So if safecoin is a means to preserve data and safecoins are data, then why don’t safecoin deserve the same rights.
You contradict your own reasoning here.
The reason it is stolen is that you want it to decompose. There is NOTHING in SAFE to decompose data (INCLUDING SAFEcoins which are DATA) So you have to create the artificial process to to do this and thus you are stealing from those who are saving their coins.
And don’t use the argument they were told. Who reads non-existent T&Cs And even if you put in big writing when people create an account that their coins decompose when such and such happens unless you do something special. People will ignore it as “who would do such a thing - its stupid”. And those who do read it will either be able to keep up with all the artificial conditions imposed and then there will be those who ignore it, forget it, make a mistake that one time, and lose all their coins.
Doing an artificial so-called decomposition becomes theft. It will steal more coins than are actually lost because how many times have you perfectly jumps through hoops, stuck your toungue out the right way so that your coins are not lost. Yes I am using some sarcasm as this idea belongs where composition is done.
SAFE will be bypassed because safecoin are not reliable, likely to disappear if you do not do things perfectly or forget. So why trust any data as being reliable if one of the important pieces of data (the coins) is not reliable.
If I were to put a gateway across a road and then placed notice in a paper that unless you pass through the gateway across the particular road than your thugs (decomposers) will come and take your life savings. Would you consider that theft? How many people read that paper, had a car to drive through the gate, believed the stupid notice, or made a mistake of which road. Of course any sane person would. Doesn’t matter what you call yourself.
I’m not so much advocating for the status quo in this thread, but rather trying to draw on everyday natural phenomena as analogies that can be used to help solve an (admittedly unquantified/unacknowledged) deflationary issue at this point with regard safecoin evolution. I thought that this was the philosophy @dirvine wanted to promote within the development of this project? Many people opposed to the composting idea are in love with the idea of deflation. I called this “blockchain bias”. Deflation is fine if prices are falling due to technological improvements. For example, it now costs much less for a 1TB HDD compared to a what a 100MB HDD did 25 years ago, which is a wonderful advancement. Getting more PUTs for the same SC or {your native country’s currency}… now that’s good deflation. However, there is no point in celebrating the fact that the currency in your wallet is worth more today because last night half your city burnt down in riots and many people lost all of theirs in the fire. Doesn’t that scenario, as well as all typical blockchains by analogy, just a case of celebrating someone else’s misfortune?
Yes, and they deserve to lose their life savings. Same concept as having a yard with a dog in it with a sign saying “Beware of Dog.” If they don’t respect the sign and get attacked it’s their own damn fault. Who cares if their stupid. We need less stupid people in the world. If you go skiiing and don’t respect the signs and end up in an avalanche what are they going to say? You were a dumbass for not obeying the signs and staying in the safe areas. If we set conditions that people know about and tell them explicitly that their coins will expire after a set period of time if they don’t do anything with them then it’s their own fault if they lose their money for not listening to the warnings.
Something that lasts forever doesn’t equal sustainable. A nuclear reactor lasts millions of years but it isn’t exactly safe to be around. Plastic takes millions of years to break down but that’s precisely WHY it’s so ecologically unfriendly. Take a look at the great garbage patch sometime. Something that is “reliable” doesn’t equal sustainable. Or perhaps one should look for reliability in something that IS sustainable and renewable rather than something that lasts forever.
Your argument is largely an appeal to emotion. People will feel like SAFE isn’t reliable because they feel like safecoin isn’t reliable. They will make an emotional decision rather than look at the technical reasoning behind it. Much like you are reacting emotionally and totally ignoring the security concern I brought up.
But do all the decomposed coins go to my personal wallet? No. The whole network is artificial. It’s a COMPUTER PROGRAM!
Those who would take property that rightfully belongs to another without that person’s permission.
You are just playing word games here and it is dishonest. The way you’ve defined “decomposition” is no different than outright theft, that is, you (or the network) are taking that which does not belong to you from another person who earned it. The idea that any data should decay within theSAFE framework is utter BS. SAFE coin is supposed to be just another form of data.
Why is that? Please justify your assumption. Also, your argument here makes no sense. If you were the only farmer, the coins would have no value. Who would you trade them with? Excusing this and working from your example, why is the ease of acquiring a new safe coin increased just because a coin expired? One coin should be just as easy to acquire as the next. The loss occurs with the expiration of the coin. It doesn’t matter whether the network or another person is doing the taking, it is still theft.
