Bitcoin Forum

What are the best criticisms of blockchain technology? Specifically, what are the top flaws of the blockchain as it's currently implemented by Bitcoin?

A few I can think of:

• double SHA-256 proof-of-work isn't computationally useful for anything (GridCoin is still affected by this criticism to a certain extent);
• it's not as immune to centralization as it could be (control of blockchain is proportional to computing power; it should be proportional to number of human users).

(cf. related Bitcoin StackExchange question (https://bitcoin.stackexchange.com/q/37764/4334))

Blockchain technology costs much more resources (energy) per transaction compared with centralized systems (if we are comparing systems with equal reliability)

Blockchain technology has less reliability compared with centralized systems (if we are comparing systems which take same energy)

Are you talking about Gavin?  ;D

You will see doublespendings when difficulty drop down.
You can doublespend right now - just take one of shit-coins.

And what about the HUGE blockchain size?
As time passes, it becomes more and more difficult to handle the blockchain.
Bitcoin have been up and running since 2009.
That is more than five years!
It is currently 34.5 gb, and it was 10 gb less 6 months ago!
This is a 25-30% increase in six months!

I've said above about the cost of resources. This means all resources and total costs. Traffic, hosting relays, disk storage, mining hardware - these things are resources.
The product of Bitcoin network is transactions.
Bitcoin network (as a bank/payment system) converts resources to transactions (transferring rights).

Depends on what's in fashion. This year is blockchain size so far, last year was PoW and 51% attack. I project in the few years, it will be weakness of SHA256.

If the computation was useful, it would be easier to attack the network : it is "useless computation" for security reasons.

- It is an app, not a protocol
- Development is slow
- Development is centralized
- Community is extremely conservative
- Inflation is high
- Transactions are slow
- Transactions are irreversible (fixed by multisig, which is too complicated)
- Mining is complicated and annoying
etc.

Personally using blockchain wallet. It's a good btc wallet for us to use. But I've experienced shut downs so many times, it ruined my transactions. So it's not a perfect wallet for us btc users. They definitely should improve  >:( :o

@Clint: Think he meant the technology, not the company named Blockchain.

Disks space has been increasing ever since that time.  Satoshi had the same question in 2009 and he wrote that disk space will increase over the next decades.
I do agree that a system which would not require this amount of disk space would be beneficial, but how would you implement that?

You can use pruning mode to save disk space.

Not sure where your statistics are coming from, but this is nowhere near true.  6 months ago it was around 27 GB. ( see eg. https://blockchain.info/charts/blocks-size?timespan=all&showDataPoints=false&daysAverageString=1&show_header=true&scale=0&address=  )

The blockchain size is not unmanageable and hard disk space is cheap.

p.s.  Plus pruning.

What's pruning mode? Is there a flag to turn it on?

See 'Block file pruning' in https://github.com/bitcoin/bitcoin/blob/v0.11.0rc1/doc/release-notes.md and also http://bitcoin.stackexchange.com/a/19928.

Oh, that's bleeding-edge then.

It's a step in the right direction but it's not perfect.


I have found myself using reindex quite often and would hate redownloading the blockchain each time it was necessary.

Nevertheless work's started on dealing with the huge blockchain and that's good.

1) Irreversibility, ie if someone sends to an incorrect / dead address / loses keys then its gone. Or people injection malicious stuff into the blockchain leaving it there forever.

2) Theoretical anonymity. While we all like privacy, there is a problem when something goes from 99.99% untraceable to 100% untraceable (TOR based tumbling). There ARE cases were its important that criminals are traceable even with great efforts, else it makes Bitcoin users a target. Imagine how attractive users are when the goods they take are untraceable, unmarked 'bills' with no risk of ever getting caught. I don't like it.

3) Vulnerable to attacks. With no central body that can react quickly, things like transaction floods can temporarily damage the network. Because it takes a community effort or dev census to nullify the attack, its much slower than someone who can implement needed changes instantaneously. Ie Theymos pulling the plug on the servers when he noticed the intrusion.

You don't find bitcoin useful?  I disagree. 



It looks like you are saying, it should be vulnerable to Sybil attacks.  I disagree.     

Well, computationally un-useful for things besides Bitcoin; that's what I meant.

Well this is hardly a criticism of blockchain tech, wouldn't you agree?  Consider this criticism of headwear tech:

A helmet is only useful for protecting the head.  It is un-useful for things besides head protection.  

The criticism is a species of the "Bitcoin-wastes-electricity" criticism.

Good point.  I would say that particular argument is indeed just as much complete and total nonsense as the earlier one. 

Streetlights waste electricity.  Chargers plugged in when not in use waste electricity.  Air conditioned casinos with doors and windows wide open waste electricity. 
Bitcoin in no way whatsoever wastes electricity. 

Sure, wannabe miners using CPUs to mine BTC waste electricity.  So do wannabe miners overdoing their fan use, or using hairdryers to make tea.
Bitcoin however rewards efficient miners only.

It's too bad there is not such a mechanism to also make streetlights, chargers, and air conditioning so damn efficient.   

Agreed, electricity is neccesary for any digital system and it is a non-argument. The same argument could be used for the hundreds of terabytes to store dog pictures on the internet.


