Safecoin divisibility

Development
61

I’m not sure you need to mint new Safecoin, they get signed over to the network when split, and signed to a new owner when pennies are redeemed.

Pennies can be minted determinustically - they are SD.

  1. every element of information involves (at least) one group; so converting 1 safecoin to a 100 pennies, requires minimally 101 (or a multiple of that if it’s a multi step process) of groups; so we’re talking 3200 nodes involved;

This is the same for any SD, and so one way to handle this is to charge for pennies on the basis of the SD allocated for them. So, when asking the network for 100 pennies, it costs both the 1 Safecoin for the pennies equivalent, plus the current network charge for 100 SD.

  1. for the network maintaining a safecoin is maintaining one structured data element; maintaining one penny is maintaining one structured data of similar size. The overhead cost for pennies is 100 times bigger than for safecoins. This is not a question of overloading the network, but a question of inefficiency.

I think this would be accounted for by charging the user for the SD needed to create each penny they want.

I think it auto balances. People will only want pennies because Safecoin have become too valuable and they need to trade smaller units of value - which means pennies become more valuable and worth the overhead.

62

Simple Numbers

(SCPP = Safecoins Per Person)
(NUSN = Number of Users on the Safe Network)

NUSN = 10k SCPP = 430k
NUSN = 100k SCPP = 43k
NUSN = 1m SCPP = 4.3k
NUSN = 10m SCPP = 430
NUSN = 100m SCPP = 43
NUSN = 200m SCPP = 21.5
NUSN = 300m SCPP = 14.3
NUSN = 400m SCPP = 10.75
NUSN = 500m SCPP = 8.6

How many users does it take, before this becomes an issue?

How many users do we anticipate the network will have at launch?

How much Safecoin will a user hoard/save on average?

63

The network does not have the ability to sign an SD; nodes in the network have signed messages; and groups can form a consensus, but the decision of a group can only be validated at a brief time frame after the decision. Churn will cause the nodes of that group to change, so a SD can only be created at a single (or mutli-signed) Client/Node and a response needs to return to a single (set of) client / node. Group consensus only exists in flow; it’s not a persistent state for a signed Structured Data.

Groups can persistently keep state in the persona databases, but this information cannot be signed, so is not a StructuredData

3 Likes
64

More Simple Numbers

In a deflationary system people have a tendancy to hoard/save. With that in mind let us run some simple numbers.

(SNPP = Safecoins Not-Hoarded Per Person)
(SHPP = Safecoins Hoarded Per Person)
(NUSN = Number of Users on the Safe Network)

@ 10% average hoard/save

NUSN = 10k SCPP = 430k SHPP = 43k SNPP = 387k
NUSN = 100k SCPP = 43k SHPP = 4.3k SNPP = 38.7k
NUSN = 1m SCPP = 4.3k SHPP = 430 SNPP = 3.87k
NUSN = 10m SCPP = 430 SHPP = 43 SNPP = 387
NUSN = 100m SCPP = 43 SHPP = 4.3 SNPP = 38.7
NUSN = 200m SCPP = 21.5 SHPP = 2.15 SNPP = 19.35
NUSN = 300m SCPP = 14.3 SHPP = 1.43 SNPP = 12.87
NUSN = 400m SCPP = 10.75 SHPP = 1.075 SNPP = 9.675
NUSN = 500m SCPP = 8.6 SHPP = 0.86 SNPP = 8.514

@ 25% hoard/save

NUSN = 10k SCPP = 430k SHPP = 107.5k SNPP = 322.5k
NUSN = 100k SCPP = 43k SHPP = 10.75k SNPP = 32.5k
NUSN = 1m SCPP = 4.3k SHPP = 1.075k SNPP = 3.225k
NUSN = 10m SCPP = 430 SHPP = 107.5 SNPP = 322.5
NUSN = 100m SCPP = 43 SHPP = 10.75 SNPP = 32.25
NUSN = 200m SCPP = 21.5 SHPP = 5.375 SNPP = 16.125
NUSN = 300m SCPP = 14.3 SHPP = 3.583 SNPP = 10.75
NUSN = 400m SCPP = 10.75 SHPP = 2.6875 SNPP = 8.0625
NUSN = 500m SCPP = 8.6 SHPP = 2.15 SNPP = 6.45

