MaidSafe as an alternative to Dropbox/cloud storage?

Beginners
41

I don’t know if #4 is possible: you would have to be able to change your password/PIN and from the explanations of how self encryption works that doesn’t seem to be possible. So that needs checking. Also, if it isn’t possible, that makes password security vital, so in practice I hope you are able to change password/PIN.

#5 is possible but involves a lot of work (investment - in development and storage), and produces an inferior product (storage with latency due to layering), which means that investment is harder to recoup - it has to be done on price, yet it can’t be cashed in quickly because it will take time to sell it all, which in turn adds marketing costs. This is not an honest business model, but an attempt to cache in on a freak event (storage cost spike) that is against the proven and most likely tend, and might simply not happen. I think this is a very poor scam, so I can’t imagine it being attempted by anyone with significant resources. Surely speculating in the price of Safecoin is far more attractive based on #3.

I think the low chance of #3 makes this a very risky and therefore unattractive proposition. I think #5 is impractical even if 3# happens. Even if #4 is possible, I still think people with these skills and resources can make money in much less risky, less investment intensive ways.

I judge that attacks of this sort are likely to be limited because of this, and so not a real concern for the network. The risk is therefore not serious: the risk is that the price of storage might be ever so slightly inflated because a tiny proportion of the network storage was subject to speculative hoarding.

1 Like
42

I agree. Those profit opportunities are unlikely to happen for the reasons you mentioned.

However, I still prefer a (Consumption Model = Pay on PUT) compared to a (Subscription Model = Pay for Blocks).

IMO, it makes a significant difference in consumer savings and real-time Network accounting. I’ve already made arguments for consideration. And I’ll concede subscription is more familiar and maybe easier to code.

Please consider this… Would SAFE storage be used efficiently if people are able to reserve space they don’t use? Or would it be better for them to pay for what they need in real time?

Apologies for beating a dead horse.

I think it’s important to set correct expectations. If subscription is as some believe… buying blocks reserve a set amount of space. Then the Network MUST account for that or suffer the possibility of systemic failure. Subscription works with static storage because the seller (Dropbox) can guarantee availability at any future point in time.

SAFE storage is especially dynamic and uses real time availability. IMO it is a bad idea to guarantee storage availability in the future when the Network cannot know what will be available later on. If storage is not guaranteed, as I believe, then we shouldn’t even give the impression of a (GB) amount. Instead, just let the consumer fill up their PUT Reserve (Safecoins), and use that portion of Safecoin like fuel.

I’m sure the Devs have considered this already. So I’ll leave it there.

2 Likes
43

@dyamanaka You make some good points that I think make pay per PUT worth considering.

I agree with your point in favour of pay per PUT, and have two concerns against which this needs to be balanced:

  1. The relative computational & storage costs of one model versus the other, or some halfway house. My guess is that doing a Safecoin transaction per PUT us a lot more costly, but I don’t know.

  2. What users want! I think we need to sit in the user’s chair and see how the different user experiences compare. My feeling is that people like to know up front what they are paying and what they are getting for it. Buying blocks gives a clear (familiar) way of understanding a purchase scenario, which makes the choice easier to weigh up and act on.

I can see ways to make pay on PUT more transparent (e.g showing the current price, rate of depletion etc, somewhere easy to check) but think they compare poorly in UX with: being prompted to top up, deciding how much to buy for now, making the transaction, and then being able to forget it until next time.

Many people are very familiar with pay to top up due to pay as you go mobile phones, and other pay in advance methods for energy use, travel cards, filling with gas/petrol and do on, so I think the model has a lot going for it from third point of view.

There are pay as you use models too of course, but I’m not aware of ones where the price is subject to continual adjustment, and occasional spikes. If we can find any, we should look at how the concerns I’ve raised affect them.

It’s a pain to be watching the cost and worrying that I might suddenly be paying a lot for storage and using it up without realising. Here a price spike, even a very short lived one, could potentially hurt a lot of users all at once and get SAFE a bad reputation. This is a serious risk that I think we’d need to eliminate.

1 Like
44

Agreed. :smile:

I would go with “best” user experience and Network sustainability overall, where ever that leads.

  • Computational cost for granular accounting is also my concern. I mentioned previously… when we switch to decimal, this would not be a problem. Until TestNet3, we won’t know definitively. Would be nice to test a “copper” version of Safecoin (1/1000) and see if charging per MB bogs down the Network.

I was thinking about making the copper Safecoin version a “decimal type” from the start. It’s already so tiny in fiat value and not really worth stealing. Could be a clever way to test decimal security along side the more valuable Safecoin. In fact, when a user allocates Safecoin to their PUT reserve, they convert it to decimal. That’s brilliant!

  • Yes, users ultimately decide what they want. I think people are already familiar with a consumption model. All we have to do is explain it the same way they consume gas for their cars. Or how people pay for minutes used on their phones. I think people can relate to this very easily, even non-techs. And the market for SAFE will be much more than just “cloud storage” consumers. A Different Perspective made it pretty clear the next generation want a Social Network.

I like your suggestions for transparency in UX and it would help a lot!

