# Is farming viable?

I read the forum often but now I registered because I did some calculations and now I’m concerned if it can work.

If I understand correctly, the number of Safecoins is fixed and final: 4 billion. Incidentally, that’s about the number of internet users at the end of 2017 (https://www.internetworldstats.com/stats.htm). Moreover, as the network fills up with data, less and less coins will be available to farm. I’m concerned about this because a large number of farmers will compete for a small number of coins which reminds me of Bitcoin, where nobody can mine a new block at home anymore. Will Safecoin suffer a similar fate?

I couldn’t find answers about the expected future scale of the Safe Network.

• What fraction of the Safecoins will be burned/farmed a day when equilibrium is reached?
• What is the expected market cap in USD or the expected value of a Safecoin?
• What fraction of internet users are expected to use it if it’s a success?
• What fraction of users will have a vault?
• Will vaults follow the Pareto principle?
• How many vaults are expected to exist?
• How often can a home user farm a coin?

I did a quick back-of-the-envelope calculation with my own estimations and found that if 10% of the internet users have a vault and 20% of the vaults belong to a home user and 1% of the coins are burned and then farmed each day then users will have to wait 50 days to farm a coin. If the Safe Network becomes more popular this can grow even longer.

Moreover, another problem is about the market cap of Safecoin. I use the numbers from the previous calculation. If a vault stores 1TB at \$25/TB cost then the market cap can’t be much higher than 50 billion dollars and 1 Safecoin’s value more than \$12.25. If the previous calculation was correct, a vault will pay less than 25 cents/day. Why would anybody care?

6 Likes

I think there is an argument for increasing the supply for both spending and farming reasons. I would like to see lots of small rewards to encourage participation.

Considering that only a small group of nodes needs to reach consensus to transfer a coin, I think the network could cope with smaller units. It will be good to do some experiments on this when we get test coins. The recent talk of benchmarking transaction throughput could also give us good insight.

2 Likes

Adaptively scaling the supply could help but I read that the supply can’t be touched for legal reasons because it would defraud the ICO investors who were promised 10% of total supply. What can be the workaround for that?

I just realized something else as well. \$12.25 coins are unfit for sending large amounts if each coin need to be sent separately.

2 Likes

Division of course.

These problems could be solved by applying a division of safecoin that takes into account the farming and where the reward decreases when the number of farmes increases. This would keep the reward ratio and don’t need to touch the total number of safecoin.

3 Likes

Why can’t they just allow the network to decide to start rewarding farmers with some divisible amount of coin?

For example, if they always tied the farming reward to the purchasing power of 1TB of data, the network could supply whatever amount of coin that would be, whether it be 10 Safecoin or .0001 Safecoin.

2 Likes

I read the RFC and I have the impression that farming attempts work with the 4 billion addresses of the whole coins, so division can’t help to solve this problem.

Division can also not solve the problem with sending large amounts because they will take thousands or millions of coin ownership transfers.

1 Like

Well it hasn’t been coded yet, so that can easily change.

No. The 4 billions addresses are only 32 bits of 64. We have space to design several types of divisions.

Transferring large quantities is the exception and not the usual. For each transfer of a large amount, millions or billions of small value will be sent. If, from time to time, someone needs to do extra work to send this kind of transaction, I do not think it’s worrisome.
In fact, it is just the work necessary to sign a ownership transfer which will allows safecoin transactions to be free. Otherwise, spam will require payment for every transaction.

2 Likes

Doesn’t farming work with whole coins alone? I looked at the RFC again and it’s ambiguous about that part.

It is worrisome because it splits the Safe Network into separate modes, such as microtransactions with divided Safecoins, full Safecoins, and many Safecoins. Where the split occurs depends on how much a Safecoin is worth. That will keep changing, and so will the dynamics of payments. The network can’t scale seamlessly.

You are saying millions or billions of small values will be sent for each large transaction but what is the reasoning behind that confidence when we don’t know how many people will use the network, what the value of a coin will be, and what kind of transactions, other than buying storage, Safecoin will be used and at what transaction values?

