Both sides (small Vs big) have economic incentives behind them, making the argument extra vitriolic. But larger blocks allow on chain growth, keeping it peer to peer without requiring any 3rd parties. (And the answer that 8x more bandwidth and storage makes it infeasible for enough people to run nodes rings false.)
I have to be honest, I was keeping a cold wallet with some ~1-2 BTC in it until the price reached record high territory in the last few months, and it took me several days to download the entire blockchain against my wallet.dat. I had to free up well over 100GB to get the coins out.
If I hadn't done this dozens of times already (when the blockchain was smaller) and had some degree of confidence that it was going to succeed and grant me access to the coins, I probably would have been freaking out as I reached about the 36 hour mark into the process.
I can't imagine it's going to get better in the next couple of years. Maybe "Moore's law for storage" will outpace economic growth in the Cryptocurrency sector, and in 20 years my mobile phone will be so powerful that it will no longer be a concern. That sounds like a bit of a gamble, doesn't it?
I don't know how I reconcile that with clients like electrum, is that using an implementation of UTXO? Does this just keep a record of any unspent transaction outputs at any given time, scan them, and try your keys on all of them?
I think you can do a similar trick with bitcoin core if you have all of the blocks, but you have a new wallet.dat. You can delete the index and it will regenerate it with the new wallet in mind, recovering all of your unspent transactions.
(You do not need to literally download all of the blocks again, unless you actually don't have them anymore. It is sufficient to rescan them. But this takes time too, and it does require that you must store a copy of the chain somewhere on an ongoing basis.)
And if you need to view specific blocks or transactions, use an online viewer or download that specific block and check the hash of the block. 10 years of 10 minute blocks is only 16MB of hashes (32 bytes per block). So it's trivial to get all block hashes and then freely use other sources to check specific parts.
Storage is cheap. 2TB drives are a bit over $50.
I have a 14TB array that I bought with those Bitcoin I cashed in. It was cheap. But the next time I want to sync the whole chain, I'm really sure I'm gonna have to use it...
Thanks for your comments though, I will look up UTXO as it's another one I have not heard of before!
> Before reading this page, users should note that directly manipulating ECDSA private keys is dangerous and can result in losing bitcoins[1][2].
Hmm.
> Use the dumpprivkey command to get the private key. Repeat for as many bitcoin addresses as have money in them.
OK, so how many bitcoin addresses do I have with money in them? examining 3.5 year old wallet.dat, checking Oh, THAT Many? And are you sure that is all of them? (It was quite a lot of addresses, and not sure how I would know if the listing was complete.)
So, is there some way to safely automate this? Try it... the command outputs are all in JSON (configurable?), so you most likely won't be solving this with just pipes, awk, and grep. I tried exporting a few private keys this way and they are not exaggerating when they say:
> This guide is [not] for the impatient or time-stressed, it is recommended that users of Bitcoin Core simply wait out the blockchain synchronization. There are many benefits to running a Full node including trustlessness, security and privacy which lightweight wallets cannot match.
I have to use Bitcoin-Qt to get the keys out, and it does not appear to be fool-proof at all. Might as well just sync the whole blockchain then.
The idea that bitcoin can exist only as a settlement layer is a dream. It was pitched as something you can buy a cup of coffee with. Now the core team wants to call that a "spam transaction". Yeah right. Time to fire Core.
it's open source, there's no one to "fire". if you're not happy with it, fork it like bitcoin cash did.
Not to mention, if Bitcoin becomes a settlement layer then no one can run a full node since most transactions will be off chain. Plus the enjoyment of running a full node is greatly reduced if you're just seeing settlement data between large payment systems.
being able to fully verify the blockchain and the transactions. sure, if all you care about is making a payment you can use a SPV client or even a webwallet, but being able to not r
>if Bitcoin becomes a settlement layer then no one can run a full node since most transactions will be off chain
that's what the lightning network is for. in which case you don't even need the huge blocks.
Whichever path it ends up going down, centralization is guaranteed.
While there is certainly centralization pressure on mining, I do not agree that 8MB blocks make it impossible for everyone except for 'powerful companies' to run a full node, especially a pruned one. An enthusiastic early adoptor or investor holding, say, $100k of the token is more than capable of spending $500/yr to run a colo'd node. Would they prefer to spend $0 and host it at home? Probably. But the expense of a high-end VPS or low-end colo box is not totally prohibitive. We're not talking about more than a few rack units worth of HW.