Safecoin VS SAFE Storage

This is a Brainstorm…

Continuing the discussion from Safecoin divisibility

I was thinking about our two currencies (Safecoin) and (SAFE Storage) and tried to see 10 years into the future. The more I thought about it, one question remained. Can the SAFE Network pay vaults with (SAFE Storage)?

Extreme Hypothetical Scenario
The year is 2025 and the SAFE Network is mainstream with 99% of all Safecoins farmed. Safecoin has reached stellar fiat value of $100 due to many killer APPS, as well as mass adoption.

The Network is at target capacity 80%.
It cost 1 Safecoin to buy 100 SAFE GB.
Unfortunately, most people refuse to spend $1 per GB.
The Safecoin farming pool remains dry, yet the GET request activity continue to increase.

10% of the farmers leave because it’s too expensive to run a vault with so little payout.
The SAFE Network wants to increase the payout but there is not enough Safecoins in the farming pool.

The Network hits 90% capacity, and raises the storage price.
It now costs 2 Safecoin to buy 100 SAFE GB… causing even less people to PUT… but draws some new farmers looking to get lucky in the farming lottery.

Death by Success
The above scenario is extreme and over the top, which is the point. The goal of this exercise is to look at what can we do to keep the Network going if this very unlikely situation happens.

How do we enable people to use the Network if…

  1. Safecoin fiat value becomes very expensive $100.
  2. Safecoin becomes more valuable than the SAFE storage it buys.

Can the SAFE Network pay the vaults with SAFE Storage?
In other words, if a vault provides 500 GET requests, then it receives 500 PUTS. This enables a mobile phone (tiny vault) to use as much resources as it gives to the Network without Safecoin farming scarcity being an issue.

Then what is the purpose of Safecoin?
I suppose it will become more like a stock for the SAFE Network, that also buys SAFE Storage.

Farmers could convert SAFE storage into fiat on an exchange. They take their fiat profit and add more vaults, thereby increasing total storage capacity.

How does Safecoin get distributed?
It will be farmed alongside SAFE storage, adjusted by supply/demand.

Does this mean everyone (Devs, Apps, Content) gets paid in SAFE storage?
SAFE storage is only paid to vaults, Quid pro quo. Safecoin can be used to fund everyone else as normal.

1 Like

This shouldn’t be possible with mesh, etc…

This means its more like a futures contract for network capacity than a stock in a company.

This 100$ price thing would be a glitch. Only a person with 100$ * total network coins (430 billion dollars) could afford to keep up a charade where it costs a 1$ per gigabyte on SAFE Network.

1 Like

Seems like your whole analysis is based on a fixed price for X bytes of storage, and is wrong.

Why do you say this.

The protocol/Network adjusts the PUT costs and GET rewards according to activity on the network.

As has been said before if # PUTs drop so does the farming rewards.
If Farmers leave then Farming reward rises
and a number of other conditions affect the price & rewards

Look up @dirvine posts concerning the cost of PUTs per unit going down as storage space increases.

The working of the SAFE network already adjusts so that this should not happen if the network/protocol is working correctly.

The only time 2 would occur is if there is nowhere to store the data, but as david says the farming rate will increase until enough people start farming again.

[EDIT: example where David talks of price being variable at any specific time and the dynamics worked out by the system The Safe Network's economics - #15 by dirvine ]

1 Like

Make SAFEcoin divisible and the problem is not a problem. We tossed that around on another thread… I think it is “in the brainstorm hopper”

i’m sorry, but i don’t understand this divisibility solves problems. say we start with 1 safecoin is 1 data bit. because you can’t have half of a bit. then multiply the maximum numbers of safecoins with the divider. now to store a song on the network requires 2873468725623693 safecoins. the maximum pool of safecoins is 4347498374837498374947…8734874793874 (i don’t want to do math right now). from what i understand the problem is what happens when the max is reached, and i don’t think divisibility can solve this, because we will have always a minimum and a maximum for safecions, and just a minimum for space. if we had a maximum on space, another problem kicks in, which is the data has to be deleted

