60% of hashrate including 2 major exchanges agree to raise block size to 8MB - page 2.

TierNolan

legendary

Activity: 1232

Merit: 1094

Quote from: shorena on June 18, 2015, 02:30:45 PM

Latest as in 0.11? 0.10.x does not allow to run a node in pruning mode AFAIK.

Right, sorry, meant 0.11, so next release.

Quote from: spazzdla on June 18, 2015, 02:34:54 PM

I was under the impression this would make it difficult to follow transactions from the begining?

Yes, but each node only needs to download it once and doesn't need to keep everything.

[Edit]

There is a suggestion on the mailing list for each node to store some of the blocks.

If everyone stored 1% of the blockchain, then you can download each block from lots of different nodes. Once you are synced, you can prune your block store.

spazzdla

legendary

Activity: 1722

Merit: 1000

Quote from: TierNolan on June 18, 2015, 02:21:06 PM

Quote from: spazzdla on June 18, 2015, 01:39:46 PM

20MB is way to much of a jump.. the Blockchain is already very large which prevents people from running a full node. This is just kicking the can down the road.. we should not think of this as a fix to the problem. We are suppose to be better than the central banks.. I am on board for 8MB but after that we need to stop the idea of just kicking the problem to our children.

They have added pruning to the latest release. That reduces the amount of disk space required to run a full node. It stores at least 288 blocks or 500MB, whichever is larger. At 20MB per block, that is 5.7GB.

I was under the impression this would make it difficult to follow transactions from the begining?

shorena

copper member

Activity: 1498

Merit: 1528

No I dont escrow anymore.

Quote from: TierNolan on June 18, 2015, 02:21:06 PM

Quote from: spazzdla on June 18, 2015, 01:39:46 PM

20MB is way to much of a jump.. the Blockchain is already very large which prevents people from running a full node. This is just kicking the can down the road.. we should not think of this as a fix to the problem. We are suppose to be better than the central banks.. I am on board for 8MB but after that we need to stop the idea of just kicking the problem to our children.

They have added pruning to the latest release. That reduces the amount of disk space required to run a full node. It stores at least 288 blocks or 500MB, whichever is larger. At 20MB per block, that is 5.7GB.

Latest as in 0.11? 0.10.x does not allow to run a node in pruning mode AFAIK.

TierNolan

legendary

Activity: 1232

Merit: 1094

Quote from: spazzdla on June 18, 2015, 01:39:46 PM

20MB is way to much of a jump.. the Blockchain is already very large which prevents people from running a full node. This is just kicking the can down the road.. we should not think of this as a fix to the problem. We are suppose to be better than the central banks.. I am on board for 8MB but after that we need to stop the idea of just kicking the problem to our children.

They have added pruning to the latest release. That reduces the amount of disk space required to run a full node. It stores at least 288 blocks or 500MB, whichever is larger. At 20MB per block, that is 5.7GB.

spazzdla

legendary

Activity: 1722

Merit: 1000

20MB is way to much of a jump.. the Blockchain is already very large which prevents people from running a full node. This is just kicking the can down the road.. we should not think of this as a fix to the problem. We are suppose to be better than the central banks.. I am on board for 8MB but after that we need to stop the idea of just kicking the problem to our children.

Carlton Banks

legendary

Activity: 3430

Merit: 3080

Quote from: dogie on June 18, 2015, 04:51:18 AM

Quote from: tupelo on June 18, 2015, 12:13:24 AM

Why exactly 8 MB? Should Bitcoin really be dictated by archaic and irrational Chinese superstition? Or is there more substance to this number?

Yes, yes it should, because 8MB is an arbitrary number just like 20MB. And just like 20MB, no one is predicting 8MB to be consumed any time soon and definitely not so quick that another consensus could not be reached to increase it. Other arbitrary bitcoin stuff: A block targeted every 10 minutes, block reward halved every 4 years, 50btc block reward yada yada yada. None of these things are set in stone, based on anything in particular or vitally important.

The 1 MB limit, and certainly the 10 minute discovery target, were objectively chosen, albeit still in a can-kicking guesswork category of design decisions. There are *ahem* several arbitrary numbers that wouldn't work in their place, so that description isn't very apt.

dogie

legendary

Activity: 1666

Merit: 1185

dogiecoin.com

Quote from: tupelo on June 18, 2015, 12:13:24 AM

Why exactly 8 MB? Should Bitcoin really be dictated by archaic and irrational Chinese superstition? Or is there more substance to this number?

Yes, yes it should, because 8MB is an arbitrary number just like 20MB. And just like 20MB, no one is predicting 8MB to be consumed any time soon and definitely not so quick that another consensus could not be reached to increase it. Other arbitrary bitcoin stuff: A block targeted every 10 minutes, block reward halved every 4 years, 50btc block reward yada yada yada. None of these things are set in stone, based on anything in particular or vitally important.

bitnanigans

sr. member

Activity: 266

Merit: 250

Personally, I think the hard limit should be removed completely. Remember the "640K is enough memory for everyone" quote? Look at where we are today. Technology advances at a very quick pace, and we'll have more than enough storage and bandwidth to handle the increases as adoption grows.

TierNolan

legendary

Activity: 1232

Merit: 1094

I think 8MB is just a compromise between 1MB and 20MB.

I think they are worried about the bandwidth between China and the rest of the world. Very large blocks could cause problems for them.

There are a few different network simulators that give different results and it depends on what parameters you set.

They are concerned that pools outside China might produce large blocks and it will take longer for those blocks to reach them and that would mean they waste hashing power.

Under some conditions they might benefit from lower bandwidth into China. Assuming >50% of the hashing power is using Chinese pools and a Chinese pool and a non-Chinese pool both find a block at the same time, then the Chinese pool's block will reach a majority of the hashing power before the non-Chinese pool. If the non-Chinese block is 20MB, then it would take even longer to enter China.

Mining farms and mining pools don't have to be at the same location. It would be possible for miners in China to use mining pools outside China, if it ever became a problem. This could shift the majority of the hashing power out of China, and then mining pools would have to leave China in order to have a good connection to the majority of the hashing power.

tupelo

member

Activity: 99

Merit: 10

Why exactly 8 MB? Should Bitcoin really be dictated by archaic and irrational Chinese superstition? Or is there more substance to this number?

jl2012

legendary

Activity: 1792

Merit: 1111

5 largest mining pools in China, including 2 of the busiest exchanges in the world, release a joint declaration to support raising the MAX_BLOCK_SIZE to 8MB. They currently control 60% of the network hashrate.

https://imgur.com/a/LlDRr

Chinese companies are operating under a very oppressive environment. All internet activities are strictly censored and outbound bandwidth is limited. They still agree to increase the block size.

Although one may argue that on this issue merchants' view is more important than miners, the hardfork won't be successful without miners' support. And don't forget, this statement includes 2 major exchanges, BTCChina and Huobi.

I hope this would conclude the debate around "raise block size or not" and "how much to raise".

I hope we could focus on the pathway leading to 8MB. Should that be a simple raise or a step function? If a step function, how? Should we limit other parameters, e.g. sigop count and UTXO growth?

The hard fork should also consider the pathway beyond 8MB if we don't want to repeat the debate (too early).

Topic: 60% of hashrate including 2 major exchanges agree to raise block size to 8MB - page 2. (Read 3830 times)