Topic: Can tail emmision be a soft fork - page 2. (Read 881 times)

There is no issue there. Satoshi decided 15 years ago that this fraction of a satoshi of block subsidy, which will be completely insignificant compared to the miner fees that make up most of the block reward, is to be ignored, as a natural consequence of having finite precision. Just like your computer computes 1/9 in 32-bit floating point as 0.11111111

Someone is missing franky.   


I thought that we were talking about various ways that tail emission might be achieved and/or justified.. or justifiable as potentially something that might be a good thing... It may not matter that much what lightning network does as long as it is not incompatible with the main chain in terms of how many BTC had been recognized.. but sure, we might say, that we do not need more than 8 digits if we can transact on the lightning network for smaller units, but that still does not seem to address the 2048 issue in which mining rewards will go down to 9 digits, and so should that value smaller than 8 digits be recognized or ignored?


You are using the expression "appeal to authority" with a kind of literalness that seems to devolve into nuances of incorrect usage, and sure no problem that we might want to have (invite) some forum members who better "know technical things" to chime in on this particular topic.. someone other than your good buddy, franky..


No.  You are the only one in the whole world who is confused.

Sorry to mention it.

Millisatoshis are not compatible with on-chain consensus. You can have more than one second layer.

Well, you can change a lot of rules by using soft-fork, see: https://petertodd.org/2016/forced-soft-forks#radical-changes

It is not about "current LN vs some other second layer". It is about "hard fork vs some other second layer". In that case, building a second layer is "more honest" than doing some hard-fork, because then people can explicitly see, how many "second layer coins" are mapped to how many "mainchain coins".

If you want to be less confused, then write some code to see, what you can change, and what is coded as a hard consensus rule. You could be surprised, how many things can be changed by some second layers. Also, if you start with just two nodes, forming a channel, then you notice they can agree on any rules, no matter what the rest of the LN thinks about it. They have a direct channel, so they can implement any rules they want, and then build a bigger network, if that idea will spread.

Because you said that it's possible to change the rules and add more decimal points in Bitcoin onchain through a soft fork, therefore, in that context, are you saying that it's possible to change the rules and actually make the block size bigger through a soft fork too?


Nodes that forked away from the Bitcoin network because they're incompatible?


I misread, I thought you said adding decimal points onchain doesn't require a hard fork, and that it could be done througn a soft fork?


How, currently, is there some "dishonesty" in the Lightning Network?


I'm very confused. Hahaha. Pardon me ser, I'm still learning. Plus no one else is posting, does everyone else understand or is everyone as confused as me?

🤔

It depends, what do you mean by "bigger blocks", because inside Lightning Network, you don't have "blocks", and there is no mining. But if you want to change the size of the blocks in a backward-compatible way, you can do so with sidechains, because there are "blocks", and there is a "sidechain", which is different than in LN.

Also note that Segwit already did "changing the size to bigger blocks", and it was a soft-fork, so no hard-forks are needed in such cases.

You don't have to trust me. You can see, how 1:1000 peg is implemented, and try writing your own M:N peg implementation, then you will see, how all of that is possible without any forks. And then, if you use M=1,N=1000, you will reach Lightning Network, but if you use different parameters, you can form a different network, with any properties you want. And if you change that proportions in a volatile way, you can have "tail supply", "burn supply", or any other model. It is all about building a network of nodes, that enforce the same consensus rules, and then it is all about translating "layer 2 coins" into "mainchain coins", and getting all needed signatures.

Why not? If Lightning Network will upgrade in the future to some constant 1:1,000,000 peg, and introduce microsatoshis, instead of millisatoshis, would you notice then, that you can technically introduce any proportions you want? As long as they are constant, it is not about "tail supply" at all, and everything is as good, as it is inside LN. The only dangerous thing is volatile peg, where you start taking coins from users, without their consent, by creating inflation. But then, using some hypothetical network, similar to LN, is "more honest", because then you can see exactly, what proportions are used to exchange coins between mainchain, and some second layer. While in hard-fork scenario, everything is hidden, and some people may not be aware, that newly produced coins would cause inflation, in the same way as many fiat users are unaware of that.

I don't know if that question should be taken seriously.

OK, but why? Because Lightning doesn't break the rules of the protocol/network. Would changing the size to bigger blocks not require a hard fork too? FrankAndBeans the 4D Chess Player would love to spread that disinforming narrative. Don't tell him about this topic.

 


Then why not simply use the Lightning Network?

I believe gmaxwell and achow should get in the topic and give us the plebs their opinions. I know someone will argue "Appeal to authority", no, it's merely asking and learning from the smartest people in the room.

Surely I am not going to claim to know if the satoshi might start to have enough of a value that it would be better to continue to recognize further subdivisions of it by the time that the mining reward splits in half down to such an amount that more digits become relevant.  2048 is quite a ways off, but I could imagine sats having a decent amount of value by 2048, even though surely we cannot have a lot of confidence until maybe we start to get closer to such date, and maybe there would become a kind of sense of necessity to be able to recognize value beyond 8 digits, especially if the accuracy of the mining reward might seem to justify it..

I presume the reason it exists in lightning, is that the 1 satoshi must be cut in several pieces, because nodes charging 0 sat base fee are attractive, and need to find an income elsewhere (i.e., variable fee). Transaction fees can sometimes be even 1 satoshi, so it's reasonable to have more flexibility with units.

You may well be correct, and maybe even the milli-satoshi on the lightning network is not necessary or helpful, even though it seems that some people already had decided to add such feature because they thought that it might be useful to add an additional three digits in order to divide satoshis into 1,000 pieces.

