of course, we support multiple bit main chips on the BitX, so some of whom it wasn't quite working as well. So I don't think it it sort of made it into,   the default   ESP miner release.   I could be wrong about that. Maybe Wonklow will correct me there. But,   yeah, I'd love to see the logs from this happening. I I can imagine a couple of things going on here.   You said that the number of shares increases.  

[01:19:00] Unknown:

Shares per minute. Yep. But that doesn't necessarily  

[01:19:03] Unknown:

mean you have a higher hash rate,   because the pool controls the difficulty that it wants for shares.   So if you're if you're restarting the connection to the pool, now you're going to be starting at the default difficulty.   And so maybe then it gets some shares through,   and then the pool says, oh, wait. Slow down the rate of shares. Go for a higher difficulty.   And then it slows down again, and then you restart, and it starts that whole process over. I could imagine that. Oh, yeah. One of the things you'd wanna look at there is the logs for the bid acts. What's going on? Is it all pools   go through this sort of, share difficulty   finding algorithm   because the pool wants to have every miner submitting shares   at a regular   rate. And, of course, there's all different types of miners that have hash differing hash rates. And so   when you first connect, there'll just be this default, and maybe that's too slow. Maybe that's too fast. And so it'll kinda it's called the VARDEF. It will   hone in on this correct share rate.  

But it doesn't necessarily mean that the miner is hashing faster. It's just trying to find the right share rate,   by adjusting the difficulty.  

[01:20:18] Unknown:

Gotcha.   Yeah. And then same for voltage, that knob too. I mean, maybe it's less sensitive than frequency, but there's upper and lower bounds on that too. Right? I mean, this is ultimately what goes into the the tuning that a lot of aftermarket firmwares do to find their sweet spot.  

[01:20:34] Unknown:

Definitely. Definitely. There's,   you know, there's gonna be an optimum voltage for any given hash frequency   on these chips.   If the voltage is too high, then you'll just be   burning power. If it's too low, it won't be hashing,   won't be hashing very well.   And, unfortunately, it's not like, oh, yeah. This chip just used this. There's different bins for the chip so they can go Yeah. Place.   Yeah. We should we should look into that some more. I'm curious. Yeah. I think I'll get the logs. Yeah. Pulling up the logs on the BitX and and seeing what's going on there would be interesting.  

[01:21:10] Unknown:

I do like this idea of dynamic tuning, though, because it often is a scenario where it's not just you tune it once and you're done. Right. It could be a hotter day, colder day, whatever.   Your fans, your filters get clogged up for the guys in the industrial sites. I mean,   prompting a retune  

[01:21:28] Unknown:

makes sense to do on a regular basis. That cadence is up for discussion. But I mean, even, like, in Colorado, it might be, like, 25 degrees in the morning and then, like, 59 degrees in the afternoon.  

[01:21:42] Unknown:

Right.  

[01:21:42] Unknown:

Like Yeah. Yeah. There there's actually a bunch of projects out there that, other people have made to sort of auto tune your bid acts, because all that's available over the API,   which is public and well documented.   So, yeah, there's there's some really neat projects,   and people have successfully tuned their bidaxes.   It'd be it'd be interesting to look at, you know, how often it's worthwhile to to retune it because it does take some time.  

[01:22:11] Unknown:

Is the tuning algorithm the special sauce where   some firmware   manufacturers, companies might do it better than others?   Or is it pretty standardized   based on the actual hardware you got? You're gonna get the same result over and over according to the bin, obviously, but for your exact hardware.  

[01:22:29] Unknown:

Sorry. I was getting I was blown up there. I missed that.  

[01:22:34] Unknown:

They I think I think a lot of,   firmware manufacturers will   will have you believe that that is the special sauce.   Yeah. That's how they communicate it. Yeah. I I don't know. I don't know, though.   I think that'd be a good question for Ryan.  

[01:22:52] Unknown:

Maybe a good test too. Yeah. Yeah.  

[01:22:57] Unknown:

Do you,   do you wanna go through the Hashers?  

[01:23:01] Unknown:

Sure.   I'm squinting at my laptop here. You guys should see my setup behind the camera. My camera is strapped to,   a   one of those boxes that rolls up your hose.   And then   my mic is mounted to another Bitcoin boiler with an m 64 in it. Nice.  

