The Perils of Big Farming

No, I think we are on the same page. Those things are great bonuses. But they could even be better if more of the rewards where steered that way. If most folks are willing to host vaults based on the merit of SAFE, then more of the SAFEcoin could be steered in the direction of developers and content providers, and that would be good… It does sound like the rates on these rewards might well be dynamically calculated eventually.

The support will come from crypto miners, whom already use there pc to GPU/CPU mine just as they did with burst coin. Some of there equipment will be like small data centers, I know one guy mine with a 58 terabytes hard drive storage. On release I know many small farmers whom missed out on BTC mining will have there hands up, many many people that consume torrents will also want to be involved and this is how maidsafe becomes mainstream, because people will be farming and spending there earnings to upload.
So most people would have at least 1-2 terabytes of spare space and this would be the network average from most end points.

If maidsafe finds one celebrity that has been hacked and their photo stolen and sold on the black market then millions of user will come on board and data centers can not have chance to get advantage…

I think that presented correctly SAFE will see a huge influx of these people, and hhopefully in time they can be getting original material in various formats rather than hit-n-miss pirated stuff

As you say these people could form a very significant or even large portion of the early adopters of SAFE.

1 Like

Gentlemenan farmer. lol. I love that.

And with zero interest rates, they could theoretically buy trillions of TBs and just refinance their debt ad inifinitum.
That’s how the Fed perverts (referencing your “perverse incentives”) are destroying the main street, by giving artificial advantage to those who recklessly leverage the most.

I don’t think they could do that (legally).
But some academic projects could do that.

Crop rotation, to use the proper term, will cause price swings.
You farm here, then there’s another project that seems to pay better, then you delete your vaults and go there, then there’s yet another project, etc.

It is not enough to compare market forces within the system. One has to consider what’s going on in the outside world.

Has anyone done cost benefit analysis this way?

  • I setup a 100 GB vault and it gets filled
  • About 5% of all files get requested in the month, and I get 1.25% of that
  • That buys me 1.25 GB on SAFE Network (forever, but let’s say I need those files for a period of 1 year)

Question: how much storage I can buy on another P2P storage network if I farm 100 GB on Network B?

I know MaidSafe is different, but a lot of people (more than 50%, I would say) can farm on Network B (earn whatever coin, sell it for BTC) and store data (buy SAFE for BTC) on SAFE Network.
Note that they are in fact using SAFE Network, they just aren’t farming on it. Price arbitrage of sorts.

Everyone’s making a lot of assumptions about some internal dynamics, but simple questions like these aren’t being considered.
It is already possible to tell how much you can make by farming for competing projects, so these comparisons can probably be made now and with growing confidence.

in the end, it should be about quality, not quantity. privacy and freedom, concepts which i hope the users will put in front of money. or they will have to, in the long run. the thing is this has to be accesible for every network user

i think this already is in plan(or is it not?) because what happens in the eventuality of a big part of the first investors spend big safecoins uploading data in the beginning? i probably do not get this right, but isn’t this a debt of somekind, if the investors safecoin do not have hdd space equivalent the first second the network goes live?

i encourage and would like to participate in an anonymous survey like this. i know there are test nets but a survey among the users of this forum (which we presume speak the truth because we want the network to do well) would be a plus in my opinion

1 Like

There are two effects that make farming more profitable in this case, restoring equilibrium. First, the extra demand for SafeCoin will increase it’s market price, making farming on SAFE more attractive. Second, as SAFE’s capacity is being filled up (more than 66% capacity in use), farming rewards increase to attract more farmers. It will keep increasing until there’s 33% of available capacity again.

Will farmers behave economically (whatever that means - @jreighley claims economy is not a science) or not?
We don’t know. Maybe they’ll evaluate all things you mentioned (like they do now, when they sell privacy in exchange for free storage).
We should assume that other crypto-based platforms will provide a similar mix, so the question of farming on Network B while spending on SAFE Network shouldn’t be ignored.

