[Archived] - Introducing a Sector Duration Multiplier for Longer Term Sector Commitment

[vkalghatgi]

Authors: @AxCortesCubero, @jbenet, @misilva73, @momack2, @tmellan, @vkalghatgi, @zixuanzh

Important Notice! Taking into account many of the points raised regarding SP preferences and general community feedback, we propose revising two key elements of the proposal

• slope --> 1x
• InitialPledgeLockTarget --> 50%, with gradual ramp-up time.

Please see here for further information.

Overview

• We are opening a new discussion thread for the FIP to Introduce a Sector Duration Multiplier for Longer Term Sector Commitment (formerly Discussion #386 and PR #390).
• We have received healthy and productive community feedback upon initially starting this discussion in June. Since then, we have been continuously iterating upon this FIP to better reflect community input.
• We have engaged in discussions and AMA's via Twitter Spaces to further this community based process.
• The result is a revised FIP (#420) which addresses questions and includes feedback we have received. This FIP is also presented below.

Simple Summary

• A Sector Duration Multiplier is introduced for all sectors, including Committed Capacity sectors and sectors containing storage deals.
• A longer sector will have a higher Quality Adjusted Power than a shorter sector, all things equal.
• The Duration Multiplier is multiplicative on the existing Quality Multiplier incentive (Filecoin Plus incentive).
• Sectors with higher Quality Adjusted Power as a result of Duration Multiplier and Quality Multiplier will require higher initial pledge collateral.
• The minimum sector duration time will increase to 1 year and the maximum sector duration to 5 years.
• The SectorInitialConsensusPledge multiplier will increase from 30% to 40%.
• This policy will apply at Sector Upgrade and Extension.

Problem Motivation

Currently, storage providers do not receive any additional compensation or incentive for committing longer term sectors (whether that be CC or storage deals) to the network. The protocol places equal value on 180 to 540 day sectors in terms of storage mining rewards. However, in making an upfront commitment to longer term sectors, storage providers take on additional operational risks (more can go wrong in a longer time period), and lock rewards for longer. Furthermore, in committing longer term sectors/deals, storage providers demonstrate their long-term commitment to the mission and growth of the Filecoin Network, and are more aligned with client preference for persistent storage. Therefore, the added value of longer-term sector commitments, coupled with the compounded operational/liquidity risks storage providers incur for longer term sectors should be compensated for in the form of increased rewards.

From a macroeconomic perspective, incentives to seal for longer durations affect the circulating supply dynamics of the network, since collateral is locked for longer. As the network exists in its current state, the percentage of FIL locked on the network is more likely to decline. This FIP looks to introduce more favorable percentage locked value dynamics, while simultaneously ensuring that this increase in locking contributes to network utility, stable circulating supply dynamics, and SP profitability and optionality.

Per the economic preference to increase the percentage value locked, we also propose adjusting the Initial Consensus Pledge Mutliplier to 40%. The intention is to create long-term aligned total value locked (TVL) dynamics to support stable and predictable conditions for storage provider (SP) returns. We further discuss the problem/change motivation, and explore impacts on SP profitability and network macroeconomics in the CEL analysis brief here.

The motivation to increase the minimum and maximum sector durations is decreased sector turnover and improved network stability. Given that the network expects to grow, having sectors that expire every 6 months means that the network potentially needs to find new sealing throughput to compensate for the loss of power from expiration. This hinders the network as it continues to scale.

Another perspective is that block rewards are high now but exponentially decreasing, and these high early rewards should be used to incentivize participation that’s long-term aligned. 6 months is not long-term. In particular it's short over the scale we need stability in the supply dynamics. Increasing the minimum duration from 6 months to 1 year doubles the minimum level of commitment, and smooths out locking dynamics by stretching inflow-outflow over a longer time period, all while impacting relatively few storage providers as sector durations for CC and FIL+ are both substantially above the minimum on average.

Finally, the new maximum limit of 5 years gives SPs the option to express a long-term commitment to the network which was previously not possible with the maximum sector duration length of 540 days, and also receive commensurate rewards.

Specification

Sector Duration Multiplier

The current sector quality multiplier follows from the spec here. The notion of Sector Quality distinguishes between sectors with heuristics indicating the presence of valuable data.

Sector Quality Adjusted Power is a weighted average of the quality of its space and it is based on the size, duration and quality of its deals.

Name	Description
QualityBaseMultiplier (QBM)	Multiplier for power for storage without deals.
DealWeightMultiplier (DWM)	Multiplier for power for storage with deals.
VerifiedDealWeightMultiplier (VDWM)	Multiplier for power for storage with verified deals.

