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

Safe-Node
101

The point is that people farm with their spare resources. So the cost is free (except for a marginal amount of electricity increase). Disk, bandwidth, etc is using their spare resources. Marginal electricity costs is more than covered by SAFEcoin rewards.

I disagree.

If you think of it as a pure data storage system then you might be close to having a point in my opinion. But even then the average farmer will not attract much extra cost and 1 SAFEcoin per month covers extra electricity costs due to drive usage or leaving it on for a few hours more.

Remember that only people who can do this will be farmers and those who cannot recoup the costs they incur than will either not farm or will farm anyhow because they feel the cents extra a week (after rewards) is worth the effort for helping have the secure internet rather than the mess we have now.

Also it is reasonable to expect the price of SAFEcoin to increase on the open market, so my farming rewards today might be worth a lot more in a couple of years and if the last 2 years is any indication then the rewards will likely pay for the drive outright.

So no I completely disagree that the rental is at all necessary.

So it resides on some archive nodes (eventually) and takes up a really tiny portion of their drives. And less and less as time goes on since drive space increases dramatically over time.

Also I disagree that the data is useless or lost, since when SAFE becomes used by ordinary folk then it will become common place to include in their sealed portion of their wills the credentials or where to find them. Such valuables as photos are cherished by older people which is evidenced by the interviews after bush fires.

Also this data would most likely be shared with their children so no its not unused, just not being updated anymore. So the idea that the network will just fill up with junk data is simplistic and does not reflect the little used data. Maybe allow people to delete their old private, but not by any rental or network initiated method.

Anyhow back to the point I was getting to at the start of the post (which someone made earlier), it is not a matter of the economics of data storage as if its some physical object. Data storage is a lot different than the economic models we apply to objects such as cars, houses, valuables, etc.

  • Data is infinitely copyable.
  • Data chunks/objects are not just the bits being stored. They have so much more value than that. Look at bittorrents where people get NOTHING for sharing their diskspace, except for potential nasty letters and legal troubles. Yet they still give up disk space, lots of bandwidth and we are talking of a bigger system that that which also spreads the load more too.
  • Solving the problems that exists now with rental models of the current internet where everything is rented.
  • security of the data both in personal security and the non-deletion of data.
  • Anonymity of data online
  • Extremely and I mean extremely small costs of storing each small file. If I want to store 10 MBytes of files only then I don’t have to pay anyone anything other than 10 times an extremely small amount.

So once you take in the real value of forever data then having a rental model just brings us back to the bad old internet where others own your data and can delete it on you for being one day late. Oh and did I say how do you ever keep track of the rental due date of all those chunks and MDs. Oh and did I mention you no longer own the data. And did I mention that you are no longer storing anonymously since the data has to be tagged with your ID, when you stored it and when rental is due. Oh did I mention that it will be a nightmare for the system to scan the network for data that is to be deleted because rental is overdue?

4 Likes
102

Yes, thinking like that is great, but do all BTC miners, even the majority, actually BELIEVE in the philosophy of BTC and would run their miners at break even or slight loss? Very unlikely, it’s dominated by big elites who know how to make great ASICS and get electricity both at minimal cost and are there for profit! So if you want the network to expand, it’s nice having these hard philosophical backers such yourself and me however, it’s also good to introduce a lucrative economic model that satisfied the needs of both type of people who may want to become a farmer. Do you see what i mean?

103

Its just common sense. Give me another reason why people would farm willingly.

Either they get profit, or want to do it otherwise or simply don’t do it. Which is what I was saying

Just how much electricity above normal usage do you think it will use? Its not like bitcoin, but its almost Nothing,

The cost for millions and millions and millions of farmers is going to be next to zero above their current costs of running their computer.

I am going to run a RPi look alike with total outlay of 10 dollars. Less than 2.5 Watts average power and on my unlimited internet. Oh and the drive is just an old 1TB. It will cost me less than 2 cents a day to run and that was buying special hardware to do it.

Explain why millions run bittorrents every day and get nothing for it, costing them lots of bandwidth, costs them potential legal trouble, potential disconnection from their ISP because of media companies.

Explain that and you might see how a non-rental system can actually work. Bittorrents goes against your agrument that rental is needed and the only way people will adopt SAFE in large numbers.

3 Likes
104

I disagree with this. If you look at the current centralised cloud storage services. People pay on a recurring basis, to have continued access to their stored data, why recurring? Because those cloud storage companies also have recurring expenses such as the bandwidth cost when people demand their data.

Ask any cloud storage company to allow you to pay a one time fee to store data on their servers infinitely and give you as much bandwidth as you need when you want to access them, they’d decline this offer because it won’t be sustainable for them, because they have to keep relying on new customers and people wanting to store new stuff, to pay for the bandwidth when existing customers retrieve their stuff.

However, this is EXACTLY what the Safenetwork is doing right now. Asking a one time payment for literally FOREVER FREE retrieval of their data which cost bandwidth and electricity, which unfortunately at the moment are not free, and may never be.

