Using the SAFE network without money

145

Your tone is sarcastic, which is rude IMJ. If you are being honest in your request for an explanation, you would not I think be sarcastic.

The lack of a fee doesn’t make it free, so this is incorrect - there is a cost in resources. I wasn’t seeing the bitcoin fee as significant - even if there was no bitcoin fee - I think it would still be impractical to overwhelm the network with transaction spam.

This is no more reliable than the guess that @anders makes, and which you think it proves false. Both are feasible. The fact is, bitcoin seems to be surviving this. You might then argue that this is because of a the very small fee charged on a transaction, in which case SAFE could adopt the same approach, rather than balancing the charges as presently proposed. This would then undermine the proposition of FREE, but it is speculation. So for example, the fee could initially be set at zero, so FREE and algorithmically increase if needed to stabilise the network, and then drop back to zero.

You see, making rude, off-hand criticisms obscures the debate. Its not about winning an argument here. At least that’s not the purpose of these discussions for those who believe in the goals of SAFE (Secure Access For Everyone) and wish the project to succeed. The object for us is to explore all the avenues and try to improve the solution. Yes, I’ve been rude too at times. I’m not pretending no-one should ever do so, but lets learn and improve the project by improving the quality of the debate, and no doubt ourselves.

I don’t see that one bot could overwhelm a large network, but enough surely could. So this is a valid attack on this scheme - but it doesn’t kill it because it is unquantified, and because of that, we can’t examine it properly. So if you want to explore this point, rather than just try and kill the discussion, please do some calculations that show that one bot, or N-bots, could produce and store sufficient data to overwhelm a network of given size and growing at various rates.

I don’t see how this undermines even the early ideas @Anders has suggested, and it certainly doesn’t mean they aren’t worthy of exploration. You seem intent on killing the discussion. Why? Just to “win”? Or because you want to stifle the discussion here, or discourage creative ideas? I’ve answered enough - I’d like to hear your responses to this, and why you are taking this “kill” approach before I continue.

1 Like
146

It won’t be counterbalanced on the long run. For a little while, sure. But the problem with your idea is that at times when SAFE usage is hardly increasing or even decreasing, the expansion of the money supply won’t slow down at all. That’s when we’d hit an inflationary spiral, because everyone holding and earning SafeCoin will see the devaluation happening, and knowing that it won’t stop they’ll sell them en masse. Because SafeCoins are dropping in value the network will have to increase the reward rate to keep farmers supplying resources, further exacerbating inflation. Then it won’t be long until billions of new SafeCoins are issued daily.

The main reason a fiat currency can get away with perpetual moderate levels of inflation is because it’s status as legal tender is enforced by the government, which anchors it as the default currency. A voluntary crypto-currency with perpetual inflation is doomed to hyper-inflation/abandonment.

3 Likes
147

That’s at an individual rate. But what about other parties that want to disrupt SAFE, like governments. The NSA? Surely they will be capable of massive parallelization of the ‘flood’ attack?

148

To anyone who knows:
It was my understanding that previously SAFE was to have a cap on the total amount of Safecoin to be created. Is this still the case, or, because the current plan is to consume coin with PUTs, has the cap idea been removed?

1 Like
149

Do you calculate the cost?

A cheap 1TB. hard drive is about 40$. As there are a minimum of 4 copies, a single attacker with a 100 Mbit/s upload speed can destroy 160$/day, 4800$/Month or 58.400$/year only in hard drives.

Rent a botnet of several thousand computer and, for a few dollars, you can fill with garbage millions of dollars in equipment.

Destroy the SAFE network will be accessible to any group of people, company or government that wanted to.

1 Like
150

No, it wasn´t sarcastic. It was over the top, same as “excellent and clear case” was over the top. I was really hoping for some insightful comments…it really depends on the way how you want to read it.

The whole thread was about “using the SAFE network without money”. Should be discuss that? It´s in the freakin title… There was no debate about “no transaction fees” there were debates about free storage and about the question whether “both private and public SAFE data storage can be made totally free for all users”. To me it seems you want to discuss different issues, which is fine, but that´s not what I commented on.

