Topic: [ANN] eMunie (EMU) - NOT a BitCoin fork/clone - call for beta testers - page 19. (Read 78428 times)

If its still open sir, I wan't join the beta testing.   

This proof-of-work system would be like having anyone running a bitcoin node compete for doing the "work" of a difficulty 1 hash that is attached to any transaction.  The idea to change the proof-of-work to a tree-sort is great, as tree sorting runs on a CPU better, but there are massive errors in this implementation:

1) The "work" is tied to the number of transactions
2) The "work" difficulty does not rise with the number of people competing to do it
3) The currency will get distributed to a handful of people who know how to spawn infinite hatchers since the infinite hatchers hardly have to do any "work"

Also, tell me if this one is wrong, as I'm not certain how this is done:
4) The "work" is assigned randomly instead of having every full node verify transactions, as in bitcoin.

I think it's funny that litecoin was run for months as a CPU-only coin and called a "botnet coin" by bitcoiners.  Then there was a giant uproar when someone figured out how to run it on a GPU, but now litecoiners call any true CPU coin a "botnet coin".  Yes it sucks if 30% of the new currency goes to a handful of botnets who don't process transactions, but this is far better than having 95% of the new currency go to a handful of guys with huge GPU farms.  If the currency does not get a good initial distribution, it will fail.  The success depends on getting it into as many people's hands as possible.

Coblee's thread about GPU-mining and Litecoin for people clueless about the history of litecoin:
https://bitcointalksearch.org/topic/thread-about-gpu-mining-and-litecoin-64239
TacoTime suggested switching the POW to a radix sort or tree search to make litecoin CPU-only again, but Coblee didn't want to do the development work to implement this, so nothing was done.

just my point. lets hope we just cant see the solution, but there is a solution.

Absolutely agree, I saw this weakness as well (already mentioned about 100 times in this thread)

But it's an interesting experiment non the less.

That's why:

But it may not benefit the network if there are too many hatchers. As pointed out before, you could run many hatchers on the same machine, potentially a huge number of them (billions of virtual hatchers which are just IDs?) depending on the details of the protocol to be revealed. Then you are not even providing the storage space. So there will be a million of hatchers competing for the same transaction. And then transaction spam may kick in, possibly maliciously designed to go to a certain set of hatchers in order to bump their trust.

It's true that you will not break the system by that. But the rules for money distribution are complicated and possibly flawed. As they are presented now, they could be cheated, and then the cheating will become the new, perverted rule. The advantage of bitcoin here is the simple, clear money distribution rule which has not been cheated so far (if you don't consider mining pools as cheating).

That sounds extremely boring. I think you don't quite understand how different eMunie is compared to BitCoin and thats maybe why you are trying to get it the same requirements.  So drop everything you think you know of 'security' and let the big boys play, as i'm fairly certain fuserleer knows what he is talking about and wouldn't just ramble nonsense.

This is why eMunie is having the beta, not to see if there are any holes in eMunie, but yet to see how people live it compared to the alternatives out there. If you want to try it, go for it, if you don't then just stick with your favorite currency.

One big difference as I see it would be:

Instead of investing a lot of Money in Hardware that basically does nothing but to produce hot air. people would invest the money in Hardware that actually benefits the network. Space to store the Blockchain and Bandwidth. Just imagine all this Millions that where invested in Bitcoin Mining would have been invested in this resources.

We wouldn't even need to talk about the max. Blocksize issue yet.

Interesting if this really works.

As per my PM. Let me know

You are trying to explain why the system cannot be cheated. Im not saying it can be, I sad : "even when source will NOT be given, or system was perfect in other ways."



thats exactly what I ment. So You agree?


POW : new currency generation is distributed to hashpower
1.people buy tons of unnecessary GPUs, build gigantic clusters of 10, 20 or even 100 Mhash . lots of resources is lost.
2. a lot of energy is wasted on POW.

effect = network is run by proffessional hashers. you WILL NOT get even 1usd worth of BTC by mining on normal computer for a year.

eMunie : new currency generation is distributed to transaction-verifying-power/luck
1. people buy/rent a lot of servers with VM and launch as many instances of hatchers as they can.
a lot of resources is lost, just more cpus mobos rams and drives and less gpus
2. a lot of energy is lost to run millions of unnecessary nodes

effect = network is run by proffessional hatchers. you WILL NOT get even 1usd worth of EMU by hatching on your computer for a year.


where is the difference? that are the advantages?

disclaimer:
Don't get me wrong, I LOVE the idea, thats new and refreshing. I'm NOT criticizing it. I'm simply very much interested in it and just want to understand it.

p.s. what about my P.M. ?

If you are still open to testers I would love to be considered. Thank You

You mean something like YAC ?

Not that very well distributed either ... CPU POW is *very* botnet friendly.

(Besides, YAC was abused by people renting AWS clusters at the start, and was suspiciously designed to be GPU friendly - but not publicly available - for the first months of existence)

I'm basing this on the above. If clients have a configuration for "trust level", then hatchers with higher trust are going to be favored by clients over hatchers with lower trust, and hatchers with no trust at all will probably have to wait until there are no other hatchers available to service transactions.

I'd like to become a tester, sign me up.

I know. That was the hypothetical situation that was being used for theoretical purposes. I'm just making the point that this theoretical situation is virtually impossible because as this network grows, the amount of hatcher nodes that even get to handle transactions and prove themselves honest cannot be greater than the greatest number of simultaneous transactions sent out on the network at any time... so in reality you end up with a relatively limited amount of hatcher nodes that are even relevant to this system. Slowly over time, as transaction volume grows, more hatcher nodes are required and allowed to gain trust with the system.

We already established this:

"The point is that the attacker has 1m (essentially unlimited number) nodes with equal trust to the honest nodes."

The idea to switch the proof-of-work to a radix sort or tree search is the best move an alt-coin can make right now.  Making the currency allocation and double-spend defense dependent on millions of spawned nodes is not smart though.  You have to make the tree-search difficulty go up based on how much work is getting done.  Otherwise all the coins will get distributed to only a handful of people.  The way to solve it is to have a difficulty adjustment so that the attacker gains nothing from infinitely spawning and doing nothing.  Each of the spawns must do a lot of work.  Just make it bitcoin, but change the mining to something that is GPU resistant.

I mean think about it... people seem perfectly satisfied with Ripple's UNLs to thwart sybil attacks right? Well shit, if that is good enough for them, then eMunie must seem like absolute security overkill to these people! It has both proof-of-work, and trusted nodes similar to Ripple...

Edit: except I'd want to emphasize again and again, they are *similar* to Ripple UNLs, but far superior, because they are not just some configured list of a limited number of nodes, they are dynamically determined by the network.

Once you have more hatchers than meets transaction level, you just end up with all new ones just sitting there idle and waiting for transaction volume to pick up. The majority of transactions would go to the "trusted" hatchers.

In a way, this is like a hybrid of Ripple UNL and Bitcoin proof-of-work, except it is a major enhancement on UNL, because #1, the network can provably chose who to trust most, rather than just some preconfigured list of trusted nodes, #2 the transactions end up being distributed over much broader range of hatchers, than UNL nodes, #3, the damn hatcher nodes get rewarded! So are motivated to actually do the work. In Ripple, the servers are not rewarded for their work; instead, it is just expected that people will run servers voluntarily.

This is what my eyes look like reading all this. ---> @.@

Id be interested in being a beta tester  6/MH

Sign me up to be a tester please, very interested. Thanks man.