@icpmax, I’m having a tough time following this, but I think your primary critique is about the token distribution–not the tokenomics. We can’t do anything about the distribution of ICP at this point. And that’s not what this thread is about. This thread is about tweaking the set of incentives around the ICP token to meet some desired purpose (whether it is to increase the security of the network, to increase demand for token, etc.).
2 Likes
For most, I disagree with @icpmax. Those seed investors (which I am not and have bought my few first ICP at 320$) had invested 10k, 30k or whatever (at .03 cents each ICP) back years ago when Internet Computer was only on paper, not much more then a dream. These investors had the most faith that the dream would come true. Without these investors, IC would just not exist. Now, if these investors believed enough to pour thousand of dollars in a dream, why would they sell now that IC is alive and work amazingly? These people believe much more in Crypto then fiat money. Why would they sell? To buy other cryptos? or get back to fiat? Where will they invest their fiat money obtained from that dump? What is better then ICP as investment for the next 20 years? If you have a clue, let me know, I may be interested.
I just believe (but may be wrong) these seeds will invest most of their new obtained ICP in the NNS and have great returns, while the price would go up.
I don’t see many other ways (other then Dfinity have done) to obtain funds for a project on paper only.
Our risk to invest in ICP now, a great working layer 1 network, is millions times less then for these seeds investors at the time.
What I found to be bad at launch wad the lack on clear information, which is a general statement and complain as I can see. Now, we have to live with it and go forward, if we believe in the technology of course, which I do from the very beginning I learn about IC. (When ICP was selling at 320$ back in May)
This is my thinking but respect yours.
3 Likes
If the penalty were so severe that it ate into the principal, and if the totality of said penalty were burned or distributed to, say, only locked neurons, then MAYBE I’d agree. But even then it’s a maybe.
And regarding neuron transfers… one thought is for neurons to be transferrable to other II’s that are subordinate to my master II (i.e., created by me via my master II). And perhaps there needs to be a rule that a neuron can only be transferred to a subordinate II that has been in existence for a year (or some other arbitrary time)… or perhaps the % of the neuron that can be transferred is dependent upon the age of the subordinate II.
I’m not a fan of transferring neurons, except for legacy/estate needs.
2 Likes
Anyone who believes neurons should not be transferable is making the same mistake as Bitcoin’s decaying block reward: it is a bad security assumption for the future.
If an external blockchain implements chain key crypto and private smart contracts, one will be able to create neurons on addresses controlled by the external blockchain. Then later one could transfer ownership of this neuron on the other blockchain, while on the IC it appears nothing happened. Only the buyer of the neuron would know the transfer happened, preventing NNS voters from blacklisting transferred neurons.
If the security of the IC/NNS depends on neurons not being transferrable, it faces a ticking time bomb of another blockchain implementing CKC and private smart contracts. If just 1% of locked ICP wants to be sold annually, that is still a huge market to build a product for. Then an attacker could still perform whatever governance attack they want, while only losing the discount on locked ICP. The discount will still be non-trivial, because neuron sellers have to incentivize buyers over just buying their own ICP and staking. Plus, the sale won’t happen instantly, so the attacker still risks the price of ICP declining after a bad proposal is passed.
The counter to the governance attack is for neuron buyers to be aware of the voting status of the neurons they wish to buy. Buyers will not want to buy neurons that are actively voting because of the risk of becoming a bagholder after a governance attack. I expect neuron markets to prevent listing of neurons for sale that are actively voting to save buyers the trouble of manually looking that up.
If the IC directly allows neuron transfers, it could implement some light rules/restrictions to enhance security, that are not enough to incentive the creation of an external neuron market. For instance, adding small delays to voting/following/transferring actions could prevent attackers from instantly voting then liquidating a neuron. Or create a special neuron state that allows it to be transferred but not vote. It would be better for the IC to have some control over neuron markets than for them to have free reign on external blockchains.
2 Likes
I personally think that other projects like Polkadot raise some sincere questions regarding the tokenomics and governance:
Here is the link on Polkadot’s governance mechanism.
To vote, a voter generally must lock their tokens up for at least the enactment delay period beyond the end of the referendum. This is in order to ensure that some minimal economic buy-in to the result is needed and to dissuade vote selling.
