7 ms·

There was interesting discussion here about disk usage over time: https://news.ycombinator.com/item?id=16154682 https://news.ycombinator.com/item?id=16154682 I

by shawn 9y ago

There was interesting discussion here about disk usage over time: https://news.ycombinator.com/item?id=16154682 https://news.ycombinator.com/item?id=16154682 I wanted to ask again here, directly to Bitcoin techies: How can BTC survive 100 years if no one can download the massive blockchain? Most residential IPs enforce a 1TB/mo cap. Much of the internet is also on slow connections (Australia). The BTC whitepaper is surprisingly hand wavey on this one subject, and the more I think about it, the more it seems like an Achilles' heel. Can we eventually reboot BTC back to the genesis block, but with all current BTC pre-mined and allocated to current holders? Kind of like a manual GC, or hard reset back to initial commit. But it doesn't seem that simple. Millions of people have certain quantities of BTC, and even tiny ones add up. You have to track that data somewhere. It doesn't seem possible to encode the whole state of the world into a 1MB genesis block. So the question becomes, at what point to we reboot, how, and by what mechanism can we achieve community consensus? This is combined with the fact that BTC is deflationary, so eventually miners will charge nothing but tx fees. No more block reward = the TX fees must be at least as high as the current block reward. That is a staggering cost since a 1MB block can only encode a couple thousand tx. That leads to the difficulty death spiral: difficulty re adjusts every 2016 blocks. Two weeks when every block mines at ten minutes. But if miners leave en masse, then suddenly it will take a very long time to mine blocks. It could be months before the difficulty readjusts down. This must sound identical to all the previous BTC naysayers, but believe me when I say I believe this can work. I'm trying to discuss the technical merits and set aside the question of whether it will fail.

