Safenetwork expandability concerns - Economical effects of redundancy mechanisms - farmers make 8 times less!


Even at 100x, your cost estimates are still VERY ridiculous. So, each photo nowadays is about 3-5MB. Let’s just be conservative and say it’s 3MB per photo. 1000 photos gives 3GB. Right now, each GB costs about $0.1 per year, or $0.07 per year in cold storage(no bandwidth) you’re saying with safe, even after multiplying your costs by 100x, that 3GB, stored forever, will cost $0.01 per GB. So, basically, it’s 7 times cheaper to store forever on Safenetwork than it is to even store on COLD storage on Google drive for example, for one year. Seriously??? Remember this is after already multiplying your assumptions by 100 times, so please be a bit more realistic with your assumptions.

And yes, i have thought about this, however, i highly doubt dropbox makes more than 700% margins though, they’d be lucky to make over 300% after expenses. Even if you took that away, that would make the safenetwork lets say 4 times cheaper than current companies, you still can’t ignore the residual cost of the network, remember current companies limit bandwidth, and since the redundancy is 8 times you will still have a slight problem of farmers being 2 times underpaid than their costs of running it. But you do have a good point regarding this, and my estimations may not be accurate.

Those are quite valid reasons and good points also, but you should also add, when downloading or streaming videos, people do not like to have their bandwidth strained, so they turn off the app, then may forget to turn back on etc. And some people just have a habit of not turning off computers and saving electricity and do not like leaving their computer on. You also forgot to mention how their electricity bills could be higher after a month than all the safecoins they got and they decided to stop.

Just like bitcoins creator envisioned everyone will use their CPUs to mine bitcoin in the future, when ASICS came out it was no longer the case. You should consider this for the safenetwork also, there will be delicated farms, for profit, and they will probably make up the most of the storage space of the network, and help it grow. However, safenetwork does not really reward those people as much, which is also of concern, because they will help signifcantly boost the networks growth. Recently we saw a huge orchestrated Bitcoin cash pump by the miners, it certainly gave Bitcoin cash more media, news and acknowledgment.


The network adjusts by increasing cost of storage, which i personally think is a bad mechanism, as increasing cost of storage will also have those wanting to store new data more reluctant, and since income from storing new data is the only income for farmers, making people more reluctant to store new data will not help the network when it’s already running short. As opposed to the OPTIONAL rental model i proposed, which means for those people that chose it, even at times of network storage shortage, they will have to pay the newly increased price that month, if they do not, their data simply gets taken off, hence they will either pay, or give back space to the network, which helps solve the issue. And if they so choose they had enough of paying (increased) rent, they then upgrade to paying a larger sum one off, which also help the network in the times of farmer/space shortage.

You see, it’s really not a bad model.


@foreverjoyful, the amount of incorrect understanding is getting frustrating but I’m enjoying the much-needed exposure and discussion of the safecoin algorithm. It hasn’t seen as much discussion as it should have, so thanks for the threads because the attention is worth it despite the incorrect information you’re constantly posting.

Incorrect. Farmers are only paid for performing GET requests of data for clients. They are not paid for performing PUT requests. Even though they need to do the PUT at some point to be able to do the GET, they only receive income from doing the GET. No new PUT does not stop farmers income. No new GET does stop farmers income. Please understand this.

I’d like to ping @Deadloch who is hopefully still working on a blog post of the economics of safecoin who may have something to add to this thread (and also the very long sustainability concerns thread). They’re full of incorrect assumptions and understanding, but I’d like to know how that blog post is going and whether these threads could benefit from any of the research you’ve done so far?


No new PUT will basically mean no more coins ‘returned’ to the network, which means no new coins can be paid out from GET or PUT requests. So i was right in saying that income does essentially, come from storing new data. Or is that also not right? That’s how David described it to me.



They don’t pay that way to the seller. They create a financial agreement with a bank to pay back it all in 55 days or pay it out over a few months. This cannot be called rental. Also if you agrue this then my bank account is also rental. I put money into it in bits and pieces or weekly pay check so I can buy that item outright.

Rental is a entirely different fish to credit cards or savings.

You suggest only one model, so they are better and yet you wouldn’t use them.

So you suggest 10% now instead of the 20%. My point still holds that after 3 years its 360% of the upfront cost. Even though the upfront cost is ever so tiny fraction of a coin that you need a mathematical approach to understand the amount. Yet rental is better?

Go read the RFC. One MB can cost as little as one (or will it require 3) “PUTS”

One “PUT” can cost as little as 0.00000000000000001 of a SAFE coin. Its in the RFC. I am suggesting that spare space is not as much and that its like 0.00000000001 of a safe coin. 10000000 times as much as the minimum.

