What I propose is the farmer get paid for both too. But if you are true that it will always be abundance of free space available. My system will never delete used free space data for the paid data. But my system garantee a security against an attack unlike yours. Because flooding a system with your proposal is much more worthy than mine. Your free space with that attack will always be wastefull.
Only the free data storage can be attacked in my model. And the system will continue to run even when attackers use up all the free data. And the incentive for such attacks is small since the attackers would achieve very little except wasting their own bandwidth. A DDOS random GET attack (bypassing the caching) would be much more harmful.
I have a pretty good image of how the SAFE work in my head. Bypassing is not possible as I see it.
But still, your free space have an issue and SAFE is build to be resiliant and it’s not the case with your free space. I myself will be discouraged to use the safe network with that feature on it because I know I can get stuck with garbage in my disk.
I read almost all your post about that and I end up with, for myself, a better idea that can work. If you don’t want it it’s fine with me. But I don’t want your’s.
All the data chunks are encrypted so there is no way for a farmer to know what is being stored. And as I wrote, the incentive for an attacker to store lots of garbage files is small. And the farmer reward remains unaffected.
The drawback with my model as I see it is that a user can store files for free and then suddenly it can happen it’s no longer free to store files!
What you don’t understand is attackers don’t need incentive to attack. They do it because they like it. I will myself attack the SAFE Network if I find an issue. And if SAFE is implemented with your system, I will do the attack. Why? Because I want SAFE to be truly SAFE (that’s an incentive). I don’t like attacking for no reason. But much others do.
1 Like
Even if that is the case, the worst that can happen with my model is that the networks starts functioning as it does now: all data storage costs safecoins.
To which criticism exactly are you referring?
This is actually a pretty serious drawback. Because it’s inconsistent behaviour. It’s confusing for end users.
The problem will be eliminated if when the network has become large the data storage will always be free because of overabundance of storage in the network. Even with the current always-cost storage model the network needs to have more data storage space available than is used. Otherwise the storage function would come to a halt.
And in the beginning, when the network is small, the free storage limit is a safety against attacks. And also, the free data storage allows the network to be bootstrapped by farmers being able to start earning safecoins already from the start (when no users have safecoins to spend).
And the extra inflation of safecoins can be managed by a reward formula similar to bitcoin.
Hah! My approach doesn’t suffer from that weakness.
There can never be an overabundance of such resources. If there ever was, SAFE (the network and the coin) would become completely worthless (equal to the price of air).
1 Like
Reminds of the “law” from the 70s/80s that still holds true today
Its a corollary of Parkinson’s law Parkinson's law - Wikipedia relating to computers
Data expands to fill the space available for storage.
And by definition if data is free then available storage is the whole network.
BTW Anders, if data costs to PUT then available storage is what one considers worth it.
3 Likes
In a way there will always be overabundance of storage resources in the network, even with the current model. How much storage do people use in their computers? Less than 100%. Similarly less than 100% of the available data storage on SAFE will be used. Otherwise if the storage hits 100% the network comes to a halt. Not good.
Well, everyone uses less than 100%. You can’t use 100%.
But regardless of how much is unused - 6.1% of 38% - there’s a very simple reason for that: they are the only user who can store data on the system. Strained capacity.
I already busted this argument of yours above when I mentioned the fast spammer and the slow ant. You’re using circular logic. The network isn’t full because it isn’t full.
The only reasonable way of preventing that it ever becomes full is to start charging and increase the price as it gets close to full utilization. You could also limit the network bandwidth to 1 byte per second, or you could add more capacity, but neither of these ways are good if there is no price signalling, which you seem unable to understand.
1 Like
It’s a valid model I guess to relate the value of safecoin to the storage used, but it’s a too lame model imo. Safecoin can gain value because of so many other things. Imagine if safecoin will become popular as a currency to use for ordinary shopping, both online and in stores etc. Then the demand for safecoins skyrockets and the market cap goes to the moon, well, has the potential to increase at least, beyond the value corresponding to merely the SAFE data storage space. And if the SAFE network starts to grow rapidly there will be a huge interest for safecoin on exchange markets that is largely unrelated to the value of the physical storage space in the network.
I missed that comment and can’t find it now. You mean that with the current model the storage space in the SAFE network can never be filled up? How does that work in a situation when the farming capacity is constant (or even gets smaller over time). I can’t see how you can get such situation to compute.
If the economics is done correctly (and by that I mean farmers can earn more in SAFE than it costs them to finance & operate vaults), it requires no effort to keep adding capacity.
If I spend $250 to buy an 8TB HDD for SAFE, let’s say I want to make $350 over 3 years, which is 4.3 cents per GB.
But, SAFE pays only for GETs, so at this time it’s impossible to use cost per GB - I should know how many hits I get and how much they pay. I don’t know either of those, but I attempted to make some estimates in other posts before.