And that´s also why my reference to one bot gaming the whole system isn´t a mere guess. If space is accessible without restriction then writing an exploit is not a question of time but a question of decision. There is absolutely no need to calculate that. If space is freely accesible in a network where you are not clearly identifyable as you are in a centralized network why should I make up a complex scenario. I just write one bot that creates garbage and stores it of freely accesible space. The farmer reward inversely decreases with the ratio of spam. And you are right “You might then argue that this is because of a the very small fee charged on a transaction in which case SAFE could adopt the same approach, rather than balancing the charges as presently proposed.” however, point is: that is an argument which YOU introduce - before we were talking about using SAFE “without the hassle of having to deal with safecoins” and the given premise that “money is the hallmark of a primitive society”. You made a comment against my style of argumentation, now you are discussing with different arguments then the ones that I rebutted. If something obscures the debate, then this kind of behaviour.

The fact that you don´t see it, doesn´t mean they don´t exist. A lot of premises were plainly wrong and the statement that Bitcoin can be created out of thin air is just one of them. I already refered to several, I don´t want to repeat them, but it´s still in the thread if it´s of any interest.

I never said that the early ideas aren´t worthy of exploration (where? please quote!) When I commented that there was not much substance I didn´t refer to the initial idea but about the particular argument. Axiomatic arguments based on false premises is a really common thing. Sometimes there is not much more to say then pointing out what it is (in my humble opinion). Apparently this is not en vogue in your opinion.

The question why I argue as I do is a metadiscussion, it really doesn´t belong here. However, as you ask me I can tell you quite clearly why I challenged @Anders to bet and that´s because I have a problem with the rhetoric and I found it misleading and counterproductive. The whole argument is made up based on premises such as “There will be more data resources available than the users will be able to use.” Which is where we get back to “killing the discussion”, because this precisely what´s called a killer phrase. There is really not a lot of alternatives responding to it other than killing it. Inventing premises is not at all constructive for any discussion and I don´t see why this should be considered as creative ideas when it actually predetermines the scenario.

And again: I find it quite ironic that you come along claiming a violation of guidelines and then give an off-topic “hit em in da face” post a “like”. Happens that when you point at someone, three fingers point back at you…

151

There was a thread back in August and I don´t see that the cap idea has been removed.

152

@Anders

Right, so if there is a cap, and there is no cost to the users for a PUT, then how exactly would the coins be paid out to farmers based on GETs, when we reach the network cap? Eventually all of the coins the network could issue would be exhausted, and so if the coins cannot be ‘burnt’ by a PUT request, then we would have to exceed the network cap in order for farmers to be paid. Exceeding the cap would, by definition, be inflationary.

As far as I understood it, when the network’s cap on SAFEcoin has been reached, the coin spent on PUT requests would fund the farmers, original creators, developers, and offer rewards for original content. If there is a network cap, but no payment of SAFEcoin for storage, then the coin never re-enters the network, and so these parties will cease to be paid (again, to @Anders).

I’ve often thought that this is the inherent weakness of Bitcoin. When Bitcoin reaches it’s network cap, the only mechanism built in to pay miners is the optional fee. Mining will become very unprofitable at that point, am I correct? Or do they have a mechanism to increase the fee later?

2 Likes
153

The thought experiment was that the cap would be removed.

The actual proposal for safecoin is that users pay to PUT (the coin is returned to the network) And farmers are paid to farm. That way the coin has real value based in resources and not a pretend value that is not worth the paper its written on.

[EDIT] The system dynamically adjusts how much resource one coin can buy, and how much is paid on GETs to ensure that it remains economically viable and encourage farmers to continue

Is the fee not set by the senders and the people processing the transaction decide if the fee is enough. (that is they pick the profitable transactions to process). Maybe someone with better knowledge can explain it better

2 Likes
154

@neo Agreed. If the network cap is removed, I would no longer support the network, as sound economics shows that inflation and valueless coin would be the result. I’m a little surprised that this stuff still gets proposed here, given that I thought fiat currency was what we were all trying to get away from…

I’m not sure at all. But that way of doing things has potential. Another thought I had about Bitcoin is that for it to be stable, it needs the faith of the people using it, which is almost as flimsy as the ‘faith and credit of the US government’. Well, more flimsy actually.

