0:03

There's just so much to do

0:05

just in terms of understanding the

0:07

code better and and fortifying it.

0:10

Sometimes you need to. Invasively

0:12

change something in order to

0:14

make it more skeptical. So

0:16

it's not. Don't ever change the code you

0:19

know be be. Conservative and not

0:21

try to move fast and break

0:23

things. A

0:26

D and Happy Wednesday Welcome to the

0:28

walk Because the podcast people to buy

0:30

the massive Legends of Hair and formerly

0:32

known as Paris Energy Our energies and

0:34

then next generation data centers are the

0:36

future of Bitcoin mining and Ai using

0:38

one hundred percent renewable energy are in

0:40

remains the same business with the same

0:43

goals. Now we're just a different name

0:45

I'm his People will go much the

0:47

day we have the amazing Glorious out

0:49

on the. Show. And I'm

0:51

at the end of last year. I

0:53

was lucky enough to spend some time

0:55

to galore out in Africa, and since

0:57

then we've been trying to get on

0:59

the show. Gloria is a Bitcoin core

1:02

maintain and developer and this in t

1:04

we get into the importance of continually

1:06

upgrading the protocol as opposed ossification, ditching

1:08

Googled wasn't Bitcoin. Order news and then

1:10

post Man and The Power of Bitcoin

1:12

Core Maintain as. Really enjoys in double

1:14

Cross. She's amazing. Do we get into this? How

1:16

do we like to? Well. As ever

1:19

can any questions or feedback hit me

1:21

up his hello what Bitcoin did.com. One

1:27

two, three, What? Dissect

1:30

three, four, five, six, Assists

1:37

the show though as county as. I

1:41

know numbers. Lotta

1:43

numbers. You should do. A.

1:47

New one of the most important because. I

1:50

have my argument as I'm not actually. That

1:52

I'm pouring. Really? A

1:55

thing that I always push back on as the idea that I have.

1:57

some kind of power Like

2:02

I have some ability to influence Bitcoin.

2:06

Wow, we've gotten started already, haven't we? Yeah,

2:09

we went right off the

2:11

deep end. Okay, you

2:14

don't have the individual

2:16

power because of the checks and balances. Well,

2:19

because there's nothing that I can't do, there's

2:21

nothing I can do that isn't fully

2:23

public that everybody can see.

2:26

And there's nothing I can do like, other

2:29

than create code that

2:31

may or may not be run by somebody

2:34

after some amount of time after

2:37

which we make a release. Anyway, we'll get into

2:39

this. So let's set it up.

2:42

Explain to people listening what you do or

2:45

what you are. Okay, should

2:47

I just intro? I haven't started. We started

2:49

five minutes ago. Okay, so

2:53

hi, I'm Gloria. I work on Bitcoin Core.

2:55

I've been doing that for a bit more

2:57

than four years. And

3:00

I've been what's called a maintainer for almost

3:02

two now. And

3:04

that just, it means nothing other

3:06

than I'm able to

3:09

push code to the

3:11

repository in

3:14

which Bitcoin Core is hosted.

3:16

A live repository. Yes, yes.

3:18

We have a GitHub where we have

3:20

a repository of all the source code

3:22

in Bitcoin Core. And

3:25

it's developed much like many

3:27

other open source projects. And

3:31

typically open source projects

3:33

have a small number of

3:36

people who are able to have right access so

3:38

they can push code to it. And

3:41

we have, in addition

3:43

to kind of the GitHub permissions, we

3:45

have a script that checks that there's

3:48

a PGP signature on all the

3:50

commits that are pushed, all the

3:52

diff that is pushed. And

3:54

so this is all, I remember when I said this

3:56

to you, you were like, oh, how do I find

3:58

out? There is a public file. He

4:00

is a minute less. Pgp. Case

4:03

of the five people who who's signatures

4:05

are up dead for these comments that

4:07

are pushed to the supreme start see

4:09

one of the five Yes I am.

4:12

However, another thing that I say. Is.

4:16

That we have all these dependencies,

4:18

right? So people, On

4:20

the Austin focus on like. The Bitcoin

4:22

core repository but they don't talk about

4:25

like the Lid at P Maintain or

4:27

as to the people who can plus

4:29

Code Suda Cryptography Library which is very

4:32

very important parts of how signatures are

4:34

treated and verified within the quality a

4:36

sense or naturals. But I did.

4:38

All this is moot because my

4:41

a man's is yes I I

4:43

recognize that there is. Like. announcers

4:45

I guess privileges with being able

4:47

to push coat of course. On,

4:50

but there is nothing we

4:52

can do that isn't fully

4:55

publicly verifiable and. Like.

4:57

There's so many steps. Between us

4:59

pushing code and it's being

5:01

eventually run on the network

5:03

by a note operator and

5:05

may be late in coming.

5:08

A widely used. On the

5:10

thing on the network. so slates. That's

5:12

my arguments. Citizens. Of

5:16

use it you know the new Star Wars

5:18

with the council to generalize usa months now

5:20

in several her and sit as well as

5:22

I can assure you photo was i think

5:25

of you the maintain as as like the

5:27

council said eyes the. Boarding.

5:29

Overall, Bitcoin network has some.

5:31

Ice Okay, sorry, I think we

5:34

understand the code very well. And

5:36

we are may be we spent a

5:38

lot. Of time thinking about the security,

5:40

his shoes and maybe we know things.

5:42

That about what has put bugs

5:45

have been found and how they

5:47

than parts ah but like all

5:49

this is public's. Okay,

5:52

But can you push something? live on

5:54

your own or or the he sighs

5:56

if you have to agree some so

5:59

we it's. You don't do it as

6:01

like a committee. Style is so

6:03

yes, I can push without.

6:06

Like. Asking the other ones for premise

6:08

and however I'm like we can play

6:10

out the scenario in which I pushed

6:12

something that everybody agrees with. What

6:15

would happen is it would be reverted

6:18

pretty much immediately. and I mean depending

6:20

on how crazy it was. Like they

6:22

can also remove my key and spits.

6:24

It all happens. usually a. Very long

6:27

time before the software's actually released and

6:29

put out in the public for people

6:31

to download. Of course of their and

6:33

someone wants to download that they can.

6:36

But like. What? Has happened in

6:38

the past As you we've had

6:40

situations where maintain our has done

6:42

something very crazy and. If

6:45

they they were as Gavin oh my long

6:47

time ago, various Islam as and so like.

6:50

Is a be were worried about

6:52

like all what happens if maintain

6:55

her unilaterally look merges like a

6:57

consensus change or something. That

6:59

doesn't mean that the software. Cast on

7:01

Lied on the network Immediately

7:03

right. There's many steps between

7:05

code being, post code being

7:07

released, code. Being. Downloaded by individual

7:09

the network then be coming after

7:11

you know and. He. Asked his

7:14

answer your question yes I can but.

7:16

Like there are, like he said,

7:18

checks and balances are many reasons

7:21

why. That doesn't mean that I have

7:23

any time does unilateral power or anything.

7:25

So she says, you. If

7:28

he were being sneaky, And. The

7:30

rest of a counselor July's all asleep and

7:32

you push something crazy. Live from. Is.

7:34

Funny, rather be law, they will

7:36

be checking the code. yeah after foreign

7:39

airlines. Typically. This.

7:41

Would just be. This. The become

7:43

since on a drama on social media

7:45

has gone crazy shit on this crazy

7:48

shit yeah and I'm people would download

7:50

that I went to Philadelphia and this

7:52

is hop as has happened in the

7:54

past straight so it like I'm not

7:56

doesn't like theorizing like this, fantasy worlds

7:58

like. Meal we have this. The

8:00

robot open source can has

8:02

community and thought that verifies

8:04

everything. So as that that's

8:06

my point. Did

8:09

you study computer science? I did. Yeah,

8:11

I went to school for computer Science

8:13

and Berkeley. And it

8:15

was is great at it. I loved

8:17

it there and I think those agree

8:19

on a sense I'm and. At

8:22

the.where I found decline so. At

8:25

a nose already told me about this when we

8:27

ring. And. Malawi

8:29

Easier You told me about a conversation

8:31

with your parents. Oh

8:33

yeah, I'm. Yeah.

8:36

I think. So.

8:39

I've been doing this. That was this.

8:41

This was in like twenty twenty. Or

8:43

so I'm where I think

8:45

kind of the grant programs.

8:48

Word Times new bring didn't exist.

8:50

Yeah, I saw the first. Grantee

8:53

a prank on which is a non

8:55

profit that. To at

8:57

that collect donations from the Bitcoin

8:59

community and give them to Bitcoin

9:01

core developers on.in A was founded

9:04

basically. Out year and I think you

9:06

interviewed on an for the Us. But

9:08

basically when I first. Got started, it

9:10

was it. It was kind of more

9:12

status quo what tends to happen with

9:15

open source and so a lot of

9:17

the internet and a lot of the

9:19

things that we use id say basis

9:21

or else on open source software which

9:24

like someone wrote for free and then

9:26

put on the internet under some open

9:28

source license. Or like our you're free

9:30

to use a whatever on. Allen's.

9:33

You. Know Austin. There's like use these

9:35

bills on open source. Software as

9:37

let the person who wrote like a prisoner

9:39

of the code is not is not paid

9:41

for and they may or may not so

9:44

be maintaining at I'm a lot of times.

9:46

They are still maintaining and are doing all this work

9:48

to kind of support the features that users. Ask

9:50

for us to push an like

9:53

security fixes to like updates for

9:55

their dependencies that you know operating

9:57

system changes or compiler changes. Or

10:00

library in a updates You also have

10:02

to. Like. Up with us

10:04

updates and as downstream. Projects

10:07

and often times that

10:09

just. Like nobody pays

10:11

for because. You. Know,

10:13

I guess our culture is not really

10:15

a by it's. You we usually it's

10:17

and. I guess this is like a

10:19

social commentary thing like we usually wait

10:22

for an intermediary to stop and and

10:24

like exploit the producer and and extract

10:26

rent from the that the consumer ends.

10:29

Sorry. To. Say

10:31

that, that's on. You

10:34

know, typically the donation model hasn't sent

10:36

very sustainable in the past but that's

10:38

kind of the only the only way

10:41

to do it with like Bitcoin core

10:43

ah like Bitcoin open source protocol Stars

10:45

to this on the to do a

10:47

half million dollar salary assists Had a

10:50

brand new Tesla yeah so I'm around

10:52

the same time I found that claim

10:54

eyes ah I was applying to intern

10:57

said say I ended up sign of

10:59

doing an internship google and I was

11:01

gonna do on with. With facebook but

11:04

I or night on that's. So

11:06

cool by at the I had a

11:08

fantastic time and google to be to

11:10

be clear I'm and learned a lot

11:13

and learn about like. Cloud.

11:15

Micro Services architecture, Cooper Net

11:17

Ease, terraform, whatever and it

11:19

was great time. ah but

11:21

I decline. Was like this

11:23

per se. It's. Like.

11:26

Balance between. Like something that

11:28

was trying to solve a problem

11:30

in the world that I cared about

11:32

and also very very technologically interesting

11:34

on so I think like within the

11:36

first few weeks I was looking

11:38

I like the Glencore on his enema

11:41

we're it was It was like

11:43

oh there's like this you know either

11:45

for thing about thread safety because is

11:48

a multi threaded saying and all

11:50

that. I learn about that and class.

11:52

Arm or. A I

11:54

think around that. Time Peer Peer will. I

11:56

had written that like many sketch library

11:58

which is like this. Fancy mass

12:00

and all that all my abstract algebra

12:03

class like text books like comes in

12:05

handy of like a first. Timer

12:07

often or since excited the

12:09

class and illness. Really cool.

12:12

A Cryptography Distributed Systems like

12:14

all. These. Things and and plate. To.

12:17

Will is a great example and a lot of other people.

12:19

That I work with who have been coding for.

12:21

As long as I've been alive or in of these.

12:24

Amazing. Like a computer Scientists and

12:26

Engineers kind of about the at

12:28

least the highest level of this

12:31

cross that I've seen I'm working

12:33

on. This cool open source like.

12:35

Censorship resistant money. thanks. That

12:38

is solving a real problem for the world's. I

12:41

I personally think that it is.

12:43

A super underrated project for software engineers

12:45

that that more like more people said

12:47

consider it something more attractive to work

12:50

on then like Google I'm I mean

12:52

I was a personal decision of mine

12:54

and my part of it yeah out

12:56

or else I usually wear. My parents

12:58

were like. Why?

13:01

Why would you? so? Do you

13:03

mean you gotta work on this?

13:05

You're getting a grant from

13:07

Human Rights on Display. What

13:09

does that? Have to do

13:11

it for you know, regeneration

13:13

senses and six. Yeah, yeah,

13:16

but you know, My.