It is possible to vote without locking at all, but your vote is worth a small fraction of a normal vote, given your stake. At the same time, holding only a small amount of tokens does not mean that the holder cannot influence the referendum result, thanks to time-locking. You can read more about this at Voluntary Locking.
Example:
Peter: Votes
Nowith 10 DOT for a 128 week lock period => 10 * 6 = 60 VotesLogan: Votes
Yeswith 20 DOT for a 4 week lock period => 20 * 1 = 20 VotesKevin: Votes
Yeswith 15 DOT for a 8 week lock period => 15 * 2 = 30 Votes
So that raises for me the first question: Why are we afraid of shorter lock-up periods? Polkadot shows exactly how this could be solved.
If one doesn’t lock up his DOT for a vote, he’ll only get voting power with a factor of 0.1. = > 128weeks against zero delay equals a factor of 60(!) in voting power weight difference.
That means someone who locks up 1Dot for 128weeks has got the same voting power as someone with 60DOT.
In my opinion that should be enough to actually counter the argument: “Only long-term holders should be able to vote/benefit from staking etc.”
The huge advantage of a system like Polkadot is the that it actually let’s everyone benefit. ICP feels more like a punishment with the 6month lock-up atm. I understand the arguments, but in comparison with Polkadot we should ask ourselves: Why?
I don’t have a good answer for this. Polkadot differs between staking and governance voting. I know that it’s not the case here. However even staking is possible with a lock-up delay of 28days on Polkadot.
And I actually don’t see any argument why that shouldn’t be brought to ICP.
If we don’t find proper solutions to these questions, we’ll probably end up like cowards. The only argument I can find is that a minimum lock of 6month, brings stability to the price, since ICP are longer out of circulation.
That’s also the complete opposite of what we should oppose to the fiat system: If we make it hard to liquidiate, because we’re afraid that the price could go down, what does that say about the system and ourselves? Are we that afraid that the free market could punish us for some flaws?
It’s like a country embracing capital controls in order to prevent capital flight and help the exchange rate…
That’s normally a completely desperate act. Look at some countries with high inflation and ask yourselves how that turned out for them.
Next thing: 6month minimum lock up in combination with 1ICP treshold for spawning a neuron
Who benefits more? Someone with 1000 ICP or someone with 10ICP?
Someone with 1k ICP and an annual inflation rate of 10% would generate ~8.3 ICP a month or 2 ICP a week. With 7 days delay for unlocking the new neuron, he’d actually have much liquid ICP available. He could restake them again to earn interest on his interest (making him benefit even more), or he could also sell them, which completely counters many of the arguments. I’ve already described it here.
The staking guide explains it:
There is a “Spawn Neuron” button that is grayed out until I have at least 1 ICP in rewards. You can compute how many newly minted ICP you have by taking your maturity % and multiplying by the number of ICP you have staked in the neuron. Once your ICP rewards are greater than 1 ICP it will let you spawn a new neuron with the newly minted ICP staked in it. Once you spawn a new neuron it will have a 7-day dissolve delay after which you can either send it to a different wallet or add it to your existing staked wallet.
Someone with only 10ICP would have to wait much longer to be able to earn interest on his interest. Futhermore he would also have less liquidity than someone with a high stake, earning the 1ICP really fast.
What did the person with “only” 10 ICP do wrong? Right, he didn’t have as much money or wasn’t as lucky as someone else to get such a high amount of ICP that early. Now he is punished, and the whale benefits exponentially. I don’t have to explain that even the current banking system is more equal since we all get the same low interest rate.
I personally see it this way: A while could dump on the market with his freshly minted ICP (staking reward) while I still can’t spawn a new Neuron. Here you are clearly prefered if you have a high stake or to turn it around:
=> One actually gets punished for only having a small stake.
IMO this is an oligarchy on steroids.
Neuron age is an interesting concept, however if we bring it back to the inequality discussion it’s not clear if that isn’t exactly the opposite of the wanted effect.
Let’s take Dfinitys stake (Idk how high atm): Probably 20% of the Genesis supply
If that is staked => 10% a year.
20% on ~400Mio ICP => 80 Mio ICP
10% a year => 8Mio ICP a year (8Mio ICP @60$ are 480 Mio$ a year from staking only(!))
Now additional benefits due to Neuron age (one of the first stakers and 1ICP treshold liquidity, earning interest on interest) => Dfinity can easily pay staff from staking only. How should that thing ever become decentralized with that whale tokenomics?