105

if I upload a movie to safenetwork, then it will be publicly available? does the users who want to stream the video would have to pay any safecoins? or accessing the data will be free like bittorrent?

106

SAFE is about reversing the trend of ever increasing rental costs. They incurr charges because they are not using zero cost storage system. SAFE uses virtually zero cost storage due to their farmers using their spare resources.

While you remain in denial of farmers being maily people using spare resources you will not be able to get past your need to keep the status quo of paying again and again and again and again.

Again bittorrents disagrees with you too.

107

Only if you store it as public data. If you store it as private data then you can share it with some others by giving them the datamap for the file.

Accessing data is free.

108

Indeed, bittorrent works, I agree with you. But i am not sure if it is entirely against my argument. Because you know when you download stuff using bittorrent you’re actually contributing by uploading it too as soon as you’ve downloaded even just a little bit of entire file. And once you finish downloading Bittorrent quietly uploads in the background without your knowledge. until you realise it and close it or keep it open cos you’re downloading other things. So if a content is popular enough, You can have multiple people starting at different %s, in their download progress,lets say 10 people is on 1%, 10 is on 2% and 10 people is at 3% lets say. So the people on the top with higher % in the download progress can keep helping the bottom, and people at the very top(90+% or have finished their download) helps those just a few % below them, lets say those help others at 80%, and they help others at 60% and so on.

So if the content is very very popular, it’s very fast and always available. however, when a data isn’t popular, this is where you can keep the torrent client open for hours and not get a single KB because their data is offline. But Safenetwork is different, all data is online all the time, hence the safenetwork IS NOT like the Bittorrent. And Bittorrent works for popular data only pretty much, for other data it sometimes works, sometimes doesn’t. If that’s how you want the safenetwork to be as well then you can have the non-recurring payment model… Safenetwork doesn’t even make people who use the safenetwork to browse upload data back at the same time, i think that can be another suggestion for improvement, what do you think?

109

Fair, maybe storage is free, but bandwidth also? I don’t think so…especially not long term when ISP realise what’s going on…

110

GBH makes sense to me, I am in the middle of starting a cloud business and ISP as well, what if I build a full rack only for SAFE, does the RACK will cover the expenses that I pay for electricity(Per KWH= 0.12$) and IP transit(monthly=10Gbps ~2500$)?

111

Glad it does, and even current cloud storage companies can implement the concept of GBH, or Gigabyte-days or any other unit etc. It’s not a bad concept, and make people pay per GBH or GBD. But cloud storage servers can be cencored or shut down, then people may lose their data, so the GBH concept on the maidsafe network would work wonders! You can be almost CERTAIN the GBH you’ve paid for with your hard earned money will stored your data in a way that that data completely belongs to you, this goes along well with David’s golden rules

When you buy(or farm) a Safenetwork-GBH you will be confident that you can use them to store data that process these properties, with the safenetwork in place, AND it makes the network more sustainable and more incentives for farmers, which will accerlate its adoption and expansion exponentially.

1 Like
112

We are touching on this many times now I think.

@foreverjoyful, you propose lowering the ambitions of the project, to the degree that most of the point of it is lost (that you own your data).
Would you possibly have some other idea to mitigate the risks you perceive?
I think it is a very valuable discussion to evaluate the robustness of the economical model. However, assumptions are always difficult to play with, and quickly lead in to areas where we are simply wasting time discussing implausible or moot concepts. (It’s always a risk, not a reason to stop exploring).

@dirvine, I think it will be possible to simulate the economics model, expose it to many many different environmental configurations, and learn more about its behaviour. It will give directions on how to design it. I am sure that has always been intended, just lifting it up as something that I think can be done in so many ways, that it will need extra care.

But the behaviour of people, the impact on society of new ideas… The likelyhood of a new idea to grow foothold and new ways to do things to evolve…

Much of what we are discussing now comes down to behaviour of people.

It is very hard to simulate or predict. Much of it can only be based on an idea that it is possible. Things that have never been done before and which relate to new societal behaviour.
Ok, time for my first cup of coffee… :smile:

1 Like
113

I’m not lowering it i feel, like i said, you own your data, but it’s governed by a ‘smart-contract’.

What exactly do you define as the point and philosophy to this project? In general, at least to be sustainable with our currently understanding and business models, which i am using to make predictions and assumptions, Data, if can be accessed infinitely, should have either a timeframe linked to it(GBH) or a bandwidth limit linked to it.

1 Like
114

Nope, as my son explained, it is as easy as turning off the feature. Common knowledge and used by the majority who only download the file they want and the app automatically exits and no uploads.

So yes bittorrents shows there are plenty of people who willingly allow uploads and are happy to do it for free.

SAFE will use spare resources like bittorrent, and reward the farmer as well. This is why SAFE can succeed where a datacentre model (like the cloud storage) cannot.

