Safecoin divisibility

While a mchine might need divisibility a machine does not need individual coins for easy calculation purposes. It could simply use straight up math of safecoin itself instead of using a coin to represent an amount of safecoin. Why do we use smaller denominations of currency in the first place? Because it makes things easier to symbolize. But a computer doesnā€™t need this. Your calculator can crunch numbers just fine without this added symbolization but a HUMAN does need this. Therefore creating added symbolization is not for the IoT but rather for human use rather than machine use. Therefore you donā€™t need 100 billion coins or whatever for additional machine use cases. A machine can handle 0.3545842925423542578714754 just fine. A human on the other hand cannot. A human might be able to hand up to say up to 4 or 5 decimal places at most. There is a reason we calculate things out of 100 or 1,000. Iā€™m sure math gurus can do more but Iā€™m thinking average joe that just wants to pay for his groceries or whatever.

Also correct me if Iā€™m wrong but isnā€™t it a rule of math that before you calculate a fraction that you make sure itā€™s all converted to the same type. Something about using a common denominator? Wouldnā€™t it then make more sense to use safecoin itself as the common denominator before doing major calculations for IoT calculations?

@neo I see your point that the load of transfering all these different currencies back and forth could rival data storage. However I think itā€™s worth noting that currency creation and transfer is a valid and well used use of the internet. And people WILL do it anyway regardless of itā€™s affect on the network. I donā€™t think the question is IF people will do it but rather how to allow people to do it effectively and efficiently so it DOESNā€™T lag up the whole bloody internet?

But it still a major workaround to try and implement a lolly idea (Candy idea). The idea has good sounding features that tickle the ear but involves increasing complexity, increasing network load and taking up valuable data space that could be used for something valuable, I dunno like information. The point I was making is that SAFE is designed for storage and security/anonymity. Not designed as a payment gateway and the purpose of storage being for coin storage. Especially when it is not needed.

And like the internet the payments and ā€œcurrencyā€ will be streamlined and straight forward using the available features built into the protocol. We simply donā€™t need this idea for SAFE to be a payment sysem as well as its primary purpose - storing valuable data.

Not convinced. But @jlpellā€™s idea takes care.

Hereā€™s another Catch 22. If a web page is important, many pages reference it. But only if the referencing pages are important should they matter. But we only know if they matter if important pages reference them. I misunderstood. You didnā€™t ask which set of coins to enable. PageRank solves similar problem. But you think the mechanism is complicated. It isnā€™t.

An address is tried. If not good, try again. Until one is good. It is simple because addresses are selected deterministically. Add a counter, like bitcoin nonce, but must always start with zero, step by 1, use first good address. Cost is unnoticeable. Because: a) Hashing is cheap, often accelerated by CPU. b) For a long time, most good addresses are the cheap ones, so need few tries for good address.

Future-proof. When safecoin becomes very valuable and need cheaper coins. Add more bits for more cheaper coins after currently cheapest in top 23% of the 4 billion. Itā€™s a natural process.

Transient escrow blocks. Payer puts in coins, states demand for change. Payee puts in coins, references payerā€™s escrow block. Validating group checks, agrees it is okay, performs ownership transfers, deletes escrow blocks.

Yes, but occurs for a variety of reasons. Sometimes because sellers are abnoxious and sell an item for $9.99 instead of $10 for no good reason. If you consider the hypothetical interaction with your bar tender, you are not purchasing anything from them for $4.99 (USA), you are trading your crisp $5 paper bill for a beer and a beautiful shiny copper penny. If he doesnā€™t have the penny, you can either say ā€œno deal bub!ā€ or ā€œno worries, friend.ā€ I donā€™t think the network really needs to worry about this. Inside someoneā€™s ā€œwalletā€ they could have a bunch of different safecoin denominations. If they want to buy something, the wallet could easily calculate which combination of denominations need to be sent over and suggest that to the user, the user could just hit ā€œsendā€ or the user could adjust things manually if they wanted to pay 10x 0.1SC instead of 1SC.

Sometimes I donā€™t have exact change, so I need to ask around until someone is willing to trade me their 4 shiny quarters for my beat up $1 bill. Sometimes the store doesnā€™t have exact change and I say donā€™t worry about it.

Complete misunderstanding of the way SAFE is doing farming rewards and the elegant way of accounting for the scarcity issue. Since it does not work that way your example is no good.

But why introduce a more complex system just because it has some nice sounding featuresā€¦

And then you still cannot solve the legal issue of the ICO selling MAID on the promise of what SAFEcoin will be. Its a legal contract really and much danger in trying to fiddle with it. If you do not think this is important then just look at a couple of the recent failures due to not honouring important ICO conditions.

Your idea has a lot of work arounds and really changes the concept a lot. Remember SAFE is not a network to suport a currency like nearly every blockchain project is. It is a data network foremost. The coin is introduced to pay for resources and to be rewarded with. It is an incidential byproduct that SAFEcoin can be used as a ā€œcurrencyā€. But your idea has a significant portion of the storage supporting the coins and a significant amount of data flow (bandwidth) to support coin movements. Whereas a blance system for sub-safecoin transactions solves all that without any data storage overheads and the transaction load (data flow) is just one transaction per payment.

1 Like

I have a question :stuck_out_tongue:

If thereā€™s a set amount of SafeCoins, and theyā€™re not divisible, and over time people forget their keys etc and SafeCoins get lost,

Couldnā€™t that lead to the same problems of BTC where itā€™s harder to farm, need to pool together in the slim chance of farming one super valuable SafeCoin, etc etc?

