I thought it was pretty clear how the composters allow for a measure of safecoin age, which hypothetically can be linked to real world time through the network composting rate. I’ll admit that my understanding of the inner workings of the safenetwork might be too naive at this point, but I think if you take a second look at the “The Composting Node” and “Network Theft Protection” post in response to @fergish located further up the thread you will get the initial concept. A notion of safecoin “health” or “age” can be determined based on the deteriorating actions of the composters. A few questions is to ask yourself in order to provide some real world analogy, “Q : How do you I know that I am getting older? A. Because my health is deteriorating.” “When will I reach my demise if I avoid major risks like bungy jumping and fast cars? A. When my health is null and void (ie health=0).” Whether you count an “age” value up from age=0 to death at age=age_max, or decrement a “health” value from health=health_max to death at health=0, the basic concept is the same.
With @neo s proposal you don’t need any composting /resurrecting /coin shuffling /coin theft…
The length of this topic proofs that this is a complex and difficult thing to built into safe safely and as @oetyng showed it is not necessary when choosing neos solution… So why would someone want it implemented?!
Just because
… How many zeros are shown is 100 percent irrelevant…
I think this adds much uncertainty to time. We need “real time” because people want to know real time. Not weird network time. I would not want killing coins, but to just send them to backup address, it is to much uncertainty. For killing coins, no way.
Still can’t see why recycle coins? If scarcity becomes too much, add 2 more bits. Easier, safer.
Agreed. But @neo stated that it was impossible to have a real timeshamp… so all this nonsense was to concoct a means to measure duration. That being said I think the I saw somewhere in the dev forum that they are considering adding the ability to have timestamps based on node consensus…
I suppose that’s fine too if one uses your 1-2-5 system. Easier and simpler yes, better ie. safer for the network long term… I’m unsure. But the whole premise I used was from @neo’s original comments was that adding bits is not allowed.
Adding bits is not for now. When many coins are lost or they are too valuable. Hypothetical future years later.
Not sure they can complain 10 years later. Not sure they could complain now. If they get equal value. Americans sue for everything. Ridiculous. Secret: Suing is not winning. But yes it is annoying.
@norimi
I would be interested in hearing your thoughts on the Bitcoin vs. Classic vs. Compost comparison I described in the “monster” post above. I was trying to point out that coins are always deteriorating, and there is always a chance that software can destroy coins regardless of which system you go with. Some threats are physical deterioration, or theft by an adversary, the other alternative is this virtual decay mechanism. What are the colloquial expressions? “better the devil you know than the devil you don’t…” I was trying to point out that perhaps allowing for virtual decay helps solve the issues and eliminate a lot of work that arises in order to protect against physical deterioration, which might also keep the network economics healthier at the same time by protecting against too much inflation or deflation…
Because the storage capacity of the network is potentially unlimited. Its that simple!
Monster post critic.
I think losing coins is no big problem. It is not “bad inflation” because it is deflation 
Less coins effects: a) Other coins are worth a little more. b) Farming scarcity is forever a little higher. Point A is okay. Point B may cause a problem. When lot of coins are lost. Needs simulation. But probably value goes up because scarcity makes harder to farm coins so coins are now more valuable.
Denominations. When restricts denomination range, scarcity is lower than classic safecoin. May cause problem. Needs simulation. If allow extending with bits, loss can be corrected when necessary. Can also add smaller and smaller denomination if coin value grows too much. Adding bits?! That drives the value of existing coins down, right? Yes. Because we have more coins and less scarcity. But not much down! They would be very low value coins. Yes, more chance to farm a coin, but the coin is less value! Those two balance each other. But needs simulation.
Note about “restricting range to eliminate lottery effect”: I don’t mean to restrict by node’s farming rate. I mean by restricting by current PUT price. Farmable coin denomination is always fixed distance from current PUT price. For example, 10x the current price, plus and minus a denomination. I think the monster post didn’t mention that. Maybe I didn’t adopt your idea perfectly (:
About composting. As I said before, I don’t like because: a) Same as @neo, it adds lifetime to thing we expect to last forever. b) Time is not well defined on safe network. c) I could like backup address for last will or personal fail safe. “I can lose my money for some time but they will be restored to my backup address after a year. I’m unhappy but not very sad.” But it adds complexity. For what reason? Nobody protects your wallet from slipping out of your pocket, but you don’t complain about it. Why think safe network should protect? But I’m arguing against my own idea (:
“Only permanence of data that matters.” @Blindsite2k is a true idealist, so very dangerous (: Idealists ignore reality but Platonic solids don’t grow on the beach. The safe network will be a very complex thing too. Almost a living being but without a soul. Small details can change its behavior much. Data store, communication, identity management, farming for business, “sustenance farming” (make enough to pay for my storage), commerce, money store, and other things. These work together in a complex ecosystem. Make mistake in one, all suffers. “Permanence of Data” sounds lofty. But shortsighted. All parts need care.