Where are they getting their competing hard drive space? If they refuse to pay $1/GB where are they storing their data? Better question: Why hasn’t the network been forked? If SAFE:Bitcoin then why isn’t there a LiteSAFE:Litecoin created? Bitcoin prices were too expensive to mine therefore people forked bitcoin and created litecoin which was easier to mine. Why wouldn’t people do the same thing with SAFE in such a scenario as you present and then just trade currencies between the networks as is done between Bitcoin ad Litecoin now? In fact if SAFE became that popular I’d imagine a few forks would have been created already so really at that point the question really is why isn’t there more trade and activity BETWEEN networks rather than disatisfaction with a single network for whatever reason?

In fact there is much talk about how Bitcoin is going to fail however I suspect it isn’t. Much more likely is the possibility that it’ll split into two competing forks and communities. Either one or both of them will survive much like there are competing distributions of Linux. The value of bitcoin might dip a bit but I doubt it’ll die off completely. Same principle goes for SAFE. If SAFE gets to the point that people are so unhappy with it then it’ll be forked and either form into two distinct communities or one fork with gain dominance while the other loses popularity and either becomes a smaller group or dies off. It happens all the time.

This is basically about whether the network can sustain itself without an inflating currency supply, right?

Also, it seems to be a scenario which wouldn’t make much sense economically. Why would the price to store data be so much higher than what people are willing to pay for it? Normally, the two are approximately the same (otherwise people wouldn’t buy storage devices).

Additionally, storage is likely to fall in price , if progress continues to be made with storage devices. This we mean that storage should continue to drive down storage costs over time. This should add a useful dynamic of change for farming/put costs.

Personally, I am not too concerned with this. I think the bigger problem is the £100 minimum storage fee when Safecoin is too big a unit. Much smaller many units would resolve that though.

1 Like

That’s so far in the future that I can’t even think about it.
I am more concerned about the simpler scenarios like farmers switching off to farm whatever brings more money at that point in time.

It is incompatible mass adoption and a high cost storage.

More people in the SAFE network → more network capacity → less cost GB

In fact, the cost per GB will not be linked to safecoin but the cost of existing technology from the moment. The last decades, and nothing indicates that it will not continue, follows the Moore’s law (approx. doubles every 18 months)

The Moore’s law doesn’t apply to capacity, but yes, farmers will accept less per GB as they upgrade disks to a larger capacity and take lower capacity disks offline.

I don’t think that anyone can predict what problems SAFE Network may face in the future, just like in 2009 no one could predict that Bitcoin will become so popular.

I don’t believe the network will be able to sustain itself if by 2025 there is no subscription style payments (ie you need to pay per year) because the proportion of garbage data will hit 99.99999%. That’s a far bigger concern for me, but we discussed that in other topics.


Yes, I agree, there will need to be garbage collection / archive / tertiary etc. well before then. I hope in first 1-2 years this is progressed and hopefully concluded.


True. Moore’s law is for chip sizing/capacity

But as we move to SSD and other silicon storage media this may see storage start to follow Moore’s law

For the last 30 years storage has increased at a rate of 10 times every 5 years. 1985 around 5MByte and 2015 around 5TByte For reasonable PC drive.

That is welcome news. Be interesting to see how you achieve it. Hopefully without losing the pay once that is currently designed in.

Although the counter consideration is that at 10x storage increase every 5 year’s it won’t be as critical as it feels it should be.

Maybe have a class of storage that is called temporary (for want of a better word) which allows the group to not move the chunks (as vaults turn off) after the date specified when PUT. Yes I know you don’t like timers etc and I agree, but this could be the group which decides that a chunk needs to be stored in a vault can do that decision. The current date/time is agreed on by majority of group, so a vault with a faulty date/time does not erase “temp” data early. And maybe allow more temp storage to be stored per coin than for “forever” storage.

Of course Moore speak about transistor but the data storage follows the same path.

The mathematics do not agree with you. If storage follows the exponential growth of recent decades, and everything indicates that it will continue at least another few decades, more than half of the total data will be relatively new, possibly less than two years.


