The Barry Silbert segwit2x agreement with >80% miner support.

[d5000]

OH, you mean the post where you ignored that it would take 50 2M blocks just to clear the current mempool if no new transactions ever occur, that "everyone" claims there are transactions that don't occur because there's a backlog, and where "everyone" ignores the amount of transactions that they claim aren't happening because of congestion?  

Yeah, it would take a couple of blocks to clear the current backlog. But as I wrote in the previous post, the 2MB blocks with the actual transaction rate (3,5 tps) would be only full for 60%. Let's add 10% more for the "transactions that [..] aren't happening because of congestion" (I repeat: the 3,5 tps are the transactions added to the mempool including transactions that are never confirmed).

So we would have 600 kB per 2MB block to reduce the backlog. Assuming that per MB of block capacity about 2000 transactions can be processed (at an average size of ~500 bytes), a backlog of 100.000 transactions would need about 150-200 blocks to be cleared. Bitcoin processes ~144 blocks per day. So basically in a day and a couple of hours the backlog would be reduced to zero.

If there is a fallacy in my argumentation, please help me to find it
 

[ComputerGenie]

...If there is a fallacy in my argumentation, please help me to find it

Remember that time, oh so long ago (like last year), when there weren't enough transactions to fill all 144 blocks with 1MB worth of transactions every day......
The problem with so many of the rationalizations I see are all based on tiny, limited scopes and the fundamental change of everything that is Bitcoin.

Segwit + 2MB + LN = all problems are fixed and the world is a happy, shiny new place where growth will no longer occur.
Not!

What if the current masses don't want their transfers "off-chain"?
What if Bitcoin grows to something even in the earshot of "mainstream"?
What if the new masses understand the tech don't want their transfers "off-chain"?
What if we fix what's broken about Bitcoin, in a way that it remains Bitcoin, and the fix isn't just some half-assed bandage that's going to need to be changed in 6 months, 1 year, or 2 years?
What if we say, "Core devs we accept no half-way measures, any fork must be future-forward looking"?


Edit: Have I mentioned, at any point, that LN is an external service; not needed (and even pointless) in the vast majority of current Bitcoin usages; and, in the end- still requires the "final" transaction to go into a Bitcoin block the same as a "standard" transaction?

[d5000]

The problem with so many of the rationalizations I see are all based on tiny, limited scopes and the fundamental change of everything that is Bitcoin.

Segwit + 2MB + LN = all problems are fixed and the world is a happy, shiny new place where growth will no longer occur. Not!
[...]What if we say, "Core devs we accept no half-way measures, any fork must be future-forward looking"?

I partly agree - I had favoured a conservative, but dynamic approach like this one for future blocksize scaling. I also agree that LN is not the Messiah some are preaching about - it's a solution suitable for certain kinds of micro- and minipayments, not more. I regard it as the equivalent to a (general-purpose) prepaid card in the cryptocurrency world.

But the dangers of too big blocks cannot be ignored. I know the importance of full nodes is sometimes exaggerated, but for the individual users they are the most secure (and the only really trustless) option to participate in the Bitcoin network.

And sometimes, I think a little bit of realpolitik is needed. It's obvious that 2MB+Segwit (or Segwit alone) isn't a final solution. It's a dirty hack, but it solves most short-term problems we actually have (that were my previous posts about). We will very probably (no mathematical proof, but I'm 99% sure) have some air after it to find a better governance model and a long-term scaling solution.

[ComputerGenie]

...But the dangers of too big blocks cannot be ignored...

And they are....

...

...It's a dirty hack, but it solves most short-term problems we actually have (that were my previous posts about)...

Which goes back to my post that offended you; dreams, wishes, and delusions are what support a short-term fix rather than an actual fix. It's delusional to think that the dreams of longevity are supportable by the wishes of breaking it all again in the future to almost fix it now.

[d5000]

...But the dangers of too big blocks cannot be ignored...

And they are....

Come on, that really has been discussed ad nauseam here

The danger (I implicitly mentioned it) that common people - and above all, small businesses - could not longer be able to host a full node with consumer-grade hardware is enough for me. We would hinder people to use Bitcoin in a trustless way -  they would be forced to use SPV clients or even centralized online wallets to participate. In this case we could also switch to LN, it would be the same level of "trustlessness" (I don't know if that's an English word, but I think it's understandable what I mean). There is no need for conspiracy theories about evil miners and banksters behind them to come to the conclusion that big blocks alone aren't that good.