[01:23:21] Unknown:

That sounds heavy.  

[01:23:23] Unknown:

Yeah. It's heavy.   Alright. Pulling up LinkCoin.   Let me get the   last seen.   Shout out on LinkCoin to our hashers supporting the two five six foundation. We've got Schnitzel's Fish Tank Axe,   Scott Offord from Bitcoin Mining World, Scott Offord from Open Hash Foundation,   and Tubaloo.  

[01:23:48] Unknown:

I think those are the same as last week. Oh, what's Open Hash Foundation? What are those guys doing?  

[01:23:54] Unknown:

I don't know, actually.   But Scott's been hashing with that bid axe for a while now. We should take a look at it. Alright.   Then on solo c k pool, shout out to Schnitzel's wall axe. Why is he using different pools for different bid access? Decentralization?   He just does all the things. All the things.   Rock paper bitcoin.fm.   Jeffrey Epstein did not kill himself.   Seed signer, BiddaX. Kuma is excited to meet y'all in Nashville. Kuma, you gotta update that. We met you.  

[01:24:30] Unknown:

Keith, we we met Kuma. Let's go. Gotta change this.  

[01:24:34] Unknown:

Wood miners, BiddaX.   FreeCAD is pain. Life is misery. Damn you, Scott. It's all your fault. JK still love you.   Oh, that's a new one. Woah. FreeCAD is,  

[01:24:47] Unknown:

pretty difficult. I tried that one.  

[01:24:51] Unknown:

Is this   for PCBs, or is it just a regular CAD software?  

[01:24:56] Unknown:

It's like regular CAD software. I I wanted to do it for, like,   looking at three d printed files,   and,   I didn't make it very far. FreeCAD is,  

[01:25:10] Unknown:

it's it's rough. It's like mechanical CAD software,   but free and open source.   It's   it's still a little bit early on in their development,   and there's a lot of, user interface stuff that they could improve on. But   this is important because It is very important. Like,   electronic CAD software, like KiCad,   there's there's just a bunch of bad   software out there that's you know, it's like you have to subscribe to it. And The licenses are insane. The licenses are insane. And and a lot of people turn to,   Fusion three sixty, which is from Autodesk,   which is   dramatically easier to use,   but   they make you save all your files, like, on their cloud service. Right? This is not   what we want for for open collaboration and sharing That's terrible. With others. Yeah. And then, you know, they get you roped in with some evaluation version and,   you know, inevitably, there's some feature that you need and so now you gotta subscribe to it and now you're paying them,  

[01:26:17] Unknown:

yearly, monthly, whatever and then they don't need to share. Pay them because they'll just limit your access to their server with all your files on it.  

[01:26:25] Unknown:

Yeah. Once the solar store the files in, like, you know, a cloud service, at least one that, you know, you don't control, like, that's that's not cool. Wonder if they just own your design too by default.   I don't know. Autodesk has been around for a long time. Yeah. And they they,   I don't know if they're such a good actor. They have lots of of history of   doing weird things and,   you know, do you have to ring people?  

[01:26:53] Unknown:

Are there open source slicers out there for three d printers or tools that map g code for CNCs? Do you guys know of any?  

[01:27:04] Unknown:

I know there's open source three d printers out there.  

[01:27:07] Unknown:

Yeah. The,   Okay. The RepRap, right, was the open source project for three d printers. I think they have a slicer too. What's the name of their three d printer?   I'm blanking on the name. It's a very popular one. It started as a project to make the self replicating three d printer.   Oh, is it the per Permusa or whatever? Prusa. Prusa. Yeah. Prusa. Yeah. That Prusa Prusa. The orange one? Yeah. That's the guy's name.   I think   I don't know. Don't quote me on this. I bet you that their software is, their slicer is open source, though. That'd be sweet.  

[01:27:44] Unknown:

Okay. On public pool,   shout out to hardestblocks.org,  

[01:27:50] Unknown:

Stalin's bid acts. Thank you for support. .Org. Have you have you all seen this website? It's pretty cool. Uh-uh.  

[01:27:56] Unknown:

It's not the flute one. Right?  

[01:27:59] Unknown:

It's not quite as cool as the flute sheet music one. That is pretty dope. But Hardest Blocks   is a it's a neat website and they just rank all the Bitcoin blocks by,   the difficulty of the share that solved it. That's sweet. Yeah. That was sick. It's really neat. And there's some crazy ones. There's some real overachievers out there. I think it was like,   you know, because what what's the current difficulty? It's like a 100 and something trillion.  