Again a little research on the model would change your views a little.

My hard drive (not USB drive) costs around 60$ AUD per TB or about 6 cents a GB. In 5 years it will cost $0.006 and in 10 years it will cost 0.0006. Well actually the SSD developments could mean in 5 Years 1GB will cost $0.0001

My example WAS an example and NOT meant to be actual TODAYs costs. So it was 5 or so years in the future, it still does not diminish the effects.

If I can be charges 6 cents for my batch of photos from a wedding or other big event then why in the hell would I rent. Can I not afford the 6 cents per batch.

And that was $7,000 or more 10 years ago. Whats your point. People think this project won’t be live for 5 years so my examples are actually approximately correct.

AND HELL YES the price can be a lot cheaper than Google cloud storage. Its all part of the new model, read it up. Using spare resources has ALWAYS been cheaper than commercial resources.

How much do you pay to store (incrementally) to store goods in your spare space int he garage? What if you made a distributed system that allowed people to store stuff in others garage and pay into a pool and then the garage owner is paid when the guy came to get some stuff out of their box. Yes I agree its not a direct parallel but one that shows it can be cheaper to use spare resources than a commercial storage provider who charges by the month. Its a different model and comparisons only show that its comparing apples to oranges due to all the differences.

I didn’t say that. Microsoft want people to pay in increments so microsoft can earn more money. The purpose of creating biased studies is to attempt to change the attitudes of people to be more favorable to the idea of renting software and hardware (game boxes). That way they can earn 10% (your figure) each month thus getting 120% per year of the outright up front price Microsoft (used to) get

Of which you claim is out there. Ignoring that research builds up on itself. How many scientific facts have been shown wrong decades or centuries after ordinary experienced folk knew it was wrong. Your claims go against what we see in society. EG credit cards usage rising so people can pay upfront. OR that even you don’t want to rent from radio rentals because you don’t like their rental plans.

I am not ignorant of using research, I had to rely on a lot of research in my R&D job years ago. But back then we critically examined the research and at times rejected peer reviewed papers because of a variety of reasons, One being it did not actually answer what was needed. Or simple critical thinking showed flaws in the research. Or later research showed a different result. So no need to lecture about research, it has ots place and needs critical thinking and your claims don’t add up and if the reasearch truely agrees with your claim that people WANT rental over upfront then I suggest it is biased research or is answering a different question to what you thought it was. I agree though that people are being channeled in to rental models.

I will call BS on this.

I pay 100$ for unlimited 100MB/s cable plan.

TPG I paid 59.99 for 20MBits/sec unlimited quota ( nigh on 200 GB/day experienced) when I lived elsewhere. Now TPG offer unlimited for either 69.99 or 79.99 on the NBN at 25Mbits/sec

Oh if you want perfect accuracy then its 500K to 2MB for most JPG phone photos and even at 3GB my point still stand because who takes 1000 photos at a wedding or honeymoon so my 1GB was an overestimate using 1000 photos which more likely is less than 250 photos.

In any case the point still stands and see above for costs.


This is correct.

This is incorrect

  • mainly because there is no way for all coins to be consumed, so there may be times where a farmer must do a lot more GETs before eventually receiving a coin, but it will always be possible to receive a coin.
  • also because coins are never received by farmers for PUT requests. Farmers can only receive a new coin from the network for GET requests.

This is not right. There’s a subtle difference between ‘coin supply’ and ‘farmer income’. You maybe think they’re more closely related than perhaps they really are.

On one hand (the ‘supply’ hand), you’re right that “increasing the supply of coins” does come from storing new data. On the other hand (the ‘income’ hand), “increasing the supply of coins” is unrelated to “whether or not farmers have income”. Supply affects the amount of income but not whether income exists or not.

To rephrase that last point, the “the way the number of available coins changes” is a different concept to “the amount of income for farmers”. But it’s quite a subtle difference. It’s not right to say that ‘a reduction in the supply of coins can stop farmers being paid’. It is right to say ‘a reduction in the supply of coins can reduce the amount that farmers are being paid’. But farmers are always paid.

Consider the concept of “the number of available coins”. This changes in several different ways, sometimes increasing, sometimes decreasing. But now consider a different concept, “the farming income”, which may pay more sometimes, it may pay less other times, but there is no point in time where farmers do not have an income.

It’s a subtle difference, but important to understand. Hopefully this clarifies, if not we’ll keep discussing because I’m sure this is a difficult distinction for many other people too!


The point is not that a farmer is paid for any one thing. The farmer is paid for maintaining a vault that does all of its work, including everything it’s supposed to do. That includes PUTs, GETs, intercommunication with other vaults, passing messages, etc.

