WTF is wrong with Bitcoin Core supporters?

WTF is wrong with Bitcoin Core supporters?

Attached: Untitled.png (1079x1188, 273K)

Other urls found in this thread:

bitcoin.org/bitcoin.pdf
blockchain.com/en/charts/utxo-count
twitter.com/NSFWRedditImage

>WTF is wrong with Bitcoin Core supporters
Low IQ

They're all paid shills. Unironically.

so it's inconceivable that different people have different ideas about scaling? for example simply increasing the block size manyfold is a short term patch for the problem long term it doesn't really work without serious centralization or not even then. it causes all sorts of problem with network propagation. even tho bitcoin actually needed that short term patch some felt a longer term solution was better served by not increasing the block size and driving segwit adoption with fees faster. i personally disagree but see the logic behind it. alos blockchain is full of spam that is also true adoption rate is abysmal.

Sub 120 IQ people like you just don't understand the thought process of geniuses like Core supporters.

Low block sizes = no CP onchain.
This is the only solution that makes it feasible in the eyes of public that's supposed to use this long-term.

Not enough turmeric in their diets

i still feel they fucked segwit up tho. this bullshit softfork is a bad idea. it's actually worse than a hardfork would have been.

What if the hard fork would be rejected be the network?

Define centralized

with 95% miners signalling acceptance? not bloody likely.

centralized as in full nodes are only ran by a few miners or not even them, like where sv is headed.

btw if a fork is rejected by the network i guess then you don't fucking need it that's an other solution.

>a few
3? 2?
And by full node you mean a complete copy of the blockchain?

Non-mining nodes using the network would suddenly find themselves not being on the 'proper' network.
>It's their problem
The long-term signal for this is that the devs run things and own your shit - you must do their bidding. This jeopardizes long-term credibility of the network, making certain risky node utilization installations less possible because of the uncertainty.

Opposite of decentralized.

And distributed =/= decentralized.

There is a difference.

Define decentralized then

This

>no CP onchain
literally all you need to beat craig

>core
Nice fake news faggots just because your scamcoin founders decided to unofficially rename Bitcoin and steal its name for their theaving get rich quick schemes doesn’t mean you Nigger NPCs have to follow suit. Fucking cringe and blue pilled

Attached: 411E6554-3472-42A0-846B-FDB13BF87B5F.jpg (225x225, 7K)

They're a subhuman dead end of evolution incapable of basic logic with regard to economics bred over millenia of negative selection by government slavers.

interesting question that, since i have been long a proponent of pruning the fucking thing regularly. but yeah if you don't include full utxo on the blockchain regularly and prune the old blocks after a while say then you got to keep every fucking block mined ever. the miners power is indeed offset by the network propagation maintained by validating nodes (despite what the stupid cashies say) exchanges notably should always run a full node and ln hubs also. this greatly limits the miners incentives to defect in collaboration even if they could sustain a 51% attack.

Define centralised in a quantitative manner
Is it a ratio of miners having the full blockchain? is it all miners having the full blockchain?

hard definition: if it's impossible for anyone to run a full validating node it's centralized. end of story.
if only mining pools can validate the blockchain fully and maybe google and amazon then it's fucking centralized. that nightmare craig is ranting about his data warehouses is pure bullshit.

>if it's impossible for anyone to run a full validating node it's centralized.
Which means you would have to believe btc is currently centralised given the percentage of people that can run a node is tiny

Attached: forkies.png (1477x563, 1.28M)

and the funny thing is all any fully validating node would need to keep is 200 megabytes of data.
say you keep the entire merkle root tree from the genesis block and prune the blockchain ever 100 blocks to the block that includes full utxo (which can also be verified from the last before it), it's a far cry from the fucking hundreds of gigs where we are at and thousands of terabytes we are head to.

few hundred megabytes is all this shit needs, there is no real reason for most people not to run a full node unless they are using a mobile device.

i have deleted bitcoin core more than a year ago. i actually need my fucking hard drive thank you very much!

it is less decentralized than it should be, but still thousands run full validating nodes and all spv servers run full validating nodes and all exchanges and services and ln nodes. which is the point. if it gets any worse then it's crap.