Bitcoin:
- How can ISP attack lead to bitcoin coin theft? I don’t think that is possible.
- Brain wallets don’t need protection from elements. Except a brick falling on the head. I know brain wallets were attacked because common scheme was faulty. Low entropy. People used stupid passwords. They can be done safely too.
Classic and Composting:
- Main problem: Composting maybe solves something small but adds more problems instead. That is why it is not a good deal. Losing coins is no big problem. Value does not depend on number of coins but total value of network. Less coins mean more value for 1 coin. No problem. Less coins mean more scarcity, so more value for 1 coin. But network can function the same. I don’t think it is unhealthy to have less coins. Half of the coins lost? Still 2 billion!
To be real honest, losing coins will be a problem in long time scale. I think adding coins will be important at some time. Add another denomination? Not for a long time! But it’s good to build in the mechanism. When needed, it is there. Naturally.
Thank you for the thoughtful consideration of the topic. You brought up some interesting points that help further the discussion. I had a few other thoughts that came to mind from your post:
You’ve made it clear that you are a fan of adding bits, and I would agree that it can seem like a simple and efficient way to solve issues that may or may not be a big deal in the future. From a philosophical standpoint you are suggesting the use of inflation in order to solve the problem of deflation. Your point of view is in the majority, as nearly all banking systems around the globe agree with you. Although I will admit that the virtual inflation caused by adding more bits to coin addresses within the safecoin network would be more egalitarian than real world inflationary pressure because it would effect all users instantly and because safecoin is a virtual-asset based currency rather than a virtual-debt based currency. In the divisibility thread I thought adding bits would be helpful too, but I considered it a case where you would pick a larger bit range than 2^32 at the start. We talked about 48 bit or my craziness of 64bit. But in my mind, I considered it the case that whatever bit depth is chosen at the start of the network should stay the same for all eternity. After all of our discussions I think it should be more than adequate to stick with 2^32 bits via the use of denominations and just fundamentally fix “the lost coin problem” rather than use inflation to counteract deflation.
No, I like your idea. I think my version would be very easy to implement in a p2p fashion because it wouldn’t require much change to what they do now, just a bitmask based on the farming rate of an individual node. Yours might indeed work way better and be even easier to implement. I have no idea, needs simulation as you would say. The summary was trying to just point out that the “the lottery effect” criticism against using denominations can be side-stepped by different means.
When did the SAFE acronym change from “Secure Access For Everyone” to “Storage Accessible For Ever”? No offence intended, but I think your point a) is also true idealism… which as you said, “is dangerous” and let’s someone “ignore reality”. It appears that from discussions in the dev forum, actual timestamps may be possible… 
which would alleviate your concern in point b).
Yes, agreed. Which is why the mechanisms at work in natural ecosystems to recycle underutilized or “abandoned” resources to maximize productivity (allowing for the growth and success of living communities over millennia) should be considered during the creation of a virtual ecosystem. No soul? I’m not so sure. Will it not have a “hive soul” from the amalgamation of all real humans who use and interact with it on a regular basis?
Well, double spending at least. Lost of performance too. See the following:
(Original conference paper at https://btc-hijack.ethz.ch/)
Skiing mishaps, car accidents and bricks aside; brain wallets and meat memory do have a known decay rate related to age.
Ok @jlpell is your major concern with lost coins to do with the reduced farming rewards success rate? (no reference to %age just in general)
And do you define lost coins as truly lost (owner actually lost access to them) or do you define them as coins that appear abandoned (coins not touched for a period of “time” or other metric)?
Because I might have an idea that might solve it for you without causing people grief.
I reject the resemblance of that to my oz of gray matter. 
