Bitcoin Forum

As the title says, that way we won't ever bother with this question again.

Oh and before you talk shit, just remember that your porn stack is +100GB.  :D

Interesting. We can start the discussion if you answer a simple question.

Could you please explain the benefits of increasing the block size to 50 MB instead of 20 8 MB (aside from higher tps) or using side chains or something else (e.g. lightning network)?

There aren't only benefits about having big blocks... And huge blocks aren't the definitive solution :)

common misconception: "20MB is too big of a jump"

answer: It's just a maximum block-size limit. It does not mean we are going to instantly start making 20mb blocks. Blocks will continue growing at their current pace. This is perhaps the biggest misconception that we will somehow suddenly start seeing 20mb blocks just because the limit is raised.

It's better to make a move when we don't have to than to make the move when we need to. Imagine having these debates when blocks are already pushing their max size? Nothing would actually get done. There'd be too much fear and debate about how to do it.

only a moron would act when the shit hits the fan. so i agree 100% here.

You are so right about that. And yet, as a species, we tend to react when the shit hits the fan. lol

I am not sure what are the advantages of 50MB block from technical point of view over say 20MB. But why people are so sceptical and bitcoin community is divided,
sure blockchain will be bigger but we have sufficient drives to cover it up. Isn't it our future and we will have to do this at some point? Why not do it now.

in some years we will probably have 50 MB blocks. no problem. but as i wrote:

common misconception: "20MB is too big of a jump"

answer: It's just a maximum block-size limit. It does not mean we are going to instantly start making 20mb blocks. Blocks will continue growing at their current pace. This is perhaps the biggest misconception that we will somehow suddenly start seeing 20mb blocks just because the limit is raised.


i hope that there is also an automatic increase over time.

Here we go again. Looks like there will never be an end to the explaining. 20 MB is too much, especially for China where some large exchanges and miners are located.
Even though the general consensus seems to be that the blocks will get filled instantly (which they won't), there is potential for someone to try filling some of them for the wrong reasons. China can't handle this right now. (https://bitcointalk.org/index.php?topic=1081959.0)

Besides 20 MB blocks are talk of the past, it should be 8 MB with the correct calculations. There is more consensus or 4 or 8 MB blocks as this would give the network more time before other solutions are found.
Increasing the block size is not the solution that we're seeking. Here is why:
It is believed that Visa is peaks at around 45, 000 tps (transactions per second). Bitcoin currently supports 7 tps (in theory; actually it is more around 2-3 but we're going to disregard this now) with a 1 megabyte block limit. If we use an average of 300 bytes per transaction and unlimited block size, an equivalent capacity to peak Visa transaction volume of 45,000 tps would be ~ 8 GB per block. This is every 10 minutes on average which would result over 400 petabytes per year.

Let me create a clearer picture:
400 Petabytes = 400 000 Terabytes = 400 000 000 Gigabytes per year.
You've read that correctly. That's 400 million Gigabytes per year if we only factor in what Visa is processing. Obviously this is not feasible (today) and this is why the block size increase is there only to buy time, nothing else.

It is possible that Moore's Law will continue forever, and that multi-gigabyte blocks may be a possibility of the future, however it is not a certainty. We can't rely on this scenario.


Note: Calculations have been done assuming a block is mined every 10 minutes totaling ~52,000 per year.
Update: Slight corrections.
Update 2: Thank you for informing me that 20 MB blocks are going to roll out.

because there is always the possibility to find a better alternative, this is one of the reason why ,many do not prevent things, before they happen, even if they know they will happen soon


this is a good thing that could actually make the block size scale with the demand, instead of throwing unneeded space...

The most immediate answer would be: "Because it would take less time until a further fork is required". This is only theoretical tho, since I know there are calculations involved to decide on what amount of MB need to be raised.
I wonder what Gavin is exactly planning with 8MB, because 8MB alone fix nothing. I want to know what he has on store after this.