The scalability problem is not trivial to solve. That's why there are so many discussions about it. And now we're caught in a stalemate. That's why I favor a short-term fix if there is no consensus regarding a long-term fix, rather than discussing the long-term fix the next years.

[ComputerGenie]

...The danger (I implicitly mentioned it) that common people - and above all, small businesses - could not longer be able to host a full node with consumer-grade hardware is enough for me...

Perhaps they shouldn't be continuing to buy 17 year old Quantum Fireball 4500RPM 40GB HDDs? Beyond that, I still fail to see the legitimacy of such a claim.

[d5000]

HDDs are the least important problem. I think you know that

The problem is the initial blockchain download (which consumes bandwidth, CPU and RAM usage) and the block propagation/keeping in sync. Keep in mind that not every potential user lives in highly-connected global cities in the first world.

[ComputerGenie]

...Keep in mind that not every potential user lives in highly-connected global cities in the first world.

Keep in mind that not every potential user needs to run a full node and that you're trying to make a bandwidth argument to someone who will never have more than 10MB down (or 0.75MB up) anytime in the next 15 years, runs a full node of 15 different coins, streams HD video 16+ hours every day, and who's ISP's legal name is "Hillbilly Wireless" (because he lives in a barely-connected small town of ~600 in a county of less than 15,000).  

[d5000]

Keep in mind that not every potential user needs to run a full node

People that don't run a full node can also use LN for their transactions. In terms of "trustlessness" it's comparable.

and that you're trying to make a bandwidth argument to someone who will never have more than 10MB down (or 0.75MB up) anytime in the next 15 years, runs a full node of 15 different coins, streams HD video 16+ hours every day, and who's ISP's legal name is "Hillbilly Wireless" (because he lives in a barely-connected small town of ~600 in a county of less than 15,000).  

Altcoins do not count. If you run Bitcoin + 14 other coins, the restant 14 altcoins (even if ETH is among them) would probably not even eat more bandwidth than one single Bitcoin node, so you are barely using the same bandwidth needed by full Segwit blocks with 1,7-2,2 MB.

Try to run 8 parallel Bitcoin full nodes, then your argument becomes more credible. And don't cheat interchanging the blockchain files And: Even your humble 10 MB are only a dream for >70% of the world population.

And you didn't mention CPU and RAM. (Moore's Law argument incoming ...)

[iluvpie60]

I am seeing this is back on the front page. Doesn't seem like this went anywhere... Unfortunately.

Where does this stand now, anyone have a good link or info?

Last time I thought to myself I will believe it when I see it, I still don't see it happening.

What more can we do beyond making posts? Are we ready to start firing up miners to get some percent into a pool and push it over the top?

[ComputerGenie]

...People that don't run a full node can also use LN for their transactions. In terms of "trustlessness" it's comparable...

I was tired of any, and all, LN arguments even before they were made. LN and nodes have virtually nothing to do with each other and the existence, or non-existence, of LN is moot to any part of this conversation.

...Altcoins do not count. If you run Bitcoin + 14 other coins, the restant 14 altcoins (even if ETH is among them) would probably not even eat more bandwidth than one single Bitcoin node, so you are barely using the same bandwidth needed by full Segwit blocks with 1,7-2,2 MB.

Try to run 8 parallel Bitcoin full nodes, then your argument becomes more credible. And don't cheat interchanging the blockchain files And: Even your humble 10 MB are only a dream for >70% of the world population.

And you didn't mention CPU and RAM. (Moore's Law argument incoming ...)

• Given that a pruned full node is @ or < 2GB, your math is way off.
• Be careful of your disdain for alts, because if some folks in this thread have their way, NotBitcoin is the new path forward.
• Even Moore has said that the end of the "law" has come (not that I ever use it as a claim).
• Moore's Law was an observation of something that has no relevance (not that I ever use it as a claim).
• I didn't mention CPU and RAM, because per required unit they are even less expensive than storage.
• No one in the random 3rd world country you are positing needs to run a full node.
• No one in the random 3rd world country you are positing needs to care about the hardware requirements to run a full node.
• Your argument about nodes vs the "underprivileged" (for lack of a better term) is still moot, because users of Bitcoin don't need to run full nodes.
• If you still can't grasp the last one, then you likely don't understand Bitcoin as a protocol.

[d5000]

I am seeing this is back on the front page. Doesn't seem like this went anywhere... Unfortunately.

Where does this stand now, anyone have a good link or info?

The Btc1 repository seems to contain the coding work that is being done. I've not checked it but according to other users, the code based on the agreement is progressing.