This. Bitcoin core is full of crypto geniuses like Adam Back. Non-bitcoin core is full of suburban and rural retards that scam and run patent trolls like CSW

Your definition was if anyone cannot run a node it's centralized.
Do you want to change your definition to >thousands run full validating nodes and all spv servers run full validating nodes and all exchanges and services and ln nodes

btc can run a full node on raspberry PIs last one can remember.

Is your definition of centralised when a btc node cannot be run on a rasberry pi?

that's bollocks tho, peter rizun is a fine example that the cashies have some brains too.

Im just trying to understand. You claim on chain scaling centralises a blockchain but you cannot give me a metric by which to measure centralization

i could still run a full node if i really wanted to. it wouldn't break the bank i would just need to buy a harddrive for it, it would cost me like $200 tops to host the blockchain for the next decade or so. for now at least... but shit is getting worse and i don't like any change that makes this worse. and cashies are not helping one bit with their over the board bullshit. they removed the hard limit off 1mb fine, not that bch has enough transactions to legitly fill 100k but okay, but 32 megabytes and 128 megabytes and 4 gigabytes blocks? that idiotic. that's just plain on going full retard on purpose.

i gave you, if i can no longer run a full node with minimal effort it's very bad. if you end up only miners (pools actually) running full nodes it's completely centralized.

The people that can afford a $200 node and have the competence to run one are a vast majority which by your definition makes btc centralised. Or do you define centralised another way?

when you stop thinking that these people are trying to improve bitcoin and its adoption and realize that their goal is to slow bitcoin adoption or stop it altogether, it all starts to make sense.

Attached: happymerchant.jpg (600x600, 42K)

So if you personally cannot run a node it is centralised? What ratio of pools to individuals is not centralised?

What constituted "minimal effort"

>so it's inconceivable that different people have different ideas about scaling?
some ideas are just pants on head retarded and makes you wonder if they are paid to use their credibility to support them. 1mb blocks can't even support any significant rate of people opening their own LN channels. LN can only become a (shitty) scaling solution if people can onboard their BTC. Now they want to reduce the block size and make that even more difficult and expensive.

There is no logic behind it. These people are too smart to believe this bullshit they are spouting. They are compromised. Maybe Luke is just some weird kind of mentally ill, but the others definitely know they are crippling not only bitcoin, but also LN.

rephrase that please you are contradicting yourself there. btc is not exactly centralized yet but it's not as decentralized as it would be healthy. i don't have hard metric but i think still a few 10.000 full nodes are running worldwide roughly to like 0 minding pools. that's more than the entire cashie userbase btw. and not centralized as of yet.

10000 to 0 means you define having more than 0 minning pools as btc being centralised.
That is the 5th definition of centralised youve changed to is thay your final answer?

>requiring precision in definitions that have to allow a certain degree of ambiguity
>muh precise number of nodes

a minimal expenditure even a favella monkey >So if you personally cannot run a node it is centralised?
to me it would be the signal that shit is over. because sooner or later it would mean no private person runs a full node. it's the tipping point really i'm definitely not rich but not impoverished.

but if you want a hard metric, when the number of full nodes drop drastically where the miners are suddenly a serious portion of the network then it's fucked and over. since the mining pools are roughly constant in 4-5 major ones and about 10 minor pools in the past years, that would mean below 1000 full nodes is the breaking point.

