What are the downsides to 8MB blocks?

[olli]

8MB per 10 minutes is about 13.6 KB/s. What was the problem again?

Not that I think that simply increasing the block size limit to 8MB is a good solution (I'd rather see something more dynamic) but if 13.6 KB/s is actually a problem, maybe you should fix your internet, rather than debate over Bitcoin.

Miners cant wait 10 minutes to have their newly mined block trasfered to the next node.

But they can wait 1.25 minutes to have their newly mined block trasfered to the next node?

[1715443733]

1715443733

[Scamalert]

Arg, you know. I call the whole thing for the "Cafe Latte problem".
If everybody start using bitcoin and everybody want to buy a cafe latte everyday paying with bitcoin then will the 8MB block not be big enough. But there are tons of shitcoins to take care of the Latte transactions (hate to admit it, some of the shitcoins are better at these types of transaction than bitcoin). Bitcoin should/will be the backbone of the world economy.

[fonsie]

All the people who are in favor of keeping 1MB blocks are afraid they will not have the required bandwidth to download bigger blocks and their favorite DivX movie for their state of the art 20$ player.



Perhaps we should keep another vote for 1024bytes(1kilobyte) blocksize, this way it will become more decentralized and the amount of nodes will triple in a day. .....or not.

[ajareselde]

Downsides to 8MB blocks are that the full node count will be even lower than it is today, and the main arguments for that are the bandwidth increase,
and actual size of the blockchain (storage), but imho, in today's world, none of those should represent a problem to 99.999% bitcoin users.

cheers

[jc12345]

Ease of use, mass adoption and having tx processed as fast as possible is the name of the game. This can never happen with a local large blockchain. The future is in chain providers or custodians and mass users with thin clients making use of the chain custodians. Bitcoin is becoming so centralised in anyways from a mining point of view that the pools and the nodes out there can have local chains. It might even come to it to not have local chains on PCs at all but having a network of VPSes in the cloud with the chain on them and most VPSes are on thick and fast pipes that wouldn't feel 8MB blocks at all or whatever sizes BIP101 will throw at them in future.

[johnyj]

Uncertainty. The larger the change, the larger the uncertainty, and for financial system you want changes to be as small as possible to avoid uncertainty, uncertainty increase the chance of totally unpredicted disaster

[tadakaluri]

Chinese mining pools seem to have played the decisive role in the debate. As CoinFox wrote recently, a joint statement of BTCChina, Huobi, F2Pool, BW and Antpool favoured a compromise – an increase of the block size to 8MB instead of 20MB.

China’s five largest mining pools gathered at the National Conference Center in Beijing to hold a technical discussion about the ramifications of increasing the max block size on the Bitcoin network. In attendance were F2Pool, BW, BTCChina, Huobi.com, and Antpool. After undergoing deep consideration and discussion, the five pools agree that while the block size does need to be increased, a compromise should be made to increase the network max block size to 8 megabytes. We believe that this is a realistic short term adjustment that remains fair to all miners and node operators worldwide.

Why upgrade to 8MB but not 20MB?

1.Chinese internet bandwidth infrastructure is not built out to the same advanced level as those found in other countries.

2.Chinese outbound bandwidth is restricted; causing increased latency in connections between China & Europe or the US.

3.Not all Chinese mining pools are ready for the jump to 20MB blocks, and fear that this could cause an orphan rate that is too high.

The bitcoin miners of China agree that the blocksize must be increased, but we believe that increasing to 8MB first is the most reasonable course of action. We believe that 20MB blocks will cause a high orphan rate for miners, leading to hard forks down the road. If the bitcoin community can come to a consensus to upgrade to 8MB blocks first, we believe that this lays a strong foundation for future discussions around the block size. At present, China’s five largest mining pools account for more than 60% of the network hashrate.

