Imbalance between PUT-s and GET-s can cause trouble?

When I liken the presently planned economic model to a Ponzi I do it in one regard and one regard only:

  • if the present value of the costs collectively incurred by all the farmers to store the 1Tb of data I give them today exceeds the dollar payment I give them today

  • then the shortfall will need to be covered either by future customers or by the farmers themselves

Let us imagine that

  • farmers all become at one point economically motivated and prefer to go out of business rather than incur a loss

  • the annual cost to the farmer per 1Gb stored per year stabilises

  • banks continue paying miniscule interest on deposits like they do now

Would the survival of the network then not depend on future customers paying for the upkeep of the past data? Assuming constant dollar price per Gb stored this according to my calculation requires the number of Gb stored per day to grow proportionally to t*t where t is the age of the system (say in days or years).

Now t*t is a Ponzi, is it not? And in case the number of puts does not grow this fast the dollar price per Gb stored will need to continuously grow.

Financially this is exactly the same as if the network collectively promised to eternally pay each depositor a fixed amount of money per year per GB stored. And the price of buying such an annuity would be determined by the network based on its current outgoings (and perhaps some prediction of the future).

I haven’t read all the back and forth below, but had to comment on this.

Let’s hope the SAFE Network is good enough that people won’t be scared off very easily. If it’s not, our time is wasted anyway.

Secondly, It’s not possible to retain the secure and private attributes of the network and put a time limit on storage. That would mean that all data would have to be tracked and time-stamped, and tied to an identity. Not to mention the confusions caused by multiple identities owning the same files, deduplicated in the same way. What about 500,000 people who store The Matrix Reloaded? Four copies of each chunk, 500k owners. Tough to sort out.

Not a workable or desirable to have a time expiry on this network, so enough said on how it ought to be that way.

1 Like

“It’s not possible to retain the secure and private attributes of the network and put a time limit on storage. That would mean that all data would have to be tracked and time-stamped, and tied to an identity”

Hi, its a very fair concern and a big challenge indeed. But can we possibly be sure now this is indeed impossible? That requires proof!

Further, do you expect completely anonymous untrackable payments to be possible on SAFE? If so it should be possible to pay for the upkeep of a certain file anonymously, right?

“Not to mention the confusions caused by multiple identities owning the same files, deduplicated in the same way”

Well an excess of people willing to pay for the upkeep of a file should be possible to solve some way, right? To start with the network could store more copies making it even less likely the movie will ever get lost. Then more bandwidth could be allocated to serving the file (in a future version of SAFE), copies may be geographucally distributed to make access faster, etc… Then people can say - if there is an excess money left use them to upkeep another file, some website or give them to a charity… Heck you could finance coding like that :slight_smile:

That is not ponzi, but akin to the local club selling lifetime memberships (except its for each piece of data). Are they ponzi?

If the world markets collapse or hard drive technology becomes extremely expensive, then a number of factors will come into play and like the rest of technology SAFE will have problems. So yes SAFE could fail under those conditions.

But do we stop because of what is really equates to a technology collapse. Computing in general will have major issues if what you suggests happens.

The SAFE economy is actually quite complex when you start looking at the conditions you propose and you would be better reviewing the threads where this is discussed in depth. Your questions are answered in there.

On a side note, we can predict future drive pricing. Yes the world economy can move that pricing up/down, but within those constraints we have a good degree of confidence for a number of years in advance.

1 Like

I understand your concerns, but unless you dig in and actually find out something more about how the network actually works, it’s easy to toss off a demand for proof, when you haven’t invested the time to find out how it does work.

As I say, I understand your concerns and there may be a situation to which they apply. But without a context, you haven’t stated a problem or solution which is meaningful. Who knows, if you get the context to frame your proposed solution, you might find that it’s not really a problem.

The Safe Network Wiki has a lot. I’m laying the groundwork in the SAFE Network School, episodes of the Safe Crossroads Podcast (Ep18-20 currently and more to come) to give the background to try to get as many as possible up to speed. It’s a big subject with a lot of interdependent components. I’m still not there, but learning with the rest of us, and sharing what I do get.

Also, if you search the forum, I believe David Irvine has addressed some aspects of the problem you’re stating. Can’t remember everything, but it allayed my concerns to a large degree.

Archive nodes will be possible in the future. Deduplication should also actually save enormous amounts of space overall. The emergent properties of the network will prove immensely valuable, as they replace data centers, etc. Things like that should release a lot of value that is tied up in huge corporate profits, which network participants will partake in directly. Etc. A lot to figure on.

Anyway. I don’t mean to slap you down. Your concerns are valid, but not in context enough. Hope this helps.

1 Like

“That is not ponzi, but akin to the local club selling lifetime memberships (except its for each piece of data). Are they ponzi?”