[01:28:25] Unknown:

Wait. So it says the hardest one here is 50.54   e.  

[01:28:30] Unknown:

Yeah. Okay. So that's   54   That's 50,000,000   t. Right? You only have to get a 100 t or a 120   Holy. T. Right? So they got 50,000,000 t? Like, that's   that's a good share.  

[01:28:42] Unknown:

Block 756,951.   That's That's wild.   And the next the next hardest was 18.7   e. So, like, there's there's quite a mode around that hardest block.  

[01:28:57] Unknown:

Unfortunately, you don't get any more Bitcoin for getting a really good share, but, you get   bragging rights on hardest blocks.  

[01:29:05] Unknown:

There's a that miner,   they should save it. It's out there. That ASIC chip, right, Scott, that exact chip  

[01:29:11] Unknown:

that solved this is out there. They should. They should. It's a lucky one. Unless it got,  

[01:29:18] Unknown:

crushed by a steamroller in Myanmar or whatever country that was in. Oh, yeah.  

[01:29:23] Unknown:

Yeah. What block did you say it was? Like, it's it's a while back.  

[01:29:26] Unknown:

Yeah. 2022.   756951.  

[01:29:32] Unknown:

Who knows?   Wow. It'd be interesting to look through and see   I guess we don't have any way of knowing necessarily what minor it was, but, like, has lightning ever struck twice?  

[01:29:46] Unknown:

It's interesting that all the   hardest or these highest difficulties are all coming in   in the last couple years. I haven't seen one older than 02/2022.   Oh, there's a 2019.   Yeah. That is There's 71765,  

[01:30:07] Unknown:

two because you should be able to pull this from   all of the   blockchain's history.   Right?  

[01:30:17] Unknown:

I would think so.  

[01:30:20] Unknown:

Yeah. There's some statistics there, but,   I mean,   if you get a share above the network difficulty, you're gonna submit it no matter what. So there should be some some higher ones early on. But   Mhmm.  

[01:30:35] Unknown:

That's sweet.   Alright. And our hashers on Ocean Pool,   Bible HODL,   two bits, one block.  

[01:30:44] Unknown:

I like that one.  

[01:30:46] Unknown:

Pisandy proof of print, BITAX one,   Reckless   apotheosis   mining.   I got snipped last week for pronouncing that wrong. Sorry.   Virginia Freedom Tech sells average bidaxes.   Why just average?   Circa say hash,   hub,   reckless systems,   watch this space, seventeenth white paper day. That's true. Coming up.   Borst   and bidax just wanna be in a 17 pro. Thank you very much for your donation to the two five six foundation.  

[01:31:21] Unknown:

Hell, yeah. Huge.  

[01:31:25] Unknown:

Seventeenth white paper day. Wow.   You guys gonna party?   For white paper day? What day is it? Is it Halloween? Is it around then?  

[01:31:39] Unknown:

Yeah. I think it's October 31, isn't it? Yeah.  

[01:31:44] Unknown:

Yeah. Let's party.  

[01:31:45] Unknown:

Yeah.   Nice.  

[01:31:50] Unknown:

Get dressed up. Carbs up. I mean, I'll probably be in bed by, like, 08:30.  

[01:31:54] Unknown:

It's it's a whole day. We can party all day. Yeah. A little bit early.  

[01:32:00] Unknown:

Oh, lastly, did you guys see that,   that rendering from NVIDIA for their five gigawatt   HPC data center in space   with a four kilometer wide solar array?  

[01:32:13] Unknown:

What? Oof.   Yeah. That sounds retarded.  

[01:32:17] Unknown:

Yeah. That doesn't sound I I all the comments,   like, Jason from Ocean and people were   talking about how are you gonna cool this thing, five gigawatts   in space.   Not to mention There's no air. You can't blow things away  

[01:32:33] Unknown:

out of white paint. How much does that weigh?  

[01:32:37] Unknown:

Yeah. A lot of star trip. In the vacuum of space, it would be weightless.  

[01:32:42] Unknown:

You gotta get it there.  