The formula for calculating Sector Quality Adjusted Power (or QAP, often referred to as power) makes use of the following factors:

• dealSpaceTime: sum of the duration*size of each deal
• verifiedSpaceTime: sum of the duration*size of each verified deal
• baseSpaceTime (spacetime without deals): sectorSize*sectorDuration - dealSpaceTime - verifiedSpaceTime

Based on these the average quality of a sector is:

$$avgQuality = \frac{baseSpaceTime \cdot QBM + dealSpaceTime \cdot DWM + verifiedSpaceTime \cdot VDWM}{sectorSize \cdot sectorDuration \cdot QBM}$$

The Sector Quality Adjusted Power is:

$sectorQuality = avgQuality \cdot sectorSize$

Proposed Protocol Change:

Introduce a multiplier based on sector duration

Name	Description
QualityBaseMultiplier (QBM)	Multiplier for power for storage without deals.
DealWeightMultiplier (DWM)	Multiplier for power for storage with deals.
VerifiedDealWeightMultiplier (VDWM)	Multiplier for power for storage with verified deals.
SectorDurationMultiplier (SDM)	Multiplier for power for storage based on promised sector duration

This SectorDurationMultiplier function proposed is linear with slope 2. See below for the function proposed.

The rationale to select this linear slope 2 function is based on a principle that the selected parameters should maximize the effectiveness of the duration incentive, subject to SP’s collateral availability constraints, while taking into account micro and macroeconomic consequences with minimal added implementation complexity. Further analysis/simulation is shown in the analysis brief CEL prepared linked above.

Therefore, the new suggested Sector Quality Adjusted Power is:

$sectorQuality = (avgQuality \cdot sectorSize) \cdot SDM$

Change to Minimum Sector Commitment

We propose a minimum sector commitment of 1 year (360 days). This is a 180-day increase from the current minimum of 6 months (180 days). This will not change the mechanics of sector pre-commit and proving, it will just adjust the minimum sector commitment lifetime to 1-year.

Change to Maximum Sector Commitment

We propose a maximum sector commitment of 5 years. This is an increase from the current maximum sector commitment of 540 days. Note, the protocol currently sets a maximum sector lifetime to 5 years (i.e sectors can be extended up to 5 years). This FIP would not adjust that. So, the maximum sector commitment of 5 years would now equal the maximum sector lifetime. Therefore, upon sector-extension, the maximum sector-extension of up to 5-years remains the same.

Change to PreCommitDeposit

With this FIP, sectors can get higher quality and expect higher expected rewards than currently possible. This has an impact on the value of the PreCommit Deposit (PCD). From the security point of view, PCD has to be large enough in order to consume the expected gain of a provider that is able to pass the PoRep phase with an invalid replica (i.e. gaining block rewards without storing). The recent FIP0034 sets the PCD to 20 days of expected reward for a sector of quality 10 (max quality). We now need to increase this to 20 days of expected reward for a sector of quality 100 (the new max quality) to maintain the status quo about PoRep security.

Initial Pledge Calculation

The change we propose to status quo Initial Pledge Calculations is the change to the SectorInitialConsensusPledge calculation as detailed below.

The protocol defines Sector Initial Pledge as:

$SectorInitialPledge = SectorInitialStoragePledge + SectorInitialConsensusPledge$

Currently,

$SectorInitialConsensusPledge = 0.3 \cdot FILCirculatingSupply \cdot \frac{SectorQAP}{max(BaselineTarget, NetworkQAP)}$

We propose changing the calculation to a multiplier of 50%:

$SectorInitialConsensusPledge = 0.5 \cdot FILCirculatingSupply \cdot \frac{SectorQAP}{max(BaselineTarget, NetworkQAP)}$

Impact on Fault and Termination Fees

We currently propose no change to status quo Fault and Termination Fee calculations. Fees continue to be based on expected daily block rewards. In the future it may be valuable to re-examine the 90 day duration for the termination fee.

Design Rationale

Supporting Longer-Term Commitments

The current maximum commitment of 1.5 years limits the ability for SPs to make a long-term commitment to the network (or get rewarded for it). We expect that increasing the maximum allowable commitment of 5 years, while also introducing thoughtful incentives to seal sectors for longer, can increase the stability of storage and predictability of rewards for SP’s. This is further discussed in the sections below.

Incentivizing Longer Term Commitments

Longer term commitments are incentivized by a rewards multiplier. The multiplier increases the amount of FIL expected to be won per sector per unit time based on the duration the sector is committed for.

The proposed rewards multiplier is linear in duration. This means sectors recieve rewards at a rate linearly proportional to duration.

Example:

• Storage Provider A commits sectors for 1 year that generate on aggregate 2 FIL/day on average. By the end of the commitment, Storage Provider A expects to have received 730 FIL.
• Storage Provider B agrees to store the same data, but makes a commitment to store it for 3 years. Since their commitment is three times as long, they receive 6 FIL/day on average. At the end of the first year of their commitment, they are expected to have received 2,190 FIL.

The rationale is that operational burden and risk to storage providers increases with duration, and it is fair that they’re commensurately rewarded for this.