Well if the present value of the cost incured by the club in relation to the membership it sells exceeds the price it sells it for then something is going wrong, is it not? The club then either hopes to finance its obligation by recovering the cost from future customers or is planning to take a loss. If such sales continue year after year after year then yes, this is a ponzi - the club will only be able to meet its obligations if the number if memberships sold increases each year. Though a club has a reprieve - people die, move out of area or become uninterested. Not so with safe. It promises to keep data FOREVER.

“If the world markets collapse or hard drive technology becomes extremely expensive…
But do we stop because of what is really equates to a technology collapse”

Well I can suggest a combination of fairly mild conditions under which the planned economy has an issue:

  • average real world cost to a farmer to store 1Gb of data per year stabilises
  • you can not make decent money by keeping a deposit in a bank - same as now

That’s not a tech collapse, right? And under an additional assumption that dollar cost of 1Gb stored remains constant we can solve a simple differential equation to find out that the number of PUTs per day should grow as t*t. Isn’t that worrying? Sufficiently motivational to search for an alternative economical model?

“The SAFE economy is actually quite complex when you start looking at the conditions you propose and you would be better reviewing the threads where this is discussed in depth. Your questions are answered in there.”

Links welcome :slight_smile:

Safe wiki is out of dated. It still has transaction manager. lol

I would delete it but I don’t want the community get all antsy on me. :wink:

1 Like

Fergish, thanks for the link to the wiki, I will see what I can get there. Of course the subject is so important that one would think it deserves a coherent writeup in the form of an article.

Re the feasibility of pay for “keeping of 1Gb of data for 1 year” and its compatibility with privacy and anonymity my question still stands:

do you expect completely anonymous untraceable payments to be possible on SAFE?

1 Like

If this were to happen, then yes its a problem to the computing world. So many projects rely on future storage costs greatly reducing per GB as their data requirements grow. Many businesses will have budgeted for this and for it to stabilise will spell trouble for their profit margins. It may not be a collapse, but its not a collapse for SAFE either. There is threads that go into depth of why not.

But I can guarantee that unless there is some economy collapse that is not happening for many years.

But yes a stabilisation of storage costs per GB would cause some noticeable changes in the PUT pricing to account for GET costs needing to rise.

Do not view it as PONZI, but as averaging. The economics say that for every amount of data stored that the costs can be averaged over its so-called lifetime. So for some data its well under priced and for some way over priced. The dynamic nature of the pricing allows for adjustments (major & minor) to account for changes to that averaging. Its like the club working out that for every life membership sold it will cost the club x$ to service it. Some of those members get many times the worth and others rarely use it.

As for SAFE storing FOREVER, my guess (really a certainty) is that a new better SAFE will be built long before then. I am no spring chicken now, but before I die I am sure that SAFE will either be rebuilt or evolve to suit the conditions that are present then.

I will try, but I can only do the same searches you can using the search button.

At this time there is not a way to do this and keep the current model for storage. The network does not time or know who own the chunks. The datamaps that a user has tells the client s/w which chunks to collect and reconstructs the file. The network does not do the reconstruction or connect chunks together, so timed storage according to a users piad amount is not there. If it were built in then it links data to a user and reduces anonymity considerably.

That is the planned model. The weak points is exchanging to fiat and when purchasing and info exchanged then.

2 Likes

BTW, here’s the equation:

let V(t) be the amount of data stored at time t
let CV(t) be the cost to all the farmers per day
let V’(t) be the first derivative of t - this is how much data you store per day
let P
V’(t) be how much money external users pay per day to the network (no internal users in this model)

Then we have

P*V'(t) = C*V(t)

Okay shame on me, I’ve forgotten how to solve this :slight_smile:

1 Like

There are all sorts of scenarios and even the simple ones aren’t that well understood (i.e. whether something would cause a drop or raise in price of PUT).

If number of PUTs drops, the cost of a PUT should drop too.
But you also assumed that demand for reading the data would go down (because of the crisis) but we can’t know whether that’s the case.

People (farmers or users?) leave: farmers leave, the cost of PUTs goes up, more farmers join in. That’s the theory. But maybe also: farmers leave, the cost of PUTs goes up more that people care to pay, so the use of SAFE drops. (That should make the cost of PUTs drop, but we don’t know which would drop faster, the cost of PUTs or the number of farmers, or users, so I think we won’t know until the network is live with real coins and real files).

1 Like

Cost of memory plummets by more than half every year. Maybe this helps.

EDIT: sorry if people already mentioned this, I didn’t read all the posts here yet -disclaimer-

Here are 2 topics that I found quickly, I am sure there was another but cannot find it at the moment


Can I ask where you got the equation from?

BTW you integrate (over a day) to go from a slice in time to a daily figure

That equation assumes a lot (even when we fix it up). Firstly it assumes constant rates because that is what those equations equate. 1st modification has to account for flux in PUTs due to cost changes. 2nd modification has to account for reduction in costs because of drive costs reducing for the foreseeable future. 3rd modification has to account for the rising costs in electricity (above whatever is used anyhow == often $0.00) 4th modification is the cost to buy coins to use because often PUTTers are not (large enough) farmers. The 4th is important because humans are placing a worth on costs that the network translates in to cost for PUTTing. [5th] … [6th] …

The SAFE economics is very dynamic and not reliant on constant rate maths.