Not sure about exchanges, but I am dubious that they would have problems with large blocks, since there are few exchanges and they can be located near major Internet switching points. 

The argument about miners is bogus.  Mining farms need to be located in locations with low cost electricity and inexpensive cooling. However, they require minimal bandwidth to a full network node. For example, a huge mining farm could run a single stratum connection to a full bitcoin node located close to a major Internet switching point. The bandwidth to the farm would be a few hundred bytes each way every 30 seconds, and this bandwidth has nothing to do with the block size or even the transaction rate of the bitcoin network.

Not quite, the next release of XT (in about a week) will increase the maximum block-size limit to 20MB which is the next step towards 20MB blocks.

https://www.reddit.com/r/Bitcoin/comments/394fn1/mike_hearn_in_about_1_week_bitcoin_xt_will/

Gavin Andresen: "A lot of people are pushing me to be more of a dictator (like Mike) ... that may be what has to happen with the block size. I may just have to throw my weight around and say this is what it's going to be. If you don't like it, find another project."

https://www.youtube.com/watch?v=RIafZXRDH7w&feature=youtu.be&t=1672

Slightly disturbing tone from Mr Andresen....

i have good news for you:

you can still run a full node next year with 2 MB blocks, and even in 2017 with 4 MB blocks and that year after 2017...

it will take years until we reach 20 MB  ::)

The opinions of the huge miners are the ones he should be going after... if 90% of the hashrate refuses to move, XT will never come to be.

Let's make blocksize a maximum of 50 gb then. In the future we can even host movies on the blockchain! Real decentralized p2p file sharing live on the blockchain?? :)

The problem with the farms is that most of them are in China where labor and electricity is cheap. A large chunk of the hash rate comes from China, and their bandwidth is limited due to the Great Firewall of China. Because of this limitation caused by the government, there could be more delays between them getting a block and them publishing a block. This could lead to a higher orphan rate, which miners want to keep low.

common misconception: 20MB maximum block-size limit is a viable scaling solution for Bitcoin.

answer: Transaction spam will increase and fill up blocks even faster, because there is no fee pressure.

solution: Allow a fee market to develop by limiting increases in maximum block-size. Move dust-transactions off-chain or side-chain.

ya.ya.yo!

What makes you think there's no fee pressure? A transaction with no fee will still take forever to validate.

Some fee != fee pressure.

The fees are too low, they encourage transaction spam. I do not think that Bitcoin should process pennies if that means we have to sacrifice on decentralization. Dust transactions can be conveniently handled off-chain or side-chain. They do not require the same level of security as ordinary transactions.

ya.ya.yo!

in 10 years , why not. the bandwidth and storage for that would be trivial by then.

Again, where do you get this? Spending one penny costs more than a penny, so you shouldn't worry about transaction spam.

I vote that we make the blockchain one ginormous block that can never be solved!  :P

No. If we can't handle 20 MB blocks now, what makes you think we are going to be able to handle 50 GB blocks in 10 years?
This is wrong. If you look back to 2010 the biggest HDD was 2 TB (according to some sources). There is no way that we will be handle 50 GB blocks in 10 years.
We most likely won't be even able to handle 8 GB blocks as I've described in my previous post.

https://bitcointalk.org/index.php?topic=1085476.msg11574987#msg11574987

Only a trivial software change is required to handle 20 MB blocks today, at least from the perspective of storage, which is what your post is discussing.

144 blocks a day,  52596 blocks a year.
At 0.020 GB per block,  1052 GB per year.

I recently bought five 3 TB hard drives for a  12 TB RAID array to store music files. The price was $90 per drive.  With the absurd assumption that every block is full, this amounts to an annual expenditure of $30 to keep up with blockchain storage at the 20 MB block size.

