What happened at Crypti is that standby delegates were suddenly voted up one day only to discover not only was there no running node server, the person himself had long since left the party and was nowhere to be found.
Wait, a delegate that isn't running a node can be voted in? So, you mean that I can sign up as a delegate and not setup any node, keep my server down in order to save on electricity, etc and then when I get notified that I have been voted up then I will have my chance to turn on my server? In the other hand, if I get voted in and I am not running a server, will I lose my chance right away and the next delegate in stand-by will get the vote?.
At Crypti, a delegate that wasn't running a node could be voted from Standby to Active status. With the current Crypti / Lisk blockchain explorer written by Olivier, there is no indication whether there is a server running or not behind a Standby Delegate. So at Crypti when you voted up a Standby Delegate, you didn't know if there was a computer or even a still-active person behind that Standby Delegate. This was less than ideal, of course, to say the least. One problem was that there wasn't even an email address on file to tell somebody they had just been voted to active status and were now expected to stand up a server node. So basically these new Active Delegates would sit there not forging after they had been voted up and everybody would wait a few days to give them a chance to get their act together before voting them down and trying somebody else.
Look, DPoS is not a magic bullet. It's hard to do it right, and it takes lots of people cooperating and communicating to make it work. Crypti had a good system in place but not enough people involved to really make it work. Lisk will add lots more people to the mix than Crypti ever had, and significant forging rewards to Active Delegates that Crypti never had. Are these two changes gonna be enough to make DPoS Active Delegate forging work perfectly and smoothly right from the start? My bet is no way. But Max and Olivier (and many others including me) care enough to keep slogging through what ever problems show up and keep Lisk and its DPoS forging going.
You don't want 5 people running 20 nodes each in secret. You don't want 5 people running 20 nodes openly, either, if there are other good candidates who could take some of those slots as Active Delegates. But despite this ideal, you can certainly show up and say you've got 20 servers ready to go if you want. Who knows how people will vote? Maybe you will get them all on line, maybe only some, maybe only one on line. It all depends on the votes.
The reality of the situation is that two people, Max and Olivier, will be setting up the 101 initial Lisk Foundation servers to get the Lisk blockchain going. Max and Olivier will also control around 15% of the Lisk votes in the beginning with the Lisk Foundation funds. This is a huge amount of centralized voting power.
So I see that its not favorable for one single individual to be running like 20 servers because this will be seen as the network being less secure. Okay, so here is my question, instead of one individual running the 20 servers physically in his house under the same ISP IP address, would it then be favorable if this same individual were to run 20 different virtual servers on 20 different data centers and IP addresses? I know its the same individual running the operation, but because each node is on different locations, if something bad happens to node A, node B, C, etc continues to run, I dont see how the network would be less secure despite these nodes having been created by a single individual.
I would be tempted just to run two servers in my house instead of one, just two delegates, would Max and Olivier know that these two servers are under the same IP address? Or, do you know (going by the Cripti experience) if IP addresses gets passed over the network and becomes public knowledge or at least knowledge to anyone whatsoever. I am not asking this question because I am planning to run lots of servers "Secretly" here on my house, I fully understood that this would be view as making the network less secure and because of this I understand and would be refraining from attempting to run lots of servers here physically in my house (so I guess I could say bye bye to what I was thinking -- purchasing 101 $9 computers - even though if I get approval from the Admins I am more than willing to do it since my house is a very secure place and there is nothing to worry about here and with me)
Running lots of servers from many locations under a single mastermind makes the servers themselves more physically secure and more reliable, but still requires greater trust in the mastermind. What if we go from 101 Lisk Foundation nodes to five community masterminds running 20 nodes each, and these five guys are Ethereum supporters who have organized a setup to gain control over Lisk? What if these five guys all shut their servers off simultaneously? The name of the game in blockchain building is trust no one. If you've got to trust somebody, then give a small an amount of trust as possible to as many people as possible, and hope that a majority are worthy of that trust.
Any attempt by you to get more than one node going may receive votes, but it is contrary to the decentralization goal is that Lisk is trying to achieve. I'm not saying that to discourage or demonize you, just to lay the cards out on the table. If you want to try for the votes to run multiple nodes, go for it.
I actually think spending $900 to buy 100 CHIP computers and dispersing these worldwide to run the Lisk backbone is a great idea. I have even suggested this very idea to Max as a way to organize and especially control the switchover from Foundation to Community delegates. He is concerned about getting dedicated people that would actually use such a handout for its intended purpose, and rightly points out an Active Delegate needs to self-declare their intentions and fund their own node as partial proof of their worthiness.
Also, check my post about "latency" a few pages back. Just because you want to host a Lisk node at home doesn't mean you will be able to do so. Lisk, like Crypti, has a 10 second block time. This requires some fast communication between 101 computers spread worldwide. If your home internet connection is "slow" in its ping latency, you'll never run a node from there no matter how capable the computer is that you have.
Check out the Crypti blockchain explorer, this shows what the Crypti experience is. Especially check out "Delegate Monitor" under "Tools".
https://cryptichain.lisk.io/I have a 8 ms ping to google's server (ping 8.8.8.
, also speedtest.net says my ping is 8 ms to their testing server, would this be enough? ( am hoping so, 8 ms ping response time to Google's server is incredibly reliable). My bandwidth is 25/25 at the moment.
Also, check my post about "latency" a few pages back. Just because you want to host a Lisk node at home doesn't mean you will be able to do so. Lisk, like Crypti, has a 10 second block time. This requires some fast communication between 101 computers spread worldwide. If your home internet connection is "slow" in its ping latency, you'll never run a node from there no matter how capable the computer is that you have.
I actually think spending $900 to buy 100 CHIP computers and dispersing these worldwide to run the Lisk backbone is a great idea.
If super fast ping response times is paramount within the 101 active delegates, then sending 101 CHIP computers worldwide could prove to be a bad idea. The reason is that internationally speaking a Chip computer located lets say in Africa could and most likely is going to have a terrible ping to the other CHIP computer located in the USA the same its going to have terrible ping responses between these two CHIP computers and the one that is located in China. If its crucial that all these 101 delegates shares a very low ping response between themselves, then all of these 101 delegates would at least need to be located in the same country. For example, they could be located all in the USA, but in different states, that way they will have great ping response times between all of the delegates and the one far away in California can have a ping response time as low as 7 to 8 ms to other node (delegate) that is in New York. But when you start thinking international level, you ARE going to have ping responses within the 80-350ms range.
I just finished checking the Cripti's block explorer (
https://cryptichain.lisk.io/delegateMonitor) and I see that not one single delegate there has an approval rating greater than 30% (29.x being user Max) if that were to be true, then a single member will not be able to mine the full 150,000 Lisks per month himself, but at least about 30% of that if that delegate is active 30% of the time. 30% approval rating is low, but at the same time Max' Uptime is 96% could be that since Cripti is a dead coin things there its not running that optimal anymore or perhaps is was always this way. Maybe it gets to be different here with Lisk.