Yes we are. As consumers, human beings are greatly benefited by deflation and shrinking costs. I know that this idea is ananthema to most folks growing up under a central bank-controlled financial system where “stable” prices is defined as two percent annual growth.
Thanks for the added info. In your example, how many users/vaults were there? In other words, how many users per coin? Does that even matter to the network? During this thought experiment I am viewing safenet as a replacement for the oldnet, which means billions of vaults/users. I know you see it that way too, but did this simulation consider that scale of operations?
Do you mean that only 50% of the coins (2^31) will be free to farm?
How did these simulations account for divisibility? Are these actual simulations or “back of the envelope” estimates?
Lastly, please help me understand/confirm the following:
I thought that whenever a node or browser is connected to the network it is able to determine addresses of the safecoins it owns, or does some anonymous middle man in the network determine this from a hash when queried?
Is it the authenticator which acts as the means by which a client browser, farming node, or any other node in the network and maintain connectivity so that safecoins can be assigned to the correct public_id as necessary?
You forget that both deflation and inflation can can be considered good or bad depending on the context and type of currency model. Consider deflation caused by improved science, technology, manufacturing methods, and knowledge. Computers, cars, food, clothing all become less expensive because humanity has made progress, that’s good deflation. Take the alternative situation under the scenario of the classical gold standard. Let’s put half the world’s currency (gold bars) on a rocket ship and launch them into the sun. Did all the remaining gold left on earth become more valuable? Well, yes, but not because humanity is doing any better in that scenario… we just permanently lost half our gold! As a consequence we have less resources to do great things. This is bad deflation, and the politics of scarcity, which is essentially what one is promoting via blockchain deflationary theory.
The same goes for inflation. Debt based economies that use the printing press to increase the supply of debt backed currency dilutes the value of the supply already at play in the market. This can be seen as a hidden tax on the users, thus causing friction in the economic engine and stifling progress long term. Based on your post you likely would agree that this is “bad inflation”. However, the printers love it because it gives people motivation to work a lot and take risks since they are always falling behind and must find some way to get ahead before the debts come due. (From the printer’s perspective this is good inflation since without it they get no recurring benefits for owning the printing press. Please note that I am not advocating a printers benefit for safecoin) Going back to the analogy, consider the case where a new deposit of gold is found that increases the total world supply 200%. This is also a form of inflation, but who cares because it’s great that now we all have access to more gold! Gold is pretty, it doesn’t rust, makes nice space mirrors… now humanity can make better tools, science, and innovation ie. progress. This is good inflation (real growth) because we are better off now than before from the improved access to material resources.
I think its important to view inflation/deflation discussions from different perspectives, depending on whether you are using debt based currency vs. asset based currency. The tricky part in this discussion is that Safecoin is a “virtual asset” based currency or medium of exchange that only exists inside a computer software system. It is backed by access to network services. I believe that the store of value it provides, which investors typically are concerned with, is maximized when both inflation and deflation are kept in balance. This means getting rid of limits to growth and bottlenecks such that the supply in circulation is allowed to grow/inflate as necessary. This appears to be exactly what the proof-of-resource model allows with the use of farmers (ie. good inflation). But it also means that this resource must not be permanently destroyed over time either (ie. bad deflation). Thus came my suggestion to have composting nodes in addition to farmer, client/browser, and other nodes. The fairness of it comes from the fact that the network promises the maximum total quantity that will ever exist is fixed from the start and agrees that no one will be allowed to increase this maximum in order to cause bad inflation.
Yes but no one is TAKING anything. It’s being destroyed outright. So I say again who are you accusing?
I have no problem scavanging and salvaging something from the dead. The dead have no claim of ownership. If you can prove an account is alive then there is no fear of decay. The whole point of the decay rate is to return lost coins in dead or lost accounts. Furthermore the coins are not being taken, they are being destroyed. And all the user has to do to keep their money is prove that they are still alive and haven’t lost the account.
I was being deliberately simplistic here. I would have thought that was obvious.