I won’t repeat the process here, but let’s use some random numbers: let’s say I get 10 GB of GETs a month, that each GET is 1 MB and that’s 120,000 GET’s a year. I would have to make at least $117 ($350/3 years) from that, and that is $0.00083 per GET.
If I can’t get that much (at least $350/3 per year), I ain’t going to buy another disk when this one gets full (and I may even stop SAFE-farming).
Since everything will fluctuate, it’s going to be messy, but in end if it costs you more to be a nicer SAFE farmer, you won’t play. Maybe you’ll share some data for free, but you won’t give a shit whether your vault is having problems, you’ll set the system to suspend or power down when you’re not using it, etc.
Those calculations are over my head. Well, I guess I could try to learn it but I’m too lazy for that at the moment, ha ha. The first priority is to ensure that the network remains working. That is also dependent on how much the network will be used.
I was thinking of making safecoin as attractive as possible to farm, trade and use. Free SAFE data storage will make the network grow faster I have assumed. And the inflation of safecoins as a consequence of free data storage can be used as an advantage! Because the farmer reward can be regulated over time to optimize safecoin as a competitive cryptocurrency.
Of course, storage is not the only reason why Safecoin could become a relevant currency, it is certainly the main selling point. Safecoin may have advantages over Bitcoin, but other cryptocurrencies have those as well. In your (all free) vision there is absolutely no need to use Safecoin for storage, storage is only offered to farm Safecoins. It´s a side effect, same as calculating senseless problems (or even better: senseful problems as with Primecoin) is a side effect. In the part-free vision it is at least questionably relevant for SAFE. However, there are several problems with the analogy between SAFE and Bitcoin (and others):
-
The first and most important problem is that there is no proof that Bitcoin works once the market cap is reached. Many doubts have been raised that the transaction fee alone is able to maintain a deflationary system. So to say: Bitcoin works and so will Safecoin is misleading because we do not know yet what will happen when distribution reached the cap. We do not even know what happens tomorrow with Bitcoin, not only for systemic reasons but also for feedback challenges coming from society and economy.
-
With Bitcoin there is no way to exploit or influence the sideeffect and that´s a mechanism of the system. So while you are absolutely right that a coin could work solely based on a competition of chunk-delivery, this doesn´t say anything about the useability of the storage network (which I thought we all are interested in).
-
Bitcoin rewards are given regularly once the problem is solved. Difficulty depends on competition/power of the system. This is important because the difficulty algorithm regulates system activity and cost of ressource: as long as the reward is more valuable than the cost for mining hardware people buy and provide miners. The calculation of difficulty based on competition between ressources always leads to an equilibrium between the real-life costs poured into the system and the reward. If the difficulty of SAFE would depend on storage/bandwidth it would give a comparable incentive to provide more and more storage/bandwidth (sidenote: this would cause a problem under which Bitcoin is already suffering: centralization due to the large advantage of large farms and data center with huge bandwidth over small home users). However, with SAFE the payout takes place depending on GET requests and these do not depend on algorithms but on user interaction. That means that the farmers have an interest that there is no spam because it massively lowers the rewards. The more spam, the higher the difficulty. While with Bitcoin only farmers can influence the difficulty, in a free storage model everyone could - unintentionally or intentionally. So if anyone would like to harm the system he*she would just have to pour garbage into the network. Cost goes up, farmers leave the network.
Every free space provided by the system may become exploited to manipulate system latency, no matter if it´s 100%, 40% or 1%. And what for? Free ressources and abundance don´t exist on a systemic level - they only exist in relation to a restricted body. It´s a distribution issue, not a general phenomenon.
1 Like
All other cryptocurrencies have proof of work or proof of stake algorithms as far as I know. Both of those have disadvantages that safecoin solves. So I think safecoin can become more successful than any cryptocurrency in the past, including bitcoin. Such success requires that the safecoin reward/inflation/deflation model is done correctly.
That’s a good point. I think that as in bitcoin to replace the miner reward with transaction fees is a horrible idea. Safecoin can have a much better solution where the safecoin supply is inflated all the time, but at a rate that is optimized for first-mover advantage, volatility and inflation/deflation stabilization. In this way safecoin can remain having zero transaction fees which will keep it competitive indeed and make the currency as frictionless as possible.
My idea is that when the SAFE network has become large, then all data storage will be free, because farming for safecoins will be so popular that much more resources are put into the network than users (including spammers) will be able to use. And the free storage limit will still be there, but will unlikely be needed when the network is large.
And think of the spam incentive: Why would any spammer want to waste bandwidth and put a lot of junk on the network that nobody will ever see or use? And trying to attack the network will be futile even in the beginning because of the free storage limit.