Topic: The MAX_BLOCK_SIZE fork - page 7. (Read 35619 times)

After there are 25BTC of fees per block to replace the mining reward instead of .25, by transactions that have to pay to get in a block in a reasonable amount of time, then it might be time to consider a larger block size. Until then it needs to stay scarce. I have 2.5GB of other people's gambling on my hard drive, because it's cheap.

Read my calculation above. With each transaction just paying $0.021 as fee, we have more than enough money to pay miners to handle 4000 transaction per second.

Couldn't the limit be adjusted with every difficulty change so that it is approximately in-line with the demand of the previous difficulty period? If the block size were capped near the transaction volume ceiling there would be still be incentive to include mining fees while never running the risk of running out of block space.

he, the "lead developer" is on the same side I am, nice.


How difficult would it be to implement it on bitcoind right now, without touching the 1Mb hard limit?
I mean the multiple limits and tolerance levels idea.

Miners using bitcoind would be able to set in config file a list of value pairs. One value would be a size limit, the other the amount of blocks longer a chain breaking that limit would have to be in order to you to accept building on top of it. Do you see what I'm saying?
That could be done right now on bitcoind, with the sole condition that anything above 1Mb will be rejected no matter what.

So you think the max block size is not a pressing problem, don't you?

Ummm, see this old forum thread, where Satoshi says:

There are many merits to etotheipi's writing but what he proposes massive _increases_ the IO and computational cost of running a full node (or a fully validating but historyless node) over a plain committed UTXO set for validation. The increased node burden is one of the biggest arguments against what he's proposing and I suspect will ultimately doom the proposal.

I have seen nothing proposed except moore's law that would permit full validation on "desktop" systems with gigabyte blocks.

There are plenty of soft limits in bitcoin (like the 500k softlimit for maximum block size). The 1MB limit is not soft. I'm not aware of any evidence to suggest that it was temporary from the start— and absent it I would have not spent a dollar of my time on Bitcoin: without some answer to how the system remains decentralized with enormous blocks and how miners will be paid to provide security without blockspace scarcity or cartelization the whole idea is horribly flawed.  I also don't think a network rule should be a suicide pact— my argument for the correctness of making the size limited has nothing to do with the way it always was, but that doesn't excuse being inaccurate about the history.

So, full nodes act as banks and issue Bitcoin-denominated instruments to their clients. Maybe the clients do not even have to trust the banks, thanks to some kind of cryptographical magic. Because of the economical scale of the transactions, only big companies and financial institutions have any reason to actually make a Bitcoin transaction, and these kinds of actors can run any kind of full node anyway. The clients can run something similar to what is outlined in this thread: https://bitcointalksearch.org/topic/ultimate-blockchain-compression-w-trust-free-lite-nodes-88208

Actually, that thread outlines the way that future PCs (if not smartphones) could conceivably run a full node (or "almost-full" node) even with no limit / floating limit.

I just can't see why this artificial limit that was intended as temporary from the start should be accepted as an immutable part of the protocol.

There is going to be a hard fork in any case, more likely sooner than later. It should be planned beforehand, if we care about Bitcoin at all.

Setup your own Electrum server with your computer at home and verify your house transaction with a smart-phone through it. Therefore you don't need to trust a third party

A smart-phone is never designed to run a bitcoin full-node

Moreover, the sentence "It only becomes a problem if Bitcoin is a success.  If Bitcoin is a success, by definition it isn't a problem." is self-contradicting

Let's assume a miner with moderate hashing power can mine 1 in 10000 blocks (i.e. one block in 10 weeks). With $50000/block, he will get about $5/block.

2500000tpb (VISA scale) means about 1GB/block. Currently a 2000GB drive costs about $100, or $0.05/GB. Therefore, the harddrive cost is only 1% of his mining income. It's negligible. (and harddrive will be much cheaper in the future)

A quad-core Intel core i7 is able to handle 4000tps (https://en.bitcoin.it/wiki/Scalability#CPU) at 77W. Assuming $0.15/kWh, it costs about $0.012/h, or $0.002/block. Even energy is 10x more expensive in the future, it's still negligible. (and CPU will be much efficient in the future)

1GB/block needs a bandwidth of 4.3TB/month. Including all overhead it may take 10TB/month, and may cost $300/month currently for a dedicated server in datacentre. It is $300/(30*24*6) = 0.069/block. Again, it is negligible comparing with the $5/block reward.