api 9y ago
I've always considered BTC an alpha release and a prototype. Kind of amazing it's gotten this far.
- always_good 9y ago
  I've always considered these comments as taking the easy route of skepticism since you can never be wrong on a long enough time scale. "Hah, knew it wouldn't pan out!"
  - choward 9y ago
    My proof is that I didn't buy any Bitcoin.
snikeris 9y ago
> I wanted to ask again here, directly to Bitcoin techies: How can BTC survive 100 years if no one can download the massive blockchain? Most residential IPs enforce a 1TB/mo cap. Much of the internet is also on slow connections (Australia). Presumably in 100 years, your cap will be larger, and your internet will be faster.
- madeofpalk 9y ago
  Here's hoping!
- IkmoIkmo 9y ago
  Short story is that bitcoin is a ledger that keeps track of where coins go. So if I give you a coin, and you give it back, 1 trillion times, we have a huge database with two trillion transactions. But at the end of the day, nothing changed, I still have a coin and you don't, because you gave it back each time. Much of that data (i.e. 99.9999% of it, in this case) can be reliably thrown away for all practical intents and purposes. That's the gist of it. I'll spare you the technicalities (which I'm not intimately familiar with anymore myself, in any case), but the basic idea is that you don't need all data to run the network reliably. That, plus moore's law's equivalents for storage and networking, google them. They're not as strong as for CPU, but they're quite good anyway.
- s_dev 9y ago
  As well -- only the last few blocks really needed rather than the whole chain.
lawlessone 9y ago
In 100 years the blockchain, (assuming ~2mb segwit blocks) will be about 10 Terabytes, sounds like a lot now but will it be in 100 years? It's currently about 150 gigs.
- Kurtz79 9y ago
  It doesn't sound like a lot even now... https://www.wdc.com/products/internal-storage/wd-purple-surveillance-hard-drive.html#WD100PURZ https://www.wdc.com/products/internal-storage/wd-purple-surv...
DINKDINK 9y ago
>How can BTC survive 100 years if no one can download the massive blockchain? Most residential IPs enforce a 1TB/mo cap This is specifically why the bulk of the community has stayed on a chain that limits blocks to 1 MB blocks. If you increase the shared resource, blockspace that everyone has to store, you increase free-rider problems and the tragedy of commons that goes with them. Given your question, and the properties of the Bitcoin blockchain size today, the rate of blocks being issued, it would take ~13 years before the blockchain would reach 1 TB. Since there is no central authority in bitcoin, there isn't a process where one can screen out "spam" and legitimate uses. The only way to do it this is to rate limit transactions. Periodically you'll see people come on to the network and try to issue large volumes of incredibly low-value transactions into the mempool queue. Are those spam transactions? I'll let you decide. See here: https://dedi.jochen-hoenicke.de/queue/#all https://dedi.jochen-hoenicke.de/queue/#all Second chart shows that there's about 16 BTC worth of outstanding transactions in the queue for inclusion in a block (about 1 PPM of total coins in circulation). The third chart shows that there is about 170 MB of transactions and you can see large number of low-fee, low value transactions being dumped on to the network.
- CyberDildonics 9y ago
  > Since there is no central authority in bitcoin, there isn't a process where one can screen out "spam" and legitimate uses. That's exactly what fees are > blockspace that everyone has to store Only people who sync with the chain have to store the blocks, not everyone who uses it. The vast majority of people don't sync with the chain and still reap all the benefits. > Are those spam transactions? I'll let you decide The miners decide
- t0mbstone 9y ago
  The simple fact that a currency like Bitcoin is technically limited to a set number of transactions per second should be proof to anyone with a brain that it is fundamentally flawed. There are over 7 billion people in the world. A good crypto-currency should be able to handle millions of transactions a second, indefinitely. It should be as ubiquitous and as easy to use as the internet itself. In an ideal scenario, the cost to run the currency network should be cheap enough that there should be virtually no transaction fees other than participation alone (think bittorrent), and the entire thing should be secure and un-hackable. It should be stable in value, or at least slightly inflationary (so people actually spend it instead of just holding onto it and treating it as an investment vehicle). Bitcoin doesn't even come close to meeting these goals. It's so incredibly flawed, it boggles my mind how people don't see the glaring problems. The wheel is moving, though! Bitcoin may not be the end-all, be-all solution, but it has certainly got people thinking.
fearface 9y ago
One way to reduce the Blockchain size is pruning: https://en.bitcoin.it/wiki/Scalability#Storage https://en.bitcoin.it/wiki/Scalability#Storage
DiNovi 9y ago
well IMO its silly to assume in 100 years the internet will be the same speed as today
keymone 9y ago
so first of all, you don't need to store all historical data to operate bitcoin node. it needs to be bootstrapped, but after that there is a thing called UTXO set which tracks unspent outputs (basically account balances), so you don't need to go all the way back to verify you can spend something from a tx in year 2010. so basically as you bootstrap and verify you can prune older blocks without sacrificing security. also there were some proposals related to utxo set commitments, where utxo set signature is recorded in the blockchain, so bootstrapping node can just download it from anyone and verify starting from the commitment block. this would solve the problem entirely. second - storage and bandwidth are still a concern out of which a whole bitcoin cash altcoin was born - some people thought it would be good idea to go with even larger blocks justifying it by even more handwaving. third - there is a strong push towards developing upper layers of the network, so called payment channels. one solution that already exists on mainnet in alpha stage is Lightning Network, is basically a protocol for routing payments between multiple channels kademlia-like. these channel transactions are not recorded in the blockchain, you only need to pay fees for opening and closing a channel, so two on-chain transactions, and then you use your channel funds like a debit card, paying for low cost things. and your last point - yes, halving is a significant event during which miners lost almost half the revenue. on original bitcoin network this is offset by a working fee market: right now about 20%+ of block rewards come from tx fees. obviously miners know the schedule of halvenings and those that will become unprofitable will try to get rid of their infrastructure investments beforehand - this will lower network hashrate which will send tx fees higher, which will somewhat offset losses as well. we've already been through couple halvenings, they didn't really affect much, i think next one will be crucial test for bitcoin's survivability. of course people pushing for zero fee transactions are trying to have their cake and eat it too, storage is a scarce resource and nobody can sustainably to offer it for free. bottom line is - blockchain cannot scale to micro transactions, most of economic activity will be done in payment channels, on chain transactions will be mostly for high-value settlements between larger players (banks, exchanges, etc). edit: word

by shawn 9y ago

api 9y ago

I've always considered BTC an alpha release and a prototype. Kind of amazing it's gotten this far.

always_good 9y ago
I've always considered these comments as taking the easy route of skepticism since you can never be wrong on a long enough time scale. "Hah, knew it wouldn't pan out!"
- choward 9y ago
  My proof is that I didn't buy any Bitcoin.

snikeris 9y ago

> I wanted to ask again here, directly to Bitcoin techies: How can BTC survive 100 years if no one can download the massive blockchain? Most residential IPs enforce a 1TB/mo cap. Much of the internet is also on slow connections (Australia). Presumably in 100 years, your cap will be larger, and your internet will be faster.

madeofpalk 9y ago
Here's hoping!
IkmoIkmo 9y ago
Short story is that bitcoin is a ledger that keeps track of where coins go. So if I give you a coin, and you give it back, 1 trillion times, we have a huge database with two trillion transactions. But at the end of the day, nothing changed, I still have a coin and you don't, because you gave it back each time. Much of that data (i.e. 99.9999% of it, in this case) can be reliably thrown away for all practical intents and purposes. That's the gist of it. I'll spare you the technicalities (which I'm not intimately familiar with anymore myself, in any case), but the basic idea is that you don't need all data to run the network reliably. That, plus moore's law's equivalents for storage and networking, google them. They're not as strong as for CPU, but they're quite good anyway.
s_dev 9y ago
As well -- only the last few blocks really needed rather than the whole chain.

lawlessone 9y ago

In 100 years the blockchain, (assuming ~2mb segwit blocks) will be about 10 Terabytes, sounds like a lot now but will it be in 100 years? It's currently about 150 gigs.