When new resources become available on the network the network issues coins. The value safecoin to farmers encourages more farmers to devote resources to the network. If more farmers devote resources to the network then fresh coin is issued and circulated and people can aquire it easily and upload to the network. But if safecoin is hung up in an account somewhere that means it’s not being circulated on the network. Which means if x percentage of safecoins are hung up in lost accounts then the network can’t accurately calculate what to award farmers for the resources they are providing or charge people for uploads. And at first this could seem like a very small error of margin but since there is no mechanism to correct this, since “data lasts forever” but since there are a limited number of safecoins to ever be issued this will slowly grow over time. The percentage of lost safecoin will grow and grow as time progresses unless we figure out a way to return those lost safecoins back to the network. Infinite growth, or infinite preservation in this case, is not compatible with an ecosystem with a finite resource. Safecoin is a finite resource, that is one of the characteristics that give it value in the first place. Sometime that is finite must be recycled in a closed loop. If safecoin CAN be lost then there MUST be a way for it to either a) be retrieved or b) decompose by itself. It can’t just sit there without causing a system failure sooner or later.
Loss of an asset does not equate to theft. The natural decay of an asset doesn’t equate to it being stolen. The network is not a person, it’s an ecosystem.
We have the assumption that lost coins would become a problem.
Let’s try to prove that first, before we ask others to prove why any suggested solutions would not be good.
Finding a baseline
So, all of the argumentation is based on that eventually a problem will arise. Well, eventually is very relative, since we have everything from now to eternity to take into account. A very large part of that spectrum is not even relevant to humanity, since earth won’t support life (any life) for more than about 1-2 billion years from now.
In 7.59 billion years our sun will engulf the earth. We will be long gone at that time.
People of Space Colonization Belief can of course say that we will be living somewhere else at that time, but let’s draw a base line of plausibility, and we can indulge in our sci-fi fantasies in other threads.
There are various suggestions, but roughly 5 million years ago, we were separated from apes genetically.
In 1 billion years we will have possibly passed that amount of evolution 200 times over. Will we even be recognizable as human beings at that time? Eh, no.
Truth is, there is no human venture that is designing anything for that long, and that goes for SAFENetwork too.
What is a reasonable lifetime then?
Defaitists would say we will get into serious troubles within 100 years. Even 1000 years seems quite implausible with the sustainability levels we currently proliferate at.
Let’s go wild and assume that humanity will survive for 100.000 years. We will technically, with high probability, still be humans at that time (that’s roughly the time for genetic evolution, with some caveats).
So, we do not need to design anything to last longer than that. Honestly though, looking beyond a couple of hundred years, or maximum a couple of 1000 years, would easily be considered being far to optimistic with regards to human sustainability, and far too pessimistic with regards to human ingenuity… (i.e. we will have replaced this solution with something better).
So, we will probably not be using the same technology in 100k years. Let’s go really wild and say we can come up with nothing better for a whopping 1000 years. So, the network has to last for 1 millenia.
Let’s get down to business:
Q1: What rate of lost coins will the network be able to sustain, before we see any troubles? Q2: Will this point occur before the above estimated network lifetime of 1000* years? (*NB: huge margin used)
This is an equation based on estimations.
It is a sanity check, something that should be done before hours and days and weeks are spent discussing something. I.e finding out if the question is even relevant at all?
Divisibility is something that needs to be solved, how is another question.
How much divisibility is needed?
Well, we have the extreme case, where the entire world economy is based on safecoin.
The world GDP is $74 trillion footnote: a according to the World Bank in 2017. The max supply of safecoin with 10^11 base units (up to 11 places within the confines of int64) is 429.5 k quadrillion irvines. (1 safecoin = 1 billion irvines, 1 quadrillion = 1000.000 billion)
If safecoin is to back the world’s money, markets and real estate, not including debt nor derivatives, then the valuation of safecoin might be:
Global Stock Markets: $73 trillion
Global Broad Money: $90.4 trillion
Global Real Estate: $217 trillion
Total: $380.4 trillion footnote: b
With a $380.4 trillion economy and 429.5 k quadrillion irvines in existence there will be plenty to go around.
It would give ~1.129 million irvines per dollar, which would allow for 100ths of micro-dollar payments.
a The $74 Trillion Global Economy in One Chart:
b All of the World’s Money and Markets in One Visualization:
Bear with me just a little longer
How do we define “problems for the network”?
Well, if we want micro payments, let’s define it as the point when we can no longer do micro payments. Oops, I slipped here, this is not correct: When safecoin backs more than 10% of current world economy, we will see problems with doing micro-dollar payments. This is without any loss of coins.But read on, and you'll see when it actually does give problem:
So, how much will world economy grow in 1000 years?
Actually, this is limited by what the world can output and sustain.