Lets look at the case where people are PUTting lots of data day after day after day.
initially the PUT prices are reasonable
after a little while the price starts to increase a little
then more and more.
farmers get sick of little GETs being done and their vaults filling up
farmers start leaving
SO what happens???
GET rewards increase until enough farmers again
and then price of PUT even out or reduce.

Remember that farmers earn the most when they are using spare capacity (== zero costs), so price increases do not affect a lot of the farmers.

2 Likes

Happy for you to propose a fix for it. :wink:

Well the equation is smth I just made up from my model: cost to store 1Gb of data per day is constant - this is C, price paid for 1Gb stored by external users is constant - this is P, the network is accumulating no money inside it, nor do farmers spend their own money - all money paid by external clients immediately goes to system up keeping. There are no internal clients in the model.

Yes that’s a lot of assumptions yet I find it useful to assess long-term stability.

I believe the solution has the form of

V(t) = a * exp( b * t )

where a and b are constants. To me that says: in a world where the cost to farmers per 1Gb per 1 year is constant the system will be able to maintain a constant price per 1Gb stored to external clients only under the condition of an exponential growth of data stored and revenue collected per day.

This is not the case. It is very much variable

Variable too. Electricity costs rising for instance.
But often it is zero because a lot of farmers will be storing their vaults on spare capacity that is powered on anyhow and bandwidth is usually paid for anyhow (as long as they stay below limits – ie stop vault at times when needed)

Then whose money do they spend when they buy their drives. Or are you saying that C is zero because they spend no money to store those 1GBs of data, because they use spare capacity.

It seems that you have not factored in the farmers who use spare capacity and it costs them nothing (no incremental costs over not running) to run vaults. The network is designed for this usage and really favors this because GET rewards are not expected to be profitable for those who need to spend (significantly) extra $$$ to run vaults.

1 Like

It’s a very big, coordinated subject. There is a lot written on it. The wiki is not complete by any means, but you can also wade into the whitepapers , or the educational videos at safenetwork.org

Yes, safecoin can make anonymous transactions. The point is irrelevant to what you are proposing.

2 Likes
  • indeed SAFE can grown to a certain size on free capacity
  • beyond that people will have to buy extra hardware

Besides there may be an opporunity cost here: people could choose join FreeNET instead or use that space for BitTorrent files.

…and the question which I wanted to answer is this: under the assumption of constant cost of Gb/year to farmers can the system maintain a constant cost for consumers?

LIkewise we can ask: again under the assumption of constant cost of keeping Gb/year to farmers how would the cost to clients behave should the amount of data put into the system daily remain constant?

V'(t) = K                      amount of data added per day
P(t) * V'(t) = C * V        money paid by external clients at this moment in time

so solving this

P(t) * K = C * K * t
P(t) = C * t

ok - so in this case the price grows in a linear manner

What I mean is that in my model the average costs collectively incurred by the farmers are exactly compensated by the revenue in dollars that the network collectively earns.

[quote=“neo, post:29, topic:5544, full:true”]

If this were to happen, then yes its a problem to the computing world … But I can guarantee that unless there is some economy collapse that is not happening for many years[/quote]

Indeed. But who can guarantee the said economy collapse is not waiting around the corner? In my humble opinion the goal of building an alternative internet which can only thrive in days of boom is not noble enough. I think that a more noble goal would be to build an alternative internet which would be able to live in the days of bust as well.

In fact in days of bust there may be even more need for it. The workers will need to collaborate to protect their rights. Small time earners will need every chance to pick up an extra penny. Some of the services available today for free may become unavailable because of the big Co providing them may run out of cash.

I would like to thank you again, Fergish, for useful links. However I can not but help being puzzled at your lastest statement. How can anonymous transactions be irrelevant to “pay per year per Gb” topic? Suppose we upload a big file like a movie and associate a wallet with that movie. Then I anonymously make a payment into that wallet. Does that not mean that now this movie can be kept in SAFE for another year without it being linked to my identity? And next year I pay into that wallet again - and the movie will stay online for another year?

All of your formulas figure that economics are the driving force for farming.

I don’t think it will be…

People will want to use SAFE because they want the security from it. They want their files to be backed up. They want their files to be in a place where hackers cannot get to them. They want confidence that they can reach their files from anywhere.

The cost of doing this is Farming. You can either buy a RAID array for you PC at home, Set it up, Encrypt it, Install firewalls. Get a VPN to several remote sites and create additional copies - Pay somebody to monitor and repair and maintain those remote computers — OR you can use SAFE and accomplish the same thing without any of the messy setup and maintainece. Farming is the price you pay for all of that benefit.

And SAFEcoin is what balances it out if you need more than you provide or if you provide more than you need.

SAFEcoin is there to prevent abuse - not to make massive profit centers possible. It’s a Cooperative – It should break even in the end.

1 Like