If Safe-storage space is proportional to how much space you provide, how would uptime factor in? What if I provide 5 TB of space, but only for 2 hours of the day? What if I provide a ton of space, great uptime, and rack up a lot of Safe-storage, then just disappear from the network. Are my files still guaranteed to be there?
So by uptime I’m assuming you refer to bandwidth and the way you are rewarded is how you are ranked by the network. Your vault is ranked based on availability and bandwidth mostly I believe as to provide better reliability and performance. So having your computer go on and off would not be beneficial. Many will have dedicated pc’s on constantly with good bandwidth BUT I do believe the team have had these considerations as they want it to be beneficial for ALL users but at the same time have quality resources provided reliably. If you’re on a predictable schedule then you’ll prolly be ranked as such I assume just dependent on how long. If you disappear from the network and you paid for storage with resources then you’d be downranked for sure and at that stage I’m not sure what would happen to your data. If you had a good rank for a long time then your earnings would sustain you for quite a long while I would presume.
Think of it this way.
If you’re paying for Dropbox then instead purchase safecoins on the SAFEex crypto exchange that will be available at launch and pay for your storage. It will be at least comparable in price but most likely much cheaper. (1.Should be a one time fee & 2. Best security)
If you plan on being a part of the network and using it for messaging and like a new decentralized Internet then you already see the value in providing the resources to the network and by being a part of it get rewarded safe storage of all your files.
You do raise an excellent question that to be quite frank I’ve never considered because I just want to be an active contributor to this project, it seems like the next logical step and the best alternative. I never thought to just dump data and leave haha. This might be a good question for @dirvine or @nicklambert . In the meantime I’ll search the forum for an answer to that question
First thread I found
The safe-storage space is not proportional to space you provide.
The SAFEnet pay for each GET (chunk of data) given to the network and charges for each PUT (chunk of data) store in the network. Therefore, a computer on a few hours will get little benefit. Moreover, its rank will be very low so, that in spite of the offered 5TB, the amount of stored data will be very low too.
About your files…
How do you ensure that no pieces of data are lost after a user goes offline?
The SAFE Network automatically maintains a minimum of four live
copies of any piece of data at any time. As a users turns off their
computer, their vault managers (the group responsible for looking after
network nodes) notify the network and all the data chunks being held by
that data manager are recreated elsewhere. This process happens very
quickly, in around 20 milliseconds.
Is “farming” going to get me SafeCoins which I can then in turn use to “buy” Safe-storage?
Yes, this is likely the model, for the reasons Nick mentioned.
Access to public data will always be free, only storing/uploading is charged. There is no ongoing fee, just a one off upload cost, so your data won’t be deleted if you pay to upload and leave it. No need to continue to farm etc.
Payment for storage will happen in blocks (e.g pay for 10GB capacity). Giving you a storage credit that is yours to use up all at once, or over time, as you wish. When you’ve used it up you’ll need to buy some more.
These are all things to be trialled in testnet 3, but this is the model that seems to answer the technical issues
It’s really helpful to us to hear your questions. One of the things we need to do is improve the usability/user experience of this model. How easy is it to understand? How does it compare with other models (e.g Dropbox subscription)? Do people trust that it’s pay once, store forever?!
I think simply the fact that it is a different model creates a barrier to understanding and adoption that we will need to overcome by making it easy to understand and to trust, when people who already have expectations about how cloud storage works encounter SAFE Network. Most of whom won’t ask these question, they’ll make their decision based only on what is presented.
I’m sure @N1ckLambert and the team have thought of this, and when the technical options are clearer (testnet 3) it will become more of a focus.
I read David’s quote as a prevention from storage space hoarding and your quote implies to me that you can buy and use it whenever you want to.
I don’t see the difference.
Maybe the confusion is with “pay on PUT”? Perhaps clearer as: you must pay in order to PUT.
The way you do this is by paying to credit your account with an amount of storage that is then used up by uploading, storing messages etc (each PUT will shave a bit off what is left). Once exhausted, you need to “top up” again.
This avoids the problem with the cost of a single PUT being less than a Safecoin, and similar issues.
Well, that’s my understanding!
Ah I now see why I deleted my post first . I was confused by the ‘storage credit’ what actually is safecoin. I thought you meant storage space with storage credit .
But that’s exactly what I mean by “storage credit”!
You decide how much storage you want and buy it in a batch, in units of let’s say 10GB. When you do that, you pay in Safecoin at the going rate. That is the point when your wallet gets debited.
As you store data your Safecoin wallet is not touched. If it was per PUT, how would you charge the wallet for 0.001 Safecoin (or whatever it works out)?
This does incentivize hoarding and I thought that’s something we don’t want.
You say you buy it in a batch, but according to David the storage you buy and do not use will decrease (to prevent hoarding). Buying it in a batch and store later on incentives people to buy up all available storage space and sell the account with X GB storage space for example.