What attracted me to the SAFE network was that the coin is based on real world goods: storage space (not to mention data security, privacy etc). The key point was that real world constraints are placed on the issuance of the currency, and that it would adapt to supply and demand. Break that connection, and then what’s the point of SAFEcoin?

3 Likes
155

As far as I can tell the only way that we can see moneyless (safecoin) SAFE is that we reach an era when all (real) needs of humans is being met without the need for money. This would include storage&bandwidth which by then would be considerable by today’s standards.

Until those times and while bandwidth is expensive and storage costs are significant we will have to place a value on the network resources that at least reflects some of the real world costs with an edge to be an incentive to participate in the network/farming.

In today’s conditions and the foreseeable future we have to have a realistic value on PUTing data and farming the data. And under SAFE’s model reading data from the network is free to the person getting the data. PUTs pay the network to store data and the network pays the farmer to store/supply the data

[EDIT this is not just a reply to @Team_2E16, but an attempt to also answer the OP]

4 Likes
156

That’s only a valid argument if the generation of safecoins is linear. Bitcoin has a halving of the miner reward over time. The SAFE network can use a similar anti-inflationary function.

157

The bandwidth cost of uploading huge amounts of data is significant. Even a spammer needs to choose what to spend the bandwidth (and time) on.

My idea is that farming safecoins will be much more popular than mining other cryptocurrencies. So there will be an excess of SAFE storage capacity that can handle spam and attacks like that.

1 Like
158

You misunderstand how botnets can be used.

EMAIL spam requires many sources and little bandwidth & cpu for each machine
upload attack to SAFE on the “free” model requires many machines for large bandwidth and little cpu to produce random data.

Both can occur at the same time on a botnet. if only 10,000 machines and ave of 25MBits/sec (10GB/hr) means that 100TB/hour can be uploaded to SAFE network for this small botnet. Thats 2.4PB/day

Now a large botnet >100,000 machines would be 24PB/day. And these botnets can go for days at a time.

Please make the case why people would farm beyond those who want safe/secure/etc network storage.

3 Likes
159

That’s essentially what it comes down to but it’s a bit more complicated than that. Default settings in bitcoin qt set aside fixed space in a block for free “priority” transactions and the rest of the space requires fees for txs. These low-priority txs are sorted by fee amount. So if you put in zero fee and you have a low-priority tx, it might take a long time for it to be included in a block. That said, many settings can be adjusted by miners.

Here’s a good overview http://bitcoinfees.com/

It seems like there could be data privacy & security issues if nodes in SAFE could set fees for storage. If nodes could reject chunks based on price, then you might end up with a small group of nodes willing to store chunks for free/cheap removing the randomness of where chunks are stored and making censorship easier.

5 Likes
160

Interesting. Obviously, if it could be done without compromising the networks security/privacy, it would be ideal (IMO) for PUTters to be able to set the price they wish to pay, and farmers to set the prices they wish to earn. Would it ever be possible to set criteria to avoid this ‘de-randomization’ effect while allowing direct supply and demand to operate between users/farmers? For example, if the randomization would be compromised, the network could reject the PUT request? Of course, this would create a PITA for users, and it is true that the cheapest nodes would always be filled first, skewing things in one direction. But wouldn’t they fill up quickly, and encourage others to drop their prices to earn more coin?

1 Like
161

BTW: I had no intention of suggesting this and agree completely with your reasoning on why it would be a problem.

Thanks for the info on bitcoin fees

1 Like
162

doesn’t the sia network have something like the users setting the storage costs? Or is it network determined.

163

Yes it does, but it didn’t work well (or rather, at all) last time I tried which was couple of weeks ago.
Note that it’s a different thing, Sia. It’s not meant to serve files to the world, but to the user. Yeah, they have a bigger vision for the future, but at the moment it’s largely a sort of P2P thing where I put my file on your disk and you keep it for me. And even for that limited use case, I don’t think they’ll have anything usable this year.

1 Like
164

Farmers can be rewarded both for GETs and PUTs in proportion to how much they contribute. This will inflate the total amount of safecoins in the system. In order to keep the inflation low, a declining farming reward can be used. This is similar to Bitcoin. However, in Bitcoin there is a definite limit to how many bitcoins will be generated. Instead of a definite limit like that safecoins can continue to be generated with a reward rate that is adjusted over time.