The long debate over the bitcoin block size is over. The updates being done to the bitcoin code have been published by the core developer Gavin Andresen on GitHub. The block size will change from 1MB to 8MB, with further doubling every two years – so in 2018, it will reach 16 MB, then 32MB in 2020 and so on.

2016: 8 MB
2018: 16 MB
2020: 32 MB
2022: 64 MB
2024: 128 MB
2026: 256MB
2028: 512 MB
2030: 1024 MB
2032: 2048 MB
2034: 4096 MB

Personally I believe that this well overestimating the rate internet connection-speeds and hard-disk space will grow. Just look at ADSL or even cable for instance.

4096 MB per 10minutes in a year will grow to: 205TB per year worth of storage.

I wonder if they took in account the laws of physics in this equation. It is not impossible but who can and will pay for such an investment just to be a unpaid bitcoin client (full node).

[lottery248]

in order to prevent unconfirmed transaction flooding, if we allowed to increase the size of blockchain, people would be able to do numbers of transaction and lead bitcoin into unsustainable to full node users.

[Amph]

8MB per 10 minutes is about 13.6 KB/s. What was the problem again?

Not that I think that simply increasing the block size limit to 8MB is a good solution (I'd rather see something more dynamic) but if 13.6 KB/s is actually a problem, maybe you should fix your internet, rather than debate over Bitcoin.

Miners cant wait 10 minutes to have their newly mined block trasfered to the next node.

But they can wait 1.25 minutes to have their newly mined block trasfered to the next node?

i think not, ideally the lower the better so if we assume few seconds of waiting you end already in x600(1 seconds of waiting) of 13.6 KB/s, which is around 8 MB/s and it is not small at all for many...

[gentlemand]

Personally I believe that this well overestimating the rate internet connection-speeds and hard-disk space will grow. Just look at ADSL or even cable for instance.

I don't think storage will ever be much of an issue. Bandwidth sure as shit is and they seem to be a little casual about that with the ever increasing block size proposal.

Nothing you can do about it yourself. You have to wait for your friendly phone company to come around and do some digging up and install major hardware upgrades.

My bandwidth is unchanged in over a decade and there's nothing on the horizon either. It suits my purposes fine but it's not going to like running a node in such a future.

[DarkHyudrA]

Chinese mining pools seem to have played the decisive role in the debate. As CoinFox wrote recently, a joint statement of BTCChina, Huobi, F2Pool, BW and Antpool favoured a compromise – an increase of the block size to 8MB instead of 20MB.

China’s five largest mining pools gathered at the National Conference Center in Beijing to hold a technical discussion about the ramifications of increasing the max block size on the Bitcoin network. In attendance were F2Pool, BW, BTCChina, Huobi.com, and Antpool. After undergoing deep consideration and discussion, the five pools agree that while the block size does need to be increased, a compromise should be made to increase the network max block size to 8 megabytes. We believe that this is a realistic short term adjustment that remains fair to all miners and node operators worldwide.

Why upgrade to 8MB but not 20MB?

1.Chinese internet bandwidth infrastructure is not built out to the same advanced level as those found in other countries.

2.Chinese outbound bandwidth is restricted; causing increased latency in connections between China & Europe or the US.

3.Not all Chinese mining pools are ready for the jump to 20MB blocks, and fear that this could cause an orphan rate that is too high.

The bitcoin miners of China agree that the blocksize must be increased, but we believe that increasing to 8MB first is the most reasonable course of action. We believe that 20MB blocks will cause a high orphan rate for miners, leading to hard forks down the road. If the bitcoin community can come to a consensus to upgrade to 8MB blocks first, we believe that this lays a strong foundation for future discussions around the block size. At present, China’s five largest mining pools account for more than 60% of the network hashrate.

The long debate over the bitcoin block size is over. The updates being done to the bitcoin code have been published by the core developer Gavin Andresen on GitHub. The block size will change from 1MB to 8MB, with further doubling every two years – so in 2018, it will reach 16 MB, then 32MB in 2020 and so on.