I started this thread to try to come up with educated guesses for some numbers but I can’t take your argument seriously if it ignores so much uncertainty about so many things.

Hi, which document are you referring to (the RFC)? I am interested in reading as well. The Safe Network Primer document on the website gives the impression of 1 whole coin per 1 address.

It’s 2 years old, maybe a lot have changed since then.

As has been said already, multiplying the supply prior to launch will have no impact. For example, you can split a pie into halves and have a piece each or into quarters and have 2 pieces each - the pie remains the same size, as does your share of it.

1 Like

This topic has been hashed out many times on the forum already. You’ve got some reading to do… after that you’ll find most/all of your concerns have been considered, with multiple solutions. To much for me to summarize here… you might enjoy combing through the divisibility threads… Cheers!

3 Likes

Here are some thoughts that others may have or not have mentioned before and you could know them as well and I mention them so that we understand each other.

Recycling means that there can be almost any number issued over time.

There will be some people who spend their coin they farmed for a PUT balance that might last them months or more for their pics uploads, forum replies, mail, etc

So not everyone needs to hold a coin for the network to be well and truly usable for them.

Then of course SAFE is updatable and division of safecoin will change everything in that regard.

OH The RFC is old and only a starting point. It is already in need of updating because of the way things have been implemented. Its a space to watch really.

That is crystal ball stuff. It depends on many variables, with the most unpredictable being people. The only thing predictable is that if the network works as intended then people will be uploading data, sending mail, responding in forums etc etc.

The cost of PUTs is dynamic so the coins destroyed per period is unpredictable with any certainty. Then how many people are actually using SAFE. At this time there is estimated less than 1/2 the world is potentially using the internet at some time during the year. But how many are doing things that would require safecoin if done on the SAFE network.

That will NEVER be dealt with because safecoin is not a money making coin and to give a estimate is to make promises and that takes safecoin away from being a utility coin paying for resources and more into a security.

If truly a success then the current users of the internet will become users of the safe network because so much content and sites will be on SAFE

But if the question is about how many will do PUTs (upload data, mail, reply etc) then its probably less than 3/4 of the current users since at least 1/4 do no more than search the internet for info and surf.

When vaults can run on phones (while charging) and tablets in addition to laptops and home computers then countries like Africa where the tribe people’s children are using phones by the millions then I’d say that the numbers could be extremely high.

When school children learn they can earn safecoin by running a vault on their home computer and phone then expect a large portion of families with school kids to be running vaults.

It again will be hard to estimate the proportion of so called users to be running vaults.

A principle I don’t know

Again something that is inestimable at this time. Will vaults run on phones only while charging? are you talking of 24/7 vaults or vaults on for 14/7? Will people see value in safecoin?

So many variables, so little data to work off. All I can estimate (if safe works) is that it will be at least in the millions.

That depends on the factors above and in addition the size of their vault, are they on slower internet speeds? Do they run their vault 24/7? How much spare space is on the network (affects farming rate directly)

One vault might earn 1/5th of another vault because of the vault size, etc. Again this is difficult to answer.

What was your number of people doing "PUT"s? What was the overall time period?

Why did you not account for the fact that data has to be uploaded in order for people to use SAFE? The more coins out there the more coins that will be spent (human nature).

Terminology issue: Coins are NOT burned Burning implies that they still exist but are unusable. Coins are spent and then destroyed by the network and available to be recreated/reissued.

And these are just figures plucked out of someone else’s imagination or best guessing.

Market cap is not a good figure to try and use. Since you cannot know the number of coins existing at any one time. Its unlike crypto coins where the number of coins existing is only ever increasing (usually) to a fixed number of coins. So you can use marketcap to work out many things. SAFEcoin is variable, it can reduce from one day to the next or increase. The cost of PUTs can change from one hour to the next.

A good number of the people will only be looking at how many PUTs can their safecoin be used for and ignore the fiat value. In other words a lot of people will be more interested in how much resources will their safecoin they earned buy on the network and never even look at the fiat value.

