FIP-0093: Set the Mining Reserve to Zero #1039
Conversation
jnthnvctr
requested review from
momack2,
arajasek,
jennijuju,
kaitlin-beegle,
anorth,
raulk,
jsoares and
TippyFlitsUK
as code owners
|
|
||
| ## Simple Summary | ||
|
|
||
| This proposal aims to remove the mining reserve tokens [282.9m](https://filfox.info/en/address/f090) by setting its value to 0. This does not preclude other FIPs from being proposed to create new tokens (as future incentives, funding iniatives etc). |
There was a problem hiding this comment.
What does remove mean? burning?
Edit: in the later session "removing from total supply" - lets add it here so it clearer since the begining.. "remove...... and the Filecoin Toal supply will be updated from X to Y"
|
|
||
| The [mining reserve](https://spec.filecoin.io/systems/filecoin_token/token_allocation/) was intended to be a reserve of tokens to incentivize future types of mining. In the spec, this is noted to potentially "not be enough" and that the community will "decide whether to make adjustments with unmined tokens". | ||
|
|
||
| This proposal seeks to remove the mining reserve. This does not preclude new tokens from being created - rather that any new incentives should be created at the point when new forms of mining are introduced rather than relying on a set of pre-minted tokens. |
There was a problem hiding this comment.
is the proposal to burning tokens, or are you proposing to change
- the fixed max filecoin token supply to be (2B - mining reserve)
- move Filecoin away from a fixed token supply economy to an unfixed one (unlimited minting)?
There was a problem hiding this comment.
This proposal only touches the mining reserve - I think the other discussion (on uncapping baseline minting) had less momentum behind it.
I still think thats a discussion (honestly rethinking how inflation works) that should be on the table - but not a part of this proposal
Tbc the part that does touch filecoin as not being a fixed supply currency would be normalizing the idea that FIPs might include proposals to mint new tokens (but also then require sufficient buy in from the community to accept those proposals)
There was a problem hiding this comment.
This proposal only touches the mining reserve - I think the other discussion (on uncapping baseline minting) had less momentum behind it.
I still think thats a discussion (honestly rethinking how inflation works) that should be on the table - but not a part of this proposal
I don’t see how my comment lead to any of these points. The editing suggestion I’m trying to give you here is similar to #1039 (comment) - please be more specific on what you mean by “removing”
| 1. The role of the mining reserve is non-standard - at the time of authoring, no other L1 has this amount of supply left up to potential release (~50% of the current supply, 15% of the total max supply) which creates a large gap between the FDV and Market Cap. | ||
| 2. The mining reserve reinforces an anchor around Filecoin being a fixed supply currency. Given there is always the ability for the community to fork to change the supply, this is an unnecessary meme to reinforce - but constricts discussion. | ||
| 3. The mining reserve requires a FIP to use (the same process for creating new tokens) - in the event these tokens are needed; its unclear there would be any _less_ pushback in the community if were to create new sources of dilution. | ||
| 4. Filecoin's inflation is already quite high relative to other ecosystems - removing the mining reserve should change the default from creating net new sources of inflation to bring new value flows into the ecosystem (vs optimizing the inflation already in the network) |
There was a problem hiding this comment.
Peer review: If the goal is to fix inflation, burning tokens is not the right strategy. Working on getting paid demand for filecoin usage and services and increasing FIL utilization will be a more sustainable strategy. (and remove the mining reserve may prevent new mining/service opportunities to exist)
There was a problem hiding this comment.
I think you're misunderstanding - I'm not saying we can't have new sources of inflation. I wrote at the top we could - the difference here is that we're not relying on an argument that 300m tokens were already printed therefore its the amount we should be willing to spend (which given FIL's fiat value changes substantially has been worth anywhere between $900m and 69b).
The default should not be to keep our existing high inflation and just add more tokens in the mix - but that doesnt preclude it if thats what the community really wants.
I don't think setting the mining reserve actually changes the activation energy on creating new sources of dilution - we'd still be having a discussion about how much dilution tokenholders are willing to accept.
| 4. Filecoin's inflation is already quite high relative to other ecosystems - removing the mining reserve should change the default from creating net new sources of inflation to bring new value flows into the ecosystem (vs optimizing the inflation already in the network) | ||
|
|
||
|
|
||
| ## Spec |
There was a problem hiding this comment.
I think you should include a design rationale session to compare the the pros and cons between burning tokens vs set values to 0
There was a problem hiding this comment.
Agreed but that should be part of the rationale section, where it already is -- possibly not at the time of the comment. In any case, that will need further work, and this section will need revamping too.
|
The first editor pass is ✅ - need clarification on what "remove" means in multiple places and is missing product, security and incentive considerations (see the template here); The rest is clear to me. |
|
|
||
| The proposal aims to move away from relying a fixed amount of tokens in the mining reserve that might be used for new mining, to having new FIPs be proposed explicitly creating (or redistributing existing) incentives. | ||
|
|
||
| There has been some discussion about whether this should be achieved by setting f090 to 0 or by burning the tokens in f090. |
There was a problem hiding this comment.
"There has been some discussion" etc isn't necessary, just lay out the alternatives and considerations.