GETs is the event that triggers the opportunity to receive safecoin. It’s designed that way because this will be an ongoing, regular event for ALL vaults on a fairly equal basis, at least in proportion to how many chunks they are storing. As long as the network is being used, GETs will occur fairly uniformly across the network. How often a GET results in a safecoin award will be influenced by the farm rate, which will be determined by the availability/scarcity of resources in that network address region, which will be pretty representative of the network as a whole, as data is stored basically at random across the address space.

You seem to be fixed on the concept that farming is like bitcoin mining in some way. It’s really not. One will run a vault in order to contribute to the network with existing resources with pretty thin marginal costs. It won’t take much in the way of safecoin awards to cover that margin, but as long as it covers ones use of the network, minimal safecoin rewards shouldn’t affect the up/down time of individual vaults that much.

The long game of running the vault is the most profitable play, after all. The longer you run a vault, the more data you accumulate to serve up as GETs, the better, so I doubt people will be switching them on and off. If the drain on your pocketbook to run resource you already have is hardly noticeable, why bother not running it?

There are aspects of accumulating all those PUTs, maintaining and moving them around that make me queasy sometimes when I try to wrap my finite head around it. To that I have to trust to the math and design of @dirvine and others who operate closer to the metal. But I do trust that the design will work. Faith in my technical betters, I guess.


This is paying in installments. But with data ownership is more complex and much harder to comprehend, I think paying residually for being able to access and store data works well with most people.

Yes 10% instead of 20% and in bigger amounts, many people will choose the “pay recurring” model.

Ok I will do my research but I do want to say this, if it cost $0.001 per GB or less, it will only mean farmers will get paid jack shit… so i really don’t know if they’d want to be a farmer… THINK about it, a hardrive that’s 1TB cost about $50, but on the safenetwork it just cost $1 to store it FOREVER, and have access to it FOREVER with unlimited bandwidth… like WHAT? It’ll get taken too fast for it to keep at that price. Who WOULD be willing to pay $50 to buy a hardrive only to make $1, and pay bandwidth and electricity costs forever?

It can be cheaper, but not by ridiculous amounts, like 50 times cheaper than a physical hard drive.

Well yes, of course all research are up for criticism, but at least for now, there hasn’t been any substantial ones to bring down the research done pointing to that MORE people would PREFER to pay in INCREMENTS over paying a larger sum upfront. I don’t like how you call it rent, and own. Cos really, you don’t “own” your data. It is dependant on if the network still exists, it’s not something physical in someone’s hand that they’re perceive it as they really ‘own’ the data per say. But even with your idea of ‘rent’, there are still people who will prefer to pay to “rent” per say given that the cost is significantly cheaper and that they may only need it for a short period.


No its delay payments upto 55 days or paying off. Paying in istallments is another thing again, that is where you agree to pay “X” amount each month for a period of (usually) 6 or 12 or 18 or 24 or 36 or more months.

They are entirely different things. Credit cards were and still are promoted as a way to pay upfront for those items and pay later either pay for all those items at once or as much as you can. It is not an installment plan or a rental plan. Both installment and rental have different models and effects to the credit card payments.

Yes we know you want this payment system and I and some others still wonder why.

Starting to sink in I see. See but still your “jack shit” is a huge profit margin over paying almost zero when using their spare resources. Soon the concept of spare resources will make sense.

If you read my post then its NOT 50 times cheaper. That was a cheap shot really. In 5 years it will be on par. But in trying to discredit me you missed the point. Spare resources means that the farmer is using zero dollars to supply the disk space to store the data. And the network is not paying for that space either. The network is only paying when data is retrieved. So zero cost to store for the farmer and zero cost to the network to keep that data on the farmer’s disk.

On SAFE you will own the data in the true sense of ownership. Strange concept in this day and age but true you will actually own the data in the truest sense of ownership. Saying it could be lost forever doesn’t change ownership. If you lose something because the network deletes it due to not paying rental then it is not owned by you.

And still no links to it. To something that contradicts even the rise in credit card ownership and use for pay upfront.


Exactly. I have a bunch of spare space on my computer and an external hard drive (1TB) a buddy just gave to me, so to me farming will either be free usage of storing files securely on SAFE or maybe even a little money. It’s the people’s internet not the profiteers. Even if I break even farming it’s worth it for the security and privacy to me. Pretty simple.


What? I have linked before!?

Incase you want the actual research paper it’s here( it is linked on the website )

The proposed explanation is that people compare prices subconsciously, so if they’re offered to pay 1 safecoin to store the data forever, or 0.1 safecoin monthly payment, as illogical as that may be to you and others, there will quite a number of people choosing the 0.1 monthly(we’re not purely logically creatures afterall) AND JUST AS WELL they do that, because it helps the network’s expandability as well as sustainability become less dependant on new data being stored, as i’ve explained in previous posts.