Partly, yes. And possibly other issues related to the internal network economics of PUTs and the effects of deflation on how users treat the network. I’ve already admitted that there is a lot I don’t know that I don’t know about the inner workings of safenet. That’s why you’re here to set me straight when I say something idiotic right? I’m learning.
Back in post #91 I gave a fairly long-winded description about why I thought a time based “abandonment” metric was reasonable. Search the thread for “2048”. I’ll say it again if you’d, like but post #91 also had some pretty pictures to get the point across. My composting node suggestion and all other nonsense about incrementing a health counter to determine coin age… that was trying to make the best of a bad situation when you told me that the network could not put a timestamp or “minted_date:” or “owned_since_:” date tag on a safecoin. Honestly, I don’t really understand this constraint because I would think that you could at a minimum just ask for a 28 out of 32 date/time concensus when assigning the owner_id to the coin. When it comes to the idea of composting, you don’t need a super accurate timestamp. Receiving a 28 out of 32 consensus on what the current year is doesn’t seem like it should be a difficult task, but tell me if I’m wrong. To stall abandonment, this “minted_date” datafield could be updated whenever the user checks a wallet balance or logs in with the authenticator or operates a farming node or { enter any embarassingly trivial user induced network event here }. If the user has zero interaction with the network for {enter whatever number of years is reasonable to @neo} years, then the coin is abandoned and goes back into the farming pool. This fact would be known ahead of time at network launch and would be part of new user orientation. Just like in the blockchain world where new users need to be told “don’t lose your private key or you will be sorry.” I know you say that people don’t read the fine print, but there is also this thing called due diligence.
I have no grief about forfeiting my network tokens (ie. safecoins) if I don’t access the network after {enter any ridiculously reasonable number of years for any sane person here } years.
Anyhow, sorry for all the sarcasm. Looking forward to hearing about your solution.
EDIT:
Checking my own sanity…
How long is too long to spurn the SAFE network?
>100 years? - Absolutely Abandoned.
50 years? - Definitely Abandoned.
25 years? - Abandoned.
10 years? - We’ll miss you @jlpell, but you have Abandoned us for sure.
7 years? - Maybe Abandoned? Let’s make extra sure they didn’t leave a note some where… found it. Hope the user leaves another note at some point within the next 7 years.
5 years? - NOT Abandoned (User says, “Sigh, I’m here. Leave my savings alone and see you again at some point within the next 5 years.”)
3 years? - NOT Abandoned (User says, “It was a long court settlement, time to spend some coins now.”)
2 years? - NOT Abandoned ( User says, “Round the world tour was awesome, now time for some cat videos.” )
<1 year - NOT Abandoned (User says, “I was just taking a nap, glad to be back online.”)
>
So while comfort levels are different for everybody, I would consider 7 to 10 years to be more than reasonable. Remember, this is just a dead man switch. As soon as the chosen trivial network event occurs, the X year deadline would be advanced continuously.
OK.
Actually I wasn’t after a rehash but a concise response. The question was not aimed at rights or wrongs of this.
You might not but plenty of others do. What you might not consider an issue now because of your current conditions, others in different circumstances do. Also some have a considerable amount of coin and as such they do not like the idea of losing their savings that have grown a lot. They might take 20-40 years to use that. I, like a number of others here grew up before the bad consumption culture took over which says wealth is a fleeting concept because you should spend it all now. Call me old school, but if the world keeps spending all they have and more then its doomed. Lets not create that culture for safecoin, otherwise a few will feed off the rest just like the wealthy do with fiat That is assuming anyone uses a coin (& network) that has the coin disappear on them.
So no there is no length of time that is acceptable and its not necessary as others have shown you. Lost coins are not a mojor problem. Anyhow read below and it can be solved by lateral thinking.
The idea was that if it was basically the effect of lost coin on farming reward success and you see also potential (yet you have not explored) effects on PUT cost. PUT cost is only affected by farming rate (which is the main variable) and as such it is not affected by scarcity (coin loss) at all.
So now you are concerned mainly with farming reward success.
Well that is fairly simple to solve. A factor to account for a (to be determined) estimated true coin loss can be added to the farming rewards so that correspondingly more attempts are made and thus the farming rewards success rate is then exactly the same as if those estimated loss coins were returned.
I’ve heard a knowledgeable person suggest 5% true coin loss every decade and that is certainly within reason and expectations. And remember the 5% in the first decade is less in number than it is in the 2nd decade.