Please describe what burning means: crediting the tokens to f099 so that the total (accounting) supply remains 2bn.
Some more considerations expounded in this thread.
| 1. Set the balance of `f090` to 0. | ||
| 2. Adjust all circulating supply dependencies and the invariant checks to recognize the total supply is no longer 2b given the mining reserve has been set to 0. The total supply of the token should be the total supply of FIL should be ~1717066618.96 (2b minus the current amt in [f090](https://filfox.info/en/address/f090)). |
There was a problem hiding this comment.
This section is meant to refer to the actual code, rather than re-state the spec. Delete this for now, we can link a PR later.
|
Please take FIP number 0093. Update the filenames and text to match that. |
anorth
changed the title
|
|
||
| ## Simple Summary | ||
|
|
||
| This proposal aims to remove the mining reserve tokens [282.9m](https://filfox.info/en/address/f090) by setting its balance to 0. This does not preclude other FIPs from being proposed to create new tokens (as future incentives, funding iniatives etc). After this FIP, the total supply of FIL should be ~1717066618.96 (2b minus the current amt in [f090](https://filfox.info/en/address/f090)). |
There was a problem hiding this comment.
It appears that the discussion is quickly converging on burning the balance and maintaining the total supply, so this will likely be updated. I'd anyway avoid referring to a specific value for the total supply given that's ever changing.
|
|
||
| There are a number of reasons why the mining reserve can be viewed as a net cost: | ||
| 1. The role of the mining reserve is non-standard - at the time of authoring, no other L1 has this amount of supply left up to potential release (~50% of the current supply, 15% of the total max supply) which creates a large gap between the FDV and Market Cap. | ||
| 2. The mining reserve reinforces an anchor around Filecoin being a fixed supply currency. Given there is always the ability for the community to fork to change the supply, this is an unnecessary meme to reinforce - but constricts discussion. 300m should not be assumed to be the default (in either direction as a maximum or minimum) for the amount of tokens available for new mining - any new proposal should include its own assessment for the tokens required. |
There was a problem hiding this comment.
I really don't think this has any impact. Filecoin is a fixed supply currency. Sure, that can always be changed, but I don't think the existence of a mining reserve has any impact on that perception or determination.
| 1. The role of the mining reserve is non-standard - at the time of authoring, no other L1 has this amount of supply left up to potential release (~50% of the current supply, 15% of the total max supply) which creates a large gap between the FDV and Market Cap. | ||
| 2. The mining reserve reinforces an anchor around Filecoin being a fixed supply currency. Given there is always the ability for the community to fork to change the supply, this is an unnecessary meme to reinforce - but constricts discussion. | ||
| 3. The mining reserve requires a FIP to use (the same process for creating new tokens) - in the event these tokens are needed; its unclear there would be any _less_ pushback in the community if were to create new sources of dilution. | ||
| 4. Filecoin's inflation is already quite high relative to other ecosystems - removing the mining reserve should change the default from creating net new sources of inflation to bring new value flows into the ecosystem (vs optimizing the inflation already in the network) |
| 4. Filecoin's inflation is already quite high relative to other ecosystems - removing the mining reserve should change the default from creating net new sources of inflation to bring new value flows into the ecosystem (vs optimizing the inflation already in the network) | ||
|
|
||
|
|
||
| ## Spec |
There was a problem hiding this comment.
Agreed but that should be part of the rationale section, where it already is -- possibly not at the time of the comment. In any case, that will need further work, and this section will need revamping too.
|
|
||
| One pro to setting f090 to 0 is that it functionally has the same effect as burning - without the same connotations of simply destroying tokens for the sake of it. One slight con, is that it may not have the same "shock" value as saying burn. | ||
|
|
||
| For the intended purposes, listed above in Change Motivation, setting the mining reserve to 0 achieves the main goals (and in the author's opinion doesn't substantially sacrifice a narrative goal as the total supply of the network is still decereasing). However, it should be noted that the difference between these two in practical terms seems very slight and is mostly subjective. |
There was a problem hiding this comment.
| For the intended purposes, listed above in Change Motivation, setting the mining reserve to 0 achieves the main goals (and in the author's opinion doesn't substantially sacrifice a narrative goal as the total supply of the network is still decereasing). However, it should be noted that the difference between these two in practical terms seems very slight and is mostly subjective. | |
| For the intended purposes, listed above in Change Motivation, setting the mining reserve to 0 achieves the main goals (and in the author's opinion doesn't substantially sacrifice a narrative goal as the total supply of the network is still decreasing). However, it should be noted that the difference between these two in practical terms seems very slight and is mostly subjective. |
|
|
||
| ## Product Considerations | ||
|
|
||
| This should not affect the product as these are not tokens that are being used. Any future proposal that might need new sources of inflation is not blocked from including new incentives in their proposals. |
There was a problem hiding this comment.