Yes, but I can sell a 30 day storage contract on Network B (and earn income) while waiting for this to happen on SAFE.

economy is the science of defying math

1 Like

Anyone thought about a measure for the average storage time on the network (corrected for growth)? In other words if I store 100GB, which is say 1 to 5 redundant so effectively 20GB how long would I have to keep storing this amount of data before I get a voucher (in safecoin) to store 20GB of data permanently on the network?

This only applies within the business side - commercial farming - and even there the effect is capped at 20%. Small non commercial farmers will always be viable, very numerous, and have the effect of lowering the network average.


This is precisely what I plan to set up. One question - what vault size (roughly) would this presume, and so what total storage are you positing here?

Exactly. I’m in the process of slowly building a game dev company. In order to do this as an indy and with self-funding, and to not have to wait 20-30 minutes when carrying out each build, I want to build using a distributed build system (at least for lighting cooks). Of course, I won’t be building constantly, and so the machines will be sitting there idle for the majority of the day, presenting a perfect opportunity to run the SAFE network and earn some coin as a bonus, and support the network.

The trick will be to throttle/distribute the builds appropriately to prevent them impinging upon SAFE, once the network traffic kicks up. Hopefully, though, this demand would also increase SAFEcoin profits, enabling me to expand my network to suit and provide balance between the two.

What do you mean with a storage contract exactly?

It depends on your luck. If you’re unlucky you may get garbage data and need to wait forever.
If you get “lucky chunks” (say Kim Kardashian pics) you may get enough requests quickly.

I mean a smart storage contract, where one rents a certain amount of capacity over a period of time and gets paid if he delivers (the stored data checksums OK on the day the contract expires). Some competitors will operate like that.

My thoughts are rough at best, but here they are

  • Initially when only a few thousand people are trying it out the average would likely to be around 1GB used per vault. This will wildly fluctuate because people are trying it out and vaults will going on/off line at a rapid rate compared to the total number of vaults.
  • It will gradually rise from (perhaps sub) 1GB to 2 or 3 as the vaults stabilise a bit and the number grow.
  • Then we will see it increase till its too much for the CHIP which could handle 2-3GB dedicated to vault(s) at this point adding a USB stick will enable multiple vaults (again) until it the average is like 3 times teh size of a memory stick. Then they will suffer in earning capability.

The rate the average increases will largely depend on the torrent community. If they like SAFE then the average will increase quickly, maybe a year and would need 64GB or 128GB sticks to remain as the average vault usage. Otherwise a very rough guess is 2 years. But by then USB sticks should see 256GB sticks cheap enough to be useful.

[EDIT: remember that there will be a lot of small devices from the start keeping the average lower than expected.

Okay, got it. Maybe I’m daft, but I don’t see why this would be bad for SAFE. It’s all supply and demand, and SAFE’s farming rate algorithm balances on that.

20K chunks leave very little room for luck. The law of large numbers is on our side here.

Yes, looking for a measure, the average luck so to speak. Hopefully this will be reasonably equal on nodes storing a fairly large amount of chunks.

I am not categorically claiming it will be, but that it might be (it will certainly be a factor, maybe good, maybe bad).

If I can rent my capacity at a given price today, that lowers my uncertainty (vs. the lottery effect). If the average “SAFE lottery win rate” is say 1%/month, it may pay to rent capacity at a slight discount elsewhere (so that I earn the equivalent of 11% there vs. 12% on SAFE).

For one farmer it may pay to get a large contract and power his farm off for a week, so he may profit by getting paid much less vs. SAFE because his power cost will be low.

Another concern for farmers will be whether they are network-rich or not. With MaidSafe it is OK to be network-poor but with other networks it may not be.

The impact of arbitrage is greatly underestimated. Many will crop-rotate.

I’m not so sure, it depends greatly on how quickly vaults fill up. If it takes days to fill up a terabyte, the difference in earnings between networks must be very significant to warrant changing networks, since you’ll have to start over when you come back.

So basically looking for 2 things:

a.) Average storage time on the network (how long is really permanent on average, with garbage collectors, time that the network will exist, etc…).
b.) The discount rate of storage over time.