Even if it takes a thousand years to be a problem, donā€™t these factors I said above inevitably lead to that?

1 Like

As I understand:

  • Node does something that gives right to attempt farming. For example, a block is served.
  • Group in charge determines coin address. If coin address is free, coin is assigned to nodeā€™s wallet address.

Can you quickly point out what I misunderstand?

That is close enough. But your example does not follow this. So that suggests you do not understand the consequences of that process

I like the idea of having the network ā€œresurrectā€ safecoins so that they can never be permanently destroyed. This would work by having a timeout period. If a safecoin has not changed hands for letā€™s say 7 years then it gets returned to the network. All someone would need to do to keep them fresh is move them or spend them. I think a way to view this is the same as lost treasure. Someone forgets about a pile of coins they buried in their back yard, years later you buy the house and find them when you are planting the gardenā€¦

That sounds super dangerous

Wake up one day and all your life savings are gone.

Also kinda sounds like inflation which I am totally against :stuck_out_tongue: Global population needs a strong, reliable, deflationary currency so they can stop being so poor around the world. Thatā€™s one of my life goals, to help support efforts to make that happen.

1 Like

But it does. One additional step necessary. When the group determines coin address, it adds a counter. It starts at zero, it is incremented until address is in acceptable range. Counter is like bitcoin nonce, but no free choice which numbers to try. Only thing changes is which address range is filled up first. Adds little higher scarcity too.

No additional storage. Coins stay 32-bit address with owner. And previous owner, is that how it works? And signatures. Address determines value uniquely, so no need to store that.

No additional bandwidth. Individual coin can be sent to other by changing owner. Multiple coins may need simple escrow service, but only if giving change is necessary. My previous post outlined it.

The SAFE netowrk does not know time. So this cannot be done.

Also its is rude at best to take coins that someone is ā€œsavingā€ up for a big purchase in time. Like buying a house or something. Then find their coins are disappearing because they were saving them. And at worse it is considered thief. No no and no again to this idea.

@whiteoutmashups, Iā€™d say like BTC which has a significant amount of their coins lost or unusable is surviving very well. SAFEcoin has 4 billion coins and even if 20% are lost the network will operate as if they were not lost. Only if billions of coins are lost is this an issue.

2 Likes

These ideas of denominational coins are just one workaround after another and frankly Iā€™ve explained the problems and the recent discussions are ignoring the fundamental reasons why the Candy ideas are not workable here for sub-safecoin use.

Iā€™ve been thinking about some of neoā€™s critique about how to keep the original investors from the ico happy in a scenario where divisibility is introduced through denominations.
Although I really liked the 1-2-5 progression the @norimi calculated further up the thread, I think that 40 or denominations might be too difficult from a user perspective. Hereā€™s an example where there are only 10 denominations, a total of 2^32 safecoin objects, and a total of (2^32) safecoins that would ever exist. This is just an example. I wish I could have found I nice geometric function to relate the denomination ratios but I ended up needing to fiddle with them in order for the total safecoin market cap to stay fixed. I fixed the quantities so that the number of 1SC is enough to provide 10% of the total coin stock for the ico folks (Just in case someone who purchased 10,000 MaidSafe wanted to be paid out in 10,000x 1SC instead of 1000x 10SC etcā€¦).

1 Like

My idea doesnā€™t try to replace your idea. It was started as another thread. Your idea may solve IoT. This one doesnā€™t. It increases expressive power of safecoin. Larger range transactions with O(log n) coins per transaction.

If farming restricted to range of coins, your concern of inequality is solved. Itā€™s easy to do. One additional step to farming. As explained above. Doesnā€™t conflict with original idea.

Adds possibility to extend when safecoin becomes valuable. Add few bits for additional range for coins with lower value. If done, it can do any range transactions with O(log n) coins.

ICO would get equivalent value. Many things changed since ICO. Unreasonable demand for no change? Project was in early stage that time. What was ICO wording? How to argue without knowing that.


Please list fundamental reasons because I canā€™t see. My idea:

  • Coin address determines value. Storage and transactions stay as before. (Change of ownership. Escrow service to help giving change could be useful. Simple to implement, explained before.)
  • Only a range of coins farmed at a time. (borrowed from @jlpell) This fixes inequality. (Averages out after handful of farming. Shown simulation results before.)

This is the whole idea. Can you point out fundamental flaws with one or other? Some bullet points please. Forgive Iā€™m not smart.

Please donā€™t call other peopleā€™s ideas lolly and candy ideas. Disrespectful. We put work and thought in it, too.

Itā€™s not much different than if you lost your paper bitcoin wallet, accept that the coins would go back into circulation after enough time had passed to be sure that they were in fact abandoned. Itā€™s not inflationary because the total number of safecoin is fixed from the start, if you donā€™t have resurrection then coins can be burned and the supply would decrease over time. This is just a brainstormā€¦

Yeah IDK itā€™s just not really doing it for me but thanks for adding ideas! Letā€™s keep em coming :slight_smile:

Just to clarify, you have no problem with safecoins dissappearing over time?
What happens if 50% of the safecoins get burnt? Hypothetically the remaining coins would double in value, but now youā€™ve lost half of the ā€œfuelā€ to run the system. It just makes paying for little things more difficultā€¦

Definitely 100x less of a problem with that then anyone losing their money if they forget to shuffle it around

3 Likes

I donā€™t like this. I played with similar idea. Could be used for smallest coin with denominations. Then no real damage. But I donā€™t like it because people donā€™t like uncertainty. Interesting, but no go for me.

1 Like