Topic: Archived Content - page 2. (Read 12169 times)

This message was too old and has been purged

Can post the thread link here?
Thanks.

This message was too old and has been purged

In comparison to commercially available computation clusters this will be much more expensive won't it? Because decentralization has efficiency costs. As a result the hashing power will be close to zero -  at most a few volunteers who don't mind making a big loss.

As soon as you gave up on user-supplied PoW this whole thing ceased to make any sense to me at all. Am I missing something here? Is there some magic bullet which is going to make miners work for less than cost or make customers pay substantially more than they need to?

This message was too old and has been purged

63 BTC raised so far, 10% of the coins been distributed.

I apologise for not being clear in my explanations


I said "no such blocks", not "no blocks".  There will be a blockchain secured by PoS.

nihilnegativum wrote


If "block reward" is synonymous with "PoS reward" then the two clauses do not cohere.  I tried to construe his question in such a way that the two phrases were not synonymous.  Since "PoS rewards" refers unambiguously to the generation of the blockchain, "block reward" had to refer to something else.  I could think of nothing other than the PoW rewards earned by executing a 10ms segment of the program, which might reasonably be referred to as a "block" of instructions.


That's exactly what I mean.


One might reasonably call it a "run once, hash many times" attack applied to a 10ms segment.

But these (and other) superficially different attacks are really just variations on a theme.  The attacker finds a faster way of doing something - the precise "something" varies from attack to attack - which profits him at the expense of subverting the system at the expense other participants.  Since the phrase "way of doing something" more or less defines the word "algorithm", "faster algorithm attack" seems appropriate.

Note than not every use of a faster algorithm will be an FAA.  The key phrase is "subverting the system".  A person who find a way of genuinely running the customers program faster, even at the expense of other workers, is not subverting the system, and I think most people would regard his actions as legitimate.

I like the idea of Elastic.
How's the ICO doing?

Ah, you overestimate human nature, even one click is not cost-free. To have the wallet open is too much of a hassle for people. Lets look at blackcoin, as I think it has the lowest staking %, right now only 17% of the network is staking, (of course this is only a problem when there is more money on a certain exchange than in wallets, because exchanges have a free attack). But I get it, if its pledged to be fixed thats that, I think its unwise to set in stone economic parts of the system before the the whole is known, but I guess secondary solutions can always be found, like the ones you proposed and more.


Saying that a coin has a value that is a function of the size of the economy it supports (+its velocity) is the opposite of saying it has inherent value. Nothing has inherent exchange value, it depends on the exchange and so on the economy, thats exactly why trade is important. By block rewards I meant the rewards for winning segments fed by transaction fees, because I agree those disincentivize trade.

Elastic has its own economy, and I think thats where the focus should be, as a specialized currency for computing not a general currency ( I mean of course that could happen accidentially). I imagine Elastic and hashpower as envisioned now are economically separate, right?  I mean that its value is not a direct function of the hashpower being traded, but only a market price of its exchange for other currencies, like normal coins. I doesn't matter if you pay 1ELC or 0.000001ELC for the ammount of hashpower that mines 1Bitcoin, not to the market and not to the network, but perhaps this aspect could be exploited, what if the network considered that x amount of hash = y amount of ELC.

How is this person relevant?

My observations so far (my 0.00000002BTC):

Interesting idea. Cagey practices. Seems like a total gamble (and the house always wins).

Looking at the git repo, not much action yet for something that's 'working fine'. I'm also wildly guessing that Lannister is currently running 's/NovaCoin/ElasticCoin/g' on the elasticd repo.

The video states "elastic is currently in an advanced beta stage". How is that possible with no code, no wallet, and devs barely starting to work on the project?

On the bright side two people possessing higher than average intelligence are at least somewhat involved in the project.

To gamble or not to gamble. That is the question.

Now we're really going down the rabbit hole...

No blocks? So something other than a blockchain then. Interesting.

I think my question to you before got buried - you mentioned the FAA earlier saying there was a devastating version you'd thought of, but I don't understand why its an issue at all if the work being done isn't for securing the blockchain?

Do you mean the worker pretending to do useful work but actually faking it entirely? If so I think that needs a new name because its not entirely the same as the faster algorithm attack discussed earlier.

Why bother not staking, given that the activity is essentially cost-free to the user?

We could build staking into the reference wallet - active by default - and encourage any third-party wallet developers to do the same, also the mining package, which would have to have access to a wallet (or have wallet functions built in) to be able to collect the fees, could stake by default.


I assume that by block rewards, you mean rewards for successfully winning 10ms segments.  I see nothing else that could reasonably be described as a block reward other than PoS rewards.

Under the "different approach" I have alluded to, there will be no such rewards because there will be no such blocks.

As for why they PoS rewards should be negligible:  First, part of the original crowdfunding covenant was that the total number of coin would be fixed, and entirely distributed to the crowdfunders.  This was not EK's idea originally; it's what he inherited when he took over the project, and he has ruled out changing it.  This means that the only source of coin available to fund PoS rewards would be transaction fees, and I want those to be as low as possible (consistent with the need to suppress transaction spam), because high transaction fees act as a disincentive to trade.   Similarly high staking fees would be an incentive to hoard, hence a disincentive to trade.

Why am I so keen to incentivise trade?  Let me ask another (rhetorical) question:  What is the intrinsic value in a coin?  If the answer is "nothing", then we have a problem, because a commodity with a high and increasing price but no little to no intrinsic value is tulip mania.