[01:32:44] Unknown:

Lots of Starship shuttle runs. They're they're just gonna bring it up there one piece at a time and assemble it.  

[01:32:51] Unknown:

Very interesting.   I'd rather mine Bitcoin in space   than put the data centers up there.  

[01:32:58] Unknown:

Yeah. But the lag, I think, is the issue.  

[01:33:05] Unknown:

Low Earth orbit. No lag.  

[01:33:08] Unknown:

Sorry. Well, some I, was talking with,   our friend, previous podcast guest,   Bob Agalrath.  

[01:33:18] Unknown:

Yes. Oh, yeah.  

[01:33:20] Unknown:

I was talking with him a bunch at Tap Conf in Atlanta this last week.   He is working with a team of people that are   pretty seriously,   pursuing   launching a bid x into low earth orbit. That's right. Yeah.   In fact, I just got the URL,   dyson-labs.com.   Oh, yeah. I chat with Calvin from there. He was a really cool guy. Yeah. Yeah. I've been chatting with Calvin too. It's it's a neat project. I mean, it   I'm not sure it it makes a ton   of, sense to launch a BitX   into space. It's certainly not a practical place   to mine, but   it's a fantastic idea. And I think if it serves as a platform   for getting more,   Bitcoin   stuff into space, then it it's worth doing. I mean, this is a   it's a hard problem,   but it's not impossible. Like, mere mortals can launch things into space. It's been done before. Like, sorting out the communications,   and the practicality of launching this,   out there is absolutely worth doing. And,   you know, I'm stoked that it's a bit axe. Right? And but it makes a lot of sense because it's open source. You can you can modify it to,   adapt to the the needs of operating in space of which there's a lot. But, yeah, pretty cool. Check out dyson-labs.com.  



[01:34:41] Unknown:

How are they gonna cool it?  

[01:34:44] Unknown:

So   what is it? You have you don't have convective cooling, but you have plenty of Right. Radiating   radiative cooling   in space.  

[01:34:54] Unknown:

Yeah. Because it's it's really cold in space. Right?  

[01:34:58] Unknown:

Yeah. But the sun if it's in the sun, it'll cook it.   Oh.   We when I used to work in the aerospace industry, we would model, like,   rotating the upper stage like a rotisserie chicken one degree   per second.   So the deep space facing side is, like, negative 300 degrees, and the solar facing side is, like, 300 degrees.  

[01:35:22] Unknown:

That's crazy.  

[01:35:23] Unknown:

Yeah.  

[01:35:25] Unknown:

Interesting. So you could, like, maybe shade it with your solar panel or something. Right?  

[01:35:30] Unknown:

Yeah. It was actually a big problem for all the small electronic boxes that controlled the upper stage. That's what I was formerly doing.   A lot of them would have resistive heaters in them   even, if they were in a   a shadowed spot or   yeah. You would have to add resistive heaters to a lot of these,   radio frequency   boxes for communications to keep them from freezing.  

[01:35:54] Unknown:

Just fail under that extreme cold.  

[01:35:56] Unknown:

Yep. What's better than a resistive heater?   A Hashrate heater. There you go. Boom. It just came in with a new application. Just all these fancy satellites just put some Bitcoin miners on there to keep their   their shaded  

[01:36:12] Unknown:

Totally. Yeah. It is a tough problem to solve too because you don't really want I mean, you can do conduction too. So when things are touching, like, heat will just transfer to the the thing it's touching.   But you don't really want, like, fluids, liquid cooling. I mean, you want less moving parts. Right?   And so you you prefer passive cooling,   passive thermal management, heating and cooling because they're both concerns. That's gonna be really interesting to see what they do for the BID AXE.  

[01:36:37] Unknown:

What is, like,   the optimum way to, like, radiate heat?   It's not like a heat sink or is it a heat sink? I mean, what do you have here? Surface area and,  

[01:36:46] Unknown:

the surface finish.   Okay. So every every material has what's called an absorptivity   and an emissivity, so how much it absorbs   radiant energy and how much it emits it. And so that ratio, like your a over   e, you wanna dial that in, and   that will help define, like, what your steady state temperature is in a solar position, in a shadowed position.  

[01:37:09] Unknown:

So you can  

[01:37:12] Unknown:

Really expensive paint.  

[01:37:14] Unknown:

Yeah. Get that really expensive paint. Yeah. Put it in the shade of your solar panel.  