13:19

Time In. I. At

13:21

us another angle to it is like and

13:23

so I'm so privileged by I got the

13:25

spacing. Computer science has a case and I

13:27

have parents that were supportive of this to

13:29

says and. And now I get to work on. This.

13:32

Really cool piece of software and

13:34

I feel really, really lucky. To the

13:37

else to do that, to the understand and now.

13:39

I'm I've I've done likes of. Flip

13:42

Chart Sometimes when I go home and I

13:44

feel like the flip chart said I was.

13:47

In the garage or his legs and

13:49

for a bitcoin was oh yeah they

13:52

do. They get a bit more now.

13:55

My mom. Who

13:58

like. Trades. There. A

14:00

retirement account is like buying a

14:02

quantity of stuff addresses. So since

14:04

his yeah about mom to know

14:07

how the nance Alaska could be

14:09

wrote yeah I mean so ah.

14:12

I have some. Opinions

14:15

about this. Where

14:17

I saints. Yeah,

14:20

I think state. Ideologically

14:22

why we're getting into this Rs:

14:24

us ideologues yourself. Custody Is it

14:27

like that's the point rates? I

14:29

think the average person though, is

14:31

probably less good. south coast cities

14:34

in the humor, my dad, my

14:36

desk, or so much interest in

14:38

Bitcoin because of me. But.

14:40

Only does this took. The president is

14:42

like the season be zone the as

14:45

he tried to bonds of costly pretend

14:47

his looser. Seattle. Chance?

14:49

Yeah. So I'm really I'm really

14:51

hesitant to like. Make

14:53

recommendations to people that

14:55

require quite a bit

14:58

as technological literacy and

15:00

like paranoid assists in

15:03

our so. At.

15:05

This is Not a say that I'm. Indoors saying like

15:07

you know using these exchanges that are

15:10

that may or may not be reliable

15:12

a lake or these their i'm like

15:14

dude a lot It's yeah, I'm not

15:17

at like a this. I'm not trying

15:19

to advise people to do things that

15:21

are risky is this is Suzie My

15:24

mom's. How

15:26

unique is Bitcoin is? Uncertain Terms

15:28

of Computer Engineering. Oh

15:31

yeah, I mean so. To

15:34

explain this the other day a

15:36

to a lot of software engineers

15:38

I guess outward it differently and

15:40

but there's There's this crazy security

15:43

model so bitcoin is kind of

15:45

designed to work and the presence

15:47

of adversaries, right? This is like

15:49

Twitter has their own lingo to

15:52

describe this season. You know it's

15:54

a hard. You know, and

15:56

and it's supposed to be able

15:58

to light Continued. To produce. The off

16:00

continue to process payments, etc. even

16:03

if people are. Am.

16:06

I talking at and the software

16:08

that we re is supposed to.

16:10

Be like we're trying to

16:12

go for decentralization. Re starts

16:14

like a ideological like. That's

16:16

how the network brings the

16:18

valley that that it has.

16:20

Is that as a decentralized network

16:23

that is run by lots of

16:25

different and cities and organizations rates,

16:27

but what that looks like? Practice

16:29

from a code reading sound

16:31

points means that like. I

16:35

write code. That. It

16:37

is. Hopefully. Usable by

16:39

anybody, right? So ugly. Support

16:41

like so many performs Windows

16:43

Mac Linux whenever we. And.

16:46

Terms of Quantity Small complexity Like

16:48

we don't want to only write

16:50

code. That can be run on like a

16:52

to be us servers or like huge bc think

16:54

like it should be run a bull as like

16:57

a baton task on your laptop lot. Of people

16:59

run their full nodes that way. it

17:01

should work on a raspberry Pi. I

17:03

think that's a very good benchmarks where

17:05

you can run a full node that

17:07

really transactions can do Vietnamese and can

17:09

you know and you all these like

17:11

things for you on this tiny little

17:13

raspberry Pi box which a lot of

17:15

people who are bitcoin users do. Arm

17:17

Button Those two things. Are. Really

17:20

difficult to balance on and

17:22

so like. We're in this very

17:24

narrow space of operation where and

17:26

yeah, anyone should be able to

17:28

run this. but then when you

17:30

turn on. The queen t on your

17:32

laptop or whatever. What it's doing is

17:34

is connecting to randoms on the networks. Ah,

17:36

we don't know what their identity is by

17:38

design right? We want? we want it to

17:41

be an anonymous or sit on that works

17:43

at it like tries to download data, tries

17:45

to find the main. Chain It tries

17:47

to download blocks or like sit find

17:49

out about new transactions or addresses of

17:51

nodes to connect to, etc. And

17:54

that's. That's what that

17:57

means is. We have to plan

17:59

for like a. I'm attacker at.

18:01

Connecting to nodes and sending

18:03

you like garbage data.like for

18:05

example can crusher. Node or cause you

18:07

know to stall or cause you know to run

18:10

out of memory. And these are

18:12

all. Attacks that

18:14

we would consider like the higher than

18:16

one of the highest severity I'm ends.

18:19

As I bring this up because. Basically

18:22

people always ask me like still to

18:24

do or like where you work on

18:26

my isn't took the code financing also

18:29

says he I and I tried to

18:31

get this message is like there's a

18:33

lot of like he knows things that

18:36

can go wrong right? The zebra like

18:38

but sixty one percent of talks are

18:40

not economically viable and it's impossible to

18:42

guy a private key from a public

18:45

keys like glory. Don't watch this Andrea

18:47

Sons Nautilus video where he explain why

18:49

Bitcoins is the cure Alls I guess.

18:52

I agree. I preach Lights I'm

18:54

talking about like there are a

18:56

lot of there's like a million

18:58

ways that. And we

19:00

all envision like the future. For Big

19:02

Claim where we wanted to be used. Millions

19:05

of people, and I'm only

19:07

saying sir are a lot

19:09

of things I can undermine

19:11

that, including. For example the networking

19:13

attacks that I was talking about on

19:15

so if somebody is able to cross

19:17

nodes or like for to a minor

19:19

for example is able to cross the

19:21

other miners and they get had star

19:23

on like the next few blocks or.

19:25

Let's say it will tear a lot

19:28

about block censorship attacks. So like you

19:30

produce a block if somebody can prevent

19:32

everybody else from hearing about that blocker

19:34

downloading up that block from you know

19:36

that is or that is a really

19:38

serious problems. Or if you've heard of

19:41

Eclipse attacks were you know that Again,

19:43

like the whole point is we had

19:45

the spirit of her structure where everyone

19:47

can talk to whoever and lights hopefully

19:49

find a good set of tears that

19:51

is able to tell them about the

19:54

main chains is that is. Compromised and

19:56

Like Somebody is able to eclipse you as

19:58

and like be every that you're connected to.

20:00

and you don't hear about the main chains

20:02

are you don't. Hear about the

20:05

competing miners like blogs. These

20:07

are all really. I.

20:09

Mean it's are not the same thing

20:11

as. For I mean if a miner's

20:13

able to do this that's like your needs

20:15

of the one percent house race in order

20:18

to attack these people. Rates are you just

20:20

block of from hearing about all the other

20:22

has rates but as a theoretical attorneys are

20:24

theoretical as hawks just just city clear sir.

20:26

and so if if is a theoretical attempt

20:28

to try an. Attempt to do the

20:30

attacks are you know how to defend against that. ah

20:33

much at was and practice what

20:35

were awesome doing as were like

20:37

analyzing the code and were writing

20:39

tests. Were us like we're

20:41

trying to isolate the block

20:44

download logic to just throw

20:46

random stuff like just like.

20:49

Tests at much as possible. Like all

20:51

the scenarios that we can possibly happen

20:53

and you know we have lots of

20:56

monitoring, knows where where seems like you

20:58

know if we observe. Oh wow, there

21:00

is actually someone trying to send us

21:02

mutated block further. Someone's trying to stall.

21:05

Block download by like I'm being

21:07

the first guy and then like

21:09

sending a star beds and then

21:11

like when we ask them. For.

21:14

You know the correct information. They just like

21:16

you know, stall at it as. Of

21:18

the lights on. There

21:21

are like theoretical attacks. And then there's also

21:23

like will observe people doing bird stars

21:25

on and not will sometimes lead us

21:27

to go and tweak those parameters are

21:29

be like oh yeah maybe we should.

21:33

You know we we should redesign this

21:35

a little bit so that's where making

21:37

a better trade off for like far

21:40

as I'm rambling by the way to

21:42

assess assess so. With, I think

21:44

a lot about transaction really.

21:46

there was one a couple

21:48

years ago that was a

21:51

transaction. Stalling were. Ah,

21:54

Let's say. You. Know

21:56

there's a transaction being broadcast most evil know

21:58

about as I don't want. Donald a

22:00

transaction multiple times only. Stuff on at

22:02

once by. The. We have these kind of

22:04

and with saving same again Madison is important

22:07

because not everybody has extremely high wife I

22:09

see is really wanted to be accessible to

22:11

ah notes on so we're trying to save

22:13

bound with by only downloading was so we're

22:16

only gonna ask. One pure at

22:18

a time to send us a transaction. However

22:20

if we try the first percent of the

22:23

first pure I mean and they don't send

22:25

it to us we should try death rate

22:27

or we should like and and we shouldn't

22:30

only try to times for example because we

22:32

don't want there to be a that a

22:34

a censorship doctor were ill appear as i

22:36

okay i'll be the three the first three

22:39

people that you ask for the transaction from

22:41

and then all of on your knock when

22:43

hear about the transactions on that would be.

22:46

Transaction Really censorship problems and so

22:49

yeah, there was said. Basically.

22:51

They're of trend is a tiered surplus

22:53

tennis reconsideration and of might there was

22:56

like modular is a sin plus fixing

22:58

up some of those things to make

23:00

it more robust and make it really

23:02

easy to reason about and then added

23:05

a bunch of tests to kind of

23:07

check our understanding of like oh all

23:09

these in all these scenarios yes we

23:12

would and of downloading that transactions and

23:14

then we has what are. You

23:16

been using something a lot. I don't know if you've

23:19

heard of fussing, so I ramble. More and

23:21

more about success. So

23:23

okay. So typically the waves are a

23:25

software engineer might write tests is for

23:27

example, you're right. The tests scenario that

23:29

you that I just described right where

23:32

you're like okay, let's spin up another

23:34

like a fake nodes and then we'll.

23:36

Stand up a few extra said

23:38

nodes to connect to it and

23:40

then will simulate the attack and

23:42

will assert that like the correct

23:44

behavior is happening on and so

23:46

you have to provide exactly what

23:48

the testers for saying are automatically.

23:51

Generates tests. Am so

23:53

what you'll do as you're right. Kind of

23:55

more of a skeleton arm. And leg,

23:57

yeah, It'll second variance like it.

24:00

After each. State track. like after

24:02

each period of time you check that

24:04

all the invariance are still true on

24:06

it. Whatever those may be and then

24:08

or something and and will kind of

24:11

intelligently. it's just like it'll It'll start

24:13

by treat it like sending random stuff

24:15

rights and then it'll figure out based

24:17

on for example, coverage of the code.

24:19

like it'll look at what lines of

24:22

code his head and as they're like

24:24

oh these imparts. Like. Hit

24:27

like a further of line of

24:29

code deeper into the modules. okay

24:32

like work in a create more

24:34

and puts that look like that

24:36

and so it basically. it's similar

24:38

to like creating random and puts

24:40

but it is somewhat intelligence and

24:42

it'll. Like over time to

24:45

like signed. The pads that you

24:47

may be wouldn't have thought of if

24:49

you were just manually writing the tests

24:51

to use as all the time I'm

24:53

now like were like size everything as

24:55

episodes as and like the amount of.

24:57

It's full coverage over the source

24:59

of the source code has increased

25:01

like dramatically over the past few

25:04

years and it sounds like pretty

25:06

severe bugs. Some it's helped

25:08

design a certain certain things as

25:10

well when when you're like trying

25:12

to test the computational complexity and

25:15

you're. Like okay let's let's slight.

25:17

Pause. It and seat and of what the

25:19

average runtime of it is or of the

25:21

more fuzzy and we'll see if the father

25:23

can come up with an implant that causes

25:25

it to spin. Like for a really really

25:27

long time and we're. Like okay, it's

25:30

like instead of just theoretically

25:32

analyzing what's the worst case

25:34

runtime of this algorithm site.

25:36

Maybe. They have we missed something or

25:39

like what does that actually look like

25:41

or like? can we odds you know

25:43

as a extra piece of logic in

25:45

there that like catches these really like

25:47

worst case behaviors or skill like it's

25:49

it's helped design. Things Health tests

25:51

exists Buzzing scrapes I think most

25:53

people with some species on me

25:56

when we think about bitcoin. Development.

