[discus] Lets increase the feather coin speed and make it the best currency
-
The speed of network confirmations does not seem to be a major problem.
With system development, after release, it is better to raise an issue, prioritise it. Discus the solution - test and then deploy.
It is a common software development mistake to introduce features, because they seem like a good idea. In maintenance all design outs have bugs.
Even if we consider data from other networks, that does not apply directly in our unique situation, so we can not be sure of the outcome of such a change.
The other argument against such a change is obtaining network agreement. As people come on board and set up world wide mining it will be harder to switch the mining network to a revised protocol successfully.
I have already noted previously that it is likely we will have to adjust the network, for instance with transaction fees, or Blockchain database size, that change should be based on an actual prioritised issue or need. Planned in advance and well publicised.
In these cases, the potential faults being, miner desertion to a more profitable currency, and incredible numbers of transactions inflating the blockchain size.
-
[quote name=“FrankoIsFreedom” post=“38697” timestamp=“1385990258”]
10 mins might of been chosen because that is the time it takes to send data to mars?
[/quote]That was very forward thinking of Satoshi then :)
-
[quote name=“FrankoIsFreedom” post=“38697” timestamp=“1385990258”]
10 mins might of been chosen because that is the time it takes to send data to mars?
[/quote]Ha Ha … I love it !
Regardless of how much of a genius Satoshi was in creating Bitcoin, I do believe that 10 mins is a massive negative for the currency.
Take that away and I think it would be almost impossible to invisage a future with another crypto superseding it. With it and in the long-term I believe it is highly possible …Please be FTC ;)
-
I don’t think yet another hard fork is going to help FTC gain peoples’ trust. Let’s focus on actual features rather than just trimming around the edges.
-
Doing something like this would appeal to those who are entirely external to the crypto scene. There were a few people saying that there is not enough difference between Litecoin and Feathercoin, they do not understand that our blockchain history is preserved with ACP or appreciate the changes we have put in to the difficulty to maintain block time and help loyal miners. If we made this move there is an obvious advantage to Feathercoin. There are a lot of coins out there with faster block times but none of them have our profile.
This in itself is not a good enough reason to make such a move. It would need to be in the benefit of Feathercoin as a currency and not just a selling point. The question is, what is the best of all possible block times and how do we judge it?
With so many coins out there trying different things there is a lot of data to answer this question. 10 minutes seems an age on Bitcoin but then merchants are starting to accept transaction broadcasts as enough. As Kevlar points out this is not really good enough as people can attempt to double spend by sending one transaction twice, once without a fee and another with a fee. The one with a fee and the other gets rejected.
As Wrapper says there is a lot to consider and other blockchain data do not necessarily apply to us. This is something foar discussion that may well just end up as only that. These are the sort of conversations we should be having anyway to make sure that we have the correct choices in place. 150 second block may well be spot on :)
-
[quote]The problem arises with the fee model. You see the way it works is if a double spend is seen, the one with the higher fee gets put higher in the priority queue than the lower one. So you could send the transaction, gets your goods, then broadcast a transaction spending it back to yourself with a higher fee, and assuming you did this all before a block was mined, you’d rip the seller off.[/quote]
Is this a good way to deal with double spending? Could it not just be first come first served?
-
[quote name=“kris_davison” post=“39176” timestamp=“1386095943”]
[quote]The problem arises with the fee model. You see the way it works is if a double spend is seen, the one with the higher fee gets put higher in the priority queue than the lower one. So you could send the transaction, gets your goods, then broadcast a transaction spending it back to yourself with a higher fee, and assuming you did this all before a block was mined, you’d rip the seller off.[/quote]Is this a good way to deal with double spending? Could it not just be first come first served?
[/quote]Yes it is the right way, because it models what a rational miner would do.
Depending on how the network happens to be talking, some peers may see one transaction before another, while others see the second one first. The priority queue is how the transaction get ordered for inclusion in blocks, and even if [u]you[/u] ignored the second higher fee transaction, there’s no incentive for [u]another miner[/u] NOT to, since he’s getting the fee. He includes it in the block, and you need to mine a longer chain before he can to enforce your rules over his. The behavior is the one that anyone who’s mining for maximum profit will take.
