The Perpetual Web?

The Safe Network is designed to ensure that public data is permanent and can never be deleted.

I don’t understand how the data can be permanent. If you exchange data space for tokens, how can you prevent the data to be deleted after the tokens are spent?
Then, if you have to create/buy new space to keep the same data, wouldn’t you engage in a never ending token emission?

The network is designed never to lose data so when you pay to upload you are paying for it to be stored forever.

This cost declines over time (as storage becomes cheaper) so it doesn’t just increase indefinitely but converges on a figure (per MB) that gets built in to the storage cost at the time.

This cost in turn adjusts to ensure nodes are rewarded adequately for the resources they contribute:

  • more demand than storage, cost goes up.
  • more storage than demand, cost goes down.

Did that make sense?


Not really…
Lets take an hypothetical situation where demand meets offer, just to leave that variable out of the equation.

That is impossible, how can you pay for an infinite service?
At most, newer storage being bought will be subsidizing older storage.

You would need a constant (perfect) equilibrium of technology making the cost of mb per dollar cheaper at the exact same rate of network size increase. If the network increases at an accelerated rate, new users will be “punished” and would be buying expensive storage to pay for the older data and vice versa.

From an economic viewpoint, it doesn’t make sense… Well, it does but looks awfully similar to a piramidal scheme or the inverse of that…

Also, if the network is descentralized, how is it able to correlate with the changing price in mb/$ ?


Yes, I oversimplified but the net effect is that people are paying for what they get. Sometimes less than the cost, sometimes more, but which will vary depending on the timescale you choose to measure over.

Sections will adjust cost, so it will fluctuate as the capacity and demand fluctuate. Not every chunk will therefore cost the same across the entire network because different chunks may be destined for different sections.

There’s a lot about the impossible network that’s different from our experience and way of thinking etc so it’s not surprising for us to have trouble accepting the design can work. I do understand your criticisms and nobody here knows for sure or can prove it one way or the other without trying. Economics will be an interesting part of the later test networks.

The point is that’s the aim and if it doesn’t work out can be revised, but I think it will work.

People in general need to buy into the idea it can work and I think most will because they don’t try to understand the fine details. Those who don’t won’t use it.

The reason I think is that people will want a guarantee of permanence and won’t mind paying a bit more in that expectation - or hope - they’ll get that.

We know that nothing is 100% guaranteed after all, but something that aims high may command a premium over a system that promises this but only if you also pay again and again forever - which nobody can do in practice!

You point out that storage costs may rise in time as late uploaders pay for data that is already stored as well as their upload.

I think that makes sense but is not a problem. In fact it makes sense for early adopters to get a cheaper rate don’t you think.

The degree of premium and discount will depend on current rate of upload versus size of legacy data and the current cost of storage, the premium you pay when you upload may look like a discount if the upload cost continues to rise. An incentive, looking back a discount for taking risk and uploading sooner rather than later?

Interesting thought experiments to have!


There is something on the system that I don’t like, but I don’t know what yet :stuck_out_tongue:
I think it is the “permanent storage” option. I am trying to wrap my head around how a descentralized system like this would work, but it is hard because there are too many variables in place.

In fact, I still don’t know if it is a technical or economic problem (or maybe both) but my gut tells me that there is something wrong.
Will have to think about this.


I’m just rehashing what @happybeing already said I think, but simply:

The netowrk just looks at the supply of storage and pays more when supply decreses and requires more payment from users at the same time.

There can be an issue where there isn’t enough in the buffer in the long run and there aren’t enough buyers of new storage (e.g. if the cost goes too high) … so there is a real possibility of the network collapsing - but this is no different than any economy. For example right now as oil and gas are being curtailed in the west we are seeing economic collapse - kinda need cheap energy to keep growing.

Also Safe Network isn’t like a ponzi. A ponzi usually revolves around an investment, not the cost of a service or consumable good.

We also have to compare the cost of storage on Safe with comparable storage services elsewhere. While permanent storage is rare (there are some other’s who are offering it I believe), there is also the security of Safe (well beyond the security of other storage providers I believe) and there is also deduplication which should reduce cost for storage in the longer run for the globe.

Yes, future users may from time to time end up subsidizing early users, but the inverse could also happen as well if storage costs go down in the long run as we would expect them to in a functional economy.


By the way if I upload something that it woudlnt want to be here and I nobody send link to this file, can somebody download or look on this file? Thx

Nope. If you don’t share it, nobody will see it. And if they came across the XOR URL accidentally (impossible really) then they’d not be able to decrypt it.


Please do, but don’t use your gut. The economic model is sound. It’s simply supply vs demand for a basket of commodities (cpu, HDD, bandwidth, electricity, physical rental space, etc).