Is irreversibility not a feature of the blockchain? I think irreversibility has always been by design.

yes, indeed; it's a good thing.

I would argue that each of these is a feature, not a point for criticism
1. Irreversibility is a feature.  You can safely ship a good, knowing that you will be paid and not defrauded.
2. Anonymity is also a feature.  Eg. My landlord doesn't know I got a raise.
3. It is a dynamic system and much more resilient to attacks because there is not a central authority to attack.

I must echo this. I am not impressed with the platitudes such as "storage is cheap". As of writing, Bitcoin Core requires 40GB. I have a several laptops with varying space of about 20-40 GB spare so it will no longer fit on any of them. They all sport SSDs so huge drives are not "cheap" and if you expect me to go back to mechanical drives just to run this one piece of software; you are insane. The spare space is perfectly adequate and has been for a few years.

I tried moving it to a NAS which has 10 TB of storage spare. I had to use a symlink to fool Bitcoin Core  because it only uses a fixed local directory. Once I had managed to convince it that no, it isn't really on another drive, it was extremely problematic in that it took hours to "verify" when starting the application and that's before getting to updating the blockchain. There are many small accesses which are horribly inefficient trying to verify. On some occasions it borked and refused to continue because it somehow corrupted during the verify. The NAS solution is currently not really viable.

Bitcoin is suffering from similar problems that PGP suffered from. Nerdy software that is rigid and inflexible and unusable for many. You can, of course, put all your coins in an online exchange but you might as well keep your dollars in the bank-it's the same thing.

If Bitcoin Core can be run from a laptop, Raspberry Pi or even a mobile phone with, say a gig or two of working data but the vast majority of the blockchain reside on a remote NAS. Then it would become workable for me and probably most other people too (yes I do want full nodes). If it can reside on a remote NAS and accessed at WiFi rates, even better. This all presumes that the size is a necessary evil so the approach is to be able to split the blockchain into historic and working data sets for the blockchain with lazy updates back to the NAS.

At the moment the software is like this:
http://www.freakingnews.com/pictures/25500/Ruben-Studdard-25987.jpg

I'll leave the rant about not having a torrent client built in for the peer to peer distribution of the block chain for a later post.

why do you need to run a full client? there are many good alternatives.

Yes. MasterCard & Visa

Because I do.

I would agree that creating an immutable record is part of the big wave of innovation that is the blockchain and bitcoin, but this irreversibility also poses a big problem: people don't really want that for their consumer purchases. They want a system that has chargebacks. They want a system that some way to recover lost coins if a private key is lost or if somebody else is responsible for the loss of one's bitcoins. They want a way to reverse a transaction that they see as unfair or predatory.

For example, what about the money that more or less went into oblivion when Mt. Gox went under? Most people want a way to recover funds if anything goes awry. A non-governmental system will have to find a workaround to reassure people.

A followup issue--to me, anyway-- is that the true number of bitcoins that will come into existence will not be 21 million as foreordained. It will be a lower number because fewer coins will actually exist. I think the overall conception of the blockchain and bitcoin would be improved if the desired number--21 million-- was somehow guaranteed. Inevitably, people will lose their private keys, exchanges will collapse, Mt. Gox-like venues will fail, owners of bitcoin will die without passing on their coins to anyone, bitcoin-accepting businesses having coins in their store wallets will go bankrupt and coins might easily be lost or overlooked at that point, and any number of other things will occur that will reduce the number of active coins, or coins that actually exist.

and in a decentralized manner

Right, I agree.

It'd be great if there were guaranteed to be 21 million coins in circulation at all times, or whatever the current number would be since we haven't reached 21 million created yet.

That way, the system could somehow account for those kinds of non-existent dead weight coins and recreate them so there's fidelity to the original vision of having 21 million coins out there(or currently correct number).

Now that 21 million is the desired final number, then it's got to be that, even if that figure was picked arbitrarily.
 

Pruning should solve this in the future, even tho the paranoid within me wants to have a local copy of the entire blockchain just to be sure, but I guess that will be a placebo at some point.

Wouldn't ensuring that mean every miner would have to do a full search of the blockchain each time he wants to include a transaction in the block he is mining?

Even this is arguably a feature.  A "computationally useful" proof-of-work algorithm wouldn't be any more efficient, it would simply result in a higher difficulty.

GridCoin appears to depend on at least one third party, Berkeley (the University of California).  The purpose of a blockchain is to enable consensus without a special trusted entity.

The "original vision" was for there to be no more than 21 million bitcoins, i.e. for 21 million bitcoins to be a bound.  Where you see "non-existent dead weight" coins, satoshi saw "donations to everyone (https://bitcointalk.org/index.php?topic=198.msg1647#msg1647)".

In order to have a guarantee of 21 million coins in circulation at all times you would at least have to define "in circulation".  How long much an output remain unspent before it is deemed not in circulation?  10 seconds?  5 days?  20 years?

Strictly speaking the block chain is not irreversible, in contrary it is able to reorganize in any depth.

Actually a valid critique of the technology that convergence to global consensus is not guaranteed within a time span of any length.
Consensus just becomes less and less feasible to change with time.