2016: 8 MB
2018: 16 MB
2020: 32 MB
2022: 64 MB
2024: 128 MB
2026: 256MB
2028: 512 MB
2030: 1024 MB
2032: 2048 MB
2034: 4096 MB

Personally I believe that this well overestimating the rate internet connection-speeds and hard-disk space will grow. Just look at ADSL or even cable for instance.

4096 MB per 10minutes in a year will grow to: 205TB per year worth of storage.

I wonder if they took in account the laws of physics in this equation. It is not impossible but who can and will pay for such an investment just to be a unpaid bitcoin client (full node).

But remember that this is the MAX block size, it isn't the I MUST FILL 100% of block size. Of course it's way high, even 128MB is ridiculous if it ever get completely full, to some people it's justa matter of seconds or a minute to download but I do worry about the uploading of those blocks.

[AtheistAKASaneBrain]

Personally I believe that this well overestimating the rate internet connection-speeds and hard-disk space will grow. Just look at ADSL or even cable for instance.

I don't think storage will ever be much of an issue. Bandwidth sure as shit is and they seem to be a little casual about that with the ever increasing block size proposal.

Nothing you can do about it yourself. You have to wait for your friendly phone company to come around and do some digging up and install major hardware upgrades.

My bandwidth is unchanged in over a decade and there's nothing on the horizon either. It suits my purposes fine but it's not going to like running a node in such a future.

Actually I don't think bandwith is that much of an issue. It will take a bit longer to sync but that's that, the worst part will be downloading the blockchain the first time, then if you open it daily is not that bad.

[Derrike]

Personally I believe that this well overestimating the rate internet connection-speeds and hard-disk space will grow. Just look at ADSL or even cable for instance.

I don't think storage will ever be much of an issue. Bandwidth sure as shit is and they seem to be a little casual about that with the ever increasing block size proposal.

Nothing you can do about it yourself. You have to wait for your friendly phone company to come around and do some digging up and install major hardware upgrades.

My bandwidth is unchanged in over a decade and there's nothing on the horizon either. It suits my purposes fine but it's not going to like running a node in such a future.

Actually I don't think bandwith is that much of an issue. It will take a bit longer to sync but that's that, the worst part will be downloading the blockchain the first time, then if you open it daily is not that bad.

40 gb!!!!
With my connection speed the block chain will be downloaded in 4 to 5  days or maybe take a week.

Currently I've installed new os on my pc and electrum is also taking a while to synchronize.

[fonsie]

Personally I believe that this well overestimating the rate internet connection-speeds and hard-disk space will grow. Just look at ADSL or even cable for instance.

I don't think storage will ever be much of an issue. Bandwidth sure as shit is and they seem to be a little casual about that with the ever increasing block size proposal.

Nothing you can do about it yourself. You have to wait for your friendly phone company to come around and do some digging up and install major hardware upgrades.

My bandwidth is unchanged in over a decade and there's nothing on the horizon either. It suits my purposes fine but it's not going to like running a node in such a future.

Actually I don't think bandwith is that much of an issue. It will take a bit longer to sync but that's that, the worst part will be downloading the blockchain the first time, then if you open it daily is not that bad.

40 gb!!!!
With my connection speed the block chain will be downloaded in 4 to 5  days or maybe take a week.

Currently I've installed new os on my pc and electrum is also taking a while to synchronize.

Why didn't you make a backup? It took me 4/5 hours last time I downloaded the blockchain.

[TrueBeliever]

i have had trouble trying to get a node up and running as it is.

Yes 8 Mb is the max size and blocks won't always be that big, but in general we will start to see many blocks greater than the current 1 Mb size.

It will start to take node hosting out of the realm of the average punter lending his spare disk space and CPU cycles "for the good of the network"

I haven't figured where I stand on this yet: support or reject the increase the blocksize increase, support buying lattes with altcoins (this is not an altogether stupid idea), the whole thing is doing my head in.

