George, I have another question about the current proposed "auction" model for service nodes.
If each service node will be running the service(s) chosen by the owner of that node, how can the service node network guarantee quality of service for a particular service?
For example, if there were 800 service nodes in the spread and only 50 of those nodes provided a particular service, what could we do to make sure that 25 of those nodes wouldn't be dropped by the spread thus affecting quality of service. This is a particularly pertinent question because if there will eventually be a variety of services on offer, each service will probably provide different income to the service node operator thus affecting the affordability of running a node providing that service.
Is it as simple as saying that the model decides which services are viable to operate?
Every servicenode will be encouraged to run every service that exists. In parallel on the same machine.
Or else the node will receive negative score points for every service that is missing.
Only if you run 100% of all available services will you receive the full 30% share of the mining reward. (or whatever we will choose)
So in your example you must explain why 750 of the 800 service nodes decided to
not run that particular service.
This is something that should not happen voluntarily (but maybe because of other reasons?), because why would they want to miss the extra SPR?
)