That is why I think old chunks need to be rewarded more relative to new chunks. A vault operator needs to view each chunk as equally valuable. It’s easy to see popular chunks that provide a steady income stream as valuable, not so much for cold data that might only be accessed twice a century. This will help to instill a sense of stewardship.
1 Like
I will do it. First I will read a lot and try to understand the issue again. Then I will try to bring the point that I see a problem very specifically.
I find it strange (and not picking on anyone and esp not the OP) how economic models, none of which have served humankind for any significant period, bring people to extreme viewpoints. I mean
followed by
Then
So it’s extreme predictions using a get out of I am no expert/IANAL etc…
However
Is great, but
Can be a bit tough, there maybe community bias, but there is a lot of introspection and friendly (in the most part) debate. Many topics such as pay the producer have a lot of debate, same with erc20 and so on. So I would not say the community was that of one voice. i.e. bias may not be as much an issue as you would think.
Honestly, though the economy chatter is really like a load of folk arguing over which horse to bet on. My view is to take a very simple approach and keep iterating, staying simple. Otherwise, a complex approach will have loopholes. I also believe no system will be 100% fair to all political colours and inclinations. What I do want to see is SAFE getting data and knowledge to as many as possible. The economy is part and I am 100% certain it can be debated for infinity and the side effects will still be unknown.
10 Likes
I personally expect topics like “I’m new here, but now I’ll tell you where you’re wrong and I’ll fix everything” to increase x100 with the release of the beta. This is human nature, everyone has an opinion. That’s why we need more work and less talk. 
7 Likes
Not to go too far off on this tangent, but maybe there another way to approach this problem. I think we’ve both been advocates for some form of monitoring of vaults by the network to make sure that they hold the data that they’re supposed to. I wonder if it’d be possible to vary this monitoring rate for the older data such that in combination with caching of new popular data, the chunks are equally valuable to the vault (without modifying reward rate chunk by chunk).
I didn’t come here with any intention of changing anything in the project. Confronting is just a natural way of trying to understand the system. When the confrontation is not resolved, it can be for several reasons:
- I did not understand.
- The project contains a problem.
- I understand, but I do not agree, although it is not a problem.
At the moment the highest probability is that it is case 1). Obviously. So I will review everything.
8 Likes
Not related to you friend. I comment that this will be the normal behavior of every new person who comes to the forum and is interested in the project. I drew David’s attention to this so he wouldn’t waste time with it. We (you and I) can talk as much as we want, because talking is cheap.
1 Like
Well I’ve been here a while and I confess the economics of the ecosystem is not something I understand well at all. This may be because it’s the area of least interest to me personally, so I haven’t put in the effort to try to understand how it’s going to function, or it may be because it’s the next thing to be worked out by the team.
Either way, if it is to be by trial and error that’s fine - it’s likely to be complex balancing act and simulations almost certainly won’t cut it in the real world. I do hope the system has enough flexibility built in so that changes can be made on the fly, so we can avoid getting stuck in a cul-de-sac and unable to reverse out, as with the Eth example. How this works with a decentralised system is another area I’m a bit foggy about.
6 Likes
I don’t think you’re missing some point. It is idealistic but also infeasible. One of the issues is that people often forget that the value of Safecoin (the Network’s rewards system for both PUTs and GETs) must dynamically relate to real world value (I.e. transact-able for more fungible forms of value like fiat).
It’s not enough for the relative price (I.e. reward) of GET vs PUT to merely see-saw back and forth in SAFE terms. If GETs are costing more, then the relative price (reward) must increase in real world terms. This would require injecting more funds into the system by either 1) levying a GET fee, 2) hoping that the rate of PUTs keeps up, or 3) just “printing” more SAFE and assuming that the SAFE economy can exist entirely detached from real economy. In the third case, the value of SAFE is no more than just illusory.