If we assume that the earth’s limits are nonexistent problems, then we could have a growth of almost 100 times of our economy within 1000 years (roughly doubled every 150 years), before we see any problems with micro-dollar payments.
Well, let’s say this was due to general inexperience on account of humanity, after all, digital currency was kind of a new thing… Let’s say humanity has learned some from its mistakes, and let’s say we are a 100 times better prepared for this new paradigm now. We estimate a 0.2% lost over a 10 year period, (about 3% every 150 years).
That means that in 1000 years, we will have lost approximately 20% of all irvines.
No problem there, we would still have 343.6 k quadrillion irvines, letting us do 100ths of a micro-dollar transaction.
Let’s combine our findings.
With our economy doubling every 150 years, and a loss of 3% of irvines during the same time, we would still be able to do micro-dollar payments after 500 years.
Does it seem like an acceptable life span for the network, without even the slightest sign of problems?
(I was a bit tired when writing this, so some number might have slipped me, please correct me if so… )
The thought process was a new network at the start - maybe 10-100K nodes moving into a mature one later on.
No its doesn’t really matter much the size unless you go to extremes one way or the other.
The network rewards farmers and that is not dependent on the number of farmers really (except for unrealistic extremes)
That was the analysis at the time give or take. The figure is to give relative size rather than try and be absolutely accurate to the analysis and really is not critical to the network or thought process. Remember the farming rewards is not designed to give our billions of coins a year. It is a reward not a wage and meant to cover paying for near zero costs, so the reward rate is small actually and since MAID is already near 1$ even a small reward rate is looking very good for farmers.
The thought experiment did not consider divisibility
NOPE, it only knows it has a safecoin when someone requests for a safecoin to be transferred. Actually it doesn’t know what data objects it has till something requires it and the section uses its indexes to retrieve it or fail the request. Its like saying a database know it has x number of products. Nope not until someone requests that information that the database engine looks up its tables and extracts that information.
So the same for nodes, It can determine what data objects it holds by extracting the information. It does not keep that information in its procedural variables unless it is extracted when requested.
I said it this way because to know what safecoins are being held by the section the node has to scan its data objects for tag type==safecoin and then make a list. A list is not kept.
When a section receives a request to transfer a safecoin to a new owner it extracts the information of the supposed safecoin data object to see if it even exists and who owns it. then it processes the transfer of ownership.
Not farming node, since its a full node and the authenticator is for the client not nodes.
Your #2 point is confused so much I cannot work out where to fix it for you. Clients/authenticators are not used in the node software at all.
Nodes store data and like a database do not keep in working memory the data objects they hold. They access their indexes to extract the IDs of the data objects they hold. So a scan of data objects is needed to identify what they hold just like one would in a indexed database table.
The safecoin data object has as part of its data the owner of it.
This topic emerged from a question raised by @whiteoutmashups in the divisibility thread. When we were brainstorming divisibility @neo said that the max safecoin number of addresses are 2^32 or a uint32_max, period. No wiggle room, no 40 bits or 48 bits, or 64bits of an uint64 allowed. (Correct me here @neo so that I’m sure I understood what we were talking about.)
Small side note: total global money supply including derivatives is not really known and could be between $544 trillion to $1200 trillion or more, a few months ago… (Your reference b.)
Doesn’t this prove my point? The other perspective is that the bitcoin network is a 9 year old bank that has lost 20% of its reserves… “blockchain bias?”
I would have to respectfully disagree… when has this ever happened? Safecoin will be a new “blockchainless” crypto thing that is also a “fuel” for a creative online economy inside of a newnet that is competing with the oldnet. Lot’s of people will make mistakes, and lot’s of people will try to burn it down if they have the opportunity. I’ll grant you the fact that the MaidSafe devs are working hard to make things as easy as possible for people, but what is considered easy for some will be seen as impossibly complicated for others.
I think @neo had some good estimates stating 50% would be lost within 10 years. So continue that trend and see how many coins are left after 100 years of learning from mistakes (hint: about 0.1% ). Yes, that would be an extreme limit case, but just reduce your annual loss rate and extend the time line a bit to get the same result. Coupled with the fact that in 30 to 40 years you will have (let’s say 8) 8 billion people + cars + robots + iots online.