This then means that a minor change to one calculation is needed and no need to double or treble the specific safecoin code and then create a whole new module for recovery/composting and we know brand new modules take 12 months or more to design. build, test, etc. So to compost will put back release for well over 1 year by conservative estimates and perhaps significantly more.
The actual safecoin code will be very small since it relies heavily on the other modules for the data objects handling.
It does not lead to inflation. Denominations are a geometric series with exponent less than 1. Sum of such geometric series is finite. Add any number of new denominations, sum will not grow beyond the limit. No inflation at all.
Yes, easily solved.
Look at the oldest documents. Store once, access forever. It was always like this.
The safe network will be amoral. It will work without conscience. Bad people will use it for terrible things and it will help them to hide. “Secure Access For Everyone” doesn’t pick good or bad.
So. I hope it will have no soul. Because it would be a dark soul.
Unknown decay rate. Like composting
But we can remember important things forever. Repeat many times over long time. Spaced repetition lead to retaining information forever.
Can you honestly say there is a use case where someone will have an account with any amount of coin, which they will then never query or access the safe network to check the balance for 20-40 years?
If they actually use the network, even just one time in X years, then the coins will never disappear.
Are you sure about that? I’m going to make some assumptions, and conclusions. Feel free to correct me:
- Let’s assume for a moment that a date/time stamp can be inserted to a safecoin by way of 28/32 consensus and that it is at least accurate to the year (+/- 6months). We’ll call this the “assignment_date”.
- A user either farms, purchases, or is given a coin. To do this the current_owner_id is transferred and and the assignment_date is set by network consensus.
- In order to confirm that the transfer was completed they access some kind of wallet app to check their balance. Let’s assume this wallet app knows the address of all the safecoins owned by the user.
- Whenever the wallet app is accessed by the user, it automatically initiates an update procedure for all of the coin addresses listed in the wallet using the user’s credentials. Since the current_owner is not changing, only the assignment_date is updated to reflect the current date as decided by nodal consensus. Let’s also assume that the app could track when the last time this network sync was performed and only do it if more than a week had passed since the last sync to avoid excessive network congestion.
- When a farming node is given an address to check, one of three possibilities exist. A) the farmer gets the coin because it is not owned by anyone, B) the farmer does not get the coin because it is currently owned and the current year as determined by nodal consensus is less then the assignment_year + the X year abandonment criteria, or C) the farmer gets the coin because the current year is greater than the assignment_year + the X year abandonment criteria.
So as far as code work goes, it seems that composting could be implemented via :
- A date/time stamp nodal consensus function, which could have other benefits.
- A wallet app that keeps track of user coins and can initiate requests to update the assignment date.
- An if/then statement in the farmer code that checks assignment_date when the nodal address it has been provided is already owned before denying or requesting a farmer request.
Someone living as a monk for 30 years somewhere in the middle of nowhere but after that wants to spend some time with family and friends again 
Ps: since depending on the implementation of safecoin there is no need for this recycling to me it is very impressive how important it is to you
Oh you might check the balance which is reading the wallet MD with the addresses. This does not touch the coins. And remember every single coin has to have your magic formula done to it. The network has no link from the wallet MD and the coins where access of either affects the other.
And yes Some people (many) have over 100K coins and at an expected price of over $1 at launch and maybe 5$ then that is something they may not use up in 20 years.
To do that you have to have code in the core that touches every coin whenever you log on or access your wallet. Wow such overheads not a good solution. Just do the right thing in the first palce and you don’t have all these hoops and conditions and transaction processing when people access their account. So you are saying for someone with 200K coins in the wallet that there has to be 200K transactions (MD update) when ever they log on. That is 200K PUT cost. Wowo this network is becoming expensive for no good reason.
Yes. Have you even looked at the safecoin RFC? Doesn’t sound like it and yet you want to expand it massively with new modules and expanded safecon code. And have not even justified its need.
For 200K coins in a wallet this is 200000 PUT charge each time
This reads a wallet MD not the coins - did I need to inform you of that?
So one or 2 PUT charge for each and every coin. 200000 PUT charge minimum every time the wallet is accessed for the person with 200K coins saved…
You really don’t want people to have any coins after a while do you
AND all this is to make you feel better that no lost coins exist. Doesn’t matter that someone savings account with all their spare coins is emptied because they keep that account unused for extra security.