Although this is subjective, I think there is a product consideration, in that these tokens were set aside to support future network functionality and that will cease to be the case. We can agree that the impact is minimal, for the reasons explained further above, but we should'd just disregard it.
|
|
||
| ## Incentive Considerations | ||
|
|
||
| The proposed change requires future explicit proposals (aiming to incentivize new forms of mining) to explicitly mint new tokens. This should not change the difficulty of making such proposals, as both (using the mining reserve, creating new tokens) would have the short term effect of diluting stakeholders. |
There was a problem hiding this comment.
Likewise, I think this discounts a core argument for this proposal, which is that there is currently a _dis_incentive to hold FIL due to the perceived room for devaluation. Reducing that disincentive sounds like an incentive consideration.
|
Can we put this in last call? |
The authors have to address editors reviews for this fip to land as a valid DRAFT first. |
|
@jnthnvctr are you still on this or have you abandoned it (and why?). You have a bunch of editor feedback to address, but nothing that would prevent a perfectly good FIP coming out of this that can then be properly put to the community for discussion and governance process. My impression is that there are many people in support of this idea, but they're currently pseudo-blocked from doing anything constructive with it because you "own" it having drafted this FIP. Would you please follow through? Alternatively, if you abandon it and make it clear that it's merely because you have lost interest, the space will open up for someone else to take over. |
|
bleh - didnt mean to close this out, was trying to incorporate the changes onto the fork (makign the edits here - will try and figure out how to undo and link back here so we can keep the thread) |
|
|
||
|
|
|
|
||
|
|
||
|
|
||
| # FIP-0093: Set the Mining Reserve to Zero | ||
|
|
||
|
|
There was a problem hiding this comment.
| # FIP-0093: Set the Mining Reserve to Zero | |
| # FIP-0093: Set the Mining Reserve to Zero | |
|
|
||
|
|
|
|
||
|
|
||
|
|
| ## Spec | ||
|
|
||
| 1. Set balance of `f090` to 0. | ||
| 2. Adjust all circulating supply dependencies and the invariant checks to recognize the total supply is no longer 2b given the mining reserve has been set to 0. |
There was a problem hiding this comment.
We have a TotalFilecoin that's used for a few things:
- Invariant check as mentioned here, after summing up all the accounts it checks that it's equal to this number: https://github.com/filecoin-project/go-state-types/blob/aff1b8dbe7fd175eb2d9e7ca7e5da00e125a5650/builtin/v15/check.go#L203-L206
- 3 exported methods:
DealClientCollateralBounds,DealPricePerEpochBoundsandDealProviderCollateralBounds- each of which return a min and a max, where the max is thisTotalFilecoin. In Lotus at least, we no longer use the first two, but the last is exposed asStateDealProviderCollateralBounds, which is being codified in the Common Node API @ Common Node API #1027
Should we find out precisely what this new number should be and replace it with that? Or do something else with these APIs?
There was a problem hiding this comment.
The new total supply would be exactly 1717066618961773405230063046 attoFIL. That's current TotalFilecoin minus f090's balance.
So we could do this?
| 2. Adjust all circulating supply dependencies and the invariant checks to recognize the total supply is no longer 2b given the mining reserve has been set to 0. | |
| 2. Adjust all circulating supply dependencies and the invariant checks to recognize the total supply is no longer 2b given the mining reserve has been set to 0. The new total supply value used for invariant checks and other uses of "maximum Filecoin", should be set to `1717066618961773405230063046` attoFIL. |
There was a problem hiding this comment.
I believe total supply is defined to include the burned balance. It's the 2bn sum of all address balances.
There was a problem hiding this comment.
The invariant check does a total of all address balances to get to 2b. If we're not burning, but setting to 0 as per this FIP then we're expecting a new total of 1717066618961773405230063046.
| ## Implementations | ||
|
|
||
| 1. Set the balance of `f090` to 0. | ||
| 2. Adjust all circulating supply dependencies and the invariant checks to recognize the total supply is no longer 2b given the mining reserve has been set to 0. The total supply of the token should be the total supply of FIL should be ~1717066618.96 (2b minus the current amt in [f090](https://filfox.info/en/address/f090)). |
There was a problem hiding this comment.
| 2. Adjust all circulating supply dependencies and the invariant checks to recognize the total supply is no longer 2b given the mining reserve has been set to 0. The total supply of the token should be the total supply of FIL should be ~1717066618.96 (2b minus the current amt in [f090](https://filfox.info/en/address/f090)). | |
| 2. Adjust all circulating supply dependencies and the invariant checks to recognize the total supply is no longer 2b given the mining reserve has been set to 0. The total supply of the token should be the total supply of FIL should be `1717066618961773405230063046` attoFIL (2b minus the current amount in [f090](https://filfox.info/en/address/f090)). |
|
With some precision in details around new total supply and minor whitespace cleanups this lgtm ✅ 🚢 |
|
Is there anything blocking this from being merged? Seems that two editors have approved? @jnthnvctr |
Abstract:
The mining reserve was intended to be a reserve of tokens to incentivize future types of mining. In the spec, this is noted to potentially "not be enough" and that the community will "decide whether to make adjustments with unmined tokens".
This proposal seeks to remove the mining reserve. This does not preclude new tokens from being created - rather that any new incentives should be created at the point when new forms of mining are introduced rather than relying on a set of pre-minted tokens.
Discussion:
Link