You have to factor in bandwidth and syncing into this. Good luck syncing to a 1TB blockchain on a regular PC. I'm not sure where you got those prices from. Usually on Amazon.de it is between 50 to 60 euros for a 1TB drive.
You also have to consider that this is a lot of money for quite a number of people. Obviously it is cheap for people with good jobs and people who live in 1st world country. What about the rest?

China would definitely have problems if we had 20 MB blocks today.

Yeah, 20MB is ridiculous and only viable in the future where 1TB is as cheap as 1GB, otherwise nodes will be limited to a small minority of people that can afford it, this weakens the overall network.

50mb blocks not necessary for a loong time

A penny's worth is much too small.

Until network capacity has grown significantly, Bitcoin should be limited to transactions > 1 USD worth by higher fee pressure. All other use cases should find an alternative solution (off-chain, side-chain).

It's simply wrong to cut back on decentralization just to make extremely small transactions viable. If it can be achieved without bloating the blockchain then good.

ya.ya.yo!

this lead me to a question, it not possible for bitcoin to skyrocket before the halving? nothing say otherwise, what if we assist at an acceleration in adoption, and we are still here stuck with 1MB?

it would be a nightmare for big merchants, since all the tx will cause a long queue chain that at some point will overwhelm the network

it is better to work ahead of time on this subject, to avoid big issues in the future

thought the same  :D

The maximum value of the block may be only 32 mb. You can not do more.
 I'm right? Or I do not understand something?

My network connection speed is 8Mb/s, or just undern1 MB/s. 20 MB will take more tham 20s to transmit. There will be many orphan blocks.

3 TB disk purchase came from Amazon.com.
 http://www.amazon.com/gp/product/B005T3GRLY?psc=1&redirect=true&ref_=oh_aui_detailpage_o00_s00 (http://www.amazon.com/gp/product/B005T3GRLY?psc=1&redirect=true&ref_=oh_aui_detailpage_o00_s00)

I advocate removing the block limit altogether. When Satoshi wrote the Bitcoin whitepaper and published the early versions of Bitcoin, there was no such thing as a block limit. Satoshi's vision was that most people would use SPV wallets and not run nodes. It really doesn't make sense to be running Bitcoin nodes from home connections, thats not very cost-effective when we could be running them on cheap fibre-optic datacenter connections. This doesn't make Bitcoin "centralized", the Bitcoin network will still be decentralized as it will still be a network comprised of a large number of independently owned nodes and anyone with the resources can run their own node. Just because it's too expensive to run a node from home doesn't make Bitcoin "centralized".

This is pretty much invalid. Anyone outside the US doesn't have access to these prices and that is the point that I was trying to make.

No, you're not right. Where did you pull this from?
Currently blocks can be 1MB big, but the fork is supposed to increase that limit to 20 MB.

Apart from storage costs, network capacity is even more important.

Gavin-supporters (and even Gavin himself in his initial "calculation") tend to be ignorant of the fact that it's upload, not download speed that is the most limiting factor. That said, most countries in the world (e.g. China) have far less residential network capacity available than the US or a few EU-countries.

ya.ya.yo!

Are you living in China?  Why would anyone in China have problems with that?  Not only are hardware costs here in China far less than you could find almost anywhere else, but I'm sure anyone with the need or will to run a full node in China would not find the cost to be a barrier.

They are only raising excuses, to remain competitive in their sector, they don't want to diminish even a tiny profit from their revenue, they are not even interested in bitcoin tech, but more on the economical aspect of it, basically only for the money, and one chinese confirmed this in another thread(about etherum project)

Exactly. It is not unusual to see 15/20 (or more) Mbps download with a 1Mbps upload speed. However this is interesting now. The people who used to complain about syncing and the time required to do so are actually supporting 20 MB blocks. An unexpected turn of events for me.
If it comes to the point where I have to sync 0.5 TB or 1TB, I'm probably switching ships as well.

It is actually fun to see some advocate that we might soon support 8 GB blocks.  ::)

Network, not storage.