The Visa or Paypal network transfers a huge number of transactions per minute.  If Bitcoin went that viral are node hosters going to store that magnitude of transactions on their home PC? Home PCs using ADSL are not quite the type of infrastructure that could handle anything like that, not to mention the inefficiencies that decentralization brings.

This is a serious problem that poses philosophical as well as technical questions.  Not even Andresen has presented a convincing argument on this topic.

I believe the final solution to this problem has not yet been formulated.

[Derrike]

Personally I believe that this well overestimating the rate internet connection-speeds and hard-disk space will grow. Just look at ADSL or even cable for instance.

I don't think storage will ever be much of an issue. Bandwidth sure as shit is and they seem to be a little casual about that with the ever increasing block size proposal.

Nothing you can do about it yourself. You have to wait for your friendly phone company to come around and do some digging up and install major hardware upgrades.

My bandwidth is unchanged in over a decade and there's nothing on the horizon either. It suits my purposes fine but it's not going to like running a node in such a future.

Actually I don't think bandwith is that much of an issue. It will take a bit longer to sync but that's that, the worst part will be downloading the blockchain the first time, then if you open it daily is not that bad.

40 gb!!!!
With my connection speed the block chain will be downloaded in 4 to 5  days or maybe take a week.

Currently I've installed new os on my pc and electrum is also taking a while to synchronize.

Why didn't you make a backup? It took me 4/5 hours last time I downloaded the blockchain.

Can I backup my blockchain also
I have backed up the private keys and seed but never knew about backing up the blockchain.

[LiteCoinGuy]

Chinese mining pools seem to have played the decisive role in the debate. As CoinFox wrote recently, a joint statement of BTCChina, Huobi, F2Pool, BW and Antpool favoured a compromise – an increase of the block size to 8MB instead of 20MB.

China’s five largest mining pools gathered at the National Conference Center in Beijing to hold a technical discussion about the ramifications of increasing the max block size on the Bitcoin network. ....

The bitcoin miners of China agree that the blocksize must be increased, but we believe that increasing to 8MB first is the most reasonable course of action. We believe that 20MB blocks will cause a high orphan rate for miners, leading to hard forks down the road. If the bitcoin community can come to a consensus to upgrade to 8MB blocks first, we believe that this lays a strong foundation for future discussions around the block size. At present, China’s five largest mining pools account for more than 60% of the network hashrate.

The long debate over the bitcoin block size is over. The updates being done to the bitcoin code have been published by the core developer Gavin Andresen on GitHub. The block size will change from 1MB to 8MB, with further doubling every two years – so in 2018, it will reach 16 MB, then 32MB in 2020 and so on.

2016: 8 MB
2018: 16 MB
2020: 32 MB
2022: 64 MB
2024: 128 MB

....

finally. a good and conservativ approach.  

[BitcoinHawker]

Do larger blocks give an advantage to the larger mining pools? Ex. Getting some sort of edge on mining the next block after finding one, seeing that the blocks will be larger?

[RoadTrain]

Do larger blocks give an advantage to the larger mining pools? Ex. Getting some sort of edge on mining the next block after finding one, seeing that the blocks will be larger?

Larger blocks give an advantage to larger and better connected pools, so they can propagate them faster. The propagation bottlenecks can fragment the network, where one larger well-intra-connected part (read: China) will be winning more orphan races, thus earning relatively more money -- creating financial incentives for further centralization.

[Jace]

The long debate over the bitcoin block size is over. The updates being done to the bitcoin code have been published by the core developer Gavin Andresen on GitHub. The block size will change from 1MB to 8MB, with further doubling every two years – so in 2018, it will reach 16 MB, then 32MB in 2020 and so on.

Will this make it into Bitcoin Core as well, or is there still a discrepancy between Core and XT ?

If so, exactly who or what still prevents the above patch to make it into Bitcoin Core? (i.e. so we can all stick with just ONE Bitcoin version)