25:59

We didn't exist. Oh,

26:01

and scale was a new order for

26:03

the new features com in rinsing. many

26:05

of us tend to think about how

26:08

much was Dawn. Planning.

26:10

For adversarial attacks or efficiency.

26:14

And. And how how is that? kind

26:16

of was Philip amongst core developers? Is

26:18

it someone specifically as was more attacks

26:21

someone's on efficiencies to do tend to

26:23

have like specialists in the terrorists. I'm

26:25

so there's no can is. Court. A

26:27

nice and are not as yes of course but.

26:29

There are people who

26:32

specialize and so. I'm

26:35

just going to sell my colleague Nicholas

26:37

who works with me up. Frank. He

26:39

He's the fussing guy and he's kind

26:41

of become our security engineer. As well

26:43

because naturally he becomes. The guy that

26:45

finds all the bugs assess as

26:47

and he he's grades am I

26:49

it it? Is. My

26:53

other colleagues at Michael. Ford who

26:55

I think you've met or thank

26:57

weights that he. Especially as

26:59

and build system sauce and not also naturally

27:01

leads him to do a lot of the

27:03

release process saw us and we can talk

27:05

about the releases or at least process as

27:07

walk that kind of very interesting and then

27:09

people will call me like the Memphis person.

27:11

I'm not the only person our from them

27:13

pool definitely. But like

27:15

in I spent a lot of time

27:17

this area and typically as somebody in

27:19

opens and I'm full pr and you

27:22

ask me to review I will. I

27:24

it's an easy chows. she

27:26

works on wallets. But.

27:29

I mean everyone does a bit of everything I

27:31

think am and it would be very boring to

27:33

spend all your time doing one thing but I

27:35

think there are a in always We leave naturally

27:37

than like okay this person's really good at this

27:39

as an. Awesome as expertise and

27:42

and I think part of our

27:44

job is to just know how

27:46

stuff works. I'm asking us really

27:49

important because everything is so complex

27:51

and it's so lax. Security oriented

27:54

pretty some interesting that without coordination

27:56

everything just come against. Don't. I

28:02

mean I guess this is a big

28:04

what I'm saying is Bitcoin works. There

28:07

are people on these upgrades and six

28:09

and things patent things move on new

28:11

ideas sector the happens because. In

28:13

a decentralized uncoordinated way because of you

28:15

think of as of about voting software

28:18

business bump british who googled when have

28:20

are thus I used to pick what

28:22

you want to work on and when

28:24

you ready just uploaded and will check

28:27

in say would be entirely coordinated. Yeah,

28:29

yeah, I think. I'd

28:32

I would just say saw a lot

28:34

of the people. That work on

28:36

the concourse so like ultra

28:39

set and want to bees

28:41

fall on and. Like.

28:45

I. Agree, there is a little bit of

28:47

like oh, you do whatever you want

28:49

and therefore. Like nobody wants to review back.

28:51

Put like nobody wants to do the boring

28:53

stuff says the bomb as. Legs

28:56

release process they are so

28:58

I guess one As I

29:00

said backwards I'm as organ

29:02

to it and so as

29:05

part as maintenance. Right will

29:07

release something and then bath

29:09

has like two years of

29:11

for be stopped maintaining s

29:13

so every sorry every six

29:15

months. More. More or less

29:17

or release a new major version of

29:19

the software. I'm and not have new

29:21

features like the to transport or it

29:24

relay or whatever. And

29:27

then six months later will release a new one.

29:29

Right? But the one that we released before

29:31

we're still pushing security updates to the will

29:33

will push out minor minor. Really says

29:35

I'm and so basically anything that

29:37

place is usually about six. Will

29:39

get merged into Master with our

29:41

development branch and then we'll back

29:44

port that to the the branch

29:46

that contains a release that's currently

29:48

out and then we'll cut another

29:50

minor released a way. This this

29:52

thing where if you're just revealing

29:54

the same shit or on a

29:56

different branch. Is incredibly boring and

29:58

most people don't do it. The

30:00

end of the process of cooking, that

30:02

of the process of choosing which bugs.

30:04

Sexist. Back for it as as

30:06

well is. Like. Something that's

30:08

ah, takes a lot of work

30:10

because you basically have to have

30:12

a birds eye view of every

30:14

single thing that's happening. Like.

30:17

In. His as being were

30:19

smart lights like this. hundreds of prc

30:21

have to decide oh see I start

30:23

something that with a sex like a

30:26

previous release and then. You. Know

30:28

all you do is like you

30:30

do cherry picker answer on said

30:32

the the older branch and like

30:34

nobody wants to do assists. And

30:36

and the end of factors like yeah, that thing that. We

30:39

released before it still was. Answer

30:41

is. Yes

30:43

yeah and so if you if you

30:45

know of treating your know software isn't

30:48

you eventually have things on sports as

30:50

yes yeah because one is that remains

30:52

is crazy remaining due to refuse to

30:54

go anything beyond the original version one

30:57

of the gone. So. I

30:59

I think I heard something like

31:01

die and I think they're Siegel

31:03

who ran like zeroed i was

31:05

sixteen or whatever and slight people.

31:07

It's people who say like law,

31:09

let's say theoretically riot like it's

31:11

a soft forks. So I don't need

31:13

to update to the new rules and

31:16

I should still be able to use

31:18

that Be no, that's a very as

31:20

is ideologically a very important thing to

31:22

us. And on for I would say. Unfortunately,

31:26

And the consensus on jan is

31:28

intertwined. with all the other pieces of

31:31

of the coin core but it does.

31:33

And so you aren't really able to

31:35

be like I want the modern version

31:37

with ah you know, the modern wallet

31:40

you I and all those things. but

31:42

I don't want to upgrade. my consensus

31:44

and gents is a I think it

31:46

is not good that these are kind

31:48

of intertwined and we have. A projects

31:51

that we've been working on for years

31:53

and years. I'm at to kind of

31:55

modularized the will be called a Colonel

31:57

or the Consensus and and so that.

32:00

Or you could get the new to each

32:03

other. New wallet features are the new kind

32:05

of more on. The front say

32:07

saying like not consensus you

32:09

know, not protocol. Changes.

32:14

Without. In A Changing. Colonel

32:18

Am. So.

32:20

My point is. Don't run

32:22

on Maintain. Software It contains bugs that

32:24

we do know about and has may

32:27

or may not have been disclosed. Already.

32:29

It's so you know that there

32:31

are like it's public knowledge that.

32:34

Been. Like someone can send. You x

32:37

and then you might lot of memory

32:39

or someone might pressure note or sundance.

32:42

I very much do not recommend running

32:44

on maintain software as a citizen Scientists

32:47

to the entire as if Bitcoin what

32:49

from in cool or if you were

32:51

Bricklin usury download or is it reference

32:53

and other software things that exist else

32:56

article. Yeah, so we

32:58

have. This came up a lot with

33:00

the recent might have see you tools

33:02

like sour sauce a bug. So

33:05

yeah, we do have a

33:08

season dependencies and we have

33:10

were very. Very very conservative as

33:12

about ads and one of the things

33:14

again like an like that Michael does

33:16

at as part of built the Some

33:19

as he. He's also very about like

33:21

reducing the amount of. Dependencies and

33:23

so like what? For.

33:26

Example: Michael has a sound over over the

33:28

time as the he's. Like oh, we

33:30

have a dependency and nobody's maintaining

33:32

its or alive and. It

33:35

is they. They're just like gathered on

33:38

maintaining it. Basically, or like is not

33:40

reviewed to the same standard that we

33:42

would expect a Bitcoin core code to

33:44

be reviewed to Arm. and now that

33:46

code is. Like. You should treat

33:48

it as it equivalent of of your

33:50

your. Code. Where with a

33:53

dependency sits on it wouldn't it

33:55

would be another open source library

33:57

Onda hub and would that be

33:59

nice? Analyzing force.

34:02

On that. So like when I say

34:04

it's an open slates civilian force as

34:06

an essay, delete it is. So

34:09

we just have. A. Copy of it right? and

34:11

we would just use our own. and I

34:13

am a lot. Would you want everything to

34:15

sing? Cool said his node. External dependencies or

34:18

sudden? not possible. Ah,

34:20

I'm sorry, I just. I. Got

34:22

it? So dependencies wise. First of all that

34:24

I was talking about like other code that

34:27

gets pulled in when you build that claim

34:29

no one of our stay home when is

34:31

operating. It like operating systems like the like

34:33

you have any to not brings. A suitable

34:35

for did not notify me once Home Decor

34:38

does it for the world has ever have

34:40

to refer lived all the system sources of

34:42

work completely on the phone. Oh

34:44

I'm ah as like. You

34:47

load! Lakers see library

34:50

when when you run a split

34:52

I'm I guess it's hard certs.

34:55

What? Are we talking about here? Like is sick

34:57

when he mean by refer to others as

34:59

I got on. There are the. Software.

35:02

Particles are a relies on for

35:04

is everything within cool cool. It

35:07

needs to it needs internet yeah

35:09

like that's another that I know

35:11

how like myself last I'm on

35:13

about to have to maintain like

35:16

another protocol and that's a sold

35:18

on a server a W s

35:20

somewhere that has to refer to

35:22

be. oh. I

35:25

I'd say there is an example of

35:27

thought in the Dns seeds. Answer: when

35:29

you start up there node and how

35:31

do you know who to connect to?

35:33

Honey know addresses? We can't put that

35:35

directly or we have a little bit

35:37

of thought directly and source code but

35:39

usually that changes day to day rights

35:41

so you. Will talk to

35:43

answer some Dns Cedars. Say

35:46

between five. And in aid of I'm

35:49

Five Defense Of. Them I'm that run

35:51

by various people and they'll be

35:53

like oh yeah. I've been scraping

35:55

the bitcoin network for for addresses

35:57

and here's a list of addresses.

36:00

That are very diverse.

36:02

Their the represents. Various

36:04

network side in the world or whatever.

36:06

And and then you know her ears

36:08

City of a Murderer and you'll talk

36:10

to a the Bees and then you

36:13

know York City of starts acting. So

36:15

that is an example of Alive saying

36:17

that you're connecting to. I would say

36:19

there's very little of that compared to.

36:22

Most. Most. Things that you

36:24

might think those. Are it's the

36:26

took, the Are and who are next

36:29

generation, the data's and business power. In

36:31

the future? Bitcoin mining, an Ai using

36:33

one hundred percent renewable energies of these

36:35

guys? Not done in. I've been working

36:37

with the Sound Down and Will for

36:40

quite some time over a year now

36:42

with a super impressed with a Valleys

36:44

especially their commitment to local communities and

36:46

sustainable computing power and we collaborated with

36:49

them on everything we made. seals are

36:51

dumb podcasts, the even sponsor my. So.

36:53

Between Robin Hood and leveling sponsor

36:56

for a conference Chico's now. Iran

36:58

is the response. Always power bitcoin

37:00

mining. A I am beyond and

37:02

we will find out more about

37:04

them. Please do head over to

37:06

Iran which is irpn.com. The

37:08

some about the casinos, pioneers and bitcoin

37:10

game in Sao they were established in

37:13

two thousand and Thirteen and because he

37:15

knew at that point. Became.

37:17

The beacon of innovation and game and

37:19

and have been ever since. They have

37:21

trusted by tens of thousands worldwide and

37:23

better seen a office cotton as security.

37:25

Of course they support the best money

37:27

in the world. They support Bitcoin. Now.

37:30

With the casino you get. Lightning. Fast

37:33

withdrawals. Vip experience and they have

37:35

over four thousand came. but they

37:37

also believe in a fair game

37:39

and of arm and offer no

37:41

transaction fees. They promote responsible gambling.

37:44

and their power players with their live on

37:46

t p feature to and and five bonus

37:48

and bears games for i to come to

37:50

find it more join be to see no

37:52

please head over to because in a to

37:54

iowa where top most game a mess on

37:57

a core service at is be you know

37:59

the i would be i t G-A-S-I-N-O dot

38:02

I-O. And also please remember to

38:04

gamble responsibly. When

38:07

you split up your node for your first time and it's

38:09

trying to find other nodes to talk to, is

38:11

there a limit? It's like you only need to be connected

38:14

to five or six and once you've, is

38:16

that like the, that's enough to create the

38:18

network effect or can it connect to all of them?