In fact we are more looking at fixed time scale till a new system emerges to replace SAFE. Maybe 20, 30 years or so since technology/networks/systems in this arena are evolving over time.

Now that isn’t to say SAFE is relying on that for the economic model, but it does account for infinity situation.

Its actually related to the very old thought experiment of the frog that hops half the distance to the wall on each hop. The distances when put into a sequence are 1 + 1/2 + 1/4 + 1/8 + 1/16 + 1/32 + 1/64 + …

It was not possible at the time to calculate the infinite series and most considered it to be a forever increasing number.

When the solution was formally devised the result is approaching 2.

So how does this apply to the infinite storage economics problem?

Well we have to made an assumption that has held true since the invention of the rotating disk media. And that is the single drive capacity has increased by around 10 times every 5 years barring 1 or two technology problems that slightly reduced that. Now we have SSD technology increasing single drive capacity by a much greater rate. (EG one time case of max 8TB to 100 TB in 2 years.) and expected to continue at an average rate greater than 10 times every 5 years. This is a certainty for the next 20 years as we know technically how to do it and costs is the only factor keeping the rate to around 10 times in 5 years. Additionally the number of drives/units manufactured is also increasing each year by large amounts.

The other assumption which also has held true for all this period and into the foreseeable future and that the cost per MB/GB/TB/PB will continue to drop at similar rates. This is the important figure for storage costs in Safe. I don’t have the accurate rate of dropping at hand but its large.

Using these 2 assumptions which technically will hold true for 20-30 years minimum (barring “acts of god” or war) we can be applying geometric series for cost per MB/GB/TB/PB dropping and realise there is a determinable cost per MB for “infinite” storage. For instance if price dropped by half every 2 years then only need to charge the double the cost to store the chunk for 2 years and you can store it for infinity.

As long as the economic model for charging to store uses the value from the geometric series as a multiplying factor then yes we can confidently say data can be stored infinitely.

Obviously there are scenarios were this will not hold true and most of them would also see the world’s data storage at risk including SAFE. For instance global war, earth impacted by very large meteorite, global power loss, manufacturing collapses in various sectors and so on. These of course are extreme and affect any storage of data and costs for data storage.


Won’t a perpetual web mean that illegal content such as classified military documents, revenge porn and child pornography will be accessible for all forever. Apart from the moral and ethical implications of creating a perpetual web, won’t the safe network be open to legal challenges from the govnt and law enforcement. How do safe network developers plan to address the legal and ethical implications of a perpetual web? (eg victims will have no recourse to stop abuse) I have concerns that it will will create a monster that can’t be put back in the box. What are the dangers here? Have they been thought through?

This might be of interest for you.

You may need to get to trust level 1 to be able to read it, but that only takes reading the forum for a bit.

1 Like

I think those are valid concerns when thinking about this using current dominant mindset, which is one of trusting those with power to look after things like censorship and not abuse that power. The reality is what drives those building Safe Network I think, the reality that those with power do abuse it and that we see this has increased steadily with centralisation and increasingly authoritarian surveillance, control and censorship.

So an alternative mindset is to say, all content is allowed to exist. Individuals then have the power to choose who is in charge of their personal censorship or filtering, whether government, a community, a friend or to eschew censorship altogether. I think most of us will select apps with some kind of filtering or censorship, but with the option to tweak, tune or turn that off if it gets in the way, or to switch to something that suits our needs better.

The second mindset is very new to most of us and it will feel different which can make the change hard to accept and it will no doubt face resistance. It may also garner support, and provide people with an alternative to what is increasingly toxic to individuals and communities, and incompatible with a world where individuals have the right to communicate with anyone and everyone rather than only with the permission of gatekeepers, billionaires and governments.

Whether Safe Network can survive in this environment is unclear. Maybe it will be niche like Tor, but incredibly useful nevertheless. Or maybe it will become accepted and help to push back against the abuse of those with centralised power.

I’m sure MaidSafe and others in the community think about these kind of issues, but nobody can know exactly what is the best or how things will turn out, so it is also an experiment, but an important one if you look around and see how a billionaire can now just buy up the closest thing we had to a public square while claiming to be a free speech absolutist, and still be cheered on by his fans as he arbitrarily censors people he doesn’t like.

What Must did with Twitter is show a lot more people that the status quo has been broken and getting worse for a long time, and Safe Network is just one of the projects which arises out of that recognition. I’m sure there will be others, and I hope together we can move things in a better direction.

Nothing is perfect, there is only an illusion of perfection and it is useful to look behind that, behind what people resisting change say and why they are saying it. I’m sure you can think of disingenuous examples that tap into people’s real fears and concerns, but which don’t address them and instead have both intended and unintended negative consequences for the population.