As for your last point, price spikes could hurt people, the same way gas prices hurt the world globally. At the same time, it also calls new farmers to “action” and helps stabilize the Network. So it works both ways.

1 Like
45

As for your last point, price spikes could hurt people, the same way gas prices hurt the world globally. At the same time, it also calls new farmers to “action” and helps stabilize the Network. So it works both ways.

Just to clarify, I’m not against price spikes, but against constructing a system that would hurt users before they realised what was happening, and a lot of users all at the same time.

1 Like
46

You can still “top up” with the “pay on PUT” model. However, instead of topping up your storage space, you’d just top up your SafeCoin.

This is kind of like how DigitalOcean and other services do it. You can do with your battle.net or Steam account, your Xbox account, and many others. You deposit money into the account and then any monthly fees are debited from there until you run out. Optionally, you can have them charge to your credit card once you run out or just have them send you an email saying you require more funds.

I don’t think this is a difficult concept to understand. Pre-buying space isn’t necessary.

Ideally, we’d also have safeguards put in place to restrict the amount of usage. You could put in a rule that says “don’t debit me more than $10/mo (or X SafeCoin/mo)”. Going beyond that would require your authorization. This way people don’t get slapped with a nasty bill if PUT prices sky-rocket for whatever reason.

As a user, I don’t think I mind the pay-on-PUT method, particularly if prices keep going down, then I know I’m always getting the best rate. With the safeguards in place, I know I won’t accidentally spend more than I intended.

Again, this is similar to DigitalOcean. They charge by the hour, but they won’t exceed a set monthly rate. This way if you cancel/shut down your hosting early, you pay less, but you also know what to expect. Similarly, with SAFE you wouldn’t be paying for storage you aren’t using.

Moreover, the “buying storage” concept might actually be more confusing to users. Because you’re not actually buying storage. With traditional storage services (like Dropbox), you get that storage space back when you delete a file. If I’m not mistaken, with SAFE, you’d have pay for that same space again. So we really shouldn’t conflate the two models, because they aren’t the same.

3 Likes
47

What about only allowing pre-buying space up to the worth of one SafeCoin? If a SafeCoin gets you 2 GB of space, you can’t buy more space until your available space drops below 2 GB. It doesn’t have the inefficiency of pay-per-put, where you are paying per MB, but at the same time it stops excessive hoarding.

2 Likes
48

2GB doesn’t even allow me to store a large movie or a medium sized truecrypt vault.

If you are going to force a maximum storage I’m allowed to buy, it must be greater than any file I would reasonably want to save. I would think it would be unreasonable to have 10GB files, so perhaps 10GB would be a good limit. We could discuss 5GB. But 2GB seems tight.

1 Like
49

[quote=“Tom_Carlson, post:48, topic:3229, full:true”]
2GB doesn’t even allow me to store a large movie or a medium sized truecrypt vault.[/quote]

In that case it would be pay-on-put for however much you may need for the current put.

1 Like
50

That’s close to what I was suggesting…

Example using a PUT Reserve with (Integer Safecoin)

Date: 3/6/2015
Wallet Balance: 11,000
PUT Reserve: 0
Storage Credit: 0 MB
Storage Rate: 1 Safecoin = 1 GB

I allocate/activate 10 Safecoins to my PUT Reserve…

Date: 3/6/2015
Wallet Balance: 10,990
PUT Reserve: 10
Storage Credit: 0 MB
Storage Rate: 1 Safecoin = 1 GB

I PUT a 100MB file, and my balance will update to this…

Date: 3/6/2015
Wallet Balance: 10,990
PUT Reserve: 9
Storage Credit: 900 MB
Storage Rate: 1 Safecoin = 1 GB

A month later, the storage rate changes…

Date: 4/6/2015
Wallet Balance: 10,990
PUT Reserve: 9
Storage Credit: 900 MB
Storage Rate: 1 Safecoin = 1.5 GB

I decide to PUT a 2GB file…

Date: 4/6/2015
Wallet Balance: 10,990
PUT Reserve: 8
Storage Credit: 400 MB
Storage Rate: 1 Safecoin = 1.5 GB

This is how I got 400 MB storage credit: (2GB - (900MB + 1.5GB)) = 400MB

The user only allocates Safecoin to their PUT Reserve… that’s it!

The Network deducts only what is needed at the time a PUT is committed. If the costs keep going down, this will translate to savings for everyone. If costs go up… better get out your old hard drives and start farming.

If the user does not have enough in reserve to cover the PUT cost, they will get a message letting them know to add more Safecoin. Most people will allocate 1000 Safecoins and do something else more fun.

Things to consider…

  • Is the “storage credit” guaranteed? It can be. Network capacity deviations have less impact than allowing users to reserve large amounts of storage. Basically, the consumer cannot claim more storage credit than what 1 Safecoin bought at the time.
  • Storage credit should always be used first before deducting from the PUT Reserve.
  • Total transactions are (1 per Safecoin) instead of (1 per MB)… less load for the Network.
  • This stuff looks complicated. The only thing the user needs to focus on is the PUT Reserve. Storage Credit and Storage Rate can run in the background, or revealed only in advance settings.
3 Likes
51

I like this model. It prevents you from spending more than you intend (only allocate what you want to your PUT reserve) and prevents wastage (if we can’t split a Safecoin, at least it turns into storage credit).