38:20

How does that kind of stuff work? I

38:22

don't know any of this. So

38:25

there's a couple of design goals

38:28

here, right? One is bandwidth,

38:30

right? You can't connect to every

38:32

single person. First of

38:34

all, you'll probably, like your computer probably can't

38:36

handle it. And also this,

38:38

yeah, it's just not feasible. Because it

38:41

means it's taking a constant communication. There's

38:43

like tens of thousands, right? I'm

38:46

pretty sure you just like, you have a

38:48

limited number of slots essentially that you could,

38:50

connections that your computer could be making. Or,

38:54

sorry, I'm trying to like make

38:56

it easy to understand. It's

38:59

just not, like it's not gonna be feasible

39:01

to connect to like, let's say 30,000 like nodes,

39:03

right? But

39:06

the other design goal is like, you definitely

39:08

really, really want to find, first

39:10

of all, honest peers, as in peers that

39:12

are gonna tell you about the main chain

39:14

as soon as they know about it. So

39:16

that you can find the most proof of

39:18

work chain, right? So like we understand proof

39:21

of work consensus, right? Like you find the

39:23

chain that has the most proof of work

39:25

on it and that's your idea of what

39:27

the chain is. But you have

39:29

to hear about it. So

39:32

yeah, what

39:34

nodes will typically try to do is maintain

39:37

like eight to 10, yeah, let's

39:40

just say, eight

39:43

to 10 like outbound

39:45

peers that you connected to

39:49

and are hopefully diverse and hopefully

39:51

very responsive and telling you about

39:53

main chain. So We'll collect statistics about everyone

39:55

that we're connected to. We Have When's the last

39:57

time they sent me a block? When's the last

39:59

time? The family a transaction that was solid

40:02

ones like how fast do they pong

40:04

when I send them a paints Like all

40:06

these metrics we used to assess like how

40:08

good our peers are and then when you

40:10

know when we're managing all of the the

40:13

addresses of the notes that we know about

40:15

will also try. To go for diversity

40:17

right? So it it is very

40:19

easy for examples for someone to

40:21

spin have a bunch of Ip

40:23

addresses from like one. I.

40:26

Speak or never on, but it's much harder to

40:28

do that across like a lot of different Isis

40:30

answer that so I dare you not like we

40:32

have all these ways of managing. Ah, I'm. Managing

40:35

ways for us to make sure that

40:37

we are. Connected

40:39

to a healthy set of peers

40:42

and so yeah will like regularly

40:44

monitor like all regularly trial at

40:46

Oculus. like pick a random one

40:48

from address book and like see

40:50

a stop here is gonna respond

40:52

to us and. And

40:55

like gifts. If we have liked that one of our

40:57

peers has not sent us anything for. You know

40:59

days of lean on nada points by

41:01

that you know is is the city

41:03

fears like was x them and war

41:05

and will try any once on. And

41:08

so yeah, there's there's so much

41:11

that goes into the software. That

41:13

is Not just. You. Know sort

41:15

of fun modes group of work, chain

41:17

or legs as the signature Valid. And

41:20

a very complex as very interesting very

41:23

very interesting I think like I didn't

41:25

I didn't know about this when i

41:27

was taking a that working class and

41:29

I'm in college and uni. Ends.

41:33

Sorry it's a cool i'm I'm would

41:35

have to motivate share with each other.

41:38

Is it is the primary information

41:40

be said to me than there

41:42

were words, transactions and blocks. Soon.

41:44

Or I'm at. While I would

41:46

say it's I primarily. Transactions, blocks

41:49

and addresses of nodes

41:51

suggest on. All. There's.

41:54

A death. Because. Once again, sort of

41:56

humans with this. Most people. Probably

41:58

have no idea how bitcoin one. Her.

42:01

To the little people don't even download in the

42:04

hood for a small censored people who are very

42:06

nice if I run. A know you'll learn

42:08

so much about their client itself. Fine And

42:10

it's so easy. It is so so easy.

42:12

To run a known make a brain blaze

42:14

yeah we you don't need like eat. Only

42:17

money like six hundred gigabytes. Everybody like this

42:19

is a myth and like a very common

42:21

method. people think you need six hundred gigabytes

42:23

to to store the block chain a meal

42:25

in his. Thirties are set by

42:27

Guidoni outlets. It can on

42:29

it on a laptop. But.

42:32

The primary thing is addresses. Transactions.

42:35

And blocks Yeah. And so when to try exemptions

42:37

created that is relays across the whole networks? Yes

42:39

is didn't ever be a time when. When.

42:42

My know just doesn't pick up that

42:44

trans Hudson will eventually get that reads

42:46

as so we're talking about transactions I

42:48

haven't and mind and yet aka cracked.

42:51

The Ah so I'm yeah, this is

42:53

like. When I love talking about the most.

42:56

Simple Air

42:58

Ssssss. So the

43:00

question was it or you're guaranteed to

43:03

hear about hands are awfully eyes as

43:05

a as some some things that I.

43:08

Basically. The process that I work on

43:10

a lot of them and of censor around

43:12

transactions. Answer sets. I'm on various different levels.

43:15

I kind of described a peer to peer

43:17

example, but there's also an employer. Examples: And

43:19

there's also lights. Don't design your things in

43:21

a way. That

43:24

I'm You know it. It

43:26

can get sniped or whatever and it at the

43:29

sorry their it can happen animals as well. I'm.

43:32

But. The the goal is like

43:35

we're trying to build a

43:37

censorship resistant payment. Network where

43:39

anyone can pay each other. And

43:41

nobody can block that. I think this is

43:44

like. Top Three. Best.

43:46

Things about Bitcoin of things that we

43:48

should strive for and but clients on

43:50

ends and three i you aren't guaranteed

43:52

so so I talk about like mumble

43:54

stars yeah all again and say look

43:56

if if if I phone and pick

43:58

it up some some. Won't be supernova.

44:01

See every minute constructing block soon

44:03

as he's every transaction Butts. It's.

44:06

All their chances that a mine have

44:08

seen transactions. it's the bus or should

44:10

the meant to be the same? Should

44:12

everybody have the same month since? Thank

44:14

you for for asking that question. This

44:17

has been our topic that a lot

44:19

of people talk about. I'm. Obviously.

44:21

A bit quain. Could.

44:24

Sunk sense as a

44:26

okay so. There was

44:28

a lot that was sad to kind

44:31

of teach everyone that not everybody has

44:33

might have the same them for there's

44:35

no such thing as the Man for

44:37

riots. There's. Each note on the network. Risk

44:40

that has their own i'm fool and they may or may

44:42

not be the same. I think

44:44

when things are working really well,

44:46

hopefully everyone. Like. Does and

44:49

I'll explain why I'm so.

44:52

Like. We have this

44:54

decentralized transaction really network each

44:56

node has. An I'm Reason with

44:58

them and bullet okay subject is it

45:01

The music loses his own interests. And.

45:03

Occupant of the for me as if they got

45:05

this all load on how you to talk about

45:08

was glorious over the bus. I guess the

45:10

as sorry as okay. So let's

45:12

start from the beginning. We want

45:14

censorship resistant payments. The way that

45:16

we achieve that is through the

45:18

decentralized. Structure of the network right? We

45:20

don't just has you know the minor

45:22

has like a website that he said

45:24

nets payments to that would not be

45:27

sense or it's the government's website works.

45:29

or a bank right? That's that. That's

45:31

obvious it isn't Hopefully or bandstand that

45:33

the saw that works. I'm so what

45:35

we have is decentralized network. Instead you

45:37

can add to one of the tens

45:40

of thousands or hundreds of thousands of

45:42

nodes you send your peers, your transact

45:44

sense and they cost a bit and

45:46

eventually it gets to everybody. That's how

45:48

it's supposed to work. On how we

45:50

implement thought and process is each. Node

45:52

first of all as down to hear about

45:54

transactions and. Validate them and will put

45:57

them into their memory pool or man

45:59

pool and that's. The cache of

46:01

unconfirmed transactions is just a cast.

46:03

And you spam access and therefore

46:06

the echo Good good question. Says

46:10

it's an okay. Yeah

46:12

so you as a note operator why

46:14

do why do you do this one

46:16

is so that you can toss everything

46:18

for one block like the one of

46:21

the reasons is so that block relay

46:23

works so when you receive a block

46:25

on what you what might happen if

46:27

season. See any transaction before. His

46:29

you have this huge networks are

46:31

like bound with Spike where you're

46:33

downloading the whole block and then

46:35

you're like spinning, spending, spending your

46:37

like loading Cgx those front desk

46:39

and you're like by doing the

46:41

computationally heavy saying of verifying all

46:43

the signature the center. it's very

46:45

com additional expenses and like it

46:48

might take you a while to

46:50

do this. so that's one. Possible.

46:53

Reality and the or possibly ah yes,

46:55

I just already downloaded all these transactions

46:58

already verify them. I have the cash

47:00

where I like Castile, The signature is

47:02

that a child's I'm already loaded. All

47:04

these he excels in some memories and

47:06

I'm done and I store it on

47:09

the next block. Bread and not is

47:11

a much much healthier for small situation

47:13

for you as the node operator but

47:15

also for the network because every single

47:17

hop is so much faster. Sorry so

47:20

block latency goes from like this to

47:22

this because each hot. Like. Multiplies that

47:24

he is hop rates on. And.

47:26

So the healthiest network the ones with

47:29

the most the least three org swear

47:31

it like reorder bad because the for

47:33

example he won't block latency to be

47:35

very very fast because it's to people

47:38

find the next block at around the

47:40

same time hopefully everybody. Hears about one.

47:43

Like. Quickly. Otherwise

47:45

you might end up with you know half the

47:47

network knows about the Spock and for the other

47:50

half know about this blocks and then at some.

47:52

Point: You have to resolve That rewards.

47:54

The fewer of those you have, the

47:56

better. When I was into the album. I

47:58

think like once a week. Compared to

48:01

regular. That

48:03

we have every thousand box or something?

48:05

Yeah, yeah. I'm but it used to

48:07

be. More frequent and every time

48:10

that happens. Like every time there's a

48:12

real hard like it's it's messy. It

48:14

means you are confirmed. Payment may

48:16

or may not still be confirmed.

48:19

Legacy you I issue. it's like as

48:21

it is it's a usability as yeah

48:23

right us. So as a security issue

48:25

it can be at least And so

48:28

we. Really, we really want to minimize

48:30

that as much as possible. So anyway,

48:32

we're talking about life. Mumbled introduction

48:34

really are useful for block relax.

48:38

And. And the

48:40

other uses obviously so that we

48:42

can have a nice censorship resistant

48:45

payment systems. And.

48:49

When. We are participating in

48:51

transaction relay. This is that.

48:54

We. Don't just like validate whether

48:56

the transactions consensus valid. Jokinen

48:59

Mum for. And and continue

49:01

to answer your question will. Shoot

49:03

Us would translate can use down the match

49:05

your mom for. On yeah, there

49:08

are consensus valid transactions. I can

49:10

take you several minutes to validate

49:12

spray and they may or may.

49:14

Not be. A

49:16

valid transaction. Red lights. There's

49:19

this really great post I think by Rusty

49:21

Russell on which. Is from a long time

49:23

ago a non sequitur on fox and with

49:25

like a bunch of them putts do so

49:27

like quadratics that tossing. It took like

49:29

several minutes to validate. You can imagine

49:32

that being like a state transaction right.human?

49:34

a block which had proof of work

49:36

on at and for who worked as

49:38

a psych. Really easy way to avoid

49:40

spam because it's so much harder to

49:42

produce of proof of work solid blocks,

49:44

that it is to like download and

49:47

validate at i'm whereas transactions you don't

49:49

have a group of work requirements so

49:51

it's much more like people and just

49:53

send you this transactions maybe spend minutes

49:55

validating it and then it through those

49:57

be invalid completely useless not even going

50:00

be confirmed, right? Thanks. Thanks,

50:02

right. And

50:04

so we have rules where we're like,

50:07

okay, we don't take really large transactions.

50:10

Like, that's a very simple rule. So

50:12

consensus rule, you could have like a 400, 400,000

50:14

like way in it, transaction, but we make

50:21

that 100,000 in in

50:23

transaction relay. Another

50:26

really good example is we

50:29

use, like we want our mempool,

50:31

it's a cache, right? We want it to be useful.

50:34

And so we use a lot of

50:36

incentive compatibility metrics to assess whether or

50:39

not this transaction is useful. So the

50:41

best example is if you get two

50:43

transactions that conflict with each other, they

50:45

spend the same UTXO, they're double spends,

50:48

you should probably pick the one that's

50:50

higher fee rate, because

50:52

that is more incentive compatible for a minor

50:54

to mine. Even if you're not a minor,

50:56

you like you're expecting

50:59

to use this for when you're validating

51:01

blocks, you're like, okay, this is more likely to be

51:03

mined. And therefore, I'm going to take it.