@wpb
I have to disagree with your statement:
we should not make significant changes the tokenomics.
@BrianGaller
You’ve written, that you wanted actual data first. Fair way to go. However I don’t read that there is sensitivity on this matter. I read many comments which say the following: Wait, let’s watch it. etc.
And that’s what really concerns me.
Having bad tokenomics is one thing, but not seeing the flaws is a whole other story. I’ve actually never seen a project like that: Without doubt the greatest tech. Without doubt horrible tokenomics, and I clearly stand to my opinion (my provided facts should speak for themselves). @Trevor wrote another great comment in here about some more flaws.
If ICP wants to succeed it will need great tech, great tokenomics and community support. Putting it on the long bench and thinking that this will somehow fix itself, will let the problems appear squared in a few years. By embracing an oligarchy, ICP loses points to tokenomics and this thing will come back like a boomerang after years. Benefiting the whales will let us lose community support (think of our fiat system). Maybe not today, but it will get worse with each year. Look at the wealth distribution in the US. ICP can do this within 5 years. If you don’t believe me, take some numbers and calculate them in Matlab via interest of interest on high stakes.
tldr: IMO ICP tokenomics is horrible. Without major governance/staking upgrades, the system could set itself up for failure.
Proposals:
- Polkadot governance should be studied and partly copied to ICP
- 1 ICP treshold for spawning a new Neuron should be lowered, or completely abolished (better).
- 6 month minimum lock-up period should be significantly lowered
- Neuron Following should be decentralized to more parties
- More transparancy regarding staking stats is required (already in discussion in here?)
- Discussion on how to avoid oligarchic structures is required
- Discussions about the treasury should be established (e.g. treasury votes with ICP?)
1 Like
Thank you for sharing your thoughts on these topics. I respect your opinion. I just wanted to say that I agree with one of your proposals, which is that the 1 ICP threshold to spawn a neuron does disproportionally benefit larger token holders who are compounding their voting rewards and are willing to manually spawn, dissolve, and top up as quickly as possible. For anyone that is not compounding, it makes no difference as the reward rate of return is equal (assuming equivalent dissolve delay and age) no matter how many ICP are locked. In practice, I suspect a lot of folks are compounding and want the compounding to happen automatically. It is my hope that the automatic compounding feature that Dfinity foundation is working on will be implemented on a daily basis so this issue goes away. If it doesn’t address this issue, I think a strong case should be made for enabling compounding to equally benefit everyone no matter how many ICP are staked. There are probably multiple ways of accomplishing that goal and I would advocate for Dfinity foundation to pursue that type of proposal regarding the tokenomics.
Edit: Another way to give large and small token holder equality from compounding would be to change the definition of when the Spawn Neuron is allowed. The fact that it is pegged to 1 ICP is what creates the inequality. If it were pegged to Maturity (the ratio between ICP reward and ICP staked) then that inequality would go away.
4 Likes
I believe that in order to achieve the maximum number of stacked neurons, it is necessary to offer a non - proprietary value. Let’s think about it? What can be valuable for a true fan of the IC ecosystem? Allocations in startups. We need to somehow encourage true stakeholders by providing guaranteed allocations in projects that are interesting for them. What can be the model? Let’s fantasize. You can add the Noblesse parameter equal to the Maturity parameter. From which it will be possible to spawn a sub-neuron. This sub-neuron will be able to vote for a certain project and get a proportional guaranteed allocation. After voting, the sub-neuron disappears. It is important that these neurons cannot leave their host to avoid speculation and market valuation. I believe that it will also be beneficial for the project owners.
First of all: this will give them the power of public recognition.
Secondly: the owners of strong sub-neurons cannot be weak hands
There is an expression Canadian people are saying: “Don’t try to use the shovel before it snow”. We can assume and try to solve tons of problems that may never arise. Good for discussion though. My stance, for what it is worth, is wait and see how it goes and solve problems as they arise. So far, I am only impressed about IC technology, Dfinity and every one involved (developers, etc…). Neurons rewards were design to rewards more investors with long term view. But short term are quiet well serve as well to me. Whales (I am not at all) are whales for a reason. They believe and invested in a paper only system. Now, I would think they have all incentive to have IC to work at his best, to get their investment at his best as well.
