Quote from: davidoski on February 06, 2017, 10:30:37 PM

Quote from: SatoNatomato on February 06, 2017, 08:22:59 PM

Quote from: davidoski on February 06, 2017, 06:08:56 PM

Quote from: CryptKeeper on February 06, 2017, 07:09:36 AM

Quote from: davidoski on February 06, 2017, 05:33:08 AM

Witnesses are the single point of failure of the system. They essentially control the network and there are only 12 of them. You can imagine that if the rogue government (bankers or whoever) wants to take down the byteball system all they have to do is to take controll over 12 computers running witnesses nodes. This seems to be rather easy to do, especially at gunpoint. Moreover - this can be done without the rest of the network to even notice - if witnesses after being taken over by the rogue party are operated without interruption. Anybody who controls the 12 witnesses can do whatever he wants with the network - for example censor certain type of transactions. All of this is a contradiction to censor resistant trustless network that bitcoin is.

I can follow your arguments and respect your opinion. Bitcoin was created as a decentralized platform and that was a great invention - in the old days when everybody could easily take part in the consenus with their CPU or GPU miners, this system was still intact. But nowadays bitcoin has become a total different thing. Expensive asic miners drive bitcoin to centralization and the need for low energy costs favor some countries.

I will ask you a question: how many mining pools do you need to cross the 50% consensus barrier in bitcoin? I guess it's a lot less than 12.

It's not so simple as you imply. Mining pools does not necessarily decide about the state of the network (that power lies in miners hands). Let's assume that a rogue party took control over mining pools controlling +51% of the hashrate. If these mining pools would try to impose their will over miners (e.g. implementing changes to the protocol not supported by miners) miners would quickly drop those pools and switch to other ones which would lead to the rogue party loosing control over 51% of the hashrate. The bottom line is - to take over bitcoin the rogue party would have to take control over 51% of miners calculated as hashrate. It's not the same as 51% of mining pools as mining pools are not the same as miners. For example F2pool one of the biggest mining pools does not have its own mining hardware - it only facilitates mining for hardware operators (miners). Consequently it's much more difficult to control bitcoin because miners are more dispersed than mining pools. There are many more miners than mining pools. Definitely more than 12.

Indeed.

It is easier to take over Byteball - if you are a global adversary, global power.

That is also its strength, that we the users can replace witnesses, and that there are few of them. Because we cannot replace the miners. We cannot even compete with them at all. But in this network, a few clicks away for any users aaaaand you're gone.

As far as I understand it if a rogue party takes over all 12 witnesses it can prevent anybody from changing witnesses by more than 1. If you can't change more than 1 witness you can't change the rogue party from controlling the 11 remaining. That's called cartel in the byteball white paper. The only way to get out of this situation is to make a "schism" - which essentially is the same as hard fork in bitcoin. We actually have this kind of cartel run by tonych at the moment... Seemingly out of necessity and presumably in good faith but you need to coerce only one person to overtake byteball at the moment.

If you take over all 12, everybody can still make the 1 witness change. And when that is stabilized, the MainChain is moved some amount of set points (think this is a constant in the code), the next witness can be replaced.

The wallet/user is always able to change 1 witness.

Yes the schisms, mentioning it in the whitepaper shows honesty, that is a fork indeed, when enough is enough - users can fire up a new byteball network and transfer/link their funds to the new network - as we are doing today. Cheesy

EDIT: Why I like this despite the schisms mentioned, because we are humans we will fuck up. Totally fuck up, and the best way forward is to move away sometimes. No software and no organization and trust-exchange mechanism will ever be perfect, since human needs will also change. The best ones allow grade change, and are open for schisms.