51:07

So that's like, that's

51:10

another example of like, what

51:12

we do in transaction validation. And that's what leads

51:14

to being able to RBF to replace by fee,

51:16

right to fee bump. Sorry,

51:21

I'm getting to your question about

51:23

should everybody have the same mempool? Like

51:26

hopefully, in an ideal world,

51:28

things work best if everybody has the same

51:30

mempool. Of course, like all these rules

51:32

that I described, they're not consensus. So

51:34

you can have whatever you want, right? Like

51:36

let's say you're like,

51:39

I mean, the example is like, I

51:42

hate inscriptions, or like,

51:44

my mempool, like doesn't

51:46

allow spending on confirmed

51:48

UTXOs or my mempool, like you can do whatever you

51:50

want. And that, yes, like, that's

51:53

allowed. But it's

51:56

not, but like people are like, therefore, everybody

51:58

should have different mempools. It's like, well, no.

52:00

no. No. That's

52:02

not where we were going. On but

52:04

I think it was like kind of this over

52:06

correction of like people used to think there was

52:08

the men pool that you would submit. Transactions

52:11

to. And then now everybody's and

52:13

all. Actually everybody has their own man pool,

52:15

all the members of a different look. I

52:17

used think this isn't his face of

52:19

things there was just like one member

52:22

of the soil the question other our

52:24

side dependencies because I imagine there was

52:26

a dismal of transactions when he realized

52:28

the transactions are self again relayed as

52:31

towards know when we go to the

52:33

member website is able to they see

52:35

as that just their own. May.

52:37

I that's awesome are a number. Faith

52:40

and man for different. Yeah, there's like

52:42

ma'am form the memory. Pool that each

52:44

node has yeah and then numbers. That

52:46

is an answer that to theoretical that's

52:48

just. One,

52:51

What. It should be. Ah,

52:53

I'm. Not necessarily they

52:56

think this from so

52:58

the I'm. My. Aunt. I

53:01

mean, they run. Their own oakland know yet

53:03

what? My senses? yes. But therefore, this

53:05

could theoretically be different from us. Yes,

53:07

Yes, absolutely so. anyway run a few

53:10

different nodes first of all and so

53:12

they They have like of. A

53:14

huge know that has like all

53:16

the transactions. And then they have like

53:18

the default node or it was a three

53:20

hundred megabytes and so like when see rates

53:23

are really Hitler's that a ton of term

53:25

fox sense like the huge man pool will

53:27

have all of them. But for example, your

53:29

three hundred megabytes sample is gonna start per

53:32

Jane the. Lower see raised us that.

53:34

Is less likely to get mine.

53:36

Were using incentive compatibility to assess.

53:38

Once and when they get worse, where did they go

53:40

to? He's a strong when. They're gone from

53:42

mine, node for you or node or just from

53:44

away, but a minor would never throw any way

53:46

to the some point they might my mom's. Ice

53:49

I would imagine that they do something very similar

53:51

to what mental space as much as they have

53:53

a huge know that just keeps track of all

53:55

of them. but they have like a default. Know

53:57

because the larger your mom for the more

53:59

hobbies. Like value of everything gets. I like

54:01

building a new block template. for example, he

54:03

wanted it out on a smaller mom for

54:05

of a huge one is just the store

54:08

and all the stuff later. It's like okay

54:10

if we go through a very intensive period

54:12

whether that's a low value transactions a problem

54:14

for later and then see rates dropped them

54:16

them bring wasn't yeah exactly so much to

54:18

the. Still yeah, yeah

54:20

exactly. And so I think some part

54:22

of like Them Full Face as a

54:24

school audit. seats are right where they

54:26

print, like they build their own the

54:28

Park tumbler and they compared to what

54:31

got mind and you can see like

54:33

or this enough stonily censorship like there's

54:35

There's many reasons. Why the transactions

54:37

might be different but it like

54:39

you can kind of see what.

54:42

What? it would be are like what

54:45

the difference was but. Of course liked your

54:47

node might be different from ample space is known

54:49

him you know and that may be different from

54:51

the minute node or whatever. I'm. And.

54:53

So it's like very difficult. Third,

54:56

It's very difficult prescriber reason why turn

54:59

box and might be different because i

55:01

have opposed to transaction where the fees

55:03

too low and it never get mind

55:05

and just sits there and. Forever

55:08

for is like everybody's bidding

55:11

higher than here for box

55:13

space than sorry that he

55:15

saw them and amusement centers

55:17

and eventually just. To. Cancel

55:19

cause who must have do as

55:21

to somehow case the myself to

55:23

I bring by myself as I

55:25

would say I'll need encapsulates. I

55:28

think once you find that term

55:30

fox and the for example minor

55:32

could just be logging. Every single

55:34

transaction that they see and then like

55:36

six months later you thought it was

55:39

council and and maybe are like paying.

55:41

Your friend or a merchant or something and

55:43

you're like oh, Kansas and any transaction ounce.

55:45

Cans Old Lady my

55:47

your mind sets of.

55:49

Lives are as if you bring your back

55:51

is really where you three with no no

55:53

as a this and sorry sir event will

55:56

you could. double spend at gray

55:58

this is still not guaranteed

56:00

that someone might not mine it.

56:03

It's a consensus valid transaction. You've

56:05

already broadcasted it. Can I increase the

56:07

fee? Yeah, yeah.

56:10

So yeah, so you could like, I described

56:12

before, you could create an RBF where you

56:15

double spend it in a higher fee rate

56:17

transaction. And theoretically a miner

56:19

should mine the higher fee rate one. But

56:21

again, they could decide if they

56:24

wanted to, if they're being weird, like they

56:26

didn't hear about, like for some reason they

56:28

didn't hear about the fee bump or

56:31

by the time you fee bumped it, they had already

56:33

started working on a block with that

56:35

transaction in it, for example. These

56:37

are all potential reasons why like

56:40

your RBF may or may not get in,

56:43

may or may not happen. There's a

56:45

lot of reasons why they're replaced by

56:47

fee, transaction may or may not work.

56:50

One of those, like some of those reasons, because

56:52

I've spent a lot of time looking at kind of

56:55

the limitations of RBF, is

56:58

just like our policy rules are not very smart.

57:01

So we make a trade off again

57:04

between like being resistant

57:06

to denial of service, like trying

57:08

to be computationally not

57:10

complex, as well as

57:13

incentive compatibility, right? So sometimes

57:15

we make a trade off where

57:17

we take a shortcut or we

57:19

have an imperfect heuristic for assessing

57:21

incentive compatibility, because the actual thing

57:23

would be too computationally complex to

57:25

do. And so

57:27

we might look at two transactions and we'll be

57:29

like, oh, this one's better to mine, but actually

57:32

this one is. And

57:34

so we might reject a replacement, that's

57:36

actually better, or we might accept

57:38

a replacement, that's not better. And

57:41

like, this is, I

57:43

consider a bug. It's a

57:45

very complicated bug because it's

57:48

the result of kind of a lot of engineering

57:51

and these are very like kind of

57:53

edgy edge cases. But

57:56

yeah, there are, I've spent a lot of

57:58

time essentially working on

58:00

these kinds of replace by fee policy

58:02

limitations. And we have a few proposals.

58:05

I have some proposals as well as other people.

58:08

I don't speak on behalf of anyone other than myself.

58:12

There are a lot of really exciting projects

58:14

that people are working on to try to fix

58:17

these things in mempool. So how

58:19

has inscriptions changed

58:21

the mempool game? Because

58:24

clearly it has. So

58:28

I remember we had like a

58:30

three round trip debate

58:34

where I think

58:36

it was around like people

58:38

were really upset that the fees were

58:40

so high because of all the I think they are C20

58:42

stuff. And

58:45

they were like, can you fix this? And

58:48

I was like, you're kind of like

58:51

somebody who's driving down the highway and

58:53

there's traffic and you're like, road workers.

58:55

Why did you allow all these people?

58:58

Like it's like, yeah, other people are

59:00

allowed to are allowed

59:02

to send Bitcoin transactions. And

59:05

the way that they're chosen

59:07

for confirmation is based on

59:09

fee rate because miners are

59:11

economically incentivized to include those

59:13

transactions. The block. Yeah.

59:16

Yeah. I mean, they use the scarce

59:18

block space to a demand

59:21

that's very high and choosing economic.

59:23

Right. And the price goes up

59:25

for a unit of block space, which is the

59:27

resource that you're trying to consume. But

59:29

this is political. Yeah. And

59:34

do you feel like as a developer or

59:36

maintainer either or do you feel

59:38

like you have a duty to not be

59:41

political, not have an opinion

59:43

as long as your job

59:45

is just to make sure the software works? Yeah.

59:49

So the

59:54

issue that I had when we were talking

59:56

about this was that people

59:58

are like, can you stop the spam.

1:00:02

And spam has a very specific,

1:00:04

like we've just been talking about

1:00:06

the definition of spam when

1:00:08

we're talking about mempool code and

1:00:10

transaction relay code, where

1:00:12

it's using a

1:00:15

resource, for example, your

1:00:17

computer's computation

1:00:19

or memory, that it

1:00:21

shouldn't be. And we

1:00:23

have all kinds

1:00:25

of ways of preventing that.

1:00:31

And so we have a, I'll have a definition of spam that

1:00:34

I use when I'm working, when I'm thinking

1:00:36

about this kind of code. And

1:00:38

people are asking me to apply

1:00:40

that definition based on use case.

1:00:44

And I don't think it's the

1:00:46

protocol developer, the transaction relay

1:00:48

code's place to decide what

1:00:52

use case is legitimate or not. Like

1:00:55

it, like, you know what I'm saying? Can happen

1:00:57

at a node or

1:00:59

even minor level like Ocean of Dunt, you can, that's

1:01:03

where you can make that decision.

1:01:05

Because in the decentralized network, you

1:01:07

get to decide yourself. Yeah, like

1:01:09

you, like if you don't

1:01:12

like cat photos, you don't like

1:01:14

wizards or whatever, that's like your

1:01:16

choice, right? But I don't think

1:01:18

this is not a legitimate transaction

1:01:21

or, you know, this is a

1:01:23

waste of blocks, you know, because NFTs

1:01:26

are bad or whatever. I don't think

1:01:28

that language has a place.

1:01:30

Well, I don't think that should be

1:01:32

considered when you are talking

1:01:35

about writing policy code. Yeah.

1:01:37

As a

1:01:40

developer. Yeah, as a developer, as

1:01:42

a protocol maintainer, like, it's been such a

1:01:44

tricky subject to navigate. And honestly, I've gone

1:01:46

back and forth on this. I've ultimately come

1:01:48

to the point that my

1:01:51

preference would be Bitcoin was just

1:01:53

used for financial transactions. Sure.

1:01:56

Of Bitcoin being money. And

1:01:58

I know someone said, well, the these still are. But

1:02:00

I've also I don't

1:02:03

like the idea of censorship. But

1:02:05

in a very long conversation with Bitcoin

1:02:07

mechanic, obviously he's very much involved in

1:02:09

ocean. I agree

1:02:12

with him. Like the thing I care about, the thing

1:02:14

I'm here for is the improvement of money, especially

1:02:17

with the kind of things we saw out in

1:02:19

Malawi and Kenya and Ghana. Yeah,

1:02:21

I have zero interest in in

1:02:24

quantum cats. I feel

1:02:26

like some of that some of

1:02:28

this doesn't attack on Bitcoin. But

1:02:31

I I've come

1:02:33

to really appreciate that it puts you in

1:02:35

a difficult position because people want you to

1:02:37

have a political view. Do you feel

1:02:40

that as a pressure? Yes.

1:02:44

Well, I don't know if people are

1:02:46

asking me to have a view so

1:02:48

much as they're demanding that we change

1:02:50

the software and write. And

1:02:54

I think let's to be kind of

1:02:56

fair to argument. We've identified

1:02:59

a couple of reasons why you would

1:03:01

have policy. But policy is in the

1:03:03

validation rules in mempool that aren't consensus.

