Safenetwork sustainability concerns - Bandwidth has an ongoing cost however Safenetwork is a pay once, benefit forever model

Appreciate the response, I am saying more so, would people be WILLING to pay for storage for their newly generated data on the safenetwork EVERY time? Think about it. if i took a photo everytime now i have to think about paying a bit of money to store it… I may just leave it on my phone which is free and as times goes on will have bigger and bigger storage space.

Then will caching / routing servers along the way not absorb many of these requests, for popular content, so the farmer doesn’t actually serve up considerable amounts of data (that have been cached)?

Everything costs, it’s just a matter of when you are paying (sometimes also how).

And something costing money is an abstraction of something requiring effort.

So, your question is turning philosophical/metaphysical: will people need to make effort to make things happen in life?
Probably always.

What kinds of effort/ways to pay will they be accepting?

Well, upfront, debit card, privacy loss, crypto currency micro payments, manual labour, credit etc etc, seems to all be accepted by various groups of population.

1 Like

You will always have to pay to store data. So you’ll always have to choose between:

a) FREE: not backing it up with all those risks
b) paying some other service with money or privacy, and incurring the associated risks (privacy, security, data loss etc.)
c) paying the SAFE fee and then being able to forget about that cost

Now of course, you may choose a), or b).

Clearly hundreds of millions are currently using a) and b) and hundreds of millions will continue to choose a) and b) long after SAFE is running. But that does not mean that there won’t be people storing data on SAFE every day, and that there won’t be growth on SAFE network for the short, medium or decades long term. Which was your proposition wasn’t it: that people might not use SAFE because the up front cost will discourage them?

Well I think that’s improbable, and I don’t think you’ve made a case for it though I’m glad you are now explaining your thinking.

I do accept the logic. Like @mav and yourself I have reservations and am not certain that this model will be the final one. What I don’t accept is that there’s a big risk here at least in the medium term, because SAFE network is going to meet an important need. The question is more whether people will accept this new model over alternatives in sufficient numbers for the long term. My expectation is that this approach will be pretty solid (and I’ve given some reasons why). But if you believe it is flawed enough to need changing, I think you have to make that case.

As always we need to test and evolve, but even though the model has some people sceptical I’m confident enough in it to want some solid arguments for believing it is flawed and should not be tried.

3 Likes

So, to the question if people will be able or willing to put effort…
They have always done, will always need to.

How much effort necessary is according to the network dynamics and the algorithm for farming attempts etc.

I think mav linked that, so you can read about it there.

Thanks for the response BTW! That was a very good one, as now i think even if the safenetwork isn’t popularly adopted by the massed due to possible issues such as this one and others(again, I’m not saying they’re likely, just possible), there would at least always be a bottom line, as some people(I’d probably be one of them) will be willing to do it out of their own will even if there’s no economic benefit. So fair enough. But see i was thinking, and this is actually the point i made this post, is that, can we work out a better structure/deal? Your long term rent proposal is good BTW, we need more ideas like that.

I actually have one too. Maybe i should make another thread suggesting to the development team this, but what if we made it a recurring fee to store data but with a threshold. And it’s not a maximum threshold, but a minimum one that dynamically adjusts. And that threshold is based on the amount of NEW data that’s stored on the network. So if the amount of new data is more than 5%(let’s say) of existing data already on the network, in the past 7 days(let’s say), then the current people who store data on the network do not get charged. If it doesn’t, it gets charged based on how much new data is stored proportionately. This way we can have a safety net to guard against possible failures.

Another way to do it is charge on an individual basis, every user has a ratio of new data they store on the network every month as a % to their existing data on the network. And If that % is higher than 5% every month, lets say, they get FREE permanent data storage! If it’s lower, then it’s paid proportionately, say if they didn’t store anything new in one month then they’d pay more than if they’d stored 4% new data in proportion to their existing stored data. This will constantly incentivise everyone to keep storing new data on the network. And if they stop then they’re charged a little bit proportionately, and keeps the network perfectly sustainable. I feel like this “add-on” feature if implemented, it wouldn’t do any damage to the current model, it’ll only boost it and make it self-sustaining. Because right now the safenetwork with it’s current economical model isn’t, or at least is much less self-sustaining, it has to depend on other things. Of course, you can argue those things are easy to get, but it still has to depend on them big time. It would be nice if it depended on less things and more able to sustain itself through it’s economical model, which is what i wanted.

1 Like

Oh, same here! And i am not even indicating in the slightest way it should not be tried. But it’s potential flaws can potentially affect it when it’s out, that’s all I’m saying. I’m very keen to give it a try either way. Because i think this project and technology is long overdue really. It’s how the internet should be. So definately not trying to say it’s extremely flawed and we shouldn’t try it, i’m simply bringing attention to possible sustainability issues in the economical model so we can all work together for solutions to it!

2 Likes

Nice to see ideas on different models.

The last one there though, how would this percentage be set?

It would have to be based on a notion of “reasonable” usage level, I think.
A private individual would have much less need than The Large Haldron Collider.
How do we determine who the user is and what is a “reasonable” usage for it?

Seems like we are breaching some fundamental principles of the network if we start dabble with that.

1 Like

Nothing to disagree with there. Keep probing by all means - that’s a valuable way of contributing.

One of the things that we lack (because it is hard) is people who can and will take the time to think through potential flaws in detail. It’s easy to make a statement that if X then SAFE won’t work, without explaining X in any detail, and without providing the research and reason to explain why this is actually worth people giving it their attention. Let’s face it, we’re lazy creatures, and asking others to do that work is often the default.