Im trying to understand what you mean when you say on chain scaling centralises a blockchain
To that i need to define centralised

>1mb blocks can't even support any significant rate of people opening their own LN channels.
i agree not even ln can scale anywhere with 1mb blocks. it's the timing that is in question only and the consensus behind the increase of blocksize. it looks like for cashies it's more of a self definition and political statement than anything dictated by necessity. that's what it looks to me at least from where i'm sitting saying that as i supported unlimited before the fork shitshow in the debate.
>There is no logic behind it.
of course there is. segwit adoption would have been immeasurably slow not for the fees. new thing lot's of fud untried old one works just fine who ever would switch?

20 mining pools sorry

How about you define it first so the burden of argumentation is not only on the other person you lazy bitch ass shitter.

Ok so for btc to not be centralised it needs to maintain an individual node to miner node ratio of 20:10000.

How could 10000 mining nodes affect what even 1 mining node does given bitcoin is proof of work not proof of stake?

Im asking because i dont know

*10000 non mining nodes

There it is. They've gone full retard.

come now at least he is not craigposting or spouting the usual idiotic fuds. but i admit i never bothered to put actual numbers behind my concerns so he caught me with pants off. very uncomfortable.

and since the bitcoin mining and network structure is changing constantly it would be really hard to make revelations about magic numbers that decide 1 or 0.

but yeah under 1000 full validating nodes means game over. under 10000 means it's time to panic. above 100000 would be ideal imo. but if the population explodes adoption rate changes miner etc etc these numbers shift i guess.

i'm with core in that everyone should run a full node who can, but their full node is highly impractical to run and they have done nothing in this regard. they should have implemented a fork between archiving nodes and pruning full nodes a long time ago and run it on a testnet.

>How could 10000 mining nodes affect what even 1 mining node does given bitcoin is proof of work not proof of stake?
easily bitcoin uses proof of work as a security feature not as a validating feature. the validating checks the proof of work as part of the block being valid. but just because the proof of work checks out doesn't mean the block is valid. a miner can create invalid blocks ad infinitum they would never propagate in a healthy network. an other miner may have less hash but if his blocks are valid they get propagated. this creates an incentive for miners to keep the rules.

Name a single scenario where 10000 non mining nodes disagreeing with 20 minning nodes results in the 10000 non mining nodes being the longest chain

segwit 2x

and if you ask what is the point of this all, since the majority miner can just shadow mine it actually helps the network in implementing deep reorg protection. and a shallow reorg from time to time is normal and not profitable anyways.

of course deep reorg protection has it's own problems it can cause the network to permanently split. which is why it was never implemented in the early days. contentious forks and networks were the worst nightmare of satoshi. i think once adoption reaches 1% it should be implemented. people would feel more secure knowing the miners would screw only themselves but not profit if they attacked the network. but i still got to think about this some more. for instance miners could in theory just short the shit out of bitcoin now and still likely profit on the attack.

no man that' snot how it works, how it works is nobody on the network sees a new block mined. it's like the miners all quit. that's what it would look like. until a miner comes and mines a valid block over the last valid block. see only the longest valid chain matters and what is valid is consensus. not of the miners but of everyone.

There was no disagreement with segwit 2x.
Ill put it to you another way. If 10000 non mining nodes were signalling a blocksize limit of .5mb and 20 mining nodes were mining 1mb blocks who would get the longest chain

In the bitcoin whitepaper it says they vote with their hashpower.
How can non mining nodes vote on validity if they do not have hashpower

Ill put it to YOU another way. If you have all the miners on one side, and all the full nodes on the other, and one of the miners defects to the node side. What happens?

