Topic: The GIANT Blockchain (Read 744 times)

Sounds like the bottleneck is the CPU.

---

Yeah, I read your thread and others that have problems with 32-bit builds (esp. WinXP). I know its not good advice, but it might be time to get ride of that OS and switch to Linux.


As long as there is a blocksize limit the blockchain wont grow exponentially.

I've had to resync several times due to core crashing randomly and corrupting the database.  I have an older computer that I built years ago, 32-bit running XP, lots of ram and 7200rpm HD space, good cable internet connection, and it takes me longer than a week to sync.  Even pruned nodes need to d/l the entire block chain before they can prune, so yea, this is going to be a huge issue very shortly as the block chain continues to grow at an exponential pace.

True, the computer I am using isn't top of the line by any stretch. Windows 7, 64 bit, 1.83GH. Internet connection isn't too bad: 15 down, 10 up.
HD is the old fashioned spinning type.

also you have prune mode if you don't use too many wallet or new wallet for each movements, and in any case there are spv client

Well it's a rather... Risky thing to rely on Moore's law IMHO :/

We're starting to get trouble upgrading again and again our computing power. Maybe we should start thinking about making the blockchain smaller :3
I mean what the point of keeping transactions that happened 5 years ago? Especially if the whole coins have been used again since then.

Blocksize was more of a concern 2 years ago, with the optimizations core devs have managed to do it is less of a burden.

It only recently took me 8 hours to fully sync. What system are you using? (most interesting are: OS (32 or 64 bit), CPU, Internet speed and disk type (SSD or mechanical disk (raid?))

I only noted Satoshi's opinion since he anticipated early on that data would be an issue one day and
that his opinion was what it was. He was not an idiot nor advocating acceptance of centralization.
That just was his answer at the time.

Remember that Satoshi release Bitcoin as an experiment and not as a perfected system.
He could never have known it would be what it is today or what we are trying to make it into the future.
For all he knew, it could have failed or been destroyed months after it was released.

The following are the quotes from where my Satoshi understanding came from:

http://www.metzdowd.com/pipermail/cryptography/2008-November/014815.html

https://bitcointalksearch.org/topic/m.6306


Satoshi's system was designed to be a fully decentralized money. That was the experiment.
But he thought that one day, even that important element (decentralization) may be forced to give way.

Whether that was correct or not is irrelevant and only time can tell whether we can
do that which Satoshi himself thought wasn't possible (at the time of his statements).

Wrong board my friend, rather go to BitcoinISdead.org or any of the other sites trying to prematurely announce the early death of Bitcoin. On this board, we try to get the solutions for problems and

to debate it's promising future. There is already pruning options, if you do not want to download the full Blockchain. People will download and keep 3 to 5 TB of movies and Appz and games and these

things are stored on Raid configured rigs, and now you saying people will not invest in bigger drives, as and when they become cheaper to accommodate something, that are not nearly close to the

maximum size of what is currently available out there for commercial use.  


I bought a 5 TB hdd the other day for $140 and it did not break my bank.... That is still way bigger than what the Blockchain is now.. So we are still OK for a few years.  

When that happens Bitcoin would die as we know it, thats why we must guarantee users can run nodes at home, otherwise it's just a fancier Paypal. (datacenter running nodes = basically big servers buildings from already existing corporations).

I think the quote you are saying doesn't mean what you think it means, but even if it meant what you are implying that it means, it wouldn't matter, it would just mean Satoshi was an idiot by implying we should do nothing about it, or he was actually hoping to centralize Bitcoin nodes eventually (which would mean this whole thing was a project to control everyone's finances as the anti-cash agenda continues).
But im pretty certain he didn't mean that we should do nothing. Payment channels code was there before Lightning Network.
Thank god Core devs are against the node-datacenter nonsense.


This is idiotic. No other system is going to be invented that delivers army grade security without the tradeoffs that the Bitcoin network does to run the biggest network in the world while staying decentralized. People that are dreaming with some sort of Bitcoin 2.0 are delusional.

nobody will use Bitcoin in 10 years. stay calm.

other innovative payment system will come till then. maybe you will be able to pay by using your fingerprint, or eye or...i don't know what device. BTC is a niche market and it will remain like that.

BTC is an e-currency which is now trendy like e-gold or libertyreverse were in the past.

The expectation is that the size of the block chain will grow more slowly than the ability to store and process it.

10 years ago, 63 MB was "ginormous". Now, you can send 63 MB in an email. In 10 years, 63 GB will be small and 63 TB might be considered "ginormous". It will take a block size of 100 MB to achieve a 63 TB block chain in 10 years.

One way to speed up your sync is to download the block chain. Look here: https://bitcointalksearch.org/topic/download-the-blockchain-here-updated-regularly-1310261

iguana.

this full node has promis, it plans to leverage the P2P torrent network to incress dwl speeds of the blockchain, and uses some fancy tricks to validate blocks of TX's at blinding speeds.

i believe all problems can be overcome. either with burt force, or fancy tricks.

At its current rate with no new growth for 10 years, I would estimate around 382 Gigabytes.
If growth doubled each year for 10 years, I would estimate around 65.5 Terabytes.
(Edit: as pointed out below, my 65TB estimate does not consider the blocksize limit)


There is no point in estimating since by that point in time, we would have already implemented some form of "pruning"
or other system that would help with the possible excessive file size. It could be as high as 680 trillion yottabytes.
But I highly doubt that since we would have taken steps by the time the size is around or before 1 Terabyte.


Yes. This is a known issue and there have been discussions about this with different possibilities as answers.
One was placing "checkpoints" in the blockchain where we hash all prior txs/blocks and agree that hash is absolute,
and then find the next block from that hash (instead of the prior block found) like a new genesis block.
That would effectively be like "the blockchain archived". This idea, like all others, has pros and cons.


Yes. This is a known concern to the decentralized nature of Bitcoin/bitcoin.
It is important to note, that Satoshi said himself that a day would come in the future where average users would
no longer reasonably be able to maintain the full blocksize and data centers dedicated to this purpose many need
to exist and be used. When that will be depends on how the data is handled now and in the future.

All this talk about bigger blocks. While all the talk about needing bigger blocks, people seem to be forgetting how GINORMOUS the blockchain is. Sure, it's only 63GB. Ignoring the space that this takes on a hard drive, that is HUGE! It has been almost a week now that I've been synchronizing a Bitcoin Core node with the network. I'm still 24 weeks behind. Yesterday I was 27 weeks behind. I suspect that it will probably be another week before the node has caught up.

We are only what - five or seven years into bitcoin? And it is only now starting to get serious use. How big is the blockchain going to be in another ten years? How big is the blockchain going to be when the last satoshi is mined?

Can something be done about this? Maybe having parts of the blockchain archived? Because the size of the blockchain might be another cause of centralization in a year or two. Who is going to run a node if it takes a month to synchronize?