He will still earn $5-0.05-0.002-0.069 = $4.879/block after deducting the harddrive, CPU, and bandwidth cost. It is $29/hr or $21077/month and is a ridiculous amount given he only owns 0.01% of total hashing power. He still needs to pay for the electricity bill for the mining equipment. It is hard to estimate but even if he uses 90% of the earning for the electricity bill, he will still earn $2107/month.

Lets not get ahead of ourselves here.  I expect that we will have a multi-layered system the vast majority of the transactions being made off-chain.

Additionally; who is to say that one wouldn't want to verify their house transaction with a smart-phone.

People completely mis-judge the free-market.   If you have to use alt-chains because the fees are so high, well isn't that a success of bitcoin already!  By no stage has bitcoin failed then.

The argument for larger blocks is VALID for a proticol that isn't Bitcoin.  However it is a catch 22, for Bitcoin.  It only becomes a problem if Bitcoin is a success.  If Bitcoin is a success, by definition it isn't a problem.

Currently (Feb 2013), we have about 50000 tx per day, or 0.579 tx per second (tps), or 347 tx per block (tpb). We are paying miners 25 BTC per block, or $500 per block at current rate. If bitcoin becomes the VISA scale, it has to handle 4000 tps, or 2400000 tpb, or 6916x of the current volume. To keep mining profitable, we may need to pay $50000 per block (to pay electricity, harddrive space, bandwidth, CPU time for ECDSA). As the block reward will become 0 in the future, this $50000 has to be covered by fee. Since we will have 2400000 tpb, each tx will pay $0.021, not too bad when you can send unlimited amount of money to anywhere in the world in no time.

That means mining is profitable even without any block size limit.

On the other hand, the 1MB constraint will certainly kill bitcoin. 1MB is only about 2500 tpb, or 0.1% of VISA scale. We are already at 13.9% of this limit. If we don't act before problem arises, people will start migrating to alt-coins.

The only way to do it is to get most major clients to accept larger blocks AFTER a future specified date.

That way once say "2017 Dec 31" rolls around 90% of BTC users will all accept larger blocks at the same time and the confusion will be minimal.

This is not that hard honestly just get MyWallet, Armory, the Satoshi client and Electrum programmers/distributors to agree on a date say "2020" and a new limit "100mb" and you're done.
I don't see why that many people would reject this change and as such the new standard should be rolled out way before it takes effect.

Miners are irrelevant in this, but they should welcome the change; BTC will never grow if normal people can't use it. We would be right back to trusting banks, only backed by BTC instead of gold this time. We all know what a wild success THAT has been!

What is the use of having portable devices act as full nodes if you can't (because of fees) use bitcoin for purchasing anything smaller than a house? As I see it, your argument is not valid. With 1MB blocksize limit, even if Bitcoin remains a relatively small niche currency, the limit will act as a hard constraint on the potential utility of the currency. Of course, once we start hitting the limit, it will hurt Bitcoin's public image so much that it's conceivable so many people will move away from Bitcoin that we get few more years of time to fix the issue.

could maybe future-date a larger block size allowance by doing something like at block 250000 2MB blocks become allowed .. or something like that. Why not anyway. And for tx fees in blocks, more tx with lower fees vs less tx with higher fees = same amount of fees in a block that took essentially the same resources to compute, so it seems pretty moot. Am I missing something?

guys, its not THAT urgent. the first thing that happens when the blocks start filling up is that transactions with low or no fees will be delayed so much that they are no longer possible.
right now, transaction fees are 0,011% of the current transaction volume. on average you pay 1/10000 of the transfered amount - or average 1/5 of an us-dollar cent -as a fee. so there is plenty of room before non-micro-transactions are noticably affected by this in any way. satoshi dice however might run into problems much sooner i suppose.

If we want to do it, this is the best moment. As ASICs start running, mining becomes less decentralized for a short period, which means we don't need to persuade so many people

Isn't waiting dangerous?

These other layers may indeed assume that role someday, but it should happen because they prove to be superior on their own merits, not because the capabilities of blockchain transfers are deliberately crippled.

Highly agreed, I also think all the changes could be done at higher level without modifying the original protocal, to keep the integrity of the blockchain