A contentious hardfork with 1/20th the hashpower

if one miner defects the the nodes slowly but surely he will make profit while all the others are not able to monetize their coins. mempool congested price drops off course until more miners defect back. but this is pretty theoretical, miners are not connected to each other directly that would be a very easy attack vector they all care about what the network sees (so long there is a network big enough) and base their strategies on that. that is an other reason why full validating nodes are important.

yeah but don't forget if every user is on that fork than the other is worthless. it's not bitcoin. the difficulty adjustment would be a nightmare tho, but this scenario is bad enough for miners to avoid. even if the majority defects a sizeable minority would seize the opportunity.

see that's the fun thing about consensus. actually the whitepaper says 1 cpu 1 vote which is true enough for validating nodes and ridiculous for mining nodes if you take it word by word.

the network changed a lot since satoshi published his first implementation. pools changed everything to begin with. it's a trustful centralized game on a trustless decentralized network and it worked out economically. go figure...

longest _valid_ chain never forget that never forget also bitcoin derives it's value from consensus in the most literal way possible what is called bitcoin is defined by consensus (not by the miners) by the world at large. it's the userbase that has final governance of the coin but it lacks formal decision making process. kinda ad-hoc. but everyone has a good idea what would people go along with and what not. core devs double-think themselves into impotence over this too much tho.

Defining centralised by any ratio of non mining nodes to mining nodes is beyond retarded because there is no scenario where non mining nodes affect validity as they have no hashpower. You could have 7000000000 to 1 non mining nodes to mining nodes and the mining node decides the longest chain in every scenario.

To limit btcs on chain capacity to 5tps under the justificatiin of maintaining a non mining node to mining node ratio is the most retarded fucking shit ive ever seen

Fee market or bust.

Right now your transaction security is being unfairly subsidized through inflation.

UTXO set exceeds 200 MB

>there is no scenario where non mining nodes affect validity as they have no hashpower
typical cashie fallacy stemming from the fact they don't understand bitcoin at all.

>You could have 7000000000 to 1 non mining nodes to mining nodes and the mining node decides the longest chain in every scenario.
only for himself nigga that's the point he would be living in an alternate reality only he can see.
it was 65mb last i checked.

Nothing wrong with them. We, big blockers, should support them in their vision.
They are free to pursue whatever vision they want for their coin.
But yes, we should educate newbies that BTC is not what Bitcoin, the economic system, actually is. Otherwise, mainstream media will just claim "Bitcoin has been proven useless" and people who know what's going on will also suffer a loss of mainstream faith in crypto.

Attached: 1520874274447.jpg (379x465, 80K)

65 million utxos. But each utxo consumes ~70 bytes easily (~32+4 bytes for the outpoint, 4 bytes for block height, at least 25 bytes for the scriptpubkey). So in terms of space it's more like 4-5 GB.

>They are free to pursue whatever vision they want for their coin.
normally that's true but instead of constructive arguments like again the kinds you hear from peter rizun and a few others it's just idiotic fudding baseless shitposting baiting and trolling.

and new people will run into that shitshow first without any technical understanding or research they take these crappy arguments at face value.

>BTC is not what Bitcoin, the economic system, actually is.
this is typical of said shitposting. absolutely baseless assumption stemming form the fact that you don't fucking understand what bitcoin is. it's consensus, as much by the market (which is again just the people dealing with or using bitcoin) as by the miners. and cash broke that consensus. albeit what the core crew did was despicable and truly in bad taste. but cash ended up making things worse. go figure. btc needs bigger blocks because people actually use it, cash doesn't but cash has much bigger blocks and btc doesn't. it's like watching retard children throwing their shit arguments wise.

Are you retarded? You can literally run it on RasPi with a old hdd....total cost bellow 70$
Or you could fit the whole chain on a cheap 256gb sd card

Any fork is retarded

>65 million
those are the non 0 balances?

yeah try that with sv in a few months! that's my point.