1 Like

I get your point and it is valid, however we need to consider 2 points here I think

  1. Confusion between Moore’s law and just plain exponential growth (I prefer the latter as a description)
  2. Places like CERN for instance throw away more than 98% of data from tests so there is a lot of currently unsaved data, that may be saved. So perhaps I am saying there may be much more we can do better (like lossless compression etc.)

All in all it may well be a very long way away, I suppose we will see.


We get it!

I am just telling you that Moore’s law is not the right name. It applies to transistors, not drives. And memory cell storage doesn’t benefit from more transistors.

Moore’s law" is the observation that the number of transistors in a dense integrated circuit has doubled approximately every two years.

1 Like

That is storage density, a different thing to storage capacity

Moores law

[quote]The observation made in 1965 by Gordon Moore,
co-founder of Intel, that the number of transistors per square inch on
integrated circuits had doubled every year since the integrated circuit
was invented. Moore predicted that this trend would continue for the foreseeable future[/quote]

Storage capacity per drive has been 10x every 5 years since the 80’s unlike the density. Storage capacity is dependent on more than density.

By Moore’s law storage would have increased 2^30 over the last 30 years since 1985’s 5MByte 5 1/4" drive. That would be 5,000 TByte drive in 2015. This is a major difference and will affect calculations for future storage ability.

Sorry to nitpick this but it is important otherwise

@neo & @janitor I don’t think a discussion / posting quotes about Moore’s law etc is adding value to this thread. The facts were useful. Please come back to the OP.

This scenario can be avoided if we design a proper dynamic PUT price algorithm. It should take into account the elasticity of PUT prices in relation to quantity of PUTs. It should also have a dynamic target % of SafeCoin in circulation, one that makes sure there isn’t excessive inflation and that there is always a sufficient SafeCoin reserve for the network to increase farming rewards during a crisis.

I have a half-finished concept for such an algorithm, I should finish it soon and post it so we can discuss it.


Big thanks to everyone for their input!!!

This brainstorm was somewhat of a Kobayashi Maru and many replies were very good. I mean this with the deepest respect. Hubris of a community often blinds them to dangers and this community is becoming more diligent every day.

This gives me confidence that no matter what happens we will work through it.


The Satoshi Nakamoto (SN) is often cited as having put some surprisingly deep thinking into this problem and made some impressively hard choices and educated guesses in order to provide a currency with a fixed minimum(*) and maximum size yet still able to cover all use cases that the currency might face well into the future even if widely adopted by most people on the planet (for a light introduction see under heading “The 21 Million BTC limit”, or more details on the original bitcoin crypto mailing list). Everything from where and why SN chose the decimal place to be where it is to the “lowest money of account” bit size 2^50.899 ~2.1 quadrillion Satoshis choice is no accident and really very clever in more ways than one. On divisibility however, enough room to allow pricing low value 1 bit of data, and still reasonable to use when valuing expensive mansions years from now. Divisibility is a core fundamental characteristic of all in-use currencies and extremely important since if you cannot value any good or service in the currency, then what good is it. Words you often see cropping up are “usability” and “familiarity” when it comes to this topic as well (I mentioned more on the importance of that here).

Then “Safecoin” would be as the Satoshi is to Bitcoin, as near to or as being the lowest unit of account - and people would use it just as often to value goods and services i.e. not much so probably a poor choice of what to call a Safecoin. Agreed that 4 billion total of the lowest unit of account is simply not enough to be useful in the long term and far from Bitcoins ~2.1 quadrillion.

(*) Satoshi is the first pass smallest unit of account, and other smaller divisions could be added by consensus in the future… if a consensus to do so could be reached that is. Also Bitcoin is often criticised for being unwieldy with the decimal point when pricing some goods and services but this is short term thinking IMHO, what value will things have in 50 years and where will the decimal place go then? How can anyone launch a fixed total count supply currency and realistically be able to predict what value the market will price 1 standard unit of it (a “Bitcoin”, “Safecoin”…) at a few years from now? In the end it is all an interface and naming convention problem anyway - the trick is to keep it consistent and familiar so that it remains usable to normal everyday non technical people.