@ 50% hoard/save

NUSN = 10k SCPP = 430k SHPP = 215k SNPP = 215k
NUSN = 100k SCPP = 43k SHPP = 21.5k SNPP = 21.5k
NUSN = 1m SCPP = 4.3k SHPP = 2.15k SNPP = 2.15k
NUSN = 10m SCPP = 430 SHPP = 215 SNPP = 215
NUSN = 100m SCPP = 43 SHPP = 21.5 SNPP = 21.5
NUSN = 200m SCPP = 21.5 SHPP = 10.75 SNPP = 10.75
NUSN = 300m SCPP = 14.3 SHPP = 7.16 SNPP = 7.16
NUSN = 400m SCPP = 10.75 SHPP = 5.375 SNPP = 5.375
NUSN = 500m SCPP = 8.6 SHPP = 4.3 SNPP = 4.3

65

So I guess that is maybe a point worth emphasising: StructuredData elements always have to be signed. This crypto-security nicely balances clients against network. For ImmutableData it was already clear that if the hash does not match the name, the network has failed. Structured Data fills that rule for mutable data; so we cannot break that rule: SD is owned by client(s) and the network cannot corrupt it as the signatures would break.

This is crucial, otherwise you as a client would have to trust the network - it might not be a single third party, and have fixed rules, but it still would introduce trust.

[EDIT] Having said that, having an SD owned or created directly by the network is massively attractive; and say if a client is assigned ownership by the network, then updating it as a client locks it in your control;

so such a locking mechanism where you allow the network to mutate an SD, or lock it under client control has a legion of future possibilities too

2 Likes
66

I think it will start to become a issue once the network breaks past one millions users.

67

Well here’s how I see it conceptually

We have the SafeCoin (SC) that buys certain amount of space (Allowance - A) and we have mechanisms to track the usage of A and its depletion (transfer of A–>Network) so what we need here is availability of A1–>A2 transfer.

The problem with this is that lowest denomination in this case would be 1MB I assume. But here’s where inflation actually comes to the rescue :slight_smile: as SC progressively buys more space so does the A supply grows with it and I would add that in this case inflation incentivizes spending in very friendly way - you will be getting same amount of space for a lifetime but you wont be getting same value in network services…so once you’ve converted your SC to A you better spend today rather a year from now.

1 Like
68

Just a heads up: The Minsky project is probably the tool to use for anyone with the DE skills required to model Safecoin. I believe there is some work towards modelling BTC with this award winning tool as well.

Usability is very important to the success of a currency

I think @Traktion has the right idea that whatever solution is chosen for divisibility that it be simple. I would go further and say that is is absolutely vital, in that If the average person cannot wrap their head around how the currency works regarding divisibility and if the currency does not follow normal divisibility rules then it could be argued that it will fail one of the three functions of money: as a Unit Of Account. If the desire is for Safecoin to serve humans as a Unit of Account and become more than just a hidden token used internally by the network for allocating resources, then the divisibility of a Safecoin will almost certainly need to be easily understandable by non-technical users. There is lots of good discussion around regarding crypto currencies as a unit of account and the importance of usability (such this and this to pick a couple. “unless consumers can get their head around value, adoption will be extremely limited”).

If a signed StructuredData (currently called a “Safecoin”) is indivisible without a complicated work around then isn’t it the natural “lowest money of account”, like the “Satoshi” is to a Bitcoin or the “Mill” is to the $USD? Say it was renamed (I’ll pick “Irvineoshi” for fun and giggles) then what we currently understand as a “SafeCoin” could be 100, 1000 or some other intuitive multiple of Irvineoshi’s. This is what I understood @Traktion to be suggesting above. It would follows that 32 bits ~4 billion Irvineoshi’s may not be enough, and the 1:1 MaidSafecoin ↔ One StructuredData “Irvineoshi” swap would have to be revised maybe even floated to avoid a crowdsale investor revolt. However the reward to simplifying divisibility would be give “SafeCoin” the chance to be adopted by normal people for wider use as a currency and not just serve as an internal resource allocation token.