Bah Humbug, its still theft
And you did not even give any weight to my idea which SOLVED your problem yet had no coins being stolen or extra modules taking over a year to implement and push back launch.
No your patterns for figuring out a lost coin only considers your current situation and what you are happy with. When are you going to not be self centred and consider the effects on the livelihoods of people who do as people do and make mistakes and do things contrary to what you think is the right way to do things.
- At present there is no demonstrated reason to worry about the lost coins.
- At present you only supply solutions that match your conditions and feelings.
- At present you refuse to consider solutions that are simple and take a few lines of code to implement.
- At present you do not consider the taking of people’s saved coins as anything wrong
- At present you do not consider that taking coins from people is going to hurt the image of the SAFE network as safe and secure for data including the coin data and people’s acceptance of the SAFE network. Hint it’ll die faster than that rental model
- At present you don’t consider that making one type of data object not owned by the owner and liable to be taken if they don’t abide to a set of rules you devise based on your current condition and feelings in life
- At present you are not concerned that your idea will push back launch by over a year
- At present you don’t consider that massively increasing the code base that the attack surface is increased perhaps many fold since safecoin is something worth attacking
- At present you don’t consider that people having to pay for all the coins over and over and over and over again because each coin has to be updated every time they access their account or farm a coin.
I don’t like composting but this can be done easily. When farming a coin, the group picks another coin too. Does the magic formula to it. With time, all coins will have the magic formula done to them. Yes it is more work. No I still not like composting 
But there is solution for this part. It doesn’t solve how to know when a coin is touched.
Good point, I forgot which side of the address you were adding bits to. Smaller and smaller denominations etc, yes.
The memory abilities of humans form a normal distribution and in general can’t be relied upon. Maybe you have perfect recall, Einstein couldn’t remember his phone number.
So this is the answer to why this post has been so emotional for some people. As a newcomer, all the marketing info I came across seemed to push anonymous p2p secure communication and free thinking without censorship. I can’t recall seeing the “access forever” mantra, but then again my memory might not be as good as yours, or I just missed it by being blinding by other cool aspects of the project. (Like distributed compute).
So, at this point I’m going to double down and make the statement that I also see a case for allowing user data to decay. Yup, I said it. “Not just the safecoins should be allowed to wither and decay but also private user data.” (@neo, please don’t reach through the computer screen and slap me.)
I could see the case for allowing globally shared public data to be immortalized, but not private encrypted data for users who “abandon” the network. There are potential external costs and unintended consequences of doing so. My initial thoughts are that this policy is essentially the “grey goo” of famed science fiction fear, but in virtual form. A few basic reasons come to mind:
-
You have a situation where the network is left vulnerable to attack from a well endowed adversary who can pump it full of garbage that will never get cleaned out. Think about this thread… wouldn’t you like it to disappear one day?
-
You have a model which forces constant expansions of storage capacity without taking into consideration the external environmental, economic, and other resource constraints of the real world.
-
The 100TB of encrypted cat videos that one guy uploaded late one night after one too many beers are now no longer accessible to anyone since the same cat ate the paper wallet on which his private key was written, and he was too drunk to memorize it. This makes it just that much more expensive to store medical research that could cure cancer, or my cat videos, while the original user gets no benefit either.
I acknowledge that with improvements in technology, leading to greater increases in storage capacity, the old garbage becomes a lower and lower percentage of total capacity. But new garbage will continue to accumulate.
And the reason you should listen first then read up and then come back. Not keep on and on a hobby horse because you like it over everyone objections. Word to the wise. I took a long time before talking.
So you want another project. Fine when SAFE is launched then I suggest you change it to suit yourself.
I say this because the goals of SAFE is persistent data. Full Stop, go and read the blogs about persistent data and autonomous network, both of which require data that does not decay.
And the reason people got emotional is that by trying to change the goals of the SAFE network you are passively attacking the project. Its great to have ideas but to claim that they need to be implemented or the network suffers changes it from a discussion of ideas to one where you have spend nearly 100 posts pushing and not budging and even adopting worse ideas of decay
I suggest you take some time and do some reading of this amazing network design which includes recycling of the coins and when you have spent time discovering all the consequences of the design formulate ideas that will progress the project forward.
In this case you have not even a basis for implementing your “must have” plans.