2 Likes
I’d like to point out Diego explained most ICP minted are from neuron rewards not node providers, so part of my concerns are alliviated.
I think personally there are several different topics here being discussed and they should go into relevant threads. I would call mine “Inflationary rate.”
While I’m aware the ecosystem will lead to ICP getting burned, it would be good to get some numbers behind that so the math works out. I think this ties into better understanding of costs and what it would take to have tokens burned = tokens minted with a 5-10% inflation rate plus node provider fees.
2 Likes
I absolutely agree on your view that automatic compounding doesn’t solve the issue. It actually further embraces the problem. Now you’ve to take a look on your neuron once in a while. High stakes might forget to spawn new neurons. There is also human work (still really low amount) in the process, which slows it a bit down.
With automatic compouding they can lean back while the ICP itself helps them to use the oligarchic structure perfectly with zero additional effort. That’s insane.
No wonder that whales want automatic compounding => The rich want to get richer, and that as easily as possible. I’ve written that exactly in my other thread.
Furthermore what I didn’t mention is that the period for reaching the 1 ICP gets smaller and smaller, the higher the staked amount gets. => That even puts more dynamic into the exponential curve for whales.
Imagine an automatic script which does that after a few votes/days.
Could you explain your statement more exactly, pls? It’s interesting:
If it were pegged to a ratio between ICP reward and ICP staked then that inequality would go away.
The problem with this is, that someone with a high stake, could simply split his neuron up a few times and avoid getting into the problem. That might incentivice to get smaller neurons and put less dynamic into the exponential curve. However if you take a look at my example in the other thread, you’ll see that the difference of someone having 2 neurons with amount/2 or one neuron with amount is not that severe.
It will get interesting if someone is forced to split up his neurons to an amount equal of a smaller stakeholder. However this would need serious calculations, and further analysis if there is a way to simply avoid it. If someone with 1k ICP will get the same percentual reward after the same amount of time as someone with 10ICP, the system is fair (excluding Neuron age at that point).
Hey @MichaelAnnh, I just want to clarify what I was trying to say in the previous post…
I am in favor of automatic compounding.
I am in favor of compounding methods that ensure reward rate of returns benefit large and small stake holders equally. Daily automatic compounding achieves this goal, but there are manual methods as well.
If automatic compounding is not going to be implemented, then pegging the manual Spawn Neuron action to a fixed Maturity threshold instead of a fixed ICP reward threshold also achieves this goal of equality. However, pegging to a fixed Maturity creates new problems such as how long it will take to achieve that Maturity.
2 Likes
Wow - a lot of very thoughtful discussion here. In separate posts to come soon, I’d like to offer my own two cents. But first, I thought it would be useful to all involved here to try to summarize ideas raised so far (apologies in advance for not tagging all of who have advocated support for given ideas - the forum here doesn’t seem to permit tagging more than 10 members in a single post):
Some key ideas raised to this point:
- 1. Re-calibrating ICP inflation and burning parameters - to achieve today something closer to a non-inflationary supply of ICP (@Trevor, @BrianGaller - who advocates for some study to be conducted to see whether this is needed).
- 2. Providing the ability to transfer neurons to another - to the beneficiary of an estate if need be (@wpb), but also potentially to anyone else at any time, possibly with penalties (@Alixthe, @talkingant).
- 3. Creating visibility into “liquid voting power” - both to provide general awareness of where control of the NNS lies at any given moment (@wpb), and also providing clarity regarding who owns how much ICP (@Trevor).
- 4. Finding better ways to achieve the 90% staking goal - both to decentralize faster and in the process achieving certain guarantees against 51% attacks (@wpb).
- 5. Supporting shorter ICP lockup periods in staking (@MichaelAnnh).
Some other good ideas raised to this point:
- 6. Fine-tuning staking reward mechanics - namely, automatic compounding of staking rewards, and/or changing the threshold for spawning new neurons, so as not to disadvantage smaller ICP stakers (@wpb, @MichaelAnnh).
- 7. Providing greater control over one’s own neurons - i.e., combining and/or splitting neurons (@wpb).
- 8. Ensuring that centralized exchanges reflect only circulating supply of ICP (unlocked supply), rather than all supply (@tlbalog).
Please do let me know if I’ve in any way inadvertently misrepresented anyone’s position here, and I will undertake to adjust. Thoughts on some of these topics to come soon.