You see it’s even worse than that, because someone with control of a large source of hashes could just NOT broadcast the double spend, and only broadcast the block with the transaction in it. Assuming no re-org happens in the next block or 6, there’s no going back, but with enough hashes, he could even force a re-org several blocks back. This is what ACP protects against by centralizing the authority of the blockchain, and why it’s a pointless feature if the attacker hash the majority of the hash power anyway: They can just mine the longer chain from the start. You’re trading centralization of authority for distributed consensus just so you can stop a person with large hashpower from double spending several blocks down the chain.
In a situation in which the mining power is distributed, ACP is a pointless feature, and unless the consensus is distributed, there’s no incentive to contribute to it’s decentralization. Therefore ACP de-incentives mining and devalues the network. If the currency is devaluing, there’s no reason to invest in it.
-
Ah ok so its the greed principle im happy with that. I think at some point if and when the hash rate gets higher the ACP discussion will be resurrected but I think most people are happy to feel safe from attack from now.
(this is just my personal view and to be fair pretty uninformed at that)Do you know of any other way to protect against a double spend? I guess you cant protect against anything until the transaction has been included in at least one block (other than an invalid transaction input != output etc)
-
The only way to protect against double spends is to reduce time between block generation (as is being suggested), while increasing hash rate and distribution. As I’ve pointed out, ACP eliminates the need and incentive for both hash rate and hash distribution: The network will be just fine in the absence of both.
-
Phoenixcoin had a block target of 45 seconds and usual 6 confirmations for a transaction before the last hard fork. Now it’s 90 seconds and 1 confirmation thanks to ACP. Orphan rate has decreased very significantly to about 2%.
-
Surely the mining rate will settle down to that level which the value of the transactions subsidises?
It is also technically possible to switch of the non automatic parts of ACP once mining is distributed enough to prevent attack.
At the moment there is a significant proportion of switchers, because miners rely on the block subsidy, and the frequency of block payout could be infuenced by the coin switchers manipulating the Difficulty change.
I don’t want to say this, because currently Feathercoin mining is viable. Id’ like to keep it that way for a bit, up until recently, as a steady miner I was subsidising the network.
Once Wallets are more distributed, merchants will also have incentive to mine. As nodes for the network, because distributed mining increases security and it would reduces confirmation times.
-
[quote name=“Alex27” post=“38354” timestamp=“1385909256”]
Lets increase the feather coin speed and make it the best currency :)This will be one of the key factors of making it a true global currency so it can take up mass adoption.
:)
[/quote]
Speed… All Cryptocurrencies that weren’t launched yesterday take a while to process… It isn’t that bad. Slowest thing about litecoin is the Mining an the Chain Syncing. -
This thread made me think of the idea that we should have an open market for the mining fees. Miners should be able to state their preferences for fees and they should be matched with the preferences of the sender. Sometimes when I send my coins I don’t need them to get there quickly, like if I am backing up to paper wallets. But if I am paying a merchant I want it to go there ASAP.
I would also like to see faster payments but it would have to be weighed up with with the trust issue and the security concerns.
-
It’s my understanding that once the ftc mining network becomes larger, speed of confirmations won’t really be an issue for brick and mortar adoption. Correct me if I’m wrong.
[quote name=“chrisj” post=“51860” timestamp=“1389631400”]
Miners should be able to state their preferences for fees and they should be matched with the preferences of the sender. Sometimes when I send my coins I don’t need them to get there quickly… But if I am paying a merchant I want it to go there ASAP.
[/quote]That is a BRILLIANT idea.
-
[quote name=“chrisj” post=“51860” timestamp=“1389631400”]
Miners should be able to state their preferences for fees and they should be matched with the preferences of the sender. Sometimes when I send my coins I don’t need them to get there quickly, like if I am backing up to paper wallets. But if I am paying a merchant I want it to go there ASAP.
[/quote]LTB got into this topic. If I can locate (doubtful) the discussion I’ll post it, but I believe the fact that most of the miners would opt for the higher fees and leave the rest in the dust (your send will take an astronomically long time … if it’s processed at all) was brought up. But a counterpoint was made that someone could then take on all of the lower fee transactions and corner the market (or something like that, I’ll try to find it).
