Farmers will only farm if they are given an amount of ‘value’ in return for their services that they find acceptable.
If there are not enough farmers, then the value offered by the network goes up (more SAFEcoin) until enough farmers are available.
So, strictly speaking the fact that Get’s are not directly paid for doesn’t mean that farmers aren’t receiving something in return for the Gets - they are, it’s just that this value is embedded in the Put fee.
If people don’t put enough data to the network and hence recycle their coins however, then, it seems, the network will run out of SAFEcoin.
We don’t know what the outcome will be though … yet. It depends on people and what people want and no one can know that until after the fact.
Also consider however that SAFEcoin gains value from people for all sorts of reasons – all of the community at large that uses SAFEcoin for trade of any kind is contributing to the value of SAFEcoin. Hence all of the community at large is providing a subsidy to SAFEnet and to hence to the Get-ers.
The network is paid for by Put’s and the network pays farmers for providing resources … and farmers who provide solid resources are paid more (I presume) - in line with some equation.
I think it could be the case that some data will be orders of magnitude more favorable in terms of Gets than other data.
The network is going to store highly valued data in quality farms (high bandwidth farms) …
So I’d presume that best farmers will be shouldering a much (vastly?) larger percentage of the Get load … but if these elite farmers aren’t paid enough to shoulder the ‘Get’ burden then they’ll leave the network … So these equations may need to be tweaked by the devs to facilitate optimal network functioning - yet they will have to be careful that they don’t bias the payment scheme such that it facilitates the centralizing of data into elite farms.
Maybe that’s complete B.S. … perhaps a dev will elaborate and straighten out my confusion.
The network does not play favourites. That is as long as the node/vault is adhering to the protocol and not serving up errors then the only factor is the speed that the requester gets its data requested.
Speed though of supply includes the vaults speed to serve the chunk up and then also the physical distance the chunk has to travel in its journey from vault to requester. This will obviously change for each get from a different requester.
Ranking system was changed so that non-persistent vaults were the norm. Now its more that a vault losses rep when it misbehaves or is unbelievably slow. The idea was to allow phones on charging, computers that turn off for the night, etc to have close to equal footing as persistent vaults. This way it was thought that the small farmer would participate more since they have a reasonable chance to earn. In other words to make farming more inclusive and the biggie is to improve security because a majority of vaults are turning off and data is harder to locate/track.
The cost of PUTs has a link to space available on the network and so increases when farming rate increases too. I am sure that long before the coins available be be issued becomes scarce the put & farming algorithms will be tuned to ensure the long term viability of the network.
This is key, and I hope there are sufficient ‘levers’ going forward to tweak things if there is a problem of Safe becoming very expensive due to high bandwidth demands of farming.
People do need to be able to access Safe without contributing, but it may be beneficial to have incentives to contribute other than earning SafeCoin if the equilibrium reached results in highly expensive put costs.
If farming turns out to have a fairly low bandwidth requirement it won’t be an issue, but what mechanisms could help correct things if bandwidth usage is sky high, leading to people contributing few resources to the network & high put prices?
Maybe some kind of bandwidth prioritisation for those who are also contributing to the network to incentivise people to contribute even if SafeCoin rewards alone don’t seem sufficient?
If VPS’s don’t charge for bandwidth & are cost effective, that may balance things out. I’d imagine the cost of a VPS will be higher than spare space on home PCs, so while they may help, the price that makes running those worth your while may still be above optimal for maximising wide use of the network (by allowing home users to contribute spare storage, which they won’t be able to do if bandwidth requirements are too high).
Considering they started with a pay to GET system and when working though the issues came to the conclusion that free GETs was the better way for many reasons.
(not a dev though)