4 Likes
On point 1. - recalibrating ICP inflation and burning parameters:
I believe that current parameters are quite nicely calibrated for the steady-state economics that we might anticipate for the IC. I’ve reached out to both @Trevor and @BrianGaller directly on Telegram to share thoughts, and invite anyone who may be interested to connect with me on Telegram (not sharing here, as thoughts have implications for ICP’s price, and this Forum isn’t meant for price-talk).
2 Likes
On point 2. - providing the ability to transfer neurons to another:
I definitely think it’s worthwhile to consider how to transfer to the beneficiary of an estate. This would provide peace of mind, and if designed nicely, could, I would imagine, be made hard to abuse.
On the idea of enabling neuron transfer in general, the security implications are worrisome. But I think @talkingant raises some really interesting points. It hadn’t occurred to me that, if another blockchain platform were to adopt chain key technology, users of that blockchain could create transferable neurons there, with no way to know or mitigate on the IC itself. I’m quite curious to learn more about how this would work. Because if neuron transfer really can’t be prevented, then yes, much better to prepare for the inevitable, sooner rather than later, and find other ways to mitigate the unavoidable risks of neuron transfer.
Still, even if chain key’s adoption by another platform would enable neuron transfer, it will most likely take a long time for other platforms to adopt chain key. So we have time to figure out how to manage associated risks.
So I think we’re still better off at the moment aiming to keep neuron transfer disabled, until we can find workable ways to mitigate risks - risks that really could undo the IC, if not managed properly.
Only once we’ve found appropriate mitigations - and just as soon as we have - then I think it would be great to enable neuron transfer… I think the novel creative opportunities represented by neuron transfer could be significant… just need to make sure we do this in a safe way.
2 Likes
On point 3. - creating visibility into “liquid voting power":
I think this is a great, uncontroversial idea that should be pursued without hesitation.
Same goes for point 4. - Finding better ways to achieve the 90% staking goal.
2 Likes
On point 5. - supporting shorter ICP lockup periods in staking:
I think this is interesting, and worth considering… so long as voting power is sufficiently small for trivial lockup periods so as to prevent interested parties from swooping in to successfully make a quick gain at the expense of those who are invested in the IC’s long-term success. I think this idea deserves more thought.
2 Likes
Finally, I think all of points 6., 7., and 8. are good, fairly uncontroversial ideas:
-
- Fine-tuning staking reward mechanics - namely, automatic compounding of staking rewards, and/or changing the threshold for spawning new neurons so as not to disadvantage smaller ICP stakers.
-
- Providing greater control over one’s own neurons - i.e., combining and/or splitting neurons.
-
- Ensuring that centralized exchanges reflect only circulating supply of ICP (unlocked supply), rather than all supply.
2 Likes
Point 2: @ayjayem
I think your whole summary is great and @talkingant actually put some practical insights from the free market into the discussion of this point. I can’t speak about the threat of others implementing chain key technology, since my knowledge about it is too low. Your statement was.
Still, even if chain key’s adoption by another platform would enable neuron transfer, it will most likely take a long time for other platforms to adopt chain key. So we have time to figure out how to manage associated risks.
I think the free market puts pressure on this from another side. p2p and otc deals. Let’s take a look at this reddit post.
They talk about selling locked ICP with -50% on spot. People could start to sell Yubikeys/ accounts (secure if the seed is deleted from identity managment /if device is changed) via p2p and otc trades. People might start a huge shadow market for neurons/accounts, without anyone knowing about the actual distribution numbers.
The irony on this is that many of the fears which are tried to be controlled desperately might become true anyway: People grabing cheap locked ICP and being able to vote. Of course someone would have to accumulate like crazy to start an attack, and in the end there is one question: Why should someone accumulate locked ICP, in order to vote against the ICP? There is no answer to it in the short term. On the long term there is also no answer to it with that lock-up period (playing through some scenarios would be interesting). The only instance being that insane could be a government actively buying ICP at any price in order to censor/mess with it one day. Here comes the PoS vs PoW debate again.
What this tells us is that we can’t control the free market and people’s behaviour, and we shouldn’t try to. That would be the same mistake as central banks trying to print their problems out of their way. They’ll just emerge at some other point much worse. We should just give the whole protocol a great framework, a sort of constitution which is strong enough to resist governance voting attacks.