[01:37:19] Unknown:

White paint. Yeah. Solar white.   Black is the best emitter,   but it also has a really high absorptivity.   So it's all material science. Pretty cool stuff, honestly.  

[01:37:32] Unknown:

Gotta keep in the shadows.  

[01:37:35] Unknown:

But none it's so funny because none of it really applies here on Earth   because   convection, the air, we live in Right. On land, and there's air all around us or water all around us, and so those drive   the thermals. The radiation's a tiny portion.  

[01:37:53] Unknown:

That's wild.  

[01:37:54] Unknown:

Yeah. We were talking about, like, radiation hardened,   electronics,  

[01:37:59] Unknown:

because because that's the biggest concern. Right?  

[01:38:01] Unknown:

Well,   yeah. I mean, you don't have the atmosphere to block all those nasty particles,   from deep space and the sun, etcetera,   from getting into your CPU and radio and flipping bits randomly.   So I think it is it is a hard problem.   I did work with a company that launched some CubeSats, and their strategy, I thought, was kinda crazy. But they were like,   fuck it. We don't care. Like, these things are cheap. We're just launching them. We're not using, like, exotic, rad hardened CPUs. We just use regular STM 30 twos and blast them up there. And they   they launched satellites, and I communicated with them. And they lasted for, like, four years until they fell out of orbit.  

Wow. Oh, nice. So I think if you just have you know, if you if you just expect in your firmware that you might get   flipped bits   randomly all over the place and you have the ability to error correct or to just restart and deal with it and it doesn't, like, completely break your,   your processor, then,   yeah, that's a cheaper way to deal with it. Deal with it in software.  

[01:39:02] Unknown:

Right.   Yeah. That's the way I'd go about it for a project like this with the BitX. I mean, it automatically   can restart and tries doing what it was already doing. It's not critical to, like, the mission.   Yeah. You're not navigating a spacecraft.  

[01:39:16] Unknown:

Yeah. Are they gonna, like, put the bid x payload, like, on a on a starship,   like, or a a SpaceX ship   and launch it up there? Or, like, how are they getting it to space?  

[01:39:31] Unknown:

From what I understand,   there's a there's a market for,   rides to space. Like, you just there's brokers and you, you know, you have just the requirements you need and, like, where it needs to go in space. And   there's various   launch providers, and you can book out space on their vehicles   that are, you know, assuming the rocket's going to where you need to go and they have space for your your satellite. And I think if you're in that CubeSat form factor,   it helps too because it's a very standardized,   form factor and the the way it,   like, shoots out of the rocket into orbit.  



[01:40:06] Unknown:

Wow. I had no idea.  

[01:40:10] Unknown:

I think you can get something in space for, you know, relatively low cost, like a little cube set. I don't know. Like a million bucks or something.   Oh, yeah. Just a million bucks. Yeah. Just a million bucks. No big deal. Dude, come on. Really?   I mean, it's very weight sensitive. That's why I asked earlier, like, how much does GPU cost? Like, it's it's it's very, very weight sensitive. So there's,   I think, like, one of the biggest optimizations you wanna do is just make this thing light.  

[01:40:37] Unknown:

Right.   We wish them success.  

[01:40:42] Unknown:

Yeah.   Yeah. Get that bid x up there. And they're they're looking for, you know, it's   expensive. So they're looking for investors and, partners on this. So, yeah, hit them up if you're if you're interested in being part of this. Was that the same dude that was in the Atlanta  

[01:40:59] Unknown:

BitLabs hackathon?   No. I don't think so. Oh, it's a separate okay. I think. I think.   Because there was definitely a submission that was a CubeSat with a bit ax on it.  

[01:41:13] Unknown:

Man, was it the same guy? I don't know. There's someone else working on it  

[01:41:17] Unknown:

that I've seen on Twitter. I've met him once too. I'm forgetting his name.  

[01:41:27] Unknown:

Alright, guys. I gotta, jump off to another call, but,  

[01:41:32] Unknown:

it was   fun times ripping with you guys. It's been fun. Good rip. This concludes   pod 50 pod two fifty six episode 91.   It is 03:50   Nashville time. We'll see you all next time. Let's go.  

[01:41:48] Unknown:

Cheers.  

[01:41:50] Unknown:

Okay. Don't hang