I don't know how to value a coin, or a real-world currency for that matter, but I do believe that it does have an intrinsic value, which will be a function of the size of the trade economy it supports.  One of the things that attracts me to this coin is that, if we are successful, then it will come with a trade economy built in.  Bitcoin doesn't have that.  Few altcoins do.  But I'm not content with just the built-in economy.  I want our coin to be used to trade other things, and I want to make it as attractive and useful for that purpose as I possibly can, hence low transaction fees.  (I also want escrow built into the blockchain, for exactly the same reason.)


That would require altering the crowdfunding covenant, which EK ruled out when he took over.  I didn't agree with the decision at the time, but it was the decision, and we have to live with it.  There's nothing to stop individual crowdfunders from donating a portion of their holdings to these projects

Since I don't control the funds, I can't answer that.


Three comments:

First, I said fatal to the current approach to verifiability, by which I meant the one described in the current revision (04) of the whitepaper.  I did not say fatal to the project as a whole.  It just means that we will have to adopt a different approach.  I don't believe that tweaking the current approach.will work.  I also said that I had some ideas about what that different approach could be.

Second, am I a dev?  On the one hand, I find myself writing "we" a lot of the time, as in "we could do this, we shouldn't do that", so I guess that means that at some psychological level I think of myself as part of the project.  On the other, I'm not writing any code, which ultimately is the only thing which needs to be developed.  I'm just a guy on the internet with some ideas about what could work, and what won't.

Third, if you're worrying, then don't invest, in this altcoin or any other start-up.  None of them come with any guarantee of success.  Even the established coins including BTC come with no guarantee that they will be worth anything in a year's time.

In my opinion - and you are welcome to disagree - ELC is as good a bet as any other start-up coin out there, and better than most.  And I say that, fully cognizant of all the concerns and doubts that have been raised, including concerns and doubts that I have raised myself.  My stake in this is 3.5 BTC worth at the start of the crowdfunding, amounting to 28,000 ELC.  I've never paid any BTC myself, as I don't have any.  Instead EK promised me half his own investment at the time (I don't know if he has invested more since then) to keep me on board.  I trust him to honour that pledge.

And right now, I am not sitting here thinking, I'd rather have the 3.5 BTC.  I'm very happy with my ELC stake.  If anything, I wish I had more.  I think it is possible, though unlikely, that ELC will be so wildly successful that in a few years time my 28,000 could be worth $millions, which would be wonderful.  Far more likely is that it will be modestly successful and my stake worth $thousands.  Or it could end up worthless, in which case c'est la vie.

It's a punt, a gamble.  But I'm not worrying about it.

 By making it minimal, I think you could be inviting tragedy of the commons or at least some of its ill effects. Sure there is an incentive to protect your currency, but a very limited one. A user will either trust in the network, and think why bother staking, others are staking, or he will lack this trust in the network, and more likely than stake sell his elastic. If there are no stake rewards, miners will get them as payment and sell them (if its not profitable to hold them), so they only have the incentive while they are on the job, but while on the job, they haven't gotten the coins yet and can't stake them. If the network pays the miners, the network has to stake and it has to ensure that it stakes even if no miner does.

I get that block rewards should be negligable, but why make PoS rewards close to zero?


I'd make a very large money supply with not too large inflation, make a significant premine and donate it all to specific noncomerical project that require computation (seti, folding, einstein, prime, etc.) or something like that...

I think its pretty safe to infer two things at this point:

• Since the funds are already being spent, they cannot under any circumstances be refunded.
  
  
• There is no 'organisation'.

This worries me now aswell i was about to buy some elastic coins, but if one dev says it are small things and another say fatal, then this makes no sense and worrieing a lot.

I believe this idea is great but its just so messy and unfollowble if your new, like team changes , who is in or who isnt , who is the actual dev or who isnt, who manage the funds or isnt, what if the total coin amount isnt reach will the coins be burned or distributed on the ico particepent and on what base.

And im just getting warm

Maybe a big meeting is needed in your organisation and sort some things

I think one of the issues is collecting funds while not even knowing whether the system can ever work.

What if there is a problem that can't be overcome, will the non-spent funds be returned?

If the user submitted work is not used for securing the blockchain then what relevance does the FAA still have? Would people not just be either earning their own payments back? Or is this just shorthand for an attack in which somebody fakes having done somebody else's work but actually does nothing useful at all?

Conversely, if you can find a solution to FAA for verification then why can't it be used for PoW?

I have another meeting in a few minutes, so I will confine myself to brief remarks


I'm sorry that I find it cagey too.


I think there are major - in fact fatal - issues, with the current approach.  I intend to show a devastating form of the FAA, one which I can see no way of preventing, or even significantly mitigating.

Consequently I think we will have to go in a completely different direction in respect of verifiability.   I have some ideas about what that direction could be.


I don't think the coding will be that sophisticated.  I think there could be an awful lot more of it than you anticipate.


In my opinion the block reward should be negligable or zero, but as it will be PoS, the computational effort will be minimal too.  The incentive to coinholders to do this work is inherent - if they don't, the currency collapses and their stake becomes worthless. The real rewards will come from running the buyers' programs.  These rewards will be determined by the market.

The reason for the low or zero block reward is that, because the total coin will be fixed, rewards can only come from transaction fees which I'd like to be as low as possible.  I think this could be a significant factor encouraging the adoption of our coin for general purposes, i.e., beyond the distributed computing market.