To maintain protocol incentives that are robust to consistent storage, the amount of collateral temporarily locked for the duration of the sector must also increase. Sector sealing gas costs do not increase with the multiplier. This means longer durations have higher capital efficiency, which further incentives longer commitments.

The form of the duration incentive multiplier is linear with slope 2. The factors behind this specific design choice to incentivize longer commitments are:

• Simplicity. Rewards proportionate to risk, with an understandable incentive mechanism that’s easy to reason about.
• Sufficiency. Sublinear may be inadequate to incentivize storage providers to accept the burden of risk longer commitments entail.
• Supply. In terms of percentage of available supply, a slope of 2 is identified in simulations as the lowest level that sustains circa 50% percentage available supply locked (two-thirds higher than current value). This is important to incentivize long-term commitments as it supports a stable business environment..
• Capital. A more aggressive slope was not chosen on the basis that capital may be less available in the future, and long sector durations with high rewards multipliers, which require increased collateral, should be widely accessible.

Refusing Shorter-Term Commitments

Currently the minimum sector duration is six months. A new minimum duration of one year is proposed. The rationale is based on three factors:

• Stability. A one year minimum smooths out locking dynamics by stretching inflow-outflow over a longer time period.
• Efficiency. Waves of expiration on a six month basis have the potential to waste resources resealing sectors at twice the rate of one year minimum sectors.
• Ethos. Filecoin has an exponential rewards emission schedule, with high rewards that have been useful to bootstrap the network. As the network matures this perspective should be refined to better incentivize storage providers who are aligned with the long-term goals of the network. One year is only six months longer than the current minimum duration, but it shows much stronger commitment to the long-term principles and long-term success of the network.

Furthermore, there is empirical evidence from the duration of sectors sealed that most storage providers support sectors greater than one year. Increasing the minimum from six months to one year will discourage only the most short-term-aligned storage providers.

Improving Stability of Rewards

A stable investing environment is needed to support long-term storage businesses. A high double-digit percentage return on pledge locked is not sufficient alone. To this end a sustained and substantial amount of locked supply is also needed.

In reality the percentage of available supply locked has been decreasing since September 2021. While current token emission rate is exponentially decreasing with time, the percentage of available supply locked is expected to continue to decline, at least until the linear vesting schedule completes, based on current locking inflow-outflows and network transaction fees.

This environment can be improved however. The first way to improve it is by increasing the 30% multiplier in the InitialConsensusPledge to 40%. This is a moderate increase that provides a solid long-term improvement in percentage of available supply locked. The second way is a corollary of the duration multiplier incentive. Longer sectors mean collateral is locked for longer. All else equal, at equilibrium this means the total amount of locked collateral is consistently higher. Simulations confirm both effects together can target a percentage of available supply locked that is sufficiently high and sustained to substantially improve the long-term storage business environment. See Supplementary Information for a detailed summary of the supporting simulation analysis

Rebalancing SP Profitability

Return on investment from pledged collateral provided by the storage rewards are currently substantial, with Filecoin-denominated returns in high double digits. Yet Filecoin-denominated returns are only part of what is needed to support successful long-term storage.

Simulations indicate a better balance between current and future rewards can be achieved through the proposed changes. The proposals adjust the Filecoin-denominated minting-based returns to a more sustainable level in the immediate term, while the long-term trajectory is unchanged. This enables improving the percentage locked supply to stabilize the business environment for long-term network success. See Supplementary Information for further details for percentage return on invested collaterals from mining reward.

Impact on Initial Pledge

The initial pledge per raw byte power will increase. This is by design. It intends to increase the percentage of available supply locked.

The initial pledge per quality adjusted power, which is the relevant measure for storage provider’s return on pledge invested, may be marginally higher than current to begin with, but will decrease with time. See Supplementary Information for plausible trajectories across different new average duration scenarios.

Impact on Pre-Commit Deposit

FIP-0034 sets the pre-commit deposit to a fixed value regardless of sector content. From the security point of view, PCD has to be large enough in order to cover the expected gain of a provider that is able to pass the PoRep phase with an invalid replica (i.e. gaining block rewards without storing). The recent FIP0034 sets the PCD to 20 days of expected reward for a sector of quality 10 (max quality). We now need to increase this to 20 days of expected reward for a sector of quality 100 (the new max quality) to maintain the status quo about PoRep security.