I’m probably misunderstanding something here so enlighten me
I can see problems for people trying to hoard in the way you describe - it will push up the price of Safecoin for example, making storage expensive to “pre-own”, but best for David to clarify first.
Yes this is something we are looking at now and discussing how best to explain these novel concepts concisely to non technical users. I think trust is something that we and the network will need to earn As well as explaining concepts to users, giving them access to users support is also a key consideration here.
I think you will only hoard storage promise of space, so that is OK. The safecoin will be consumed by the network and farmed out.
So putting more into storage means you just gave to farmers more who will hopefully distribute it (well it’s distributed to them). When you need to use that space you hoarded (expensively) the network will ensure you get it even if it has to increase famers rewards to do so.
Hope that makes sense I know I am very bad at explaining this. It will be much clearer when we implement it as my head won’t be in routing algorithms
I’ll try to clear the confusion with examples, assuming I understood “The Safe Network’s economics” thread correctly.
- Currently, the smallest transferable unit is 1 Safecoin… no decimals.
- We plan to switch to decimal Safecoin after the Network grows bigger… maybe in 1 year?
This means storage pricing must be expressed in the following way.
1 Safecoin = X (GB) Available
Here’s an example of fluctuating pricing, based on Network Utilization (supply/demand).
(March 5 2015)
1 Safecoin = 1 GB Available
Paying 1 Safecoin allows you to store up to 1 GB… at that moment.
(December 25 2015)
1 Safecoin = 500 MB Available
Paying 1 Safecoin, allows you to store up to 500 MB… at that moment.
Here’s where the confusion comes from.
If a consumer pays 1 Safecoin on (March 5 2015) but never actually stores (PUTS) anything, their available space is reduced down to 500 MB when they try to store on (December 25 2015).
This is what I referred to as the “use it or lose it” approach. This discourages hoarding and encourages consumers to buy only as much as they need… at that moment.
There are some points to consider.
- If the user sees 1 GB “available” today then 500 MB later, will they feel ripped off?
- Is there a way to manage this negative feedback?
- Is there a better way to do storage pricing?
1A. Probably. Most people will “assume” if they pay 1 Safecoin, the displayed storage amount is what they will get.
2A. Yes. Since the smallest unit is 1 Safecoin, that should be the default block amount purchasable at the moment a user tries to PUT data. In other words, they cannot buy more than what they upload on a “Pay-on-Put” basis.
Example. (Pay On Put)
Consumer first tries to PUT 250 MB of data. The Network gives them a message (Accept / Decline), charging them 1 Safecoin, which gives them 1 GB… at that moment. They can either use the remaining 750 MB or leave it as credit. But they do run the risk of their 750 MB reducing if prices go up.
The Network only charges for data when a consumer tries to upload (PUT). The consumer cannot arbitrarily spend 100 Safecoins to buy 100GB and upload at a later time.
This process makes it very simple for the consumer. Try to upload, get a message of how much it costs. Or better yet, have a real-time display of how much 1 Safecoin = X (GB) in the SAFE App Launcher.
3A. Yes. Once we switch to decimal Safecoin, the PUT charging can be made more granular (cost per 1 MB), instead of charging for batch amounts (1 Safecoin per X GB).
No it seems nice, I like all the differing ideas. I suspect we may need to test this all out. You spend ages on this and are very convincing. We will draw it up in the new language to see how it looks.
Is this currently how it works? If so, which part of the network is responsible to keep the balance of credits?
That specific portion you quoted is a solution I came up with just now. Credit to @Seneca as he was part of the discussion.
The discussions on this forum indicate that a consumer can buy storage ahead of time via blocks. Example spend 100 Safecoins to buy “promise to store X space” at a later time.
I was suggesting we remove that ability because it causes confusion as well as problems. None of this is set in stone. We need to test on TestNet3 before we can get a clear view of how it “should” work.
Ah ok. I wasn’t aware of the discussion around the idea of buying blocks of storage in advance. I understand it might makes it easier to grasp for user since buying blocks of storage is a familiar concept but it doesn’t really translate well to the way Safe works, I think anyway.
What’s the reason why we don’t want to split Safecoins at launch, computational overhead?
From what I understand, it was a security feature. So if an attack were possible they would only be able to steal 1 Safecoin, instead of the entire wallet balance.
1 Safecoin = 1 one vault location
10 Safecoins = 10 vault locations
Sorry if I can’t explain it in technical terms, but that’s how I understood the reasoning.
EDIT: I just realized I was answering a different question. The computational overhead concern will be resolved when we switch to decimal Safecoin. For now, it “might” be an issue if there are Massive amounts of Safecoins being transferred. But some believe it may not be an issue as the Network grows because it will have more nodes to distribute the load. Again, testnet3 will help us get a better view.
Hum, interesting, got any link on the subject?
EDIT: Actually I found a few searching for “granularity”, I’ll read that.