Kurtz79 9y ago
It doesn't sound like a lot even now... https://www.wdc.com/products/internal-storage/wd-purple-surveillance-hard-drive.html#WD100PURZ https://www.wdc.com/products/internal-storage/wd-purple-surv...

DINKDINK 9y ago

>How can BTC survive 100 years if no one can download the massive blockchain? Most residential IPs enforce a 1TB/mo cap This is specifically why the bulk of the community has stayed on a chain that limits blocks to 1 MB blocks. If you increase the shared resource, blockspace that everyone has to store, you increase free-rider problems and the tragedy of commons that goes with them. Given your question, and the properties of the Bitcoin blockchain size today, the rate of blocks being issued, it would take ~13 years before the blockchain would reach 1 TB. Since there is no central authority in bitcoin, there isn't a process where one can screen out "spam" and legitimate uses. The only way to do it this is to rate limit transactions. Periodically you'll see people come on to the network and try to issue large volumes of incredibly low-value transactions into the mempool queue. Are those spam transactions? I'll let you decide. See here: https://dedi.jochen-hoenicke.de/queue/#all https://dedi.jochen-hoenicke.de/queue/#all Second chart shows that there's about 16 BTC worth of outstanding transactions in the queue for inclusion in a block (about 1 PPM of total coins in circulation). The third chart shows that there is about 170 MB of transactions and you can see large number of low-fee, low value transactions being dumped on to the network.

CyberDildonics 9y ago
> Since there is no central authority in bitcoin, there isn't a process where one can screen out "spam" and legitimate uses. That's exactly what fees are > blockspace that everyone has to store Only people who sync with the chain have to store the blocks, not everyone who uses it. The vast majority of people don't sync with the chain and still reap all the benefits. > Are those spam transactions? I'll let you decide The miners decide
t0mbstone 9y ago
The simple fact that a currency like Bitcoin is technically limited to a set number of transactions per second should be proof to anyone with a brain that it is fundamentally flawed. There are over 7 billion people in the world. A good crypto-currency should be able to handle millions of transactions a second, indefinitely. It should be as ubiquitous and as easy to use as the internet itself. In an ideal scenario, the cost to run the currency network should be cheap enough that there should be virtually no transaction fees other than participation alone (think bittorrent), and the entire thing should be secure and un-hackable. It should be stable in value, or at least slightly inflationary (so people actually spend it instead of just holding onto it and treating it as an investment vehicle). Bitcoin doesn't even come close to meeting these goals. It's so incredibly flawed, it boggles my mind how people don't see the glaring problems. The wheel is moving, though! Bitcoin may not be the end-all, be-all solution, but it has certainly got people thinking.

fearface 9y ago

One way to reduce the Blockchain size is pruning: https://en.bitcoin.it/wiki/Scalability#Storage https://en.bitcoin.it/wiki/Scalability#Storage

DiNovi 9y ago

well IMO its silly to assume in 100 years the internet will be the same speed as today

keymone 9y ago

so first of all, you don't need to store all historical data to operate bitcoin node. it needs to be bootstrapped, but after that there is a thing called UTXO set which tracks unspent outputs (basically account balances), so you don't need to go all the way back to verify you can spend something from a tx in year 2010. so basically as you bootstrap and verify you can prune older blocks without sacrificing security. also there were some proposals related to utxo set commitments, where utxo set signature is recorded in the blockchain, so bootstrapping node can just download it from anyone and verify starting from the commitment block. this would solve the problem entirely. second - storage and bandwidth are still a concern out of which a whole bitcoin cash altcoin was born - some people thought it would be good idea to go with even larger blocks justifying it by even more handwaving. third - there is a strong push towards developing upper layers of the network, so called payment channels. one solution that already exists on mainnet in alpha stage is Lightning Network, is basically a protocol for routing payments between multiple channels kademlia-like. these channel transactions are not recorded in the blockchain, you only need to pay fees for opening and closing a channel, so two on-chain transactions, and then you use your channel funds like a debit card, paying for low cost things. and your last point - yes, halving is a significant event during which miners lost almost half the revenue. on original bitcoin network this is offset by a working fee market: right now about 20%+ of block rewards come from tx fees. obviously miners know the schedule of halvenings and those that will become unprofitable will try to get rid of their infrastructure investments beforehand - this will lower network hashrate which will send tx fees higher, which will somewhat offset losses as well. we've already been through couple halvenings, they didn't really affect much, i think next one will be crucial test for bitcoin's survivability. of course people pushing for zero fee transactions are trying to have their cake and eat it too, storage is a scarce resource and nobody can sustainably to offer it for free. bottom line is - blockchain cannot scale to micro transactions, most of economic activity will be done in payment channels, on chain transactions will be mostly for high-value settlements between larger players (banks, exchanges, etc). edit: word