What Bitcoin is actually designed for vs Gavin & Theymos vision for scalability

[ArticMine]

The problem is that Bitcoin does not work just fine. Both sides of the blocksize debate will agree to at least that.

Edit: All of the innovations above depend on being able to confirm transactions on the main chain.

If you want your tx to confirm, simply pay a competitive fee.  Or lowball and use RBF if you fail.  That's precisely how it's supposed to work.

The MPEX and #B-A types (among many others) don't agree there is a problem with Bitcoin, and I concur.

~Anyone can (in ~complete security) use Bitcoin to send ~any amount of money ~anywhere in the world, ~instantly and for ~free.

To me, that seems like a fucking miracle on the order of life, consciousness, wheels, fire, and Teflon.  Where's the issue?

In exactly what way does Bitcoin not work just fine?

The problem is that with 3 tx per second ~ 260,000 tx a day not even the wealthiest 0.004% of the world's population would be able to use Bitcoin just once a day.

Edit: Furthermore this is hard coded into the protocol so no matter by how much the price of bandwith, CPU time, storage etc falls, this limit cannot be changed. Bitcoin ignores technological change, simply because it is hard coded into the Bitcoin protocol to ignore technological change.

[iCEBREAKER]

The problem is that Bitcoin does not work just fine. Both sides of the blocksize debate will agree to at least that.

Edit: All of the innovations above depend on being able to confirm transactions on the main chain.

If you want your tx to confirm, simply pay a competitive fee.  Or lowball and use RBF if you fail.  That's precisely how it's supposed to work.

The MPEX and #B-A types (among many others) don't agree there is a problem with Bitcoin, and I concur.

~Anyone can (in ~complete security) use Bitcoin to send ~any amount of money ~anywhere in the world, ~instantly and for ~free.

To me, that seems like a fucking miracle on the order of life, consciousness, wheels, fire, and Teflon.  Where's the issue?

In exactly what way does Bitcoin not work just fine?

The problem is that with 3 tx per second ~ 260,000 tx a day not even the wealthiest 0.004% of the world's population would be able to use Bitcoin just once a day.

At present, only 0.004% of nerds use Bitcoin once a day.  I'd love for Bitcoin to already have the "problem" you described, but it just ain't so.

You are using the present tense "The problem is that..." to describe a remote future possibility I usually refer to as Cryptopia, where the wealthiest bid up to astronomical levels the price of precious space in the Holy Ledger.

It should say "The problem may be that..."

Regardless, such adversity would only strengthen our antifragile system, and the answer to the tps problem looks like sidechains, altchains, off-chains, and payment channels.

Given the disappearing block subsidy and high price volatility, the sooner we wean Bitcoin off mother's milk and onto real food (tx fees) the better.

[r0ach]

It appears blockstream finally sees the light:

"Another idea that has been gathering support is Blockstream CEO and Hashcash inventor Dr. Adam Back's “2-4-8” quick fix, which would incrementally increase to the limit to 8 megabytes in three steps over four years time."

Going to 2mb first is kind of pointless.  Might as well fork straight to 8 since it's going there in a couple years anyway.

[iCEBREAKER]

It appears blockstream finally sees the light:

"Another idea that has been gathering support is Blockstream CEO and Hashcash inventor Dr. Adam Back's “2-4-8” quick fix, which would incrementally increase to the limit to 8 megabytes in three steps over four years time."

Going to 2mb first is kind of pointless.  Might as well fork straight to 8 since it's going there in a couple years anyway.

Just because you FAIL to see the point of going to 2MB first does not imply the rest of us do not.

In fact the opposite conclusion, that there IS a point to first doing a small increase in an abundance of caution, is indicated by the fact Dr. Back's BIP248 (sorry gmax but we're all calling it that) proposal enjoys a plurality if not an outright majority of support.

[pogress]

For Bitcoin to pass whatever glass ceilings it has, it probably needs at least enough TPS to process all of the world's large, high value transactions such as houses, cars, yachts, etc.

At 7 TPS, you can do 604,800 transactions per day (probably closer to 500k in reality but whatever). 

There are 14,000 houses and 43,000 cars sold per day in the US.  1 meg blocks are probably too small, or dangerously close to maxing out if handling all of the world's house and car transactions.  We don't need to scale to gigantic blocks, but something like 8-10 meg blocks is probably what it needs to scale to for being a major player in the world, without a glass ceiling, where most things of several thousands of dollars in value are done on-chain.  This is an achievable goal without things like the Lightning network existing.

You cant expect 8-10 meg blocks will work forever. First, if you agree once there is constantly higher demand than the blocksize allows to use, the Bitcoin will become highly unrelieable (long tx confirmation times + many transactions will never confirm because simply there is constantly higher demand than the blocksize allows to use). In such scenario it really doesnt make sence to use such unreliable Bitcoin and all will look elsewhere for a working altcoin.

