Profits are nice but unless there's a business behind the coin, they're nothing short of incomplete.
Most sensible thing I have read all day.
I still do not agree with the rationale for a node score to be partially based on how many coins are associated with a node though. Me neither. In fact I had a lenghty disagreement with Mr.Spread about exactly this point. Shortly before he left us.
I think the way servicenodes will work is with a 2-Step procedure:
1) is a servicenode even within the "spread" and allowed to run?
(This will be limited by the maximum amount of possible service nodes (Total coins / 2880 = Spread),
and if a servicenode is even willing to pay the minimum collateral necessary to secure a seat.
Minimum collateral is derived by a completely free market price discovery.)
2) After your servicenode secures a seat, does it provide the services it is supposed to provide with a reasonable quality (availability, throughput,
not necessarily performance)?
If yes, good, if not, the system kicks the service node out.
Now, 1) is more important than 2), and regarding the competitive nature of the game 1) is necessary for 2) to even apply.
It is expected that people who run a servicenode will want to improve their server because it will then improve the quality of the services, which in turn attracts investors in the first place.
It is not necessary to let the protocol reward people who have access to the best server hardware (with the best performance) and hereby exclude other people who have only access to average server hardware.
(which doesn't have best performance, but still a reasonable amount of availability and data throughput)
No, it is only necessary to punish and exclude the free loaders. (people who deliberately try to game the system by installing servicenodes that get payed while providing nothing)
I think availability and throughput is more easily assessable than "performance".
What I mean with that, if a servicenode is highly available but rather slow, it should stay in the race and help contribute to the service network,
it can't be that only very fast low ping servers are allowed to play this game, because this game is not merely about speed.
There are other factors.
This really depends on the specific nature of a service.
Some need more speed, other more harddisk space, others more availability, etc...
So: no, there will never be a node score that is partially based on how many coins you have on your account.There will be scores alright, but like some kind of policing mechanism the score is only here to strictly control if the service nodes are playing by the rules or not.
Scores are not here to reward those who offer a better service.So to go back to the 2-Step procedure:
1) Competitive Pricing mechanism (how much are you willing to pay?) --->
Completely freely moving parameter.2) Strict policing to exclude free loaders (are you playing by the rules or not?) --->
Strict boolean value of true or false2) doesn't need to influence the competitive nature of 1)
2) is just a switch, while 1) is a spread.
And most importantly only 1) is tied to the amount of coins you have in your servicenode.
2) doesn't give a sh*t if you have 10000 SPR or only 100 SPR in your wallet.
It will judge you no matter what.EDIT: An example:
Suppose we have a service application codenamed "Spread the Search" which acts as a decentralized search engine.
Now, to decide if a servicenode plays by the rule we could introduce a general rule like:
"is the servicenode crawling and indexing atleast 10000 websites a day?"
If no, the servicenode is kicked out.
If yes, the servicenode is not getting additional rewards if it crawls 20000 websites a day, because this is not necessary.
No, it will simply have passed the test, no need to reward the "overachievers", because it can be expected that people will want to improve their server hardware ANYWAY,
because it will improve the service overall and therefor attract more investors.