So sure, maybe I took it too far... but surely starting in 2048, all of the digits of the block reward are not going to be recognized if the current system stays at ONLY recognizing 8 digits instead of the 9 digits that would be needed to recognize all of the block reward, and every 4 years an additional digit would be needed merely to recognize the block reward.. but yeah, if it clearly does not have any value, then sure, maybe it is not a problem to fail/refuse to recognize any digits beyond the current 8 that are recognized on chain.. and the 11 that are recognized off chain, but those recognition of 11 digits off chain would not affect onchain, unless onchain started to recognize more digits.

I don't presume to know the answer for sure, but it does seem that with the passage of time, that bitcoin is likely going to continue to go up in value (designed to pump forever), so it seems that there could be some justificiation in terms at least minimally recognizing the block rewards all the way down to the proper number of digits, which would mean starting to add one additional digit every 4 years starting in 2048.


I did not consider anything that I was saying to be related exactly to what dollar value that the satoshi might have, but sure there is a bit of a presumption that if more digits are needed, then the sub-units represented by those extra digits would likely need to have some kind of a value. .even if only fractions of a penny, just like a satoshi is fractions of a penny currently yet some folks using the lightning network believe that there is some kind of value in terms of adding three more digits to recognize 1/1000 of a satoshi, even though that is currently pretty damned small values in terms of current dollars.. and so yeah, the less valuable the units representing the further digits down, then the addition of digits might not be necessary and merely taking up band-with (space) as vjudeu had mentioned... so I suppose that there is a bit of a presumption that the dollar value of bitcoin, satoshis, milli-satoshis and any further digits added thereafter are going to go up sufficiently in order to justify their being represented, and that was part of the reason I mentioned adding 8 more digits every 32 years to merely make sure that the mining rewards are accurately depicted.. but at the same time, if those micro-units further down do not really have any value, then even my proposal to add 8 more digits every 32 years may well not be something that is worth putting into practice.

Why do you think that Lightning Network with 1:1000 peg requires no hard fork, but 20,999,999,9769,0000 : 21,000,000,0000,0000 peg would suddenly require it? It is all about proportions, nothing more, nothing less. If they are constant, it is as good as inside LN. If they are volatile, then you can reach "tail supply", or "burn supply", it depends if you increase your numerator, or your denominator in "numerator : denominator" peg.

Which is my point. BTC-without-tail-supply, which is the current currency, cannot be inflated in a no-fork way.

Because that would probably break a lot of software, and either implemented via softfork or hardfork, there will be bad consequences. Turn your question the other way around: why should there be infinite digits? Clearly, 1 sat is of little value, and it will probably be of little value in a couple of decades as well. Even if 1 BTC = $1M, 1 sat is still only worth $0.01.

Wouldn't that be just "with the presumption" that a Satoshi would be worth at least a Dollar/near a Dollar? What value would that bring to the miners if a tail-emmission or an inflationary model, assuming it could get community consensus, would better incentivize the miners, immediately, after the fork?


Are you sure about that? That requires a hard fork, no?

I am already questioning your presumption that extra digits beyond a satoshi would not have extra value, even though I can understand why infinite digits might not be a good place to stop, but instead attempt to figure out some number of digits (currently at 8 on chain and then 11 digits if we are referring to lightning network).

So it is difficult to presume in advance how far to go with the number of digits before they would not have value.. so maybe there is a need to ONLY go to 8 digits right now, but add another 8 digits every 32 years (that is 8 x 4) .. and if the extra digits do not have any value, then I can see the point that they are just taking up space, but I am having difficulties seeing the point of not being precise when we are not sure how much value sub-satoshis are going to have or how many digits we might need to go at any particular time before no more value is present. 

It seemed to me that ONLY one digit would need to be added every 4 years. but I suppose if we are considering backwards and forwards compatibility then there might need to be some kind of a forward projection of value (and/or considerations regarding how many digits to go in order to project out)..

In that regard, it seems that already on the lightning network, there has been some considerations that 1/1000 of a satoshi does have some value... and so by the time we start getting into more than 8 digit bitcoin block rewards in 2048 (as I pointed out in my earlier post), then at that point there likely would be some further needs for subsatoshis even going beyond three additional digits beyond the satoshi that the lightning network already shows.

I am starting to like my idea of adding 8 more digits every 32 years.. but then when to start?.. To me it seems that there would be some kind of a need to start prior to their being recognized in the block reward... but surely I can appreciate that at some point bitcoin is likely NOT going to be going up in value enough in order to justify adding more digits (or worthless bandwith consumption precision, as you suggested).

True, which means there are more chances to get it implemented. Because it is a difference between constant peg, and volatile peg. If you have Lightning Network, you have 1:1000 peg. It is fine. In "infinite halvings" model, you have constant 20,999,999,9769,0000 : 21,000,000,0000,0000 peg. This is also fine. In "tail supply", you start from 1:1 peg, but it is volatile, and it is changed by each and every overprinted satoshi, first from 20,999,999,9769,0000 : 20,999,999,9769,0001, and then the denominator is just incremented with each satoshi, created by "tail supply" supporters.

But anyway, it does not affect implementation that much. You could have "burn supply", instead of "tail supply" as well, and start from 20,999,999,9769,0000 : 20,999,999,9768,9999, and then decrease your denominator with each and every burned satoshi. That kind of consensus could work as well. Which means, if you can implement a constant peg, you probably can also implement any other volatile peg, because the whole implementation will be very similar, you will just change the exchange rate between the layers.