(It seems unavoidable that in Bitcointalk the thread degenerates in some form of "small blockers" vs. "big blockers" and sometimes vs. "moderate blockers" discussion, but I estimate that in this forum only very, very few development decisions are being taken.)

Given that a pruned full node is @ or < 2GB, your math is way off.

So current pruned Bitcoin nodes do not have to download & verify all blocks anymore? OK, that would be new for me. Or are you talking again about the (irrelevant) HD space?

users of Bitcoin don't need to run full nodes.

No, some don't need, but a full node should be a viable option for everyone that needs it, above all for businesses and professionals in every part of the world. Only a full node gives you the trustlessness that makes Bitcoin great.

A totally redundant blockchain where everything is stored in every node won't provide that and is terribly inefficient. People that don't need full nodes because they only move small amounts via cryptocurrencies could also use off-chain transactions/LN or sidechains. That is why I don't agree to your statement here:

LN and nodes have virtually nothing to do with each other

I am convinced that we need a form of sharding (splitting the Bitcoin transaction data in smaller sections, so not all full nodes must verify all data) - be it sidechains, childchains, extension blocks, or a traditional sharding solution like Ethereum is planning (if it's not vapourware), if we want to scale up. I, for myself, don't want a coin where only datacenter and rich users could run a full node. And yes, I want a LN-style micropayment solution to pay for coffee and other small things. But I also don't want LN as the only second layer as some Core Maximalists like it.

I'm not even against Bitcoin Unlimited, but I would really like to see EC tested in the wild - as an altcoin or a "cleanly forked" Bitcoin (based on a UTXO snapshot) to see like it behaves, until it is introduced as an option for the main BTC client.

I think that is my final post to the big/small block issue because I think we are desvirtuating the thread.

[ComputerGenie]

...People that don't need full nodes because they only move small amounts via cryptocurrencies could also use off-chain transactions/LN or sidechains. That is why I don't agree to your statement here:

LN and nodes have virtually nothing to do with each other

...

I'm still not sure you understand the basic premise of what LN is. LN only works if you have a "thin client" that creates a valid Bitcoin transaction. I'm at a loss as to why you, and so many others, want to create new clients to do exactly what current clients do. The only advantage that LN provided to that average user is the rare instance where they want/need to make voidable transactions (which is more anti-Bitcoin than I care to discuss). It is 100% pointless to use LN to make an immutable bitcoin transaction (since the end result of any finalized LN transaction is, in fact, the same Bitcoin transaction that one would have made using Bitcoin).

I'm still not sure why you're stuck on these insane ideas that:

• The protocol should be limited so that people that don't need to run nodes can run them anyway using dialup and a PC less usable than an Apple Lisa
• LN has anything to do with normal Bitcoin transactions

[buwaytress]

The problem with so many of the rationalizations I see are all based on tiny, limited scopes and the fundamental change of everything that is Bitcoin.

Segwit + 2MB + LN = all problems are fixed and the world is a happy, shiny new place where growth will no longer occur. Not!
[...]What if we say, "Core devs we accept no half-way measures, any fork must be future-forward looking"?

I partly agree - I had favoured a conservative, but dynamic approach like this one for future blocksize scaling. I also agree that LN is not the Messiah some are preaching about - it's a solution suitable for certain kinds of micro- and minipayments, not more. I regard it as the equivalent to a (general-purpose) prepaid card in the cryptocurrency world.

But the dangers of too big blocks cannot be ignored. I know the importance of full nodes is sometimes exaggerated, but for the individual users they are the most secure (and the only really trustless) option to participate in the Bitcoin network.

And sometimes, I think a little bit of realpolitik is needed. It's obvious that 2MB+Segwit (or Segwit alone) isn't a final solution. It's a dirty hack, but it solves most short-term problems we actually have (that were my previous posts about). We will very probably (no mathematical proof, but I'm 99% sure) have some air after it to find a better governance model and a long-term scaling solution.

It's funny that the rationalisations used on either side gets accused of being based on tiny, limited scopes =).

As someone who will never be fully initiated into the realm of "knowledgeable" I have to say that both you guys say what is genuinely on our minds. What we do understand of this allows us to want a way forward, but we prefer solutions that say: "this is how we will move forward, and this is what we will do when these likely issues come up" instead of "we won't do that because that sucks".

Amen on realpolitik. No one is deluded to think there is one final solution.