You contend rental is needed. I contend that rental will mean files only last as long as the first rental period is. And you lose a layer of anonymity and security, you no longer own your data since the network can delete it on you. You lose a innovative concept to rival the cloud storage systems which WILL DEFEAT SAFE if we adopt their model. Just on speed alone would they defeat us.

SAFE would of necessity become a good cloud storage system with little to distinguish it from them other than the safe pages.

  • Data ownership - cloud storage says you own it like you say. But in fact cloud own it and at best SAFE network owns it since both can delete the data. The “owner” is simply the person the system charges rent to, and gives the initial access to. (immutable data)
  • Rental - SAFE == Cloud.
  • Access - SAFE == Cloud
  • Security - even though SAFE lost some with rental it is still marginally better than cloud
  • Anonymity - About the same now as Cloud accessed though VPN. SAFE loses a lot of anonymity with rental
  • Fee collection - SAFE dramatically harder since each object has a rental due data. Cloud just charges a fixed amount for upto storage.
  • Personal Data collection - SAFE a little better than cloud since it won’t be selling your data to advertisers. SAFE will have to collect personal data to be able to track your chunks and MDs and know who to charge rental and when.

In essence SAFE only becomes a better cloud storage provider under the rental system because of how many features it loses or compromises on to implement a rental model. Plus safesites will be a lot worse than the internet sites because under a user pays rental you will see a lot of safesite data lost because ordinary users will NOT pay rent on year old comments they made. SAFEsites will have to go to a site owned data model in order to maintain any resemblance of a complete site and thus another layer of anonymity and data ownership is lost.

In my opinion we end up with a cloud storage solution and a cripple internet lookalike. Please consider the ramifications of what is needed for rentals and the ongoing effect charging ordinary people rental on their data.

In my opinion your contention that farmers will have to pay out big money to run a vault is misguided at best and incorrect. With the numbers in the world able to run vaults on very high caps or unlimited running into millions upon millions, we can have a viable network without anyone paying for bandwidth. And the costs of running a home vault is extremely small and any comparison to bitcoin mining is bad since mining costs a lot to run a rig. SAFE farming will cost at best a couple of cents a day with a dedicated rig and this includes replacing the rig+HD every year or two

No you cannot. You might have your data deleted on you because you forgot to pay rent or were laid up in hospital, or otherwise incapable.

2 Likes
115

As @neo already responded to that, and I emphasise it:

And this also in great part responds to this question:

1 Like
116

Thinking on this, i think i should introduce a new concept, Gigabyte-gigabyte(GBGB) So for one GBGB you buy you can store 1GB of data AND you have one GB bandwidth when you want to retrieve it. After that, the data is still there and you can still retrieve it, you just need to pay some more safecoins to do so. The data is still there and so you still fully own it!

And why don’t we make it such that, people can choose which one they want, whether it’ll be better for them to store it perpetually and pay when accessing, or if it’s a more popular data, they might want to use the GBH model.

The point of all this is though, to peg the data to something, either time or bandwidth, rather than nothing. Because the idea of paying ONE OFF to have unlimited time and unlimited bandwidth to access the data is really, i’d say, kind of revolutionary, and it may work, but from all the models ever existed, it’s fair to doubt there are some sustainability or scalability issues as the money that would essentially ‘fund’ all this is the demand for storing new data on the network. But you have to realise UNlimited means infinite, however we have a FINITE(hence limited) demand for storing new data. So as the demand for accessing the existing data gets more and more, the limited demand can’t sustain the unlimited demand. So the unlimited demand will be limited by how much limited demand there is to store new data on the safenetwork… Do you understand what i mean now? So the demand for accessing data on the network can potentially be unlimited, people can just keep on visiting the site, refreshing every second, but the demand for storing new data on the safenetwork, since it’s paid, is limited, hence this raises some concerns… you see.

117

I totally see what you are saying.

The assumptions you make on how it would play out, what usage patterns we would see, the ratios, probabilities etc., simply I do not see them as that well defined and certain as you seem to do.
That gives another weight to the problem than what you currently feel.
I think those fundamental assumptions are the key to make meaningful estimations that guide the design.
They are many and hard to make, so require some effort.

1 Like
118

Your son? Is that oetyng?

And ok, so you own your data, would the GBGB concept solve that though? You still get to own your data, but will pay for further access after you’ve accessed it a certain number of bandwidth units already.

1 Like
119

I have a new father :smile:
(EDIT: I must find that line of passage, but no, we are not related, as far as I know :smile: )

2 Likes
120

Come on this is a discussion not a trolling session.

You have now ignored the rest of my reasoning. Oh well.

Have you considered the effects of caching for your poplar data.

And if the farmers are not paying for their uploading to the network, why is there a need for special systems to compensate them?

The other problem is that you again remove a layer of anonymity/security to be able to tie data to owner and vault.

Pay once forever data solved a lot of anonymity and security issues in one innovative way. It reduces the modules, coding, and thus bugs in the network and allowed a economic model to rival the rental model and to allow SAFE to transcend the current non-owned data of cloud storage.

4 Likes