An explanation of what Blockchain is

If the block chain is a ledger with every transction in Bitcoin stored as a record, how much computing power will be needed to keep this data as time moves on. Will this data not eventually become enormous and unwieldy.
 
If the block chain is a ledger with every transction in Bitcoin stored as a record, how much computing power will be needed to keep this data as time moves on. Will this data not eventually become enormous and unwieldy.

I think fpalb gave his thoughts on this back on the first page

but I am concerned about the scaling, this really needs to improve soon. I don't know what the correct answer is, I can see the point of view of both sides. What I do think is clear is that putting transactions for trivial things like a cup of coffee purchase in the blockchain, forever, stored by all nodes all over the world probably doesn't make much sense, it's overkill. I think it's likely that the future of bitcoin is as more of a base layer, where further technologies will develop new layers on top of it which are backed by bitcoin. For example perhaps a lot of back and forth transactions between a few parties happen on some higher layer, then are grouped together with only the final amounts owed settled with bitcoin transactions.
 
I gotta say fpalb, this is easily the best thread that explains the principles behind Bitcoin and Blockchain. Coming from a IT background I wanted to understand the tech behind it so this is brilliant.
 
Can I ask some questions on mining?
  • Presumably there are hundreds if not thousands of people/nodes mining Bitcoin?
  • Based on the 10 minute blockchains then only 6*12 , 72 bitcoins are awarded every hour? ( Excluding fees)
  • The award of the bitcoin is given to the node that solves the hash problem first, so if a node has more CPU power, are they guaranteed to win more often than someone with lower compute power?
  • If there something in place so every miner eventually knows they'll win or is it all luck/compute power?
  • Why has the fees for bitcoin gone up? Is it tied to the amount of work a miner/node has to do to confirm a transaction? Is there a defined formula for it?
 
Last edited:
Can I ask some questions on mining?
Sorry missed this while I was on Christmas break.
Presumably there are hundreds if not thousands of people/nodes mining Bitcoin?
yes, though by now it is mostly done professionally at large scale. You need to have cheap hardware costs and cheap electricity to even hope for it to be profitable.
Based on the 10 minute blockchains then only 6*12 , 72 bitcoins are awarded every hour? ( Excluding fees)
6*12.5 actually, so 75 new bitcoins come into existence and are awarded every hour on average.
The award of the bitcoin is given to the node that solves the hash problem first, so if a node has more CPU power, are they guaranteed to win more often than someone with lower compute power?
In the short term there may be variance, but yes in the long term nodes should solve blocks at a rate proportional to the computing power they are dedicating to it. Think of it as if you and I are entering a raffle every week and you buy twice as many tickets as I do, over time you should win twice as much as I do, but in the short term through luck I could win more than you.
If there something in place so every miner eventually knows they'll win or is it all luck/compute power?
No guarantee, It's luck/compute power, see the previous raffle analogy. What small time miners tend to do to lessen the effect of risk is combine their power together into pools, then divide the proceeds of the blocks they mine proportionately among themselves - similar to a lotto pool.
Why has the fees for bitcoin gone up? Is it tied to the amount of work a miner/node has to do to confirm a transaction? Is there a defined formula for it?
Fees are unrelated to mining work. Each block (every 10 minutes) can currently only contain up to 1MB of data. So even if miners have a lot of pending transactions from users, they may only be able to include a subset. It's at their discretion which they include, but logically they tend to include the transactions with the highest fees, as they get to keep those. Fee setting in wallets is currently a bit messy as it can be difficult to predict what fee will be needed to be in a future block if the number of transactions is fluctuating a lot. Most wallets have a slider letting you move along the scale of low fee to high fee depending on how urgent you want your transaction confirmed.

One interesting and non-intuitive thing to note here is that the size of a transaction in bytes is what's relevant NOT the amount of bitcoin being sent by that transaction. The size in bytes is determined by the number of inputs (how many transactions the coins being spent were received in) and the number of outputs (how many different addresses the transaction is spending to). It's possible a bitcoin transaction sending 0.001 bitcoin could have a higher size in bytes than one sending 100k bitcons. So fees are often measured in satoshis per byte.

You can see a history of the average transaction fees here: https://bitcoinfees.info/ which really shows how it's a rather recent symptom that there is frequently a backlog of transactions and therefore people choosing higher fees to get priority.
 
  • The award of the bitcoin is given to the node that solves the hash problem first, so if a node has more CPU power, are they guaranteed to win more often than someone with lower compute power?
Which begs the question, why don't Amazon, Google or Apple, with their endless hardware resources, not completely dominate this space?
 
Which begs the question, why don't Amazon, Google or Apple, with their endless hardware resources, not completely dominate this space?
Because generic computing hardware isn't efficient enough for it. Miners these days are using purpose built machines, they are designed to be able to do the mining calculations as efficiently as possible and nothing else. They can do hashes per electricity unit at rates orders of magnitude above general purpose computers. They closest thing to that is graphics cards.

A better question might be why are chip manufacturers like Intel or AMD or GPU manufacturers like Nvidia not dominating it.... I guess it's a risk they didn't want to take. Mining is risky because you have all of the risk of cryptocurrency itself but also the risk that you don't actually make a profit mining. Whether you make a mining profit depends on unpredictable variables such as the future mining difficulty (a function of how much competition you will have in mining), the lead time of your hardware and the lifespan of your hardware.

I came to the conclusion a long time ago that if you believed bitcoin would be successful that attempting to mine bitcoin was not worth the extra risk over just buying it upfront. So you could just as easily ask, why did Amazon, Google or Apple not just buy loads of bitcoin a few years ago?
 
Because generic computing hardware isn't efficient enough for it. Miners these days are using purpose built machines, they are designed to be able to do the mining calculations as efficiently as possible and nothing else. They can do hashes per electricity unit at rates orders of magnitude above general purpose computers. They closest thing to that is graphics cards.
I couldn't agree with that I'm afraid. Amazon, Google, Apple and IBM have unlimited hardware resources at their disposal and also the finest crop of technical minds in the world. Just look at what Google AI was able to do in just 4 hours!
Any of these companies could modify an existing super computer with all the bells and whistles and easily beat a few miners IMO.

A better question might be why are chip manufacturers like Intel or AMD or GPU manufacturers like Nvidia not dominating it.... I guess it's a risk they didn't want to take. Mining is risky because you have all of the risk of cryptocurrency itself but also the risk that you don't actually make a profit mining. Whether you make a mining profit depends on unpredictable variables such as the future mining difficulty (a function of how much competition you will have in mining), the lead time of your hardware and the lifespan of your hardware.
If it was deemed worthwhile to do so (especially in the long run) these companies (like Amazon, Google, Apple and IBM above) would be mining already.

I came to the conclusion a long time ago that if you believed bitcoin would be successful that attempting to mine bitcoin was not worth the extra risk over just buying it upfront. So you could just as easily ask, why did Amazon, Google or Apple not just buy loads of bitcoin a few years ago?
Probably because they feel it is not worth anything either ;):p
 
Back
Top