|
Wages may go down, but purchasing power increase. Business profits may be reduced in nominal terms, but actually be growing.
|
|
|
|
Would you be including the prefixes of the merkle tree nodes too? These would serve as a checksum, so the receiver would at least know which transaction has the colliding prefix. In fact, with the tree nodes included, you could be much less conservative with the prefix length, sending 2 or event 1 byte(s).
Collisions could also be handled by having a re-send system. Hashes were grouped together into 32 hash groups and the merkle root for those hashes. The sender could estimate how much of a prefix is required. If 4 bytes was required, then a group of 32 hashes would require (32 * 4 + 32) bytes = 5 bytes per hash. The receiver could replace the 32 hashes with its best guess and then check the CRC/Merkle root. If it doesn't match, then it could ask for those hashes in expanded form. Assuming a block with 4000 transactions in a block. Peer 1: Sends block with 5 bytes per hash. This works out at 20kB Peer 2: Checks block and finds 2 groups with a mismatch Peer 2: sends tx request byte[32]: Block hash varInt: group count int: index 1 int: index 2 This works out as 41 bytes Peer 1: Sends full transaction hashes for those blocks This works out at 64 * 32 = 2048 bytes This is much less than 4000 * 32 = 128kB. However, it is at the cost of an additional round trip delay. If each compressed hash was prefixed by a length, then the sender could try to estimate which hashes require longer prefixes. The birthday paradox may cause potential issues here. This "groups" idea is essentially the same thing as the merkle tree, if you make the group size 2. With smaller group size, you have to resend less transaction hashes in the case of collision. Also, the birthday paradox is less of an issue; you could use smaller prefixes. Also, consider that we can compress the merkle nodes. Collisions in the nodes can be checked with the parent nodes all the way up the tree, but it becomes exponentially less likely that you'll need to resend a branch as you move up the tree. The advantage of this hierarchical checksum is that you can make the prefixes, of the nodes and the transaction hashes, much smaller. I think that re-requesting any transactions is too costly, so we should work out the prefix size so that there are "probably" no collisions, but only just. Someone needs to do the math. |
|
|
|
The stuff about the market; all true.  |
|
|
|
Gavin's Cost is independent of the trade value of bitcoins, so that wouldn't prevent a halving from improving the attractiveness of fee paying transactions. Regardless, there is much that can yet be done about Gavin's Cost before the next halving anyway, so I don't think that Gavin's Cost is really a problem so much as it is (and will continue to be) the real blocksize limiting factor whether or not we continue to hold a hard limit on blocksizes. With 'thin' blocks (block headers & merkle tree only) the one meg hard limit would easily permit an average transaction rate of 30 - 40 transactions per second right now, doing nothing further. A blocksize limit of 10 megs would get us into the 300+ per second range, which is a workable rate for a major system; and higher rates can be handled better by overlay networks anyway.
Ummm... yeah, I just realised that this 1 - e^(-(1/600) * X) is not a linear function. This was the source of my misplaced concern. If we reduce a transaction down to the size of it's hash, this cost becomes extremely small. Thanks for the feedback. |
|
|
|
One thing I note is that there are lots of ways to optimise block wire representations. Sending blocks as lists of hashes, for example, would use 32 byte hashes in our current code just because it's lazily reusing the Bloom filtering path. But 32 bytes is massive overkill for this. You only need to distinguish between transactions in the mempool. You could easily drop to 16, 8 or even 4 byte hashes (hash prefixes) and still be able to disambiguate what the peers block message contains very reliably. Indeed they could be varint encoded so the hash is only as long as it needs to be given current traffic loads. Malicious actors might try to create collisions to force block propagation times up, but if peers negotiate random byte subsets upon connect for relaying of block contents then such an attack becomes impossible.
I thought of this also. How do you deal with collisions, though? What if the receiver of a hash list knows a transaction whose prefix collides with one in the list, but isn't in the list itself? Would you be including the prefixes of the merkle tree nodes too? These would serve as a checksum, so the receiver would at least know which transaction has the colliding prefix. In fact, with the tree nodes included, you could be much less conservative with the prefix length, sending 2 or event 1 byte(s). Perhaps that makes no sense. I'm pretty noobish with the protocol specifics. I'm very interested in reducing block propagation times because I believe slow times will eventually drive transaction costs to uncompetitive levels. This is due to the inherent cost of "orphan risk" which comes from including transactions. see here https://bitcointalk.org/index.php?topic=334284.msg3669703#msg3669703. |
|
|
|
|
Dogecoin is actually pretty funny :-)
|
|
|
|
...and then the block reward halved...