As of end of July 2022, the calculations are roughly:

EpochReward := 20.53 * 5 = 102.7 FIL
NetworkPower := 17.98 * 2^60 Bytes
CirculatingSupply := 330.7 * 10^6 FIL
// Sector quality = 2
StoragePledge := 0.0196 FIL
ConsensusPledge := 0.4386 FIL
PreCommitDeposit := 50 * StoragePledge = 0.9802 FIL
InitialPledge := StoragePledge + ConsensusPledge = 0.4582 FIL
// Sector quality = 100 (values are approx 50x greater)
StoragePledge := 0.9802 FIL
ConsensusPledge := 21.93 FIL
PreCommitDeposit := StoragePledge = 0.9802 FIL
InitialPledge := StoragePledge + ConsensusPledge = 22.9102 FIL 

Note:

• per the change proposed by the FIP, the ConsensusPledge is calculated as:
• $SectorInitialConsensusPledge = 0.4 \cdot FILCirculatingSupply \cdot \frac{SectorQAP}{max(BaselineTarget, NetworkQAP)}$
• The minimum Sector Quality is now 2 per the duration multiplier function outlined in the Design Specification.

Backwards Compatibility

This policy would apply at Sector Extension and Upgrade for existing sectors.

Test Cases

N/A

Security Considerations

Risks of Faulty Proof-of-Replication

The existing 1.5 year sector duration limit in effect provides a built-in rotation mechanism that can be used to turn over power in the event we discover a flaw in PoRep. Increasing the maximum commitment to 5 years weakens this mechanism. Alternative policies have been suggested. This is something the community must be aware of and agree on.

Risks to Consensus

The proposed rewards multiplier increases potential risk to consensus. The main consideration is how long it would take for a colluding consortium of storage providers to exceed threshold values of consensus power.

Analysis indicates a malicious consortium would need consistent access to high levels of FIL+, and near-exclusive access to the maximum rewards multiplier, for a substantial period of time, for a viable attack.

Example:
The network currently has 18 EiB of quality adjusted power.
Consider the scenario of 50 PiB/day onboarding, with 5% attributed to FIL+, and that this is sustained for several months.

Now if the malicious consortium can acquire 50% of FIL+ deals and commit sectors for 5 years to gain the maximum duration multiplier, and all other storage power maintain the lowest possible duration sectors of 1 year, then in a single day, the adversarial colluding group is expected to gain 0.7% of consensus power. This follows from:

1. advPower = advFILplusPct * FILplusMultiplier * durationMultiplier * powerOnboarding * FILplusPct
2. advPower =  0.5 * 10 * (2 * 5) * 50 * 0.05 = 125

where advFILplusPct is the fraction of FILplus deals available that are acquired by the adversary, FILplusMultiplier is the 10x FIL+ power multiplier, durationMultiplier is the maximum 5 year duration multiplier (5 * 2), powerOnboarding is the byte power onboarded, and FILplusPct is the fraction of the power that is FIL+.

3. honestPower = 0.5 * 10 * (2 * 1) * 50 * 0.05 + 1 * (2 * 1) * 50 * 0.95 = 120
4. advPowerDailyPctGain = advPower/(18*1024 + honestPower) 
5. advPowerDailyPctGain = 0.7%

If this scenario is maintained, the adversarial group is expected to exceed 33% of consensus power within 140 days.

Factors that mitigate this risk are that it’s unlikely a single group could achieve 50% of FIL+ power consistently, and unlikely that the adversarial group exclusively takes up the longer duration sectors with enhanced power multipliers.

A limitation is that the above calculation assumes the malicious party is starting from 0% of consensus power. If they already control 10%, time to 33% is reduced to approximately 100 days.

Rollout Shock

Rollout shock could occur if SPs race to extend their commitments and gain a further 10x multiplier. This could be mitigated by gradually increasing the maximum multiplier 1x to 5x during an initial period (e.g. first three months)

Product & Incentive Considerations

As discussed in the problem motivation section, this FIP introduces incentives to further align the cryptoeconomic schema of the Filecoin Network with intended goals of the network to provide useful and reliable storage. We introduce the idea that longer term sectors represent a long-term investment and commitment to the Filecoin ecosystem, and therefore should be rewarded proportionally with greater block reward shares.

Note, we also introduce the possibility for storage providers to receive additional multipliers from committing CC for longer. Even this has added value insofar as it represents a commitment to the ecosystem long term that should be rewarded.

From a product perspective, we see strong support for a network more aligned with longer-term stable storage. From a recent (< 3 week old) snapshot of all LDN applications, the responses fall into the buckets below. Almost half (47%) of all applicants want long-term or "permanent" storage.

Period	Count	Percentage
1 to 1.5+ years	56	22%
2+ years	22	9%
3+ years	40	16%
5+ years	15	6%
Long-term/Permanent	120	47%

We recognize that this proposal may not align with a small fraction of SP’s who exclusively prefer shorter commitments to the network, but contend that from an ecosystem perspective, this policy on aggregate makes most participants better off. Note, regular deals can still be accepted for less than sector duration, so there should be minimal loss to flexibility for onboarding clients.

For smaller SP’s, introducing this policy could help improve their competitiveness and ability to capture network block rewards, Under this proposal, returns on tokens put up as collateral scale linearly for all SP’s (regardless of size), whereas only larger ones are able to take advantage of economies of scale for hardware. This proposal, if anything, should benefit smaller SP’s because they can still get rewards boost/multipliers without prohibitively expensive hardware costs, and termination risks associated with FIL+ data.

