Topic: Vertcoin - First Scrypt N | First Stealth Address - Privacy without mixer - page 286. (Read 1232734 times)

In the name of being diplomatic and not making a fuss, I've held back on saying anything about this for a while, but now we need to talk about it.

While we've achieved fantastic things in terms of p2pool, and Vertcoin has the highest proportion of p2pool mining of any chain I'm aware of, we do have an issue with one or two pools, as Verters great work has revealed at http://www.verters.com/block-analysis

Namely, bitcrush and vertco.in are both rather too big. I have no doubt they are good pools (after all, so many of you wouldn't be there if they weren't) but that's not the point.

Some of you may believe, because it's called a 51% attack, that you need 51% of the network hashrate to do it. Not so - it's just that if you have 51% of nethash you can _ALWAYS_ do it. If you have less, you have a smaller probability of accomplishing it, in proportion to how small your hashrate is (ie. it's incredibly unlikely, but you could in theory do it with 1kh/s).

Let me put some figures on this for you:

With 25% of nethash (which is what bitcrush and vertco.in both have at present), the probability of an attacker successfully orphaning the commonly used 6 confirmations is about 5%. It rises exponentially though, and if these pools grow to 30% of nethash, they would have a 13% chance of accomplishing that attack. Let them get to 35% of nethash and it's a 28% probability.

On the flip side however, it doesn't take a massive reduction in their hashrate to vastly reduce that probability - for example, with 20% of nethash a pool has only a 1.5% probability of accomplishing such an attack, and at 15% of nethash it's down to a much more acceptable 0.2% probability.

The other issue is, two big pools could in theory collude. I am personally satisfied this is no issue here, because I know that Bushido runs vertco.in and is not about to attack his own coin, but it still is an issue, and we shouldn't need to trust anyone here - the whole point of this is to make things trustless and secure by design.

I'm very proud of the way we've managed to get the p2pool adoption in VTC so strong, but let's not rest on our laurels - there is still much work to be done. Please consider moving if you're on bitcrush or vertco.in, preferably to p2pool.

Remember fees are lower on p2pool generally, payouts are higher and you're helping the network - http://p2pool.vertcoin.org/ will show you all the p2pool nodes that can be detected and selecting one close to you geographically is best (also setting one or two others as failover in your miner config) - the advantages of p2pool are manyfold, including that the pool operator cannot lose or steal your coins either through malice, incompetence or getting hacked, because they never hold the coins (they are paid directly to the miner), also, if the p2pool node you are on fails, you still get paid your shares because they are distributed on the sharechain, and not stored on any one node to be lost due to a failure.

If you don't wish to use p2pool, at least try to join a smaller pool. Ideally I'd love to see everyone using p2pool but that's not realistically going to happen and it's actually not ideal for very small (ie. less than one modern GPU) miners, but even if we could get all the traditional pools down to at most 15% of nethash, that would be a very significant achievement.

Up to now vertcoin hasn't proven to be anything more than a newer version of litecoin (which was also meant to be asic resistant or even proof). I am convinced vertcoin has a bright future, but now we can only speculate that a modified algorithm will be incorperated in vertcoin when the need arises. There's no integrated system to make such a change any easier than with ltc or btc..

Come to think of it; the adaptive-n schedule in fact suggests no change will be incorperated into vertcoin for the next decades!

If you don't wish to use p2pool, at least try to join a smaller pool. Ideally I'd love to see everyone using p2pool but that's not realistically going to happen and it's actually not ideal for very small (ie. less than one modern GPU) miners, but even if we could get all the traditional pools down to at most 15% of nethash, that would be a very significant achievement.