Put another way:
- PUTs are currently occurring at a stable rate of 100 PUTs per day with a reward of 0.1 SAFE/PUT (10 SAFE a day)
- GETs have been occurring at an average rate of 1,000 GETs per day with a reward of 0.01 SAFE/GET (10 SAFE a day)
- Between PUTs and GETs 20 SAFE are rewarded a day
- Suddenly the number of GETs for a particular data spikes to 1,000,000 GETs/day bringing the total average GETs/day to 1,001,000 for a month before the number of GET requests for the popular data begins to decline
- The number of PUTs per day remains unchanged at 100 PUTs per day
- The variable cost to service a PUT remains unchanged
- The variable cost to service the GET for the popular data has decreased due to caching and Network redistribution of capacity; however, the absolute cost to service GETs has now gone up
- The reward for servicing PUTs decreases to 0.01 SAFE per PUT (1 SAFE a day) and the reward for servicing the GETs increases to 0.1 SAFE per GET (100,100 SAFE a day)
- The total of rewarded SAFE has now increased to 100,101 SAFE per day
Where does this extra 101,081 SAFE per day come from if the number of PUTs has remained constant at 100 a day?
Alternatively you might say the reward would rebalance such that the total rewarded SAFE per day does not exceed 20 SAFE. In this scenario:
- Network value remains constrained by the PUT rate
- The farmer has lost in both relative and absolute terms because the reward system has not taken into account the farmer’s real world costs (real, nominal and opportunity)
- The farmer’s reward is not linked to the total amount of completed work
- In essence the amount of value the farmer provides has no real bearing on the amount of reward s/he receives
In times past, it has been suggested that MaidSafe could benefit from consulting with economists. I think that is still the case. Like PARSEC, some things work well in theory but don’t function well in practice. Some aspects of the SAFE economy as currently planned may mirror similar constraints in terms of being theoretically elegant, but functionally less practical.
I greatly appreciate your perspective. This Community has become increasingly insular over the past year or so. As said before, closed systems die. Every system needs new insight and perspective. Through the mechanism of discourse we arrive at superior solutions.
12 Likes
There is also another factor to consider.
Safecoin fiat price will be decided by free market, supply and demand.
If fiat price increases this should encourage more farmers to join or add storage capacity.
This in turn lowers put cost ( in safecoin ) as put cost is decided by free space on the network.
I dont know if that makes the problems envisaged better or worse, but certainly needs taking into consideration.
5 Likes
This is and has always been the core of the price of safecoin. The value in relation to data is to maintain enough farmers to manage that data. Both price and value are related to the supply/demand balance. So the fiat price goes up, the amount of safecoin paid to farmers reduces naturally. That is because the network continues to reduce safecoin to test we still have enough farmers, if not then Put prices increase and farmers get more safecoin. It is a balance and I recon a very decent balance to take forward.
I am ignoring Get verses Put costs etc. and just focusing on Put as that is where we start the iterations from.
14 Likes
Opinions are like arseholes, everyone has one and most of them stink.
You put it very similarly to what I think. Good explanation.
1 Like
I will try to clarify my reasoning in a very logical and objective way. If any point below is wrong, please point out the error and help me:
a) Resources for the network are desired. In other words, many quality Vaults are desired.
b) In order not to depend on goodwill or donations only, there is an economic incentive for Vaults.
c) Users do not need to purchase SAFEcoins to make GET requests.
d) Users need to purchase SAFEcoins to make PUT requests.
e) Users need to purchase SAFEcoins to use these currencies as cash and enjoy a decentralized payment system within the SAFE Network.
f) Points d) and e) are the catalysts that drive SAFEcoin’s price up.
g) If the price of SAFEcoin goes up, Vaults rewards are more attractive.
h) If the rewards are more attractive, this encourages an increase in quality Vaults.
i) If the number of GET requests increases in the network, at some point more quality Vaults will be needed.
j) If there are few GET requests and many quality Vaults, the cost to make PUT requests is lower, which encourages more PUTs.
k) If there are more PUTs, this stimulates more Vaults, as a logical deduction of points d), e), f), g) and h).
l) From j) and k), it is concluded that starting the system with many quality Vaults can represent a cycle of positive feedback and prosperity in the short term.
m) The number of GETs increases if there are more users interested in files that are already on the network.
n) The increase in the number of PUTs is likely to result in a subsequent increase in the number of GETs. But there is no way to be sure about that. It will depend on the quality of the data and the interest of the population.
o) If the increase in PUTs does not result in an increase in GETs, at first this is positive for Vaults, as rewards increase while GET requests do not increase in the same proportion.