Implementation

TBD

[f8-ptrk]

how do we avoid a pump and dump scenarios on the filecoin price with this?

• coingecko shows 246 million fil in circulating supply
• filfox shows 136 million fil pledged

there is not enough fil circulating to fully x5 the whole network. if we have one big enough player buying fil early and aggressively as soon as this is announced we might see unpleasant price movements that make it feasible to buy, extend, terminate i think

This policy will apply at Sector Upgrade and Extension.

if this is a good idea - i do not know

[edit]

in fact that player wouldn't even need to actually extend - the FOMO this creates might be already enough

[tmellan]

coingecko shows 246 million fil in circulating supply
filfox shows 136 million fil pledged

there is not enough fil circulating to fully x5 the whole network.

This is not a concern since the multipliers scale QAP. There's a QAP denominator in initial consensus pledge. Furthermore the pledge is multiplied by circulating supply. Both of these factors regulate the percentage of available supply locked. This provides an increase that is expected to saturate at the value we've pointed out around 50%.

[tmellan]

@f8-ptrk rollout type shock is mitigated by several factors
i) using a lower multipler, following on from SP WG discussion, we think 1x slope may be sufficient
ii) gradual ramp up of InitialPledgeLockTarget over 3 months
iii) 100% of SP taking 5y is unlikely. 3y sectors will present good ROI. Different SP have different risk profile and preferences causing a distribution of durations
iv) 100% change to 5 years on the first day is extremely unlikely. The incentives regulate against this. In general it is beneficial to seal early, but it is not that strong, and the more that seal early, the lower the pledge / qap for those who seal later, so seeing a step change in duration is not likely.

[cryptowhizzard]

@tmellan where have you been lately? Didn't someone give you the memo that there is a huge amount of FIL in the hands of a few people who initially participated in testnet and got all there FIL moved to mainnet? For them it is easy to instantly X10 all their sectors and push the smaller and mid sized miners out because they have that FIL available. The smaller miners don't have that FIL available.

So what you are doing here is give them staking rewards on their FIL's payed by the smaller miners blockrewards who wil go /10 and who can't afford to X10 even if they want.

[tmellan]

Hi @cryptowhizzard

General point:

• current suggested policy is 5x rather than 10x for 5y max

@tmellan where have you been lately? Didn't someone give you the memo that there is a huge amount of FIL in the hands of a few people who initially participated in testnet and got all there FIL moved to mainnet?

Concentration of tokens vs sufficient incentive to bootstrap project: standard perspective is early supporters get larger rewards and this is fair because they took early risk. This was before my time, others can comment on how close to optimal the design was.

For them it is easy to instantly X10 all their sectors and push the smaller and mid sized miners out because they have that FIL available. The smaller miners don't have that FIL available.

• SPs who've earned tokens should have the opportunity to put them to use to support the network.
• Collateral is likely easier to acquire than to scale operations, which is a factor that supports small SPs
• Large SPs may have more collateral but typically in proportion to their QAP. Relative proportions remain unchanged
• If others lock a lot, this is good your network. It increases stability. And your pledge lowers as per modeling

[cryptowhizzard]

It does not matter if it was before your time or not. They did not take a bigger risk then others, thry got testnet (free) sectors converted to mainnet sectors with value.

[f8-ptrk]

will this change the min deal duration from 6 to 12 month too?

if not - when is re-snapping planned so miners are not forced to keep "dead data" sectors proving (at least 6 month, 4.5 years in the worst case) just because it is more profitable than not doing so

not everyone wants to store for 5 years and for most of the data out there it makes no sense to store them for long periods of time

making long term commitments to keep sectors available, storage sealed shouldn't come with in flexibility in how data can be stored in these sectors.

there is data out there that MUST be deleted after a certain period - re-snapping would give us a pseudo proof of deletion in some wired way

[zixuanzh]

with 5 years it becomes even more ridiculous than it already is.

People don't have to do it for 5 years if they don't want to do it.

[zixuanzh]

Can someone give me the stats on CC being snapped last 6 months?

This is a good question but the answer might be something else which is the demand for using the network. See this funnel dashboard for Fil+, the network does not even have consistently more than 10PiB of Fil+ application everyday (and less than 1PiB of allocation/day). So the answer isn't just about the dynamic of CC vs Fil+. This FIP does not intend to change that dynamic between the two options.

A key thing here is that Filecoin as an ecosystem might need more publicity, more eyeballs, more help, more resources, more applications to do more business development (and in more creative ways) to drive up storage demand. This FIP aims to help in that direction as well.

[f8-ptrk]

yes. those are the rules right now. 6 month to 18 month. now you put an incentive on doing it longer - what do you think will happen? it's called an INCENTIVE! if i do not do it i will loose out. you guys really that naive?