1:03:06

And historically, there have been

1:03:09

kind of things put

1:03:11

in place to discourage

1:03:14

things that are bad

1:03:19

for network resources. So for example,

1:03:21

we have standard output types and

1:03:23

we have a dust limit and

1:03:25

not to try to protect or

1:03:28

try to discourage creating

1:03:30

a lot of and even the witness

1:03:32

discount as well as kind of to

1:03:35

try to discourage bloating of the UTXO

1:03:37

set. What is a witness discount? Oh,

1:03:41

no, we're getting into it. OK,

1:03:43

so let me just

1:03:45

finish that idea first. So I

1:03:48

talked about how you can run

1:03:50

a fully validating Bitcoin

1:03:52

node and you don't need to

1:03:54

store the entire block chain. You just need

1:03:57

to store the UTXO set all

1:03:59

of the UTXO. and that is enough

1:04:01

for you to validate whether each

1:04:03

transaction and each block is valid. Well,

1:04:06

you also need the header's chain. You

1:04:08

don't need all, you can prune most of it. And

1:04:12

so the size of UTXO set is

1:04:14

kind of very important to scaling and

1:04:17

to the accessibility of running a node and all those

1:04:19

things. And

1:04:22

so, what were

1:04:25

we saying? Oh, witness discount. So when we

1:04:27

are looking at the size

1:04:29

of a transaction, certain

1:04:31

parts of the transaction are weighted

1:04:34

differently from others. And

1:04:36

so the witness data, which is like

1:04:39

the signatures or the scripts that make

1:04:41

this transaction valid, which

1:04:44

can be discarded, like

1:04:47

that is

1:04:50

weighted less. That's

1:04:53

all the witness discount is. And

1:04:57

so that means that if

1:04:59

you make a large witness,

1:05:01

that does not cost, that won't

1:05:04

cost you as much in fees, for example, as

1:05:07

if you made a lot of outputs

1:05:09

and outputs are what go into the UTXO

1:05:11

set. And

1:05:13

similarly, we have a list of standard

1:05:17

output types. So

1:05:19

that's pay2witnesspubkeyhash or

1:05:22

pay2taproot or pay2witnessscriptash

1:05:24

or those like

1:05:26

standard types. And they're very

1:05:29

small and they have a standard

1:05:31

format. And say, okay,

1:05:33

we don't have to get into the death

1:05:35

limit. But these are, they're like very

1:05:37

important. What is the death limit? What's the actual

1:05:39

value? Sure. Oh, it depends

1:05:41

on the output type. So I just described

1:05:43

the different types. So I don't

1:05:45

know if I should explain what they are. But

1:05:48

if you create an output that has like one

1:05:50

Satoshi in it, well,

1:05:54

you're not like, if you are to spend that,

1:05:57

it's going to cost you more in fees to

1:05:59

spend that. UTXO then it's

1:06:01

worth. You know what I'm

1:06:03

saying? Unless you've got a special

1:06:05

Satoshi, the first Satoshi on the block.

1:06:07

Oh sure, sure, sure. That's

1:06:11

bullshit right, you can't identify the individual Satoshi's.

1:06:16

Well that's what Ordinals is right,

1:06:18

like you have an external... But

1:06:20

they're like, they've just basically made

1:06:22

up their own number and system

1:06:24

but these Satoshi's don't exist individually.

1:06:27

Because a UTXO is

1:06:29

essentially a... Like if I put a 20

1:06:32

pound note on the table, there aren't individual pennies

1:06:34

in there, it's just a 20 pound note. You

1:06:36

can't identify the individual pennies in that note. Sure,

1:06:39

sure. I guess we can kind

1:06:41

of... No, you don't need to do that. I

1:06:44

think, yeah, like you can like

1:06:46

do anything to derive meaning that

1:06:48

isn't kind of baked into the

1:06:50

protocol. So the Ordinal Theory is

1:06:52

something that is an external protocol

1:06:54

outside of Bitcoin. Yeah, that has nothing to do

1:06:56

with it. Like when people tell me to stop doing Ordinals, we

1:06:58

don't have the Ordinals. Fuck

1:07:01

it. Yeah. So

1:07:04

the dust limit, does that have

1:07:06

to be a kind of a moving beast as Bitcoin

1:07:09

value changes? Or is

1:07:11

it irrelevant? Because it's priced in Satoshi's? It's

1:07:14

priced, it's determined based on

1:07:16

like Satoshi units, right? So

1:07:18

if you assume that use,

1:07:21

let's say, you're always

1:07:23

going to have to have at least one sat per V

1:07:25

by H, the dust limit is like a three, but

1:07:27

like once that per V by H on

1:07:30

any transaction order to spend it, then you

1:07:32

can't be making like... If you make a

1:07:34

five sat output, then it's uneconomical

1:07:40

to create that output because it's

1:07:42

worth less than what it would

1:07:44

cost to use it. All

1:07:47

right, let's talk about Ledger, my

1:07:49

favorite hardware wallet in the world, the people I've been

1:07:51

using since 2017. Now

1:07:53

Ledger is a world leader in Bitcoin security, and

1:07:55

is one of the best ways for you to

1:07:58

own and secure your private keys. Now listen. If

1:08:00

you are still holding your Bitcoin on an exchange with

1:08:02

a custodian, what are you doing? I've been telling you

1:08:05

for years to stop doing this. It

1:08:07

is time to take your security more seriously

1:08:10

Remember if it's not your keys Ain't

1:08:13

your Bitcoin now with ledger hardware wallets

1:08:15

paired with the ledger live app You have

1:08:17

the easiest way to start managing your own

1:08:19

private keys and taking control of your Bitcoin

1:08:22

With the price booming, you don't want to be

1:08:24

leaving that on exchange. We've seen what happens with exchanges, right?

1:08:27

Now you can send and sign Bitcoin transactions

1:08:30

with full transparency in the ledger live app

1:08:33

And we'll see it can be easier. I've told you

1:08:35

a million times. I'm still using the same wallet. I

1:08:37

bought back in 2017 I've still

1:08:39

got that same ledger and

1:08:42

nano s also ledger has recently

1:08:44

released their Bitcoin orange nano They've

1:08:47

got a very cool initiative going with

1:08:49

it They are given $5 from every

1:08:51

purchase to the Bitcoin developer community and

1:08:53

the start now with brink Which

1:08:55

I love to see so if you want to find

1:08:57

out more if you want to get yourself a hardware

1:08:59

wallet Please head over to ledger and it's shop dot

1:09:02

ledger comm which is sh

1:09:04

op dot LED gr.com

1:09:08

Alright, let's talk about swan and their mission

1:09:10

to help millions of people get into

1:09:12

Bitcoin and they've built the fastest

1:09:14

and easiest way to get That started it

1:09:17

takes just a matter of minutes to sign up and

1:09:19

start buying Bitcoin with the swan app For

1:09:22

introducing friends and family to Bitcoin has never

1:09:24

been easier and swan now has

1:09:26

zero fees on the first $10,000

1:09:28

of Bitcoin buys so it is

1:09:30

free to get started Don't worry If

1:09:32

you are already a swan client your

1:09:35

next $10,000 of buys are also free

1:09:37

if you love to recommend Bitcoin And

1:09:40

I know you do swan is an absolute no-brainer You

1:09:43

can take a friend from zero to Bitcoin in

1:09:45

just one quick conversation Just

1:09:47

tell them to search for swan Bitcoin in the App

1:09:49

Store It really takes no

1:09:51

time and it's completely free for them

1:09:53

to get started again search for swan

1:09:56

Bitcoin, which is SWAN

1:09:58

BIT. I am

1:10:00

to download the app. Okay,

1:10:03

Let's talk about costs us. Less.

1:10:05

Talk about cold storage. Now listen that

1:10:07

the bitcoin price has been booming this

1:10:09

last year and I know some of

1:10:11

you will see you net wealth and

1:10:13

Bitcoin flying up. You may not considered

1:10:15

forty six store and see my not

1:10:17

considered Tolstoy venus. Get onto this. I

1:10:19

have been using costs for about four

1:10:21

years now. Absolutely love it so a

1:10:23

lot more comfortable. Know what Bitcoin is

1:10:25

locked up in a multi sig solution

1:10:27

because he's a very Best Buy or

1:10:29

Thorns a bustling else They have brought

1:10:31

a product of the markets that changes

1:10:33

everything. Something I. Needed something I'll be

1:10:36

bored by for long time which is

1:10:38

inheritance and as some of you plan

1:10:40

your inheritance would you be coined by

1:10:42

pretty some kind of puzzle or sharing

1:10:44

of other people but cause of now

1:10:46

that they have build this natively into

1:10:48

their app so you can handle custody

1:10:50

and you can handle inheritance directly in

1:10:52

the app is for and is no

1:10:54

mistakes as run the people you love.

1:10:57

Will. Inherit your bitcoins to Something happened to you.

1:10:59

We. Have to do is to the recipient

1:11:01

and when he's not say when you go

1:11:03

on the person you put in charge wait

1:11:06

six months and they take control of your

1:11:08

bitcoin. Also also provides you with a helping

1:11:10

hand to make sure your bitcoin stay safe

1:11:12

to this transition so you'll find out more.

1:11:15

If. You realize it is time to

1:11:17

give yourself a good cold storage solution.

1:11:20

Please do gun check out Casa at

1:11:22

Casa de I have which is see

1:11:24

a essay.iowa. Something.

1:11:26

Like having Canada they didn't They don't make

1:11:28

the the penny anymore because more than have

1:11:30

any to make him any. Yeah.

1:11:32

The ice on and so ah

1:11:35

that leads are time at a

1:11:37

certain amount. This because this has

1:11:39

been like a very very long

1:11:42

standing rule as we had like

1:11:44

standard ness tax and policies were

1:11:46

you can't. he can't do

1:11:48

that or like if i am running that

1:11:51

number i have my them pool and i'm

1:11:53

wearing my node and i receive a transaction

1:11:55

that creates a dust output i mean angle

1:11:57

hadn't rejected even as it's consensus valid That

1:12:00

was there as, I guess, the

1:12:05

idea is to protect network resources,

1:12:07

to encourage people not to do

1:12:09

things that are going to hurt

1:12:11

the scalability and whatnot. People are

1:12:13

like, why can't you do that

1:12:15

for inscriptions? Because

1:12:17

policy can ban or

1:12:19

can prevent things.

1:12:24

So people are like, why don't we have

1:12:27

a policy rule where if we see an

1:12:29

inscription and a transaction, like the NFT thing,

1:12:33

then we also ban that in

1:12:35

policy. And this idea

1:12:37

has been dressed up in many

1:12:40

different trench coats, has

1:12:42

different pull requests to the Bitcoin

1:12:44

Core repository. Oh, do you know about this?

1:12:48

And they've become like flame

1:12:50

wars on GitHub, which

1:12:53

is not Twitter. It's a

1:12:55

place where you discuss code. As

1:12:59

maintainers, has

1:13:01

this been something that's been debated, or is it

1:13:04

very obvious that you just don't get involved in this censorship?

1:13:08

We've talked about this because

1:13:10

I think people take pull

1:13:13

requests seriously. Anyone can

1:13:15

contribute. It's not like we're like,

1:13:18

we'll take every pull request seriously. When

1:13:21

I say we, I speak

1:13:23

on behalf of myself and not anyone else, of course.

1:13:26

But yeah, you think about it. And

1:13:28

actually, I published a really

1:13:31

long document summarizing one of

1:13:33

the iterations of

1:13:36

this proposal because

1:13:40

it had hundreds of

1:13:42

comments from various people

1:13:44

to various

1:13:47

degrees of professionalism. So

1:13:50

I tried very, very seriously to

1:13:52

index every single comment I could find

1:13:55

on it, summarize the arguments and weigh

1:13:57

the merits of the technical arguments against

1:13:59

each other. So yes, I

1:14:01

did very seriously review that PR, and

1:14:05

I can send you the doc if you want. So

1:14:11

the idea was like, why can't you do this for

1:14:13

inscriptions? I

1:14:17

think there's two things. One is

1:14:20

you can do this on your node, but

1:14:22

when you're asking, there's

1:14:26

a difference between I want to do this on

1:14:28

my node, and I want Bitcoin Core to ship

1:14:30

this as the default. And

1:14:34

so there's two things that I really

1:14:36

just want to put out there. One

1:14:38

is Bitcoin Core doesn't have the

1:14:41

ability to just censor things. We

1:14:43

can only, like I said, we

1:14:46

try to publish software that is

1:14:48

accessible to run, but also incentive

1:14:50

compatible to run. If I publish,

1:14:52

sorry, not I, but if we

1:14:54

create a piece of software

1:14:56

and we're like, hey, miners, can you please

1:14:58

run this? You'll lose like 30

1:15:01

Bitcoins in fees, but like please run

1:15:03

this. They'd be like, no. By

1:15:05

the way, did you see those fees in the

1:15:07

first few months? Yeah, that's what I mean. Like

1:15:10

35 Bitcoins? Yeah. Again, we

1:15:12

just publish code. We

1:15:16

don't force people to run it. We

1:15:19

can't ask people to run it. People voluntarily

1:15:21

download the software and then decide whether or

1:15:23

not they want to do that. So if

1:15:25

you publish software and you're like, well, if

1:15:27

you run this, you lose 30 Bitcoins in

1:15:29

fees, but like please, that's not how that

1:15:32

works. We just finished talking

1:15:34

about how everyone has optionality and what kind

1:15:36

of mempool rules that they're going to use,

1:15:38

because it's not consensus again. We

1:15:41

have so many clear examples of

1:15:44

like Marathon or F2 pool mining

1:15:46

things that would

1:15:48

have been rejected by the Bitcoin

1:15:50

Core default Zionist rule. This the out

1:15:53

of band stuff? Yeah. Yeah. What is

1:15:55

that all about? Are they constructing

1:15:57

transactions but that don't go into the...