p) It is possible that the number of GETs will increase not as a direct consequence of an increase in the number of PUTs. This scenario can occur if old files become increasingly popular, growing at a rate greater than the rate of increase in the number of PUTs.
q) If point p) occurs, at first this is negative for Vaults, as GET requests increase at a higher rate than the rate of increase in PUTs.
r) If point q) occurs, there is a solution that can counterbalance: it is possible that the incentive to Vaults will not be reduced if point e) is growing.
s) As a consequence of p), q) and r), it is concluded that the greatest risk of the system is the scenario in which the number of GET requests is increasing without a proportional increase in the number of PUTs or the use of SAFEcoins as cash.
If scenario s) occurs, reducing the rewards is likely to reduce the number of quality Vaults, which would make the PUTs even more expensive, generating even less revenue for the rewards. While this ripple effect occurs, the loss of performance in the system would probably reduce the use of the network. In this case, it is possible that, at some point, the number of GETs will be reduced to the point where it becomes profitable again for new Vaults to join the system.
I drew a hypothetical scenario below considering that the system started positively in accordance with point “l)” and then suffered the effect “s)”:
Although the “s)” effect is hypothetical and speculative, it does not seem unlikely to occur. I will mention below some situations that can stimulate the “s)” effect:
- Each PUT entered in the system provides a punctual remuneration, but may incur a recurring cost.
- The cumulative effect of recurring costs can accelerate the cost of PUTs very quickly.
- It is important for the system to serve as a repository of unpopular files, as they do not generate GET costs and generate PUT revenue. However, these types of data can be discouraged since the costs are all prorated (the storage price may start to become attractive only to potential GET generators).
My doubt is what approach could be taken to counteract the effect s), if it occurs.
2 Likes
The sum of single points does not an argument make.
Use of double negatives is as likely to confuse, as are assumptions about what exists, being a good baseline… Youtube advertising, might not be sustainable and could well just be an illusion.
This thread seems more about speculative economics to confirm an assumption, than Features and Advantages… and less interesting for that.
One truth in economics, is that people will draw curves and put arrows on them!
1 Like
This may be a way in which the requester of the data pays for the GET. That is to say the SAFE site or dApp charges the data requester/user. However, there would still need to be w mechanism via which the SAFE site or dApp pays farmers for servicing the request. This would still be tantamount to paying a fee for GETs.
Exactly. There is no way to be sure about that, and it is largely wishful thinking to assume that more GETs will directly and unilaterally translate into more PUTs especially in a way that does not overtax the farmer. It’s kind of like revenue vs. operating cash flow. Many startups go out of business not because there wasn’t an opportunity to make revenue but because there wasn’t enough access to cash in real time to cover operating costs. This is part of why the subscription pricing model is so popular; it functionally tames the cash flow problem.
Precisely. Then one also has to introduce the further confounding factor of the ratio of private to public data.
I think this further underscores the merit of considering paying for PUT as revenue vs paying for GET as operating cash flow.
Agreed.
I think you raise many considerable thoughts. I appreciate your measured approach to doing so.
I have found that when (new, in particular) people raise alternative points of view, the “antibodies” tend to come out as existing community members defensively shoot down the “intruding” perspective. Much as would be with a heart transplant, the host is liable to reject that which would save it. Some have found this off putting and left the community over this, but please don’t let this dissuade you. We need more diversity of thought—especially of the pragmatic strain.
9 Likes
Humbug,
OP opened with “objectively” and doesn’t mind being challenged.
Keeping it simple, is the route to resolving truth.
Playing politiks is the opposite.
1 Like
I am not clear where all the Get price verses Put price is coming from in these threads. There has never been a Get price and to start we are paying for storing or mutating data, no more than that.
This Get verses Put just adds a dimension of confusion. To understand the “economic model” it’s best to simplify the preposition. So if we stick with pay for store/mutate then the salient points might be visible, right now they are hidden to me at least.
8 Likes
I thought it was just me. I’m trying to follow along and I think he’s defining a GET as a “cost” to the network itself? I have been pretty confused by it all but I’ve also been running on low sleep lately being so busy.
4 Likes
If a GET request to download a file does not represent a cost to the network, in fact I got it all wrong. Is it the case? I would like to understand that better.