This has to do with QAP accounting and how the QAP is "smeared" over the sector duration in implementation. Not the most ideal situation either but it's out of scope for this FIP.

yes it is. but you have to ack that it might be a good idea to implement that first! you do not build buildings starting with the roof!

this FIP is a lot. it's not something small. it's changing the protocol in a way you, SPs, no one has a clue what we will end up with. graph assumptions as you want - i am sure you did that before mainnet when the initial params were decided, and still: here we are - you plotting graphs again but this time ask people to trust in your assumptions, future predictions for 5 years...

all i am asking for to not build your dream on sand but do the ground work first.

[herrehesse]

So then “keeping all things equal” it all comes down to whoever has access to the biggest amount of capital. This will surely increase price short term, but all normal miners and current healthy businesses without access to large sums of filecoin to lengten their CC x5 will be outcompeted.

It is simple economics. Normal lifetime sectors will be heavily outcompeted by 5 year sectors. All average miners will cease to exist. Deals are still capped by bandwidth limitations of the overall network and large CC operations with no intention of storing data, but access to large amounts of FIL, will win.

ONLY solution: cut back the multiplier tiers.

Why 500% and not 200% ?

[tmellan]

Long durations are good for the network but entail more risk. This needs to be rewarded to ensure a good distribution of sector durations all the way up to 5y.

Even at the max of 5y (and there will be a distribution not absolute concentration at 5y), the reward is only half (5x) the extra incentive for useful data (10x). This seems fair given the extra risk to store for 5y, and the value that SPs doing this bring to the network.

[TippyFlitsUK]

Summary of Storage Provider questions and concerns raised in the EMEA SPWG meeting 1st August 2022.

• Why is the minimum being raised? This will have a significant impact on our machines which are set at minimum to reduce hands on commitment of loaner machines?
• This is too extreme on a lot of levels. Who has done the maths? There is a benefit, block rewards, longer deals is more block rewards. Also, doing the maths, the ecosystem can not hold this multiplier. I suggest we do: x 1.1 / x 1.25 / x1.5 (not x10)
• FiL price jumps all around. In the future, we need something like an FVM contract on deals. We cannot do a 5yr deal with a fixed price per GiB that will be fixed for 5yr
• What's the timeline? should that land before the fil+/consensus decoupling?
• Has the Crypto Econ Lab queried how many SPs set sectors to minimum and why?
• It will cost 6 million USD to seal 1PiB, SPs are not banks!
• Most miners will quit, only a few large operators will remain.
• Multipliers are not workable with the current proposal.
• Rewards will NOT actually increase, this will kill the wrong miners.
• Why are multipliers being applied to CC sectors and not just verified deals?
• The risk is the CC sectors multiplier
• SPs just sealing CC sectors will be rewarded more than SPs who are actually storing real data.
• We will end up losing miners that we really don't want to lose.
• Many miners will have no choice but to cash out.
• Some FIL lenders require a $5 million or higher minimum SP valuation, this would hollow out the middle.
• Will it be possible to extend existing sectors?
• I like the proposal in general but the multipliers need to be changed.
• For approval of this FIP should there be a signed vote count?
• Most SPs will just become staking providers, they won't be able to compete with the "block mining" chain.
• This will destroy the network, 10X for CC is not viable.

Summary of CryptoEcon & PL comments:

• Rewards will increase for miners - General disagreement from SPs.
• May need to look at a more gradual roll-out.
• Nothing is finalised - open to changes even after the FIP potentially passes.
• CC sectors have utilization as they can be converted - General response from SP's is that the proposed CC multiplier is still too high.
• We need to facilitate longer deal durations - General SP agreement but not with the current multiplier proposals.
• More time needed to discuss - there is no official due date we want everyone to feel comfortable.
• We shouldn't drag our feet on this either as the current state of the network needs to be addressed.

[herrehesse]

What now happens with FIL+ is that there is a HUGE cap on the availability due to healthy reasons such as:

• It takes an actual lot of work to find, process, distribute them.
• Notaries and the community checks and controls this, no one can just "get it".
• Bandwidth is an actual necessary resource, most entities did not invest in with only CC --> they are doing now which makes the network healthy!

This makes sure that no entity can simply outgrow someone with 10x because of their capital. If we push this proposal with an x5 on CC sectors we will go straight towards and proof of stake network and the most important factor of earning more and outcompeting the competition is having the largest bag of funds.

[herrehesse]

Also, commitment means nothing if you can terminate sectors for a small (90d) fee. Everything is liquid, also the sectors pledged for 5 years. We already saw this play out by the end of last year.