I think the terminology I used was wrong. I was calling any computer connected in the network a node. And viewing a node as having one of many possible personas. A client “node”, a farmer “node”, a manager “node”, a child “node”, and adult “node” etc. So my question was whether or not the authenticator code is common to the different nodes in order to get them connected. I had assumed that it was modular and the same code was reused for each persona to connect in to the safenet. Sounds like you answered my question though and that different node types have different specialized means/libraries for communicating and the authenticator is that just for the case of a client node. Right?
I think we are talking about different things. I was asking for clarification on how a client or farmer knows exactly which safecoins they own? I would assume this information is stored somewhere as a wallet on the network and accessed by the authenticator when the passphrase is entered?
Well the current proposal used a u64 to hold the address and is why some discussions were around having safecoin in a 36 or 40 bit address space and thus more coins and then change the 1:1 MIAD → SAFEcoin as a 1:16 or 1:256 conversion. This meant same value and more wriggle room for increasing fiat price of safecoin. (and a corresponding change to farming rate would have been needed too)
Not really. Because you have to look at the causes of those losses and how they are not relevant today. When BTC was 10 thousand BTC per pizza losses were massive.
While people did not understand the need to keep their private keys backed up they lost massive amounts of coins. This was a wide spread problem
Then there is the issue of the originator’s coins that while we know the address those coins are also considered lost. And they account for a large slice of the lost coins.
Unlike BTC where we can view the address where the lost coins are, under safe a network wide search of all address space can identify the coins that exist but cannot identify the lost ones.
So no, taking BTC is really not a good example. Under safe the education for keeping ones credentials backed up safely is more widespread now-a-days partly because of sites like google and facebook. Also SAFE itself will have so many educational videos that didn’t exist when BTC started and during the time when most coins were lost.
No that is misquoting.
I was looking at a very unrealistic example just to show how little effect losing coins will really have.
I’d say less than 10% and easily <5% truly lost for at least a decade.
If you take your much loser definition of lost then those hoarding and saving will be included and you may estimate it at 20-30% as supposedly lost. But that is false and what authorities do when they want to take from the people.
Ok, so we can blame you for this thread then You gave me such a hard time for wanting to simply increase the address space to use a 64bit integer and were quite firm about the fact that nothing more than 2^32 could be used in order to keep the investors happy. So I started thinking about other options…
Not really. What I wrote doesn’t actually prove much. Considering however, that it was one of the most thorough attempts here to estimate what numbers would be necessary (albeit with a bit of humour), I’d say we’re very far away from any certain conclusions.
Actually, I just wanted to run the thought all the way, to see what I would find in the end. I would have loved to find something as crushingly undeniable, as the amount of power needed to bruteforce SHA256.
(Something along the line of 4 billion galaxies each with 4 billion earths, each earth with 4 billion people each running a thousand times Google’s worth of servers, during roughly 37 times the age of universe. Neat huh?)
Nope a client is a client and a NODE is a node and the two only met in the requests being sent between them
The client is an interface between the person (&Apps) and the networks. Nodes have no interface to client s/w or APPs except by the requests sent over the “wire”
There is only one node type. It does many functions including running a vault.
Client and its authenticator is what the person runs so that he can browse or run APPs. No SAFE NODE operations at all.
While it is possible for the one PC to be running node software and client software they don’t actually communicate to each other except as if the node is out there on the internet.
I proposed a wallet data structure as part of the divisibility and something like that will be needed for the account to “know” what coins it has. It would store the coin addresses in some MDs handled by “wallet” APPs of some sort.
The authenticator itself is unlikely to know of your wallets or the coins in them. You will need to provide a wallet ID to anything that wants to operate on the coins with that wallet’s ID as the owner.
So while the address is stored in a larger integer as part of the safecoin data object address it doesn’t mean that the usage can be more than 32 bits.
Even though I do not think it is a good idea at all to decay safecoin (for all reasons brought forward), I do not think the subject of safecoin is problem free, far from it (which makes it all the more worthy of deep and thorough discussion).
Safecoin is a huge subject, and the problem space is large and unexplored.
I also want to add that I think it is very important that ideas can be tried openly. Let people think, let ideas be tried. We can’t always know where we will find new ideas, limiting ourselves is doing us great disfavour.
I see it as a sport to explore ideas with as little presumptions as possible, even though I know what I think from start.
So I do not support the notion of some that it would be bad for the community to discuss these kind of things. Actually, more appalling to me is to even hint at such.
However, obsessing with an idea and not taking impression from arguments (not saying this is the case here), will make people lose patience quickly, and rightly so.
)