[dinofelis]

...If there is a fallacy in my argumentation, please help me to find it

Remember that time, oh so long ago (like last year), when there weren't enough transactions to fill all 144 blocks with 1MB worth of transactions every day......
The problem with so many of the rationalizations I see are all based on tiny, limited scopes and the fundamental change of everything that is Bitcoin.

Segwit + 2MB + LN = all problems are fixed and the world is a happy, shiny new place where growth will no longer occur.
Not!

What if the current masses don't want their transfers "off-chain"?

Then they don't use it.  In fact, "off-chain" is a perfect way to connect to exchanges.  LN is perfect with exchanges as central hubs, and exchange customers having their holdings in a channel with their exchange, which protects them from theft, and renders the exchanges trustless.  It would turn the current exchanges in the new banks.

What if Bitcoin grows to something even in the earshot of "mainstream"?

It most probably won't, unless, of course, in the "exchanges as new banks" world ; bitcoin's main application is "greater fool game", and of course, we need a lot of greater fools, but on the other hand, I don't know how many of them will be gullible - though there's hope, given historical records of frenzy.  When "average Joe" gets into bitcoin, sell everything !

What if the new masses understand the tech don't want their transfers "off-chain"?

Actually, it may be simpler to get "new masses" to bitcoin by giving them a central authority (their exchange) and giving them classical banking with off-chain linking to their exchange.  It would be more secure too for them, they would feel much less at a loss, it would be standard banking.  There may even be a law that only off-chain transactions are officially allowed, and that you need a licence for on chain transactions. 

What if we fix what's broken about Bitcoin, in a way that it remains Bitcoin, and the fix isn't just some half-assed bandage that's going to need to be changed in 6 months, 1 year, or 2 years?
What if we say, "Core devs we accept no half-way measures, any fork must be future-forward looking"?

Bitcoin is broken beyond repair.  The first thing to fix in bitcoin, is the emission curve.  A decreasing coin emission renders every coin unstable and/or centralized.  The second thing to fix is proof of work consensus.  These are the two principal value propositions of bitcoin, and they render bitcoin unfixable if you don't modify them - which goes against bitcoin's belief system.
Only a temporary fix, that is changed regularly, can keep it afloat for a while, and this might very well be sufficient, because bitcoin doesn't maybe need to scale to very very large sizes.

Edit: Have I mentioned, at any point, that LN is an external service; not needed (and even pointless) in the vast majority of current Bitcoin usages; and, in the end- still requires the "final" transaction to go into a Bitcoin block the same as a "standard" transaction?

Sure, but the LN could transform bitcoin into classical "banking" (but rather, classical clearing house transfers).  Bitcoin is not money, but is a speculative betting asset, and that betting is mainly done on clearing houses (exchanges).  Now, the number of players may become large, but has nothing to do with the size of payment networks.  Bitcoin being more like "complex derivatives for the people", the LN could be a way to tie yourself to your clearing house and render the relationship more trustless.

[classicsucks]

...If there is a fallacy in my argumentation, please help me to find it

Remember that time, oh so long ago (like last year), when there weren't enough transactions to fill all 144 blocks with 1MB worth of transactions every day......
The problem with so many of the rationalizations I see are all based on tiny, limited scopes and the fundamental change of everything that is Bitcoin.

Segwit + 2MB + LN = all problems are fixed and the world is a happy, shiny new place where growth will no longer occur.
Not!

What if the current masses don't want their transfers "off-chain"?

Then they don't use it.  In fact, "off-chain" is a perfect way to connect to exchanges.  LN is perfect with exchanges as central hubs, and exchange customers having their holdings in a channel with their exchange, which protects them from theft, and renders the exchanges trustless.  It would turn the current exchanges in the new banks.

A huge subset of crypto users are already transferring their wealth off the bitcoin chain - they use alts which are better proven than LIghtning, and render Lightning unnecessary. Basically the bitcoin fees got higher, and also it's believed that transferring from one blockchain to another also provides better privacy for bitcoin users.

LN is late to the party.

[hv_]

...If there is a fallacy in my argumentation, please help me to find it

Remember that time, oh so long ago (like last year), when there weren't enough transactions to fill all 144 blocks with 1MB worth of transactions every day......
The problem with so many of the rationalizations I see are all based on tiny, limited scopes and the fundamental change of everything that is Bitcoin.

Segwit + 2MB + LN = all problems are fixed and the world is a happy, shiny new place where growth will no longer occur.
Not!

What if the current masses don't want their transfers "off-chain"?