1:16:00

They're not broadcasting them. They're just putting them as

1:16:02

part of a block. So

1:16:04

it would be in their mempool and then

1:16:06

they just don't send that to anybody else.

1:16:09

Or like a lot of times they

1:16:11

do it because they're running some

1:16:14

patch that is different from the

1:16:16

Bitcoin core like rules. Is

1:16:20

there any danger to out of band

1:16:22

transactions? So let's

1:16:25

go back to the block relay latency

1:16:27

thing that we were talking about where

1:16:30

anytime you create

1:16:33

a... Okay so this I think when

1:16:36

things are working right best

1:16:39

I mean like everybody already knows all

1:16:41

the transactions by the time you broadcast that

1:16:43

block and then like it propagates

1:16:45

like immediately right. And

1:16:47

if they construct it so if they a single

1:16:51

out of band transaction is fine if they did a

1:16:53

whole block of them and then

1:16:55

broadcast that we have a potential latency

1:16:58

issue. So actually a single one

1:17:00

makes a difference as well. So

1:17:03

kind of the way that block relay

1:17:05

usually happens nowadays is with compact blocks

1:17:08

and we have like kind

1:17:12

of direct high bandwidth

1:17:14

like first round like

1:17:16

no round trip ones as well where

1:17:18

basically I just send you like a

1:17:20

skeleton of the transactions and I

1:17:23

have like short IDs talking about all the transactions

1:17:25

in the block. And

1:17:27

if you already have it in your mempool you're like cool I've got it.

1:17:30

But if you're missing something you have to be like

1:17:32

hey Gloria can I get one of those transactions and

1:17:34

I'll be like yeah sure here you go. And that's

1:17:37

an extra round trip. And

1:17:40

so yeah like a single transaction can make a

1:17:42

difference and of course a whole block of transactions

1:17:44

then it's like oh I have to download all

1:17:46

these and I'm going to validate all of them

1:17:48

and then you know. And

1:17:51

so yeah it does make a difference.

1:17:53

Is that a point of debate that should

1:17:55

the network allow out of band transactions and

1:17:58

could they be stopped? No,

1:18:00

no, no because Like

1:18:03

it's like I mean we can like we know

1:18:05

about marathon because they offer like a service. Yeah

1:18:07

do this, right? But

1:18:10

there's no way for you to enforce like

1:18:14

You can only give me transactions that I

1:18:16

already heard about it's like

1:18:19

that's why we have blocks it's

1:18:22

you know, it's it doesn't work that way and

1:18:26

So, yeah, there wouldn't be a way to

1:18:28

like and like cuz that's why we have

1:18:30

consensus in blocks is we have rules

1:18:32

and blocks and then we're like, okay, these are

1:18:34

the orders of the transactions if we had

1:18:37

a way to do that without if

1:18:39

we had a way to be like, you know, this is

1:18:41

this is the order in which transactions are

1:18:44

received and created and whatnot then

1:18:48

This wouldn't like we wouldn't have we wouldn't be

1:18:50

doing all this, right? so

1:18:52

have inscriptions Presented

1:18:57

Will agree that you shouldn't censor them will agree

1:18:59

that you know, that's not your job at a

1:19:01

policy that you want to say something Yeah, we

1:19:04

can't like sense I don't like the word censor because we

1:19:06

don't have the ability to do that Like I just like

1:19:08

I like I just said we like you know what I

1:19:10

mean Like we can we can create

1:19:12

no we don't you have to look at why I

1:19:14

started this episode Like I thought I started like right

1:19:16

off the like I don't have the power to decide

1:19:18

what other people's mental policy We

1:19:23

can't do that and and and like

1:19:26

my other point was that like It

1:19:29

doesn't make sense. Like you can't we can't

1:19:31

write software. That's just like yeah, like just

1:19:33

throw away That's not incentive compatible. But

1:19:36

have inscriptions change

1:19:39

Change the what you do in that if it presented

1:19:41

newer tech vitals, you have to think about Like

1:19:46

the inscription transactions Just

1:19:48

the fact that they existed these BRC 20 like

1:19:52

Has that changed the way you have to work

1:19:54

or do you just see transactions and therefore it

1:19:56

doesn't matter? Mmm,

1:19:59

I would say I

1:20:03

think there's two things we can talk about. One

1:20:05

is a bug related

1:20:07

to just what happened with really

1:20:10

high transaction volume. The

1:20:12

other one is more attention to, I

1:20:15

think Mononaut calls it mempool sniping.

1:20:19

So I'll talk about the first one. This

1:20:21

is not the structure of inscriptions.

1:20:24

This has nothing to do with ordinals. It's just

1:20:27

the high volume of

1:20:29

transactions. This is just an anecdote.

1:20:35

We discovered that there were a couple

1:20:37

transaction relay, data structures that were not

1:20:39

as efficient as they could have

1:20:41

been. It probably would

1:20:43

have worked for hundreds per minute,

1:20:45

but it doesn't work for tens

1:20:48

of thousands per minute. Very good

1:20:50

thing to find that out. Yeah, exactly. I'm

1:20:52

not saying like, but

1:20:55

a lot of people are like,

1:20:57

has anything happened? I think this is a

1:20:59

positive thing that came out of that.

1:21:04

Then the other one, what

1:21:07

was I going to talk? It was the

1:21:09

mempool sniping stuff. I

1:21:13

really wanted to talk about this on a podcast. It

1:21:16

is what I have been talking

1:21:18

about for ages. I

1:21:21

was always like, okay, pinning

1:21:24

is the thing that we're going to talk about. I

1:21:27

was always like, this is really

1:21:29

bad. This is

1:21:31

especially bad for L2s and

1:21:34

Lightning Network stuff, where I'll

1:21:37

just do a shorthand one. Basically,

1:21:40

you have the same security model because you

1:21:42

can always settle on chain. You have all

1:21:45

these off-chain transactions that you're not going to

1:21:47

broadcast. Through that, you

1:21:49

get the scalability and the privacy and

1:21:51

the cool features. But the security is

1:21:53

the same because you can always settle

1:21:55

on chain. You've already enumerated all the

1:21:57

spending paths, and they include like,

1:21:59

hey, if you're going to do this, I have

1:22:01

two days to respond. And

1:22:04

I respond by broadcasting this

1:22:06

transaction that spends from it.

1:22:09

And I have two days to do that until

1:22:11

the other spending pass opens where you get

1:22:13

to redeem the money. That's usually how those

1:22:15

L2s are kind of set up, is

1:22:18

you have this critical period

1:22:20

within which you need to broadcast

1:22:22

your transactions. And

1:22:24

so my argument was always like, okay, but

1:22:26

what if, like what about the games you

1:22:29

can play in mempool where like

1:22:31

I can try to censor your transaction through

1:22:35

various means. So one of the things I was talking about was with

1:22:37

like replaced by CRBS.

1:22:39

Sometimes you might broadcast a replacement

1:22:41

or you might broadcast a transaction

1:22:43

that makes perfect sense, but because

1:22:46

of the limitations in

1:22:48

how the validation like code works,

1:22:50

like it'll just be like dumb

1:22:52

and like not realize this. Anyway,

1:22:57

oh, so with kind of all this like NFT

1:22:59

stuff, like

1:23:03

basically I would say this and people like, but does this ever

1:23:05

happen? I'd be like, well, no, but like

1:23:07

theoretically it could and you could lose money.

1:23:10

But we actually like saw what

1:23:12

people call like pinning. What

1:23:15

happened in the wild? Yeah, yeah, it happened in

1:23:17

the wild because what was it like here's

1:23:19

the 20 houses, like anyone

1:23:22

can mint, the

1:23:24

coin or whatever. And then so it was like

1:23:26

basically a race to like who can get,

1:23:28

who can do it first, right? Or

1:23:31

like, I think there was one where

1:23:34

they exploited like the

1:23:36

descendant limit thing where, actually

1:23:40

it probably doesn't make sense for me to talk about the technical

1:23:43

details, but like they basically exploited this limitation

1:23:45

where we're like, oh, we're only gonna

1:23:47

consider the first this

1:23:49

many. And then after that,

1:23:51

like it'll just like we'll limit the

1:23:53

topology of the transactions that you're broadcasting

1:23:55

because we can't handle more than this much.

1:23:59

And then so they're like, okay. I'm like you

1:24:01

know I I it's about the slots

1:24:03

and now you can each other out

1:24:05

a transaction to this on. And

1:24:08

then. That. There is another oh

1:24:10

does this is the other thing is like.

1:24:13

Ah, these like swapping things. That

1:24:15

Are you saying anyone can pay?

1:24:18

Have you heard of six? Has

1:24:20

anyone can pay? Yes. Yes,

1:24:23

So you can find your time box and.

1:24:26

So that other people can change it. As

1:24:28

they can odd inputs and. Outputs to

1:24:30

ads. On. And

1:24:32

so. What's that mean? What

1:24:35

that meant was like and man, poor

1:24:37

people could replace it with. Their own

1:24:39

version of the trans oxen were like

1:24:41

they got to buy. Be honest, he

1:24:43

instead of the original person who broadcast

1:24:46

the John Fox and. And

1:24:48

us that one is not a

1:24:50

limitation in mumble policies. that one

1:24:52

was eased. Anyone can pay and

1:24:54

that's how anyone can pay. It

1:24:56

works. And so there's a kind

1:24:58

of all these like people are

1:25:00

kind of discovering the slate. I

1:25:03

think censorship is so is the wrong. Word. But

1:25:05

like things that can go wrong between

1:25:08

the time when you broadcaster transaction and

1:25:10

it. May or may not getting

1:25:12

confirmed. I'm and I mean that's great

1:25:14

that that people Alec I think it's

1:25:16

I'm so annoyed when people are like

1:25:19

all mental should sixes and I'm like

1:25:21

you know what's happening here is people

1:25:23

are buying stupid sports car that don't

1:25:25

have locked finance and people are still

1:25:28

getting their car soul and and they're

1:25:30

like what's wrong with these roads. Are

1:25:34

these roads the ones who. Was moved

1:25:36

to Slake. Start designing your thing

1:25:38

in a way that it can

1:25:40

be stolen. So.

1:25:44

Yeah, I'm. To.

1:25:46

The answer questions. About

1:25:48

me and. a

1:25:51

do one of us vacation for the

1:25:53

come to me ah my very simple

1:25:55

typical mind always sounds dumb for town

1:25:58

can you possibly or suffice I

1:26:00

don't even know what it means to also buy one.

1:26:02

We just lock the

1:26:04

code and it can't be touched.

1:26:07

That sounds psycho because what if there's

1:26:09

a critical bug? Like my basic logic

1:26:11

says, that doesn't make any fucking sense. Why are

1:26:13

we talking about it? I

1:26:16

think probably ossification means something different to

1:26:18

many people. I agree that the one

1:26:20

where it's like, just stop touching the

1:26:23

code that doesn't make any sense. Yeah.

1:26:25

The only ossification that makes sense to me is like

1:26:27

a stalemate of things

1:26:29

just don't get done because we can't get agreement

1:26:31

to get things done. That's also

1:26:35

scary. I'm

1:26:38

talking about big significant.

1:26:40

Consensus changes. Well,

1:26:42

so there are a couple consensus

1:26:45

changes that are needed.

1:26:50

Somebody said, was it John? John

1:26:53

Carvalho said, there

1:26:55

are no hard forks. He said

1:26:58

there won't ever be a hard fork. It is

1:27:00

just a copy. Did he say it's just like a

1:27:02

copy of Bitcoin? Someone

1:27:04

will always run the original Bitcoin. Right.

1:27:06

Yeah. Yeah.

1:27:08

What is we need? What consensus changes? You're

1:27:11

talking about the one that we need for like when we're

1:27:13

all dead in like 21, 40 or whatever.

1:27:17

Like running out of time for the timestamp.

1:27:22

Yeah. And there's I think there's

1:27:25

a couple bug fixes. That would be nice

1:27:27

if you've heard of the consensus cleanup. Oh,

1:27:29

we probably shouldn't talk about that. Oh, why?