>baseless assumption stemming form the fact that you don't fucking understand what bitcoin is.
bitcoin.org/bitcoin.pdf

It's a lot of things, unfortunately none is 100% bitcoin judging by the whitepaper. All I'm saying is that the proposed small block economic system which will make BTC completely collapse is completely different than what's outlined in the whitepaper. So the BITCOIN brand shouldn't take a hit because of idiots.

there are no such thing as balances on bitcoin, but I guess you could make that analogy.

blockchain.com/en/charts/utxo-count

>the proposed small block economic system which will make BTC completely collapse is completely different than what's outlined in the whitepaper.
that's true enough. i don't think it will stick. core will increase block size when it's inevitable. they won't play chicken too long. otherwise i'm out of this fucking shitshow and goodbye

alright but say 5 gigs is still smaller than any of my games of a high def movie. i find it reasonably acceptable. 200 gigs? not so much multiple terrabytes (sv) fuck no man!

hm i'm just reading about pruning in bitcoin core wallet it seems like if you become a passive node then you can prune it down greatly however you are no longer participating in the sync process it is also implemented in a weird was you can set the kept size of data but not how many blocks you want to keep. shit is not how i would have done it.

Yep, totally reasonable concern.

There are two things that can be done about the problem in :
1. You could have run your node in pruning mode, where it would free up a lot of block space.
2. To avoid people having to download the full blockchain to start their node, bitcoin devs can introduce "UTXO set commitments" that will let new nodes boostrap themselves in at a very recent block.

BTW even if you had the space for BSV chain, you would probably want to prune it anyway since you don't want child porn on your harddrive. :D

You can do full-sync with pruning. The only problem is that then you don't anymore have a full blockchain to scan (say, if you import a wallet).

>since you don't want child porn on your harddrive
every time kek

so like i said if you set to pruning, you are excluded from the network in some way i don't fully understand as of yet. they should fix this and make reasonable pruning the new default setting while only archiving nodes keep full history which should be ran by say tech giants and maniacal enthusiasts.

i will give core with pruning an other go anyone knows if ln nodes can run in pruning mode?

If you prune you can't help other nodes in their initial sync. I think that's the only way where you're excluded.

The hardcore decentralization advocates dislike pruning for that reason (blockchain needs to have as many copies as possible).

Lightning can be run pruned from what I understand, just make sure the lightning software doesn't fall behind your bitcoind.

not just the initial if i read it right that' obvious you cant send what you don't have.

>The hardcore decentralization advocates dislike pruning for that reason (blockchain needs to have as many copies as possible).
yeah this is my problem decentralization is good bt why the fuck keep the entire copy? it's statistically extremely improbable someone could run a deep reorg attack a 100 deep on bitcoin, if it happens the entire thing implodes anyways there will be no recovering from that. btc returns to 10 cents and people will cry rivers of blood.

but for now there is no way to properly make pruning trustless and decentralized from the get go. you actually need to fully sync the blockchain from the genesis block which is hard-coded. but if blocks would include at least the hash of the full utxo (ordered in a deterministic way) we could sync from a few days past with new client i think. merkle root chain still needs to be kept of course but it grows linearly independent of blocksize or transaction count quiet sustainable.

im loving this wave of FUD, just bought 2 more BTC

Basically the power to make decisions lays in the hands of a group that can be controlled in some manner

> true adoption rate is abysmal.

Let’s increase adoption by making the base later, which is the easiest to use, less efficient and force people who were already on the fence about crypto toward a more complex and less secure layer.

Makes sense

on the whole, they are mentally deficient.

yeah lets abandon a model proven secure over 10 years and go with lightning!! woo!

By niggernomics which is what forkies use to judge what is Bitcoin basically any fork of Bitcoin with big blocks is Bitcoin since it doesn't have to meet the only requeriment that matters.

Attached: bcash is not bitcoin.png (1534x828, 240K)

Since the ABC/SV fork, any chance Bitcoin Cash had at changing the gravity of the scenario is long shot. When's SVC (Satoshi's Vision Cash) coming?

Attached: 1234.png (150x145, 56K)

Cashies are too stupid and lazy to run their own nodes so they think it's not important

>lets fork to hueg blocks because thats safer than leaving layer 0 as is

Just run the whole fucking thing. It's at 233 GB r ight now. Buy a cheap as fuck 2 TB HDD and you are set for years while safe from state funded 51%'s.

U tryna teach him game theory?