[Paritor]

It appears blockstream finally sees the light:

"Another idea that has been gathering support is Blockstream CEO and Hashcash inventor Dr. Adam Back's “2-4-8” quick fix, which would incrementally increase to the limit to 8 megabytes in three steps over four years time."

Going to 2mb first is kind of pointless.  Might as well fork straight to 8 since it's going there in a couple years anyway.

Just because you FAIL to see the point of going to 2MB first does not imply the rest of us do not.

In fact the opposite conclusion, that there IS a point to first doing a small increase in an abundance of caution, is indicated by the fact Dr. Back's BIP248 (sorry gmax but we're all calling it that) proposal enjoys a plurality if not an outright majority of support.

I agree with we can increase to 2MB first to see the effect. Then we can decide what to do next.

[CIYAM]

Yet another pointless block size topic?

If you aren't interested in the *real problems* (like potentially empty blocks due to miners behind the GCF giving up on including txs at all and just trying to grab the coinbase rewards) then what on earth do you think this topic will achieve?

If you're also not aware of SPV mining and the problems it brings to the Bitcoin network security then again you are simply wasting everyone's time.

Also @theymos has nothing to do with Bitcoin development itself (so the title of this topic is basically silly - don't confuse BIP101 as being "the entire debate about block size").

[hdbuck]

It appears blockstream finally sees the light:

"Another idea that has been gathering support is Blockstream CEO and Hashcash inventor Dr. Adam Back's “2-4-8” quick fix, which would incrementally increase to the limit to 8 megabytes in three steps over four years time."

Going to 2mb first is kind of pointless.  Might as well fork straight to 8 since it's going there in a couple years anyway.

Just because you FAIL to see the point of going to 2MB first does not imply the rest of us do not.

In fact the opposite conclusion, that there IS a point to first doing a small increase in an abundance of caution, is indicated by the fact Dr. Back's BIP248 (sorry gmax but we're all calling it that) proposal enjoys a plurality if not an outright majority of support.

I agree with we can increase to 2MB first to see the effect. Then we can decide what to do next.

who's we?

[iCEBREAKER]

It appears blockstream finally sees the light:

"Another idea that has been gathering support is Blockstream CEO and Hashcash inventor Dr. Adam Back's “2-4-8” quick fix, which would incrementally increase to the limit to 8 megabytes in three steps over four years time."

Going to 2mb first is kind of pointless.  Might as well fork straight to 8 since it's going there in a couple years anyway.

Just because you FAIL to see the point of going to 2MB first does not imply the rest of us do not.

In fact the opposite conclusion, that there IS a point to first doing a small increase in an abundance of caution, is indicated by the fact Dr. Back's BIP248 (sorry gmax but we're all calling it that) proposal enjoys a plurality if not an outright majority of support.

I agree with we can increase to 2MB first to see the effect. Then we can decide what to do next.

who's we?

*crickets chirping*

[johnyj]

If satoshi set the limit to 8MB, then during the spamming attack in October many nodes would have already failed. Current mainstream computer hardware and home network simply can not handle 8MB blocks

If you look from the view of miners, a large block size limit will not affect them, since miners will just mine an empty block without including any transactions, then that empty block will be quickly broadcasted to the rest of the network, regardless of how large the block size limit is

On the contrary, if a miner would like to include 16K transactions to form a 8MB block, that large block will greatly slowdown his processing, making him impossible to verify the transactions and broadcast that block in 10 minutes, e.g. his block get orphaned

So from miner's point of view, they always prefer to include as little transaction as possible to maximize their mining income, unless the fee is so high that including those transactions will greatly increase their mining income. But anyway, beyond certain size, no matter how high the fee is, they would not be able to include new transactions, since their block simply can not be relaid to the rest of the network in 10 minutes

The only solution that scales is clearing based design like lightning network or offchain centralized local clearing center, and bitcoin network will work more like Fedwire system, only handle large transactions between thousands of banks (Which means no change is needed since Fedwire system which handles all the transactions between US banks only do 4 transactions per second now)

In fact, clearing and settlement is a very smart design in financial world, it works with all kinds of financial transactions. Being decentralized does not mean fully distributed, there will be thousands of bitcoin hubs that do clearing between them, and bitcoin network is still enough decentralized

[CIYAM]

I concur with @johnyj that Bitcoin has been designed (whether Satoshi really intended it to be or not) as a SWIFT replacement rather than a VISA replacement.

Rather than try and turn a turkey into a chicken I think we ought to simply build on what we have.

[tl121]

If satoshi set the limit to 8MB, then during the spamming attack in October many nodes would have already failed. Current mainstream computer hardware and home network simply can not handle 8MB blocks

During the "attack" a.k.a. "test" my little Atom based Nuc had no trouble keeping up with continuous 1 MB blocks and would have been able to keep up with 8 MB blocks as well, based on my usage statistics. There would have been no problem with network bandwidth, either, so long as I was just verifying blocks and not attempting to mine.  This machine was hardly mainstream computing.  If there had been a performance problem I had two other computers that I could have switched on to do this work, but I didn't see the point in burning extra electricity when a 15 watt machine sufficed.