Thanks for the reply but this doesn’t really answer the question I asked. I already understand the positive things decentralized networks have over centralized ones like Twitter, and much of what you describe is already acheived by mastodon re censorship/filtering. what I’m asking is about the perpetual web in relation to laws (which are different to the whims of corporate censorship) . It is not illegal to censor people for their political opinions as a private company (however bad we probably agree that can be) but it is always illegal to share child pornography. safenetwork will be the only network ever invented where it will be impossible to remove such images EVEN IF we know exactly who published them. Because as you said they will be permanent. This is differnt philosophical argument about whether one should or should not censor things. Even on tor illegal websites can be taken down by law enforcement (even if its difficult). Ae you sure you want this because it could unleash a monster on the world and there may be no going back? I don’t see much thought or discussion about this re perpetual web. We only hear about the benefits, but the negatives are unique and potentially quite frightening because they go beyond discussions about protecting censorship and into discussioons about the ethics of protecting active criminality even if such criminals are caught and prosecuted.

1 Like

On normal (dark) internet is also this ilegal activity

a) that content can be removed by law enforcers (if they are prosecuted and
b) the goal of SAFE is (presumably) to allow joe public to access and create content anonymously, without censorship but also in a way that is more easy than the dark web you describe (otherwise why bother with it all? )
c) as I said, the PERPETUAL web is not a feature of the dark web. It is this specific feature unique to the SAFE network that I am concerned about . Making everyone anonymous and unaccountable easily, means that even if they are caught and prosecuted for crimes of producing the content, those crimes will still be there AS CONTENT to share forever without risk.

The response from many on the forums are similar to this: ’
the simple matter is that nearly all technology from fire, through metal making, though roads, through science, through computing/electronic communications has enable both the evil and the good in this world. ’
But this is an insufficient analogy, because all the evil committed by these things CAN be stopped or fought against through laws etc. The evil committed by anonymous people posting content on the safe network that will remain there to be shared in perpetuityy is unique and cannot be fought against in any way without shutting down the network altogether (and that is almost impossible too once the cats out of the bag) . Are you OK with child porn being available easily, freely and forever (its not just that there will be no consequence to anyone who creates and distributes its also about how normalised it will seem in a society that can do nothing to stop it) and is that justified just because on the other side of the argument it can help someone blow the whistle about nefarious CIA activities or whatever. I’m not sure the knowing of one evil cancels out the committing of another evil.

Not if they can’t find the server -as is the case often enough with Tor hidden services.

The real issue is access not what is in the perpetual web. Also, as @neo pointed out earlier Safe many only be around for 30 years or so until something replaces it … now maybe all data will be replicated onto some new network, but honestly there are bigger problems in the world than some data that few can even find.

So as for whether it’s worth it or not, then 100% it’s worth it IMO. And it’s being built. So more important to concern ourselves with access and finding data, not the existence of data.

Being able to share it means that people have to know where it is. If indexing servers on Safe decide not to show it’s location, then data effectively disappears from non-dark-web services over time - particularly if deemed socially unacceptable data. But Safe will have a dark web side too and there will be indexing services hiding about that will probably provide links to such data just as they do today on Tor, I2P, Freenet, and others.

So I think Safe really changes little in regards to socially unacceptable data in the grand scheme of things. The major social media services on any non-dark-web platform are going to hide socially unacceptable data as they do today.

The real can of worms as always is who gets to decide what is socially unacceptable data … I’d prefer a decentralized approach where you can choose your “data indexing provider” for yourself and not rely upon a single centralized provider.

1 Like

often but not always! Unlike with SAFE.

Access is more important, I agree, but that access can be made good and easy by any nefarious person who wants to. There is no way to control access either. As someone pointed out elsewhere on thsi sight someone could take the domain safe/Disney and put child pornography there. Is the SAFE network designed to be used by everyone (including children?) or not? If the goal is to replace the traditional internet then we still have the problem with access as much as permanence. People could use DNS filters/browser addons but that kind of takes us back to square one: an internet filtered by a central source as we have now with google.

But who would decide not to index it? Who makes that decision as to what is indexed and what isn’t? That is what google does already.

But its not like other dark web networks because the goal is to make it user friendly, permanent, accessible and 100% anonymous. So its like the dark web on steroids for everyone. That changes how much people are exposed to dodgy and illegal content. That will effect culture at large much more as much as it will effect the victims of crime and abuse.

Google does a horrible job of it. We need competition and that will bring with it answers to the demands of users.

I am hoping to fund the development of a social media system on Safe in the future where you can do just that.

1 Like