[cryptowhizzard]

@tmellan can you please answer here on the remarks of @herrehesse . The smaller miners do not have access to unlimited funds where the bigger miners have. It is a known fact that they got their sectors beamed from testnet to mainnet by uncle scotty 2 years ago and b/c of that are sitting on huge amounts of FIL. As the same cake has to be divided ( rewards ) it simply means that the bigger miners who can X10 with more funds outcompete the smaller ones. They have only 1/10 th of their income left.

Can you explain how this is healthy for decentralisation instead of repeating your own tune? Can you perhaps also tell us how this will attract new miners to join the space who need to start small?

[tmellan]

@herrehesse

• small (90d) fee --- agree, this should be reassessed in subsequent FIP

[tmellan]

Can you explain how this is healthy for decentralisation instead of repeating your own tune?

So I think centralisation is an important consideration we should think about in general and others should comment here too pls. But from my perspective, in this proposal, SPs have access to capital in proportion to their history of storage provision. This means there's no size bias in the proposal, in the sense that yes large SPs ofc have more tokens in absolute terms, but also more QAP to spread it over if they decide to boost to longer duration. Relative proportions and advantages/disadvantages remain the same.

Can you perhaps also tell us how this will attract new miners to join the space who need to start small?

It's easier to find collateral than scale operations which is likely good for small SPs.

[Kubuxu]

I would like to propose a modification to the proposal, dealing with increased MaxSectorDuration.

The current proposal increases the maximum sector commitment length from 1.5 to 5 years. It does it by increasing the MaxSectorDuration parameter from 1.5 to 5 years. This parameter tweak seems very simple, but it has an unaccounted externality of causing mitigation of possible PoRep bug to be much more difficult.

Proposal

The core of the proposal is to separate the period of validity of a proof from the period of commitment for a sector. This leaves the 1.5-year sector extension process in place to maintain proof validity, but allows SPs to commit to longer periods.

The existing sector Expiration property would be renamed to CommitmentExpiration with a duration range from MinSectorLifetime to MaxSectorLifetime.

A new sector property is introduced called ProofExpiration . The initial value of ProofExpiration always falls into the range of (CurrentEpoch + MaxProofDuration - ProofRefereshWindow, CurrentEpoch + MaximumProofDuration] or CommitmentExpiration whichever is smaller.

• MaxProofDuration: maximum period from last commitment or refresh for which a PoRep is valid, set to 1.5y to match current behaviour
• ProofRefreshWindow : a window of time before ProofExpiration when the proof can be extended, set to 0.5y

The proof refresh window creates a trade-off: the larger the window, the more refreshes can happen in a singular batch, but more frequently each proof must be refreshed.
For example with MaxProofDuration of 1.5y and ProofRefreshWindow of 0.5y, each extension will be at least 1y and at most 1.5y. SPs will have to refresh sectors every MaxProofDuration - ProofRefereshWindow epochs.

The proof expiration is not freely chosen by the SP, but takes a value that is derived and quantised from the sector’s activation epoch.

Storage Provider can at any point call RefreshProofEpiration(SectorSelector) requesting a refreshed proof expiration. This call only results in an actual refresh of the ProofExpiration if called within ProofRefereshWindowof the ProofExpiration.

In case of a PoRep bug, the RefreshProofExpiration method would be disallowed for sectors created with insecure PoRep.
Other policies are possible too, such as requiring re-sealing or some other computation. The mechanism can support other policies developed in response to a concrete security flaw.

It also has the following benefits:

• The bug response upgrade does not need to iterate over all sectors within one epoch. The mechanism to iterate over all sectors exists in form of an expiration schedule.
• Sectors’ ProofExpirations are spread out uniformly (according to onboarding) and are not possible to manipulate on a short timescale, which could be an issue otherwise leading to non-even distribution and terminations of sectors.
• Storage Providers must take action to avoid early sector termination when a Bug Response Policy is implemented, instead of the network having to create a new mechanism to enforce the Bug Response Policy.

Limitations of this mechanism include:

• The response period must be specified now but can be modified at cost of some additional complexity in future.
• Relative to the naive approach of just increasing the maximum sector duration, this mechanism requires more SP operational costs and on-chain activity.
  But only a little less than the amount required today to extend sectors. (Less because we won’t recompute pledge, etc).
• Relative to the naive approach, this mechanism introduces operational risk to SPs, who will pay a termination fee for failing to refresh expiring proofs.
  Since the sector duration multiplier proposal greatly increases rewards for long commitments, this operational cost might be considered just part of the risk taken in pursuit of those rewards.
  On the other hand, SPs are given a generous window of time to refresh their sectors.

For more technical details please see this document.

[anorth]

Thanks @Kubuxu, I strongly support this as the implementation path for this change, over the naiive simple change to parameters.

[vkalghatgi]

Thanks @Kubuxu. We think this is largely separate to the central issue of the FIP, which, is whether and how we want Storage Providers to be able to express a long-term commitment to the network. This FIP seeks to be as straightforward of a change to address the problem motivation outlined. However, CEL would encourage and fully support your proposal as an additional FIP following this one (if accepted), as long as it does not change the core economics of the Sector Duration Multiplier Proposal.