[CIYAM]

During the "attack" a.k.a. "test" my little Atom based Nuc had no trouble keeping up with continuous 1 MB blocks and would have been able to keep up with 8 MB blocks as well, based on my usage statistics. There would have been no problem with network bandwidth, either, so long as I was just verifying blocks and not attempting to mine.  This machine was hardly mainstream computing.  If there had been a performance problem I had two other computers that I could have switched on to do this work, but I didn't see the point in burning extra electricity when a 15 watt machine sufficed.

You are not getting the point - it is not your machine but your internet connection that matters.

If you are mining behind the GCF you can add as many machines as you like but that won't increase your internet bandwidth.

[Soros Shorts]

During the "attack" a.k.a. "test" my little Atom based Nuc had no trouble keeping up with continuous 1 MB blocks and would have been able to keep up with 8 MB blocks as well, based on my usage statistics. There would have been no problem with network bandwidth, either, so long as I was just verifying blocks and not attempting to mine.  This machine was hardly mainstream computing.  If there had been a performance problem I had two other computers that I could have switched on to do this work, but I didn't see the point in burning extra electricity when a 15 watt machine sufficed.

I wish those crap machines that take 5 seconds or longer to verify and relay a block would just go away. They are just slowing down block propagation. I hope your machine is not one of these.

How many seconds does your Atom CPU pin at 100% everytime a new block is found and passes through your node?

[iCEBREAKER]

During the "attack" a.k.a. "test" my little Atom based Nuc had no trouble keeping up with continuous 1 MB blocks and would have been able to keep up with 8 MB blocks as well, based on my usage statistics. There would have been no problem with network bandwidth, either, so long as I was just verifying blocks and not attempting to mine.  This machine was hardly mainstream computing.  If there had been a performance problem I had two other computers that I could have switched on to do this work, but I didn't see the point in burning extra electricity when a 15 watt machine sufficed.

You are not getting the point - it is not your machine but your internet connection that matters.

If you are mining behind the GCF you can add as many machines as you like but that won't increase your internet bandwidth.

But But But...Newegg sells hard drives for less than a penny per gigabyte!

But But But...look how tiny this 128GB CF is!

But But But...I like to drive forward while using the rearview mirror to navigate, because past Xs being Ys guarantees future Xs will be Ys as well!

[johnyj]

If satoshi set the limit to 8MB, then during the spamming attack in October many nodes would have already failed. Current mainstream computer hardware and home network simply can not handle 8MB blocks

During the "attack" a.k.a. "test" my little Atom based Nuc had no trouble keeping up with continuous 1 MB blocks and would have been able to keep up with 8 MB blocks as well, based on my usage statistics. There would have been no problem with network bandwidth, either, so long as I was just verifying blocks and not attempting to mine.  This machine was hardly mainstream computing.  If there had been a performance problem I had two other computers that I could have switched on to do this work, but I didn't see the point in burning extra electricity when a 15 watt machine sufficed.

Every machine on the way add a delay,  10 such notebooks will be enough to make some miner not receiving blocks and keep building blocks on a chain which will be invalid later, thus in their blocks, transactions with one confirmation would become invalid after 10 minutes

I just watched Peter Todd's speak from Hong Kong conference, he had an interesting point: It is the hash power decides, e.g. when chinese mining pools have large hash power and slow network, while rest of the world have super fast network but less hash power, then the miners in western will often be working on wrong blocks due to they could not receive correct blocks from china in time and they are slower than chinese miners in finding blocks

[tl121]

During the "attack" a.k.a. "test" my little Atom based Nuc had no trouble keeping up with continuous 1 MB blocks and would have been able to keep up with 8 MB blocks as well, based on my usage statistics. There would have been no problem with network bandwidth, either, so long as I was just verifying blocks and not attempting to mine.  This machine was hardly mainstream computing.  If there had been a performance problem I had two other computers that I could have switched on to do this work, but I didn't see the point in burning extra electricity when a 15 watt machine sufficed.

I wish those crap machines that take 5 seconds or longer to verify and relay a block would just go away. They are just slowing down block propagation. I hope your machine is not one of these.

How many seconds does your Atom CPU pin at 100% everytime a new block is found and passes through your node?

Why would this matter?  The time would be infinite if the node were powered down.  It's not that I'm being paid to run a node.  Big miners and mining pools are not sending their blocks through my node.  If the presence of slow nodes has a negative effect on the network as a whole then it is only because the peer to peer software in bitcoin uses poor algorithms and/or has poor performance.

 BTW,  when I configured this node I deliberately chose the slowest new machine that I could get. I wanted to see first hand how efficient or inefficient the software is and what the possibilities of growth were.  I also didn't want a lot of high powered machines creating heat and making noise.  In any event, I am subject to a monopolistic ISP and limited to an upload rate of about 50 KB/s.  There will be what you call a long delay even were the node to be infinitely fast.