IMO that can’t be done with the permanent fear that anything could go wrong. There has to be power and courage. We should put faith in people and encourage them in governance voting. We should award them and not punish them with 6month lock-ups and rejecting neuron trading.
Everytime something is strongly forbidden/unwanted , people will find their paths, especially in crypto space. If neuron trading is forbidden like in the example above, the free market will get into p2p/otc trading.
The next possible approach for the free market: Proxy voting and lending voting rights. Not buying/selling ICP directly, but only votes to certain decisions.
What prevents someone from calling out a certain amount of money for a certain amount of ICP on the next vote? Nothing. What prevents someone from accepting it? Nothing. That’s an infinite story, and the more ICP loses its head in something that can’t be controlled, the worse it will get.
Many of the fears could disperse, via establishing Point 5 with the Polkadot approach. Governance rights via weighted lock-ups. An additional consitutional approach could be that changing the lock-up weights would not only require a simple majority, rather a 2/3 majority. Maker protocol also shows how something could be done: They’ve got an emergency shutdown which can be triggered by a 50k MKR burn deposit (only 5% of the supply). It might not be directly related in terms of topic but it’s just another example where people have much more faith in governance, and take responsibility, than our community might assume. Liberal market + strong governance.
tldr: The free market will find its paths anyway. We should focus on establishing a great framework for all these processes. ICP should find a way to incentivice people to vote for a bright future and to engage with the protocol in the long term. Not via controlling what can’t be controlled: Humans
My comparison with Polkadot in a comment above should show how that could be done (Point 5).
Point 6: @wpb
I’d really like to hear more about automatic compounding. I doubt that it will fix the issue, as I’ve described above. Just let me know more about it (Calculation, and how it could be made fair).
Your alternate suggestion is worth a try:
If automatic compounding is not going to be implemented, then pegging the manual Spawn Neuron action to a fixed Maturity threshold instead of a fixed ICP reward threshold also achieves this goal of equality. However, pegging to a fixed Maturity creates new problems such as how long it will take to achieve that Maturity.
I had time to think about it and I actually don’t find any way how that could be used by high stakes to their advantage. If spawning a new neuron is fixed to a certain maturity (e.g. 2.5%), everyone has got the same condition (neuron age not considered here). If 2 Person (A 10 ICP, B 1kICP) decide to stake at the same time, they’ll be able to spawn a neuron at the exact same time (e.g at 2.5% maturity). That’s fairness.
If there is concern that the liquidity of ICP by holders (especially big pockets) is decreased, we’re at the perfect track. => If someone fears that his advantage isn’t here anymore, there might be something we got right. Now we get to a point where the minimum 6month lock-up period could really hurt.
Assets staked for that amount of time and not getting any liquid ICP that fast because of the 2.5% minimum maturity => That might be too much and it’s rather a punishment than an award.
Here comes point 5 into play again.
6 Likes
@MichaelAnnh My definition of automatic compounding is related to making it easier to apply reward ICP to the staked balance of ICP in a neuron.
Currently it is a manual process since we have to wait for the maturity to represent at least 1 ICP of voting rewards, click Spawn Neuron, go to the new neuron and click Unlock, wait 7 days for the dissolve delay to expire, click Disperse to release that reward ICP out of the new neuron, go back to the parent neuron, and click top up neuron. It’s just an undesirable way to compound your ICP rewards.
So automatic compounding would automate these tasks as a minimum solution without changing any of the required steps or timing.
Automatic daily compounding would apply the daily voting rewards to the staked balance instead of building maturity. It would be an option for people whose goal is to compound instead of liquidate their rewards. This is an option that should be configurable any time, whether your neuron is locked or dissolving. The reason this is fair for everyone is that everyone receives the same voting reward rate of return every day no matter how many ICP are staked assuming the same neuron configuration of dissolve delay and age.
3 Likes
I would like to agree with the possible dangers of relying on non-transferability of neurons. My intuition tells me it would be incredibly difficult if not impossible to actually prevent neuron transfers, and I agree with the idea that attempting to prevent them could lead to undesirable outcomes.
I’m not sure if neuron transfers should be fully embraced, or what mitigations should be put in place…but the idea that you can fully prevent transfers should probably be put to rest, and fighting them may prove disastrous.
Deep thought on this is warranted.
4 Likes