[steven004]

Thanks @Kubuxu , This is generally a very good idea, not only attaching to this FIP, but generally a good structure for sector extension for long run. I would think this could be an independent FIP for PoRep security and deal/Sector extension.

[anorth]

@vkalghatgi The "straightforward" change to parameters introduces a very significant risk to the network. You can't invoke simplicity as an excuse for ignoring it.

We can take discussion of the concrete implementation for this over to #415, but in my view the proposed mechanism is a pre-requisite to acceptance of the sector duration multiplier proposal.

You could say that we would encourage and fully support your proposal as an additional FIP following #415 (if accepted), as long as it does not change the core storage security properties of the Filecoin network.

[dltx-asa]

This is Hasan from DLTx
I appreciate all the time and effort put into this proposal. The discussion we had in the EMEA WG on Monday, August 1 did show that the majority of the SPs that attended did not agree with the multiplier used for Collateral for the long terms deal, and also raining the minimum period for storage. It looks like it's forcing SPs to spend more.
However, the counter argument from the proposal side did not include numbers or examples that illustrate the analysis which was conducted to ensure this proposal will not do more harm than good in the future.
We need sample scenarios of showing the amount of FIL needed and awarded for different deal sizes and periods. This is the only way that we can all be on the same page and make the argument for approving or opposing the changes

[tmellan]

Hi Hasan,

We're working on change to the proposal that can adapt the multiplier to be something SPs can better support, while still hitting of our key points: sufficient incentive that a large number of SP increase duration + increased and stable % of available supply locked. More details and supporting evidence incoming.

On increasing the minimum duration by six months, the reasoning is threefold:

Stability. A one year minimum smooths out locking dynamics by stretching inflow-outflow over a longer time period.
Efficiency. Waves of expiration on a six month basis have the potential to waste resources resealing sectors at twice the rate of one year minimum sectors.
Alignment. Filecoin has an exponential rewards emission schedule, with high rewards that have been useful to bootstrap the network. As the network matures this perspective should be refined to better incentivize storage providers who are aligned with the long-term goals of the network. One year is only six months longer than the current minimum duration, but it shows much stronger commitment to the long-term principles and long-term success of the network.

Could I ask SPs to clarify their objections? Note, for sectors sealed right now, the average duration >> 6months, indicating most SPs will be unaffected.

We need sample scenarios of showing the amount of FIL needed and awarded for different deal sizes and periods. This is the only way that we can all be on the same page and make the argument for approving or opposing the changes
Noted, we will get clearer evidence beyond already presented in the Notion Simulation Summary.

(edited to fix link)

[f8-ptrk]

you cannot get evidence for events in the future. by definition. all you can do is present educated guesses.

Could I ask SPs to clarify their objections? Note, for sectors sealed right now, the average duration >> 6months, indicating most SPs will be unaffected.

Yes you can. You present a solution - and you want to implement it NOW! ASAP!!! Not a "lets make this a string and not a []byte" change kind of solution but a "lets redefine core mechanics of the chain and change values we defined less than 2 years ago".

But the problem this should solve is not clearly defined to most of the people. It would be nice if you could

• define and present the problem and what the result of inaction will be, and when
• present the assumptions about the future that you make to come to the conclusion that the problem will be a problem in the future - assuming we go on without this FIP
• what went wrong and why do you think it went wrong

this will help people understand where you are coming from. then we can start discussing the solution. for now it looks like you want to lock up more fil soon and fast for way longer times to an extent that will have consequences for SPs you seem to not fully understand.

i know this would delay the process by at least a month or two or even more. SPs are most likely willing to take the "risk" of not implementing the solution any time soon but wait for you to fully explain why you want to do this and what the results of not doing it would be.

btw:

[f8-ptrk]

the urgency is hard to understand. we all get the long term alignment incentive thing.

[DecentCr8]

@tmellan , Do we have concrete proofs for the following reasoning? What's the other impacts? As a layer 1 design, we may want to provide more flexibilities instead of have more limitation. Increasing the minimum duration will absolutely reduce flexibility, and will push out some use case scenarios. More comments below.

On increasing the minimum duration by six months, the reasoning is threefold:

Stability. A one year minimum smooths out locking dynamics by stretching inflow-outflow over a longer time period.

Stability is more depending on decentralization and the scale of the network. This whole FIP may harm the decentralization since it will wash out small/medium storage providers; and also the scale of the network from infrastructure point of view since it moves to staking game model.

Efficiency. Waves of expiration on a six month basis have the potential to waste resources resealing sectors at twice the rate of one year minimum sectors.

Are there any storage providers brings it up? A storage provider could choose one year if she does not want to "waste" resources resealing without any problem, but, anyway, there must be a reason for a SP to choose 6 months since she knows the cost is higