Quote from: Este Nuno on July 27, 2014, 07:45:16 PM

Quote from: AnonyMint on July 27, 2014, 07:39:02 PM

Quote from: canonsburg on July 27, 2014, 07:29:24 PM

I don't get this obsession with bloat. No one in their right mind is going to download the full Bitcoin blockchain so they can spend $10 buying bread at their local shop. So why the hell are people complaining about Monero blockhain bloat? The vast majority of people are not going to be using the full client running a full node. That's just crazy to expect that. Even Bitcoin development is going in this general direction, by getting rid of the QT client and moving to a Core client because in the future, the Core program will be running the full node and third parties will develop the wallet program.

100 Petabytes scaling and thus can't scale without centralization of mining. See upthread details.

So we have people on one hand saying 1TB or some more, and then on the other end you're saying 100PB(?!). Whose estimate is closer to reality?

If it actually is in the 1-2TB range I think that's pretty manageable to be honest. Most people want to use thin clients anyway.

To the previous poster:

(a) Some people are looking for any hammer they can to bash on the CryptoNote coins in general, and XMR specifically. This is the most obvious, because it does represent the most apparent tradeoff made thus far: The cryptonote blockchain is larger than the bitcoin blockchain.

(b) It's actually interesting, for the above reasons, if your goal is for BTC or XMR or BBR or whatever to "take over the world" (i.e., replace cash) and scale to handle thousands of transactions per second.
  (b.1) The block time is 60 seconds, so there is 10x more overhead from "blocks" (as opposed to transactions). An "empty" XMR blockchain is O(10x) larger than an "empty" BTC blockchain. An empty BBR blockchain is 5x larger than an empty BTC blockchain. I use the "order of" notation O(x) to note that this is not a precise answer, but it's the right order of magnitude.

  (b.2) To facilitate anonymity, transactions in CryptoNote are split into fixed-size chunks, so that the precise size of the inputs and outputs is not a clear distinguisher of the transaction. Thus, one "payment" (as the user sees it) involves more inputs and outputs than might be typically used in Bitcoin. Kind of. None of the analyses that people have posted are very careful on this front, because the increase in transaction size depends on how well your client software can match up your existing set of coin inputs (which were also sent as fixed-size chunks!) to the outs it needs to generate. This may be anywhere from 1x to 10x.

  (b.3) When using mixins, the ring signature scheme adds an extra signature for every mixin used. Thus, again speaking vaguely, a CryptoNote transaction is a few times larger than a BitCoin transaction, other factors being equal.

(c) People are guessing and seem to be biasing their answers in favor of the direction they like.

I don't think people have a good answer to how much (b.2) above actually adds to things -- my personal hunch is that it's maybe a factor of two or three, but that's a guess. Some people are pointing out stridently that the "empty blockchain" is actually only X much bigger than bitcoin, and that's true, but it's also a bogus measure, because we shouldn't care about the size of a blockchain that's being used primarily for mining coins and sending them to an exchange.

The people picking huge constants and multiplying them all together are being similarly deceptive, because they're ignoring that some overheads (such as the block header, etc.) are fixed and shouldn't be included in the multiplication.

If you pick and choose this, you can construct almost any answer you like.

Here's my guess:

- Splitting into fixed size chunks: 2-3x
- Ring signatures with a sane minimum for guaranteeing anonymity: 3x
- Block time: Negligible in the limit as the number of transactions increases
- Bigger individual signatures and addresses: 2x

Overall: Roughly 10x the size of bitcoin. Perhaps 10x, perhaps 30x. With bitcoin at nearly 20GB, the XMR equivalent would be somewhere in the 200-600 GB range.

This is a concern. Once the blockchain is too big to fit on a single commodity hard SSD, for example, the *time and knowledge* needed to participate as a full node in the system increases substantially -- RAID arrays, building your own hardware, etc. The time to download the blockchain may also be a concern for the time to bring a new node into the ecosystem.

This is also an overblown concern for the moment, given the newness of all of these coins, the experiments underway at ring signature pruning (BBR), the increasing understanding of pool management to mitigate dust transactions (XMR), and the fact that we have *no idea* what kind of fees people would be willing to pay in order to have their transactions remain anonymous, or whether and what technical solutions will work well to help keep the size down. There's also a complex long-term interplay with Moore's law and the volume of transactions.

tl;dr: People are blowing a lot of smoke to try to achieve their own aims and pretending to be a lot more certain about it than they really are. It's an important question. But you probably don't need to panic about it in the short term.