I think having decimal coins and no storage credit would be conceptually simpler, but this would do just fine.

1 Like
52

@dyamanaka, I simply don’t understand the need for all this complexity.

I conceive the simplicity this way:

When you need resource in order to PUT data to the network, whether for messaging, storage or whatever, you purchase a block of PUT resource from the network at whatever the network deems the price to be at that time. Then that amount is added to your PUT cap, denominated in kb, mb, gb, etc.

As you PUT data, whether for transient messaging or storage or whatever, your wallet/account subtracts that amount from your cap. This is done in consensus with your close nodes so that your account can’t game the data.

When you run out, you buy more, whether as a direct action or as an automatic setting in your wallet.

The purchase can be done in whatever denomination you find appropriate (depending on the current network charge) whether one safecoin, 100 safecoin, or 1/10th safecoin (when splitting is enabled).

Simple, relatively easy to code, compute and maintain–and, as importantly, think with.

Why not keep it this simple? From past input and likes from David Irvine, I believe he’s thinking along these lines.

EDIT: note first like.

1 Like
53

I’ve already expressed my concern in this statement below. The solution I proposed is meant to address that concern.

54

I really do hear this David and appreciate it as well, these debates/queries are very interesting and valuable. I should say though even if you paid with a and use it a year later the will have reduced in value a lot as well (50% every 7 years average I think). So it is all as complex as we make it out to be (you and me) as it just is.

All the investigations are good though for sure and while I see some safecoin links and think (later I will get to that later) I do tend to absorb them later at night in peace and my feeling is we are not going to far wrong with most of these ideas. If we are careful the parameters of SAFE “space” purchases can change without affecting safecoin the currency as it evolves. As long as we are careful.

1 Like
55

I looked over your post again and I think I see a bit better what you’re trying to do. Basically you’re just talking about placing a limit on the PUT cap.

That is, have a cap of no more than 10 gb (or whatever) and that can be recharged as it gets toward zero. That way you’re forcing a more up-to-date network assessment of the PUT cost, related to currently estimated resources available.

Am I restating the core your idea correctly?

1 Like
56

Yes.

The cap is actually 1 Safecoin. So whatever 1 Safecoin buys is the most (GB) a user can reserve/hoard. That amount is shown in the “storage credit” display. Additional benefits are exactly as you stated.

58

Are you sure about this? As far as I have understood the system even that is not a “reserved” amount of space, but decreases (or grows) over time.

59

I think the most important is to let the client now “at a glance” how much space he’ll get when he confirms the (safecoin) payment.

If I look in OneDrive now, I get 100 GB for 1.99$/ month, 200GB for 3.99$/month 9.99$ for Office365+1TB/month etc…
It shouldn’t be more complex than this for anyone when purchasing drive space(or whatever app credit) in SAFE network.

I would like to see in Safe netowrk something like: 100GB…1 safecoin (forever) , 200 GB…1.99 safecoin (forever), 1TB space+meesaging+mail … 9.99 safecoins etc.
Also, if I want to , 1 safecoin to change to its value is in dollars, euros (1.23$, 1E etc) to make more sense for someone not so techy(most people?) how much he’s paying in obsolete FIAT money :smile:

Extrapolating a bit, the provider of the service should include different option for paying for it:
1.Directly safecoin from Safecoin wallet
2. using bitcoins and having them changed to safecoins at market price
3. (maybe) using fiat through an automatic service which converts them to either bitcoins of safecoins

Although hard to do now,number 3 will be the killer payment implementation for mass adoption!

60

The point is that when you purchase PUT resource from the network, you really need to HAVE what you’ve purchased. It wouldn’t do to spend a safecoin to buy 1gb of PUTs, only to find that, due to fluctuating rates, some time later you only have .75 gb even though you’ve not used anything. If you purchase an amount from the network, it needs to actually be something that you have.

The purpose of @dyamanaka’s model is to force the individual to go back to the trough regularly instead of buying a huge commitment from the network which could be an outstanding liability if things fluctuate the wrong direction. I’m not sure that it’s necessary, but if it can be coded without too much difficulty and people can understand it easily, it wouldn’t hurt.

1 Like
62

In a Full Reserve Storage system, yes.

To me it reads more like the purpose is to make it irrelevant when/how frequently you buy (or hold before you spend).
As is, that Safecoin you buy is always an outstanding liability - regardless of how much network capacity fluctuates in either direction.

An easy way to imagine this is SAFE Network Day 1 Hypothetical Scenario that I described in this comment:

  • 4 300 000 000 MB of usable storage available on the network
  • 430 000 000 SAFE coins claiming that capacity
  • One SAFE buys you 100 KB ($0.11 / GB according to today’s price - btw, compare this with Amazon S3)

So there’s no question things will (at least initially) fluctuate “the wrong way”.

1 Like