Plus, there are also other reasons why people would prefer to pay monthly, maybe they have better places to invest the remaining 0.9 coins, maybe they think safecoins will raise in value and next month they’ll pay less than 0.1 safecoins. And by doing that they potentially help the network, so having an extra option for people to do that isn’t bad at all.


What happens to SAFE if there is a storage break through that does for storage what quantum computing promises to do for computing?

I mean there was that article in one of the threads where the researchers stored information in a single electron back in 2000 and speculated from the results that the capacity per electron was at least the library of congress but possibly infinite.


That is not rental, but a way to ownership

Discussion and Conclusion
We provide an assessment of our current understanding of
(1) how reference prices are formed, (2) how reference
prices are retrieved and used, and (3) the effects of refer
ence price. We offer summaries of prior findings and an
agenda for further research, which includes a set of proposi
tions. We also provide a critical assessment of the role of
customer heterogeneity, which has raised questions about
the validity of reference price effects found in modeling
based research. In this section, we discuss the alternative
domains of the reference price construct and briefly review
the normative models that have incorporated reference price
into the demand function to draw managerial implications.

This is not conclusions for rental vs upfront but for reference price research. And examination of cross brand pricing and expected price depending on time and exposure to other prices.

The studies tabled in that research is about price research in terms of expectations and time of exposure of the price (ie after seeing other prices or not seeing them)

It is not about and nowhere discusses installment or rental (two different models). This is what I meant about critical thinking and not relying on others with an agenda (marketing site) to make conclusions for you

One reason I did not attribute this link to proof people prefer rental. Even that marketing site you say claims they want installments which is eventual ownership as opposed to never owning (rental).

This is a marketing site - sigh

Rental small things like a pencil case or paper weigt is what renting EACH of your 3 or 5 or 10 MB photos would be more like. Not renting a fridge or whatever.


It would just mean cheaper storage right? Quantum computers sound awesome also but it’s a whole other ball game that will take a loooong time to come to market for us peons. SAFE is all we’ll have for quite some time, and I’m cool with that because it’s one hell of a tool. I’m not so sure that the SAFE architecture couldn’t still be beneficial to quantum computing in some capacity.


Today one maidsafe is around 0,40 dollars. When farming will start the coin certainly moove up. It’s very hard in this condition to predict the cost to store on safenetwork. If we want people join us i think (it’s my opinion) the first 500 mo or 200 mo (it’s just an exemple) should be totaly free. If don’t most of people just will use the browser.


Just for information, here in the Czech Republic, I can get 400Mbit/s for $50. This would be hosted at the provider premises in a big rack and all the electric consumption (max 110W) goes to the provider.

Looks like a great deal, isn’t it?


This was touched on in the other big thread here. The feeling of community and wanting to take part is a big driver too.

In reality, the farming reward may tend closer to zero than the price of the hardware, simply because people want to be helpful. Numerous peer to peer networks have relied on this alone in the past.

I do think having farming rewards is a good long term incentive, but the financial motivation may not be a huge driver for many farmers.


I wonder if quantum tech can help keep the cloud model alive while also accelerating its adoption in the general public. To expand a q box more space and more electricity doesn’t seem to come into play. Also seems that q communication may be scaliable across distance, maybe the speculated post quantum can breach that signaling… but sounds pretty strong.

But agree that crypto is an arms race and it still may make sense to stack tech. SAFE is it bit like a hologram and so may mirror what nature already does. If the universe at one of its turtles all the way down is a blured out quantum computer as Lloyde seems to think, SAFE seems like a reflection of that bluring out or holographic pattern. As for deeper locks, makes sense if we have to buy time against an AI that will social hack us and lock us all out. But could be pipe dream question a bit like the petrol people who think AI will find a way to save petrol instead of undermine it further.


The network should rather be secured by the greed of farmers instead of their philanthropy. Otherwise a well orchestrated campaign on social networks might persuade a significant portion of them to stop supporting the network, which could kill it.

An attacker could lured them into another supposedly better network or that frees them from the control exerted by core developers. They could also make them believe that MaidSafe is controlled by big corporations or any other absurdities.

If farmers have good financial incentives, then they will think twice before leaving the network.

To illustrate this, take the recent attacks on bitcoin network as examples: I think they have failed because miners considered their real long term financial interests.


Oh, I agree. When I say tend to zero, I don’t mean actually free - I just mean they may offer a heavily discounted price, if they feel they are also being a good community citizen.

Ultimately, the market will set the price, influenced by how much good will farmers have or don’t have. I think this is a good thing.

EDIT: To add, good will is pretty powerful too. If it takes a lot of money to overcome good will, we have extra security baked in.