Then they don't use it.  In fact, "off-chain" is a perfect way to connect to exchanges.  LN is perfect with exchanges as central hubs, and exchange customers having their holdings in a channel with their exchange, which protects them from theft, and renders the exchanges trustless.  It would turn the current exchanges in the new banks.

A huge subset of crypto users are already transferring their wealth off the bitcoin chain - they use alts which are better proven than LIghtning, and render Lightning unnecessary. Basically the bitcoin fees got higher, and also it's believed that transferring from one blockchain to another also provides better privacy for bitcoin users.

LN is late to the party.

+1

Same I would say

[ComputerGenie]

...Sure, but the LN could transform bitcoin into classical "banking" (but rather, classical clearing house transfers)...

From that, I think you understand the point and principles of LN less than he does. Care to elaborate on how you think that LN and "banking" have any relational ties?

[dinofelis]

...Sure, but the LN could transform bitcoin into classical "banking" (but rather, classical clearing house transfers)...

From that, I think you understand the point and principles of LN less than he does. Care to elaborate on how you think that LN and "banking" have any relational ties?

I think the incentives in the LN are such that only a "wheel-and-spokes" architecture makes sense.  The LN becomes profitable (can offer competitive advantage over on-chain fees) if you have A LOT of bitcoin to lock into MANY links with MANY "end point customers".  If you can only lock in 3 BTC into 3 channels with 3 other dudes with each one 1 BTC, your channels will quickly become "pushed against one side" and you'll have to settle and "reload" ; because most small nodes will not be in "windstill" places in the network, but will undergo systematic "fluxes" in one direction or another.  However, if you are a big hub (say, an exchange) and you can link your customers to you with a LN channel each, in which they put all of their holdings, then your links will rarely be exhausted (people spend, and buy coins), you will be able to maintain a link for a very long time before settling, and you can divide the "settling fee" by the large amount of transactions that go over the channel.  As such, your LN fees can be outcompeting all other LN fees others might demand and still be profitable.

In the end, you would simply have many "end customers" having all their holdings locked up in a LN channel to "their exchange" (their bank), and the few exchanges having big channels amongst themselves (like banks settle between them through the central bank).  Exchanges and other big hubs would then be the "commercial banks" and the miners/block chain would be the settling layer/central bank.  It would be very expensive/dangerous/.... for an individual customer to want to settle from his "bank" (expensive, rare, difficult to access block chain) ; so in the end, his bank has the full information of all his spendings ; can block the spendings it deems not right ; can charge customer fees.... like banks do today.  These big hubs may buy guaranteed room on the block chain with mining pools (while individuals won't be able to guarantee that they can settle in time).  As a small customer, you would be tied to your "exchange" in the same way as you are now tied to your bank, apart from the fact that eventually, you might succeed to settle.  If miners agree and let you have room.

[ComputerGenie]

...In the end, you would simply have many "end customers" having all their holdings locked up in a LN channel to "their exchange" (their bank)....

I can see your point from that specific scenario, but as a general concept, that's not what it's about.

In general principle:
Let's say you sell hot dogs..
I want a hot dog.
I create a transaction to buy 1 hot dog.
You agree to not to "finalize" the transaction because I may want another hot dog (or I may not).
I want another hot dog.
We invalidate the 1st transaction
I create a transaction to buy 2 hot dogs.
You agree to not to "finalize" the transaction because I may want another hot dog (or I may not).
I want another hot dog.
We invalidate the 2nd transaction
I create a transaction to buy 3 hot dogs.
You agree to not to "finalize" the transaction because I may want another hot dog (or I may not).
....
My fat ass has eaten 16 hot dogs and I want no more.
We agree to process the "finalized" transaction for 16 hotdogs.
The "finalized" transaction goes into the mempool like any other Bitcoin transaction (basically as if we'd only ever agreed on me buying 16 hotdogs and making a "standard" Bitcoin transaction to begin with).

Yes, there are cases where someone might put up enough servers that they control the "loin's share" of the LN and "control" fees, until some random guy throws up the LN equivalent of a node, charges nominally more than the "average" mined fees, and the bank you imagined looses any "stronghold" that they had.

While I'm not a fan of altering a protocol for non-protocol related end uses for the protocol, the principle of LN is to introduce voidable transactions into a system designed to be non-voidable. I think that this opens the door for a version of Bitcoin that is anti-Bitcoin; however, since it can actually be done without segwit (a few minor tweaks and a minor softfork) and it creates a greater use for Bitcoin in the general populous, I'm not totally against it.