So I hope you will plug away and can find some flaws that are worth the team looking into. We need to be aware though that most of them have been thinking about this stuff longer and in more depth than almost all of us! So it helps a lot if we do the leg work before expecting MaidSafe to spend too much time on our concerns.

2 Likes

Actually thinking further regarding what i said, I’ll start a different suggestion thread stating this(probably tomorrow since i had a long day and it’s midnight here), because if we do it this way, it would not only add a safety net to the network’s economical model but accelerate the adoption of safenetwork on a massive scale.! So think about it, people will be incentivised to store more new data, so that the space in the network will be more and more occupied, so that the amount that each farmer get paid will be more and more, so that more people will hear about it and join the farming or whatnot. It’ll definitely accelerate the growth of it. i think the % we should set it like Satoshi Nakamoto did with the block halving in BTC. So maybe higher % for the first years and the % slowly decays with time(as technology will be assumed to advance and it’ll cost cheaper to store and maintain data, as bandwidth and hard disks gets cheaper etc). Or that % can even be dynamically adjusted based on the networks total storage space(used AND available, basically, based on the adoption of safenetwork)

What does everyone else think about this?

P.S. you could add another feature which is if you contribute the amount of data to the network the same as the amount you’re storing, you don’t need to pay any recurring fee. So if you’re storing 100GB but at the same time you’re donating 100GB disk space to the network, your storage only needs a one time payment. THIS WILL BE MASSIVE, because essentially you’re turning 100GB of disk space, which otherwise will be a point of centralised data centre that the person will put their own private stuff in, to 100GB disk space of decentralised data centre, at really minimal cost/no difference in storage space but massive difference in everything else to the person. WHO WOULDN’T join?

A % of data stored is a very relative number with regards to storage need, depending on who you are.

If that is not considered, you would be accepting that either larger players are subsidised, or that the level is out of reach for private consumers.

As it is, that algo lacks some I would say

Welcome to the SAFE network in that case :smiley: :smiley:

This is what happens. You pay to store 100Gb, the network stores that for you and takes your payment and delivers it to farmers, who will provide at least 100Gb.

If I put this another way, without safecoin, the network can attest (close group signature) your node stored 100Gb, so as a client you can use the vault key to sign a storage request as that vault. This is pre safecion.

Now with safecoin the relation between the customer of the network (client) and provider (vault) does nto require the client has a vault at all. Also the client can use multiple devices and Id’s.

3 Likes

But do you expect it to be net positive in terms of balance, or storing will cost more than the revenues generated by farming for the same amount of space ?

The price to store a chunk will be slightly in front the the price paid to farmers (they set it in reality, but being there or not). So I mean you pay 1 unit or something to sore a chunk, the farmers will get just less than that in a growing network, a bit more in a shrinking network and just that single unit if the network is stable (it never will be).

9 Likes

What is meant by a chunck: 1 of the number of copies (now 8) or all of them? It seems to me that all 8 have to be paid for somehow.
Of course this will be compensated with the ones that already exists in the network, but that won’t be an exact ‘match’.

The network will charge for the storage of a “chunk”, whatever that means, lets call it a file. Users will not see or care how many replicants the network makes really (if we make 8 copies for instance the network factors that in). All we care about is the golden rules:

  1. I store data
  2. Nobody can delete it
  3. Nobody can corrupt it
  4. Nobody can prevent me accessing it
  5. When I demand it, I get exactly that data back, no alterations or changes, but that exact data I stored.

This is what the network charges for, regardless of the work it has to do :wink:

6 Likes

I understand.
However the question I think I’ve detected is: If you have a vault that stores e.g. 100Gig and you have stored 100Gig in the network: will you have to pay extra or do you have earned money?
To make a guess about that, you do have to know the details.
Of course I think it would not be a problem if it would cost extra, because of the points you mentioned :slight_smile:

2 Likes

If you pay first in a growing network then you would lose I think (in this scenario), but if you farm first you will likely get paid more than you need to store the same amount of data, assuming the network is growing. It is very much dynamic though.I am not fully sure we can answer the question exactly, but I see what you mean. So you get paid X for farming 8GB, but in fact that represents say only 1GB of unique data (no replicants) so for storing your 8GB in that case it may cost you more than that 1GB payment you possibly got. A lot will depend on deduplication and the lag between farming and storing. If we imagine you store at the rate you farm then it is likely it will cost more.

5 Likes

IP transit getting cheaper day by day, and I think safe network should peer at public internet exchange points around the world. as this will help and reduce bandwidth cost.
Seattle IX for example provide one time fee for bandwidth(not full route) and it works prefectly
so imagine the farmers are located in multi IXPs, the internet will be a better place and safe network will be able to manage the chunks based on XOR space. mean most of the bandwidth will flow through internet exchange points.

1 Like

But David i mean we could make it such that if they do NOT donate ANY space to the network, or donate LESS than they’re storing on the network, they should pay a recurring fee for storage, proportional to the amount of space they are needing vs amount of space they’re donating to the network.

Because like i said, right now the whole infrustrure and farmer’s incentive is depending on newly stored data. But I am assuming there will be more people, if not much more, who just want to store data and access existing data. Which may require much more bandwidth than what the new storers of data are able to pay for. Hence it’ll be a barrier to the network expanding infinitely.