... and then the price of bitcoin went up x10. Rises in the value of bitcoin are going to outstrip any of the measures mentioned to reduce the cost of block propagation. OR the price of bitcoin simply wont rise because fees are too high. Miners also receive a very large subsidy and they have an indirect incentive to keep the network going (a few less Bitcoins which are worth something is more than a few more worthless Bitcoins).
Tragedy of the commons. The block reward is WAY too high and the reduction rate is way to slow. I still think this is a serious problem. |
|
|
|
Is anything being done about this?
Yes.
The obvious solution is to reduce the orphan cost. To do this solved-block propogation time needs to be lowered.
This is being worked on by migrating from a complete block to a skeleton block, with only the headers and merkle tree included. This alone would reduce the orphan risk cost by about a factor of 10. Thanks for the reply. According to the above analysis, the fees are already about 1/30th of what they should be for miners to break even. The skeleton block solution gets us to 1/3. That's just at the current price. Don't you think this is a big problem? I assume that, at some point, economic reality will kick in and miners will start to demand much higher fees. Suddenly bitcoin isn't so great compared to alternatives. |
|
|
|
Is anything being done about this? Since transaction fees are proportional to the BTC price, this amounts to a cap on bitcoin adoption; less people will use bitcoin as the tx fee becomes uncompetitive. We are already PAST this point, imo. I feel that there is some urgency to fix this problem now. The obvious solution is to reduce the orphan cost. To do this solved-block propogation time needs to be lowered. I don't really understand it, but could CoinWitness be the key solution? https://bitcointalk.org/index.php?topic=277389.0;allI don't know of anything in the development pipeline that addresses this issue. |
|
|
|
|
Run on the platform of abolishing the Federal Government and you got my vote.
|
|
|
|
It 100% will cause rampant inflation. When rich people are holding they money, the money does not enter circulation, ie. buying up food/electronics/furniture etc..., it's just sitting in their bank account or invested in some stocks. but when you distribute these to the masses...
That is not true. Money doesn't just sit in a bank account. It is loaned out and spent. Money "invested in some stocks" is spent by the person that sells the stock. However, I believe it will cause inflation. With an income of 2500 francs per month for doing nothing, a portion of the population will stop working, and the GNP will fall. More money chasing fewer goods will cause prices to rise. Eventually it will reach a balance. Then the whole system will be cheaper than existing system and everyone have incentive to work to gain more than minimal standard of living. True, eventually 2500 francs will have so little purchasing power that people will start working again. Unless they peg the handout to CPI. |
|
|
|
Both these movements are artificially driven by the state. Basically an excuse for socialism; increased state power.
Uhh, no. The state is a vehicle by which to unify progress on those fronts. Contrary to your paranoid visions, the actual intentions of say, environmental projects and regulations are to make us more energy efficient, our lifestyles more sustainable, and preserve the environment. Yeah, keep believing that buddy. Organised crime will save us from ourselves! |
|
|
|
Both these movements are artificially driven by the state. Basically an excuse for socialism; increased state power.
Lol, more paranoia please  You might dislike it, but girls got paid less than boys and they are not more stupid. I think that it is kinda unfair. The state funds, academic positions, research and organisations. Girls get paid the same when you control for same role and same level of productivity. Broad aggregates don't tell the whole story. |
|
|
|
Causality is only meaningful within the universe.
Causality is not even meaningful within the universe. True, but if you believe that then I needn't make a case. |
|
|
|
Reads scam all over, too good to be true  What if it's not, though? Shame to miss this opportunity. |
|
|
|
|
Both these movements are artificially driven by the state. Basically an excuse for socialism; increased state power.
|
|
|
|
|
Why is it assumed that the universe was created at all? "It's here, therefore someone created it!". Well, why?
Causality is only meaningful within the universe.
|
|
|
|
|
Check out Marc Stevens. He's all about this shit.
|
|
|
|
Sounds communist at this point.
It's terrorism! |
|
|
|
Whoa, time out kids. I'm not a noob. I didn't ask about law without central government. I asked about law in a Voluntaryist society.
From wikipedia:
"Voluntaryism (or sometimes erroneously voluntarism), is a libertarian philosophy which holds that all forms of human association should be voluntary.[1] The principle most frequently used to support voluntaryism is the non-aggression principle (NAP)."
Very different indeed from simply lacking centralized government.
Government is, by definition, not voluntary. So really Voluntaryist society necessitates the absence of a central government. |
|
|
|
|
Show Posts