4 Likes

That was one of my first thoughts. 4 billion coins and they want the world to use safecoin? Not even one per person.

I agree that increasing the address space from 32 bits (4 billion) to like 34 bits (16 billion) or 35, or 36 bits would be a sensible thing to do.

And as @Traktion says if you give the same proportion then I cannot see this as a legal problem, just some insecure people might feel they are somehow been taken advantage of.

So if 34 bits then the swap is 4 safecoin per MAID

The counter point is that more coins available then the overall transaction rate increases due to needing more coins to pay for something. A point against going to any more than 34 or 35 bits.

Because the RFC is a very early draft of the principle and easier to start with whole coin. The RFC does the division by having a balance that is reduced by the PUT cost every PUT. The balance is created when the user spends a coin for it.

Actually its not quite as bad as it might seem at first. Each coin is handled by a section. So sending a million coins would perhaps involve a few thousand sections. So a little parallelism occurs here whereas in BTC it would not.

Thus a million coins being sent might only involve a few hundred of those transactions being handled by any one section.

And you never know safecoin could be implemented purely as a balance system when finally done. One MD per wallet and the balance kept there only updatable by the network. Its all to be worked out still and these issues are known to the developers.

8 Likes

Safecoin are recycled with use. In order to put data on the SAFE Network, the user must pay in Safecoin, which the network then “burns”, allowing for the creation of new Safecoin to be allocated as farming reward.

Safecoin are divisible. My sense is that price per PUT will fluctuate in response to data storage supply (i.e., resource costs and capacity) and demand (i.e., willingness to pay and resource need). Theoretically, 1 Safecoin could buy exponentially more data in year 10 then it could have in year 1.

In light of the points made above (along with other factors), it will likely take decades (if ever) before the max supply of 4.3B Safecoin exist. As has been mentioned, many community members have explored and discussed this topic. For example, I have hypothesized that.

2 Likes

I really like the idea of farming safecoins. But I still want to have free SAFE storage. Because from an end user perspective people are used to have free storage of their data on Facebook and even on Google Drive up to a limit. Similarly, if it’s possible to implement, it would be great to have a free SAFE storage quota of say 1 GB per user to start with and the limit can be increased automatically by the network over the years to adjust for technological progress.

And the registration of new SAFE accounts probably needs to be controlled for some time still, even in the beta version of the network, to prevent bots from registering loads of free accounts.

And to get back to the topic, yes farming is viable I think, even small farming by ordinary users, at least in the beginning because of the safecoin potential. Even if people can’t turn farming into a business model it will still be attractive for users to earn safecoins.

And even big tech companies like Microsoft and Google might set up lots of farms for the SAFE network, not to make profit, but to ensure a robust platform for their IoT businesses and so on.

Nothing is free. With Facebook or Google you pay with your data and personal information. Everything you upload is analyzed and used for advertising or anything that makes them money.

3 Likes

Agree. Really cheap is likely to be better for the network and users in the long run. If there are lots of farmers able to contribute easily, this will be more achievable.

To add numbers, spending £1 per month would scarcely be noticeable by most people. A cheap loaf of bread, a short bus journey, less than I pay for 1gb of 4g data, etc. If you just send a few emails, messages, posts, I doubt it would even be a fraction of that; the effort by the network would just be so small to justify more (economically speaking).

2 Likes

I know there would be some system to make farming viable on small devices with low capacity and to prevent centralization of farming in large datacenters. Do we know the algorythm how this will work?

Imagine I have huge server and want edge over home users. I already have advantage in capacity/latency/cpu power:
No more coins for vaults larger than X?
OK, I will run more vault instances
Only one instance per PC allowed?
No problem, I will use virtual machines.
Only one vault per IP address?
Can be problem for me on IPv4, but it will alco cut off massive amount of users behind NAT.
On IPv6 i can have bilions of addresses, no problem.

Am I missing something in my thoughts?

2 Likes