1:27:31

Because why shouldn't we? I

1:27:33

just I don't. People are

1:27:35

aware of it. I'm very. I

1:27:39

try not to say things on podcasts that

1:27:41

can be taken out of context. Like

1:27:43

if I say like we need ex software,

1:27:45

right. And then. Yeah. Yeah. And

1:27:48

sometimes things that I've said have like landed really

1:27:50

poorly on Twitter and that, you know, like I'm

1:27:52

like people will think like I'm trying to say

1:27:54

Bitcoin's broke and it's like, well, no, I wouldn't

1:27:56

be working on Bitcoin if I didn't believe in

1:27:58

it. No, you can

1:28:00

just fix it. I

1:28:03

don't have that power. How

1:28:05

is Big Boy broken? OK,

1:28:09

that's never what I'm trying to say. What

1:28:11

I'm always trying to say is there are, like

1:28:14

I said, all

1:28:17

these network level attacks, all these

1:28:19

like scalability, like all these like

1:28:22

ideas we have of scalability of

1:28:24

censorship, resistance, privacy, et cetera. These

1:28:27

are not like single lines of code that

1:28:30

Satoshi wrote in the original client. And then

1:28:32

like Bitcoin has that right there are design

1:28:34

goals and like we have to fortify them

1:28:36

over time. And like I said, with the

1:28:39

example with the transaction traffic

1:28:41

being really high, the requirements change

1:28:43

each time you go from like

1:28:45

one level to the next. And

1:28:48

nobody ever designs their software for a million

1:28:50

people in the first iteration. You can

1:28:52

build a POC. I'm

1:28:55

never trying to dig. I

1:28:57

got people thought I was hating on Satoshi

1:28:59

once, too. So

1:29:02

she's great. I'm not saying that

1:29:04

like the POC is bad. I'm

1:29:06

just saying like over time, like

1:29:09

the requirements get more stringent. And

1:29:11

literally come on my show and say Bitcoin is broken

1:29:13

and Satoshi is a dick. First

1:29:16

of all, I named my dog Satoshi

1:29:18

because he loves Satoshi. Charlie

1:29:21

Shredmegan named his boat Satoshi. Yeah,

1:29:23

well, I think I

1:29:25

have to be on that one. You

1:29:28

got the dog. OK, but there are some things

1:29:30

at some point that need doing. Ultimately,

1:29:33

though. That

1:29:35

hundreds and definitely require hard fork. I

1:29:39

wouldn't I'm not convinced.

1:29:41

Yeah, I think it's like

1:29:43

there are things where it's like, oh, we

1:29:45

probably need a hard fork, but we can

1:29:47

probably figure out a way to do it using

1:29:49

the software. Yeah, so

1:29:51

because we haven't really ever had a hard

1:29:54

fork. We

1:29:56

had no way to be in one. like

1:30:00

that was a bug fix the inflation bug or something.

1:30:04

I feel like I heard it was but like we're

1:30:06

not like Ethereum that has one every like couple of

1:30:08

weeks or... No I don't think we've had

1:30:10

like a well

1:30:12

I think maybe in the

1:30:15

very very early like prehistoric you

1:30:17

know Bitcoin days but no not

1:30:20

we hadn't had like a purposeful... And people have

1:30:22

fought away but... Yeah yeah yeah yeah and people

1:30:24

have worked away. But

1:30:28

ossification is just it is just

1:30:31

can't really happen right? I think

1:30:33

it's again it's

1:30:35

like what's the definition of

1:30:37

that. I

1:30:39

can speak to kind of my personal

1:30:42

thing that has some

1:30:44

things in common with what people say when

1:30:46

they ask for ossification which

1:30:49

is like I think we're

1:30:51

at the point where it's best to

1:30:53

seek to like understand

1:30:56

and formalize and test and

1:30:58

document and like isolate

1:31:00

and test the

1:31:03

code that we have as much

1:31:05

as possible and I'm not really

1:31:07

saying like we

1:31:09

should add a bunch of features and whatnot at least to

1:31:11

Bitcoin core. I think there's sorry

1:31:14

to Bitcoin core to like all

1:31:16

the people who are experimenting with

1:31:18

wallet UI and like trying out

1:31:20

new like no like I'm

1:31:23

not talking about that like there's there's there's

1:31:25

so many things that need to be built

1:31:27

so many you know more layers to scale

1:31:30

but in terms of like the

1:31:32

protocol I would say

1:31:35

that I'm I lean

1:31:37

towards the conservative side because

1:31:39

there's just so much to do that just

1:31:43

in terms of like understanding the code

1:31:45

better and fortifying it as I would

1:31:47

say and sometimes

1:31:50

that involves a protocol change

1:31:52

right like sometimes you need

1:31:55

to kind of like

1:31:58

invasively change something

1:32:00

in order to make it more scalable or

1:32:03

better or for whatever reason. So

1:32:08

don't ever change the code.

1:32:10

It's like be conservative and

1:32:14

not try to move fast and break

1:32:16

things. I'd say that

1:32:18

that's something I have in common with the people

1:32:20

who are like ossify. But don't touch the

1:32:22

code, I think, is quite absurd.

1:32:25

What do you think of this privacy stuff that's

1:32:27

happened this week with Samurai and Zabi?

1:32:30

I mean, it's obviously depressing. Has

1:32:33

that spiked any interest or discussion about

1:32:36

bringing in privacy to Bitcoin at the

1:32:38

protocol level? I know that comes with

1:32:40

all different complications, but is it a

1:32:42

discussion point now? There's

1:32:46

a lot of privacy built into the

1:32:48

protocol level. Do

1:32:51

you mean? I mean,

1:32:53

like, Monero level privacy

1:32:55

on transactions. So

1:32:58

I haven't personally had a conversation

1:33:00

about this with the other people who work

1:33:05

on the Quinn Core. We did talk about

1:33:07

it and how terrifying it is. There was

1:33:09

a tweet. Let me find this one for

1:33:11

you. Because this is where I actually think

1:33:13

it's particularly depressing. I didn't

1:33:16

think it was depressing at a

1:33:18

personal level because I'm a bit lazy

1:33:20

with a lot of my privacy. But

1:33:22

this one, Peter van

1:33:24

Vortenberg, she

1:33:26

retweeted Anna Chekhovich. I

1:33:30

can't pronounce her name, but

1:33:32

she said at the ACF,

1:33:34

which is Anti-Corruption, here you

1:33:36

go, the Anti-Corruption Foundation, we

1:33:39

relied on with Zabi when protecting our

1:33:41

donors from Russian government surveillance and risk

1:33:43

of imprisonment. That was not

1:33:45

taken into consideration by the US authorities

1:33:47

when they began attacking privacy talks. I

1:33:50

mean, that's particularly depressing. I

1:33:53

think that kind of goes. I'm

1:33:56

not an expert in regulation.

1:34:00

the history of what money

1:34:02

transmitters, how they're classified

1:34:04

or whatever. Yeah, I

1:34:07

think that it is really serious

1:34:09

that, like I said, there

1:34:11

are so many roadblocks that are

1:34:13

on the technical level, on the

1:34:16

regulatory level, on the marketing level. It's

1:34:20

like whenever I think about a

1:34:22

network crash bug, I feel like

1:34:24

a lot of it is like it'll be

1:34:26

a PR disaster, it'll be really

1:34:29

difficult to convince people

1:34:31

that they're quite safe. It's

1:34:33

like nodes crash and

1:34:35

also that it would be pretty bad. So

1:34:42

the question on how we

1:34:44

think about privacy, I

1:34:46

think there's

1:34:48

a lot of that. So

1:34:51

with the release that just

1:34:53

came out, like V2 transport, which

1:34:56

is encrypted connections between peers, that's

1:34:58

default now. Yeah,

1:35:04

we support running nodes on core

1:35:06

and stuff. So privacy is often

1:35:08

a consideration when adding new

1:35:11

features. Oh, and Basel is

1:35:13

working on broadcasting your

1:35:16

own transactions on tour connections to

1:35:19

try to improve transaction relay privacy

1:35:21

that way. There's

1:35:23

so much to Bitcoin. Yeah, so much.

1:35:27

Yeah. Do you feel

1:35:29

like there's large parts of Bitcoin you

1:35:31

don't even know or understand yourself? Oh,

1:35:33

yeah, yeah, definitely. Like even

1:35:35

within Bitcoin core, there's

1:35:38

all this cryptography stuff that

1:35:40

is like magic. And all the

1:35:42

time I'm like, oh, I

1:35:44

want to go explore this because this is really

1:35:47

interesting. But

1:35:50

so much of Bitcoin is so

1:35:53

cool. And that's kind of, I

1:35:55

think how a bunch of people end up

1:35:57

working on core. It's like, don't trust

1:35:59

very much. verify. How many people

1:36:01

who like scream that have actually

1:36:03

like verify the code works the way

1:36:05

that they think it does. But

1:36:08

that was like what I was doing where like now when

1:36:10

I'm around like how does this work in Bitcoin I'll be

1:36:12

like trying to look at the code to like see how

1:36:14

it works like that's a great feeling. But

1:36:17

like obviously like nobody knows everything.

1:36:20

Yeah there's

1:36:23

so much to Bitcoin and it's

1:36:25

so interesting so many interesting problems.

1:36:27

Is there enough developers. No. And

1:36:31

is that a funding problem

1:36:33

or is that a whatever number you

1:36:36

had you still say no. Oh

1:36:39

no I think I think I can imagine it

1:36:42

like a scenario in which I'll be

1:36:44

like I'm feeling good about this. But

1:36:47

the reason why

1:36:49

a reason why I come on

1:36:52

on on podcasts is

1:36:55

like ultimately Bitcoin is

1:36:57

like we're trying this experiment

1:36:59

of like completely decentralized organization or

1:37:01

governance or whatever. And at the

1:37:04

end of the day like you

1:37:06

or me or any other Bitcoiner

1:37:08

has like the same level

1:37:10

of responsibility to you know like if

1:37:12

there was a really big bug or

1:37:15

if you know Bitcoin didn't succeed the

1:37:17

way that we want it to it

1:37:19

wouldn't be because like the CTO like

1:37:22

didn't do his job or whatever it would be

1:37:24

because it would be because all of

1:37:27

us who have this responsibility didn't you

1:37:29

know do it properly. And if

1:37:32

Bitcoin fails because everybody thinks it just

1:37:34

you know boils the oceans and has

1:37:36

no interest and has no value and

1:37:38

you know helps criminals or whatever like

1:37:44

that would be like everybody's fault. And you

1:37:46

know you do amazing work

1:37:48

in evangelizing Bitcoin but that's not

1:37:51

your job or it like that's

1:37:53

not like it's not inherently like

1:37:55

Bitcoin said you know Peters are

1:37:57

CMO. It was

1:37:59

like you went. and did that. And just like

1:38:01

all the people who worked on core, it's like,

1:38:03

you know, maybe they held Bitcoin or maybe they

1:38:05

got nurse sniped in college and they're like, well,

1:38:07

I'm going to work on Bitcoin core. It's not

1:38:09

because like the Bitcoin

1:38:12

core company hired you to be the

1:38:14

CMO and, you know, this got

1:38:16

to be the CTO and like Michael

1:38:18

Taylor's the CEO or whatever. It's

1:38:21

like we're all equal

1:38:23

peers in this decentralized

1:38:26

community of Bitcoin. And

1:38:28

so, of course, naturally, like we don't have any HR

1:38:31

department. We don't have like, you

1:38:34

know, a thing that is

1:38:36

responsible for making sure there's

1:38:38

people working on the infrastructure.

1:38:40

Right. Like,

1:38:43

and so basically what

1:38:45

we have to do is we have to try to like find

1:38:48

people. And so,

1:38:50

you know, like Jonas at Chanko does

1:38:52

a lot of educational programs. There's Beatrust,

1:38:54

there's Summer Bitcoin, there's Kala, like Brink

1:38:57

has a fellowship program where we try to train

1:38:59

people and, you know, I'll go to

1:39:01

like, this is why I go to conferences

1:39:04

so that I can meet people who I've

1:39:06

only seen pop up a couple of times

1:39:08

on GitHub or on IRC. And I'm like,

1:39:10

hey, like, let's review PR together. Right. Because

1:39:12

if I don't like, there's no HR department,

1:39:14

right? We're all trying to figure this out

1:39:16

together. We all have an equal responsibility. You

1:39:19

listening to this podcast, don't

1:39:23

make a break. Are all. Yeah.

1:39:25

Yeah. You can also just

1:39:27

contribute by, you know,

1:39:29

if you hold a certain amount of thought, like

1:39:32

nobody's charging you to do

1:39:34

that, but there is a cost for it.

1:39:36

Pay your voluntary tax. Yeah.

1:39:40

Like, don't wait for somebody to charge

1:39:42

rent. Don't wait for, you know, like

1:39:44

we all have like

1:39:46

an equal part to play in this.

1:39:48

And like I said, I started this

1:39:50

part, I don't have any extra power.

1:39:53

Like you can come to GitHub and

1:39:55

open a pull request and contribute to

1:39:57

Bitcoin core. It's free for anybody to

1:39:59

do so. And

1:40:01

I would encourage you to do that. I'm

1:40:04

going to call this episode, Gloria

1:40:06

says Bitcoin is broken. Satoshi is