Standing by to hear all the technical reasons why this is a silly idea :-).

7 Likes
69

This view point may seem strange. I wasn’t going to post it because many might not understand or agree in this fiat generation. But oh well… I’ll put it out there as a seed idea.

I made a post called Rabbits VS Turtles over a year ago. Part 1 of the post talks about two kinds of currency: Rabbits, Turtles.

If (Safecoin = Rabbit) then (PUT Balance = Turtle)

PUT Balance
It is often referred to as: balance, allowance, ledger, account. Technically, your PUT balance is really a promissory note, meaning the SAFE Network promises to give you X GB of SAFE storage. That note is consumed once you upload/PUT data onto the Network.

PUT Balance = Promissory Note
SAFE Storage = Physical Commodity Consumed

From my point of view, the infrastructure for divisibility is already in place. If people can transfer specific amounts (1GB, 1MB) from their PUT Balance, then we have quasi-divisibility and a turtle currency.

Here’s the theory
Safecoin = Ideal Store of wealth (maybe deflationary)
PUT Balance = Ideal Spending currency (probably inflationary)

  • Safecoin can only be converted one-way into PUT Balance (promissory notes).
  • PUT Balance can be exchanged for stable Fiat Value.

Instead of sending Safecoin, you send (SAFE TB), (SAFE GB), (SAFE MB). Let’s look at that in relation to the rest of the world. I’m going to pull some numbers from this post.

Sending 1 (SAFE MB) can be understood as $0.00006 in fiat terms.
Sending 1 (SAFE GB) can be understood as $0.06 in fiat terms.
Sending 1 (SAFE TB) can be understood as $60 in fiat terms.

This value above is derived from average physical hard drives (HDD) on the current market. It should be higher when adding “cloud storage” utility. But I cannot calculate that without a live Network.

This is already a very small denomination and is likely to get smaller as the market increase storage capacity. Consider that the current market world wide already understand what GB and MB is, and can value it relatively close to their own fiat currency.

Regardless of Safecoin speculation, (SAFE GB) remains relatively stable in fiat terms… This is important because many businesses need to convert back into their local currency to pay their bills.

What’s the catch?
This only works if (SAFE GB) can be traded on exchanges for other currencies.

What about Safecoin investors?
They will cheer for this idea IF Safecoin buys more (SAFE GB) in the future. They will hunt me down if it goes the other way. I’ll have to hide in a cave if that happens.

EDIT: Based on discussions below regarding different storage tiers (TB, GB, MB, KB), it’s easier to group label them as XXX (SAFE GB) which is most common in this generation. It will change when (SAFE TB) becomes the new medium.

PUT balance example…
1TB is expressed as “1000 (SAFE GB)”
1GB is expressed as “1 (SAFE GB)”
1MB is expressed as “0.001 (SAFE GB)”

Fiat conversion example…
Multiply (SAFE GB) to the average (GB) fiat price.
1TB = 1000 x $0.06 = $60
1GB = 1 x $0.06 = $0.06
1MB = 0.001 x $0.06 = $0.00006

Please note the above examples are “baseline values” only. SAFE storage will have added value due to SAFE Network utility from APPS and Content. This gives us a good place to start.

8 Likes
70

@dyamanaka,

Very clever! I will be interested to see if others can find any fundamental problems with this idea.

P.

2 Likes
71

Hmm what kind of scenario would that be in your opinion? I cannot imagine safecoin buying today less space a year from now. :slight_smile:

edit:
So if the lowest denom is indeed 1MB it isn’t feasible to PUT anything below that? What if I want to save 20KB text file?

1 Like
72

the lowest denomination is a bit, but it does raise questions: what if the accuracy of the account balance is worth more than storing x bits; then should storing < x bits be free, or should it cost the minimal denomination?

3 Likes
73

known henceforth as the “irv”

74

As I understand it now we will have an account balance manadged by the network that will track your PUTs and deduct appropriate amount off of your acc balance and that is accurate dow to one bit? So why to not let the same mechanism govern account ti account transfers with the same accuracy.

75

One reason is that disk space should exponentially decrease in value where safecoin should do something on the opposite scale (I hope). Perhaps for fast deflation currency like things then this is OK?

PS not thinking just adding some comment, in middle of a flurry of work in house

3 Likes
76

Well then we have something for store of value that is Safecoin (provided there won’t be 1:1 swaps back on the network level and once you’ve converted your SC to account balance theres no way back other than through some 3rd party with appropriate price premium for conversion) and something liquid that you wouldn’t want to sit on and spend it right away that is account balance which is extremely inflational. The catch I guess that you would have to freeze converted SC out of the farming pool for the amount of PUTs it bought…

77

Off-Topic

This reply will get me in trouble. But it will come up eventually.

If (Safecoin) is seen as a “stock/rabbit” for the SAFE Network DAC (Decentralized Autonomous Company), then (SAFE GB) is the network cash value of that stock.

Safecoin buys less GB in the future?
The SAFE Network is designed to issue 4.3 Billion Safecoin shares, which dilutes the cash value IF the rate of storage “availability” doesn’t keep up with new issuance. This is the potential problem @BenMS was pointing out… and it’s not good for short term investors.

Safecoin buys more GB in the future?
However, de-duplication counters the above inflation scenario… but it’s incalculable. We could have people spending 90% of the cash (SAFE GB) and still only take up 10% of actual Network storage capacity. This is an investors wet dream.

I think the 2nd scenario is more likely but I always consider worst case scenarios. Since this is going off-topic please reply as new topic if you want to discuss further. I don’t have anything else to offer because it’s unknown until the Network goes live.

2 Likes
78

Is that necessary? Easier to convince people yes, but essential, maybe not. So long as turtles can be exchanged for Safecoin on the network, there is a route for paying your off-network bills. Also, as almost everyone has a use for turtles (when they buy network services) it feels more substantial than a fiat promisary note. Actually, I think there’s a hidden benefit in your catch here…

A significant difference between turtles and a true Safecoin denomination is that turtles will be inflationary - be worth less over time (storage costs will fall in real terms, and Safecoin will probably appreciate in real terms). This means is that people would be wise not to acquire and hold onto large quantities of turtles, which will make them a bit less attractive than the equivalent amount of Safecoin, which would affect their utility for some, but be of little consequence to most (who will only ever hold small amounts of turtles because they’ll normally be spending them rather than earning them).

People earning a lot from turtle micro-payments would be most affected. Anyone earning turtles would want to exchange all their excess into Safecoin as soon as possible. This has the effect of further reducing the cost of storage for everyone else (by exerting downward pressure on the value/cost of turtles and upward pressure on the value/cost of Safecoin on this exchange). Another impact of this is it makes it a little bit harder to be a high earning business than a small earning business - which means it keeps the playing field more level.

These all seem good outcomes to me. Another well thought out idea from @dyamanaka :smile:

First impression is that I like this.

Now:

  • will Safex be able to trade Safecoin with portions of SAFE storage account balances? @dallyshalla ?
  • which I guess is really: is this a feasible maidsafe API feature (ie moving portions of SAFE storage account balance between accounts)?
3 Likes
79

Very interesting idea, it almost sounds like you could go for a futures trading platform based on that. My main query/concern would be a repeat of the point I was attempting to make previously: How will non-technical users (of say some “killer-app” that exists on the Safe Network and has nothing to do with storage) understand payments in Safecoin using exotic division systems, alternative “sub-currencies” and the like? It can be surprising just how many people do not know what a MB is for example. The risk of alienating non technical crowd would most likely mean the difference between Safecoin becoming a widely used as a currency (including being a unit of account) or just some obscure technical token used to purchase space by geeks on the Safe network. Is avoiding a division system like most people who use fiat currencies are familiar with really worth that risk?

3 Likes
80

Agreed,

That point of view was representative of businesses who don’t like volatile currencies because they have to keep “repricing” their goods and services. If we solve the stability problem using (SAFE GB), they are more likely to adopt it as a spending currency and everyone is happy.

But they still need to pay their fiat bills and need to exchange (SAFE GB) for local currency.

Oh and yes, to all your bullet point questions. This assumes MaidSafe likes the idea and can make (SAFE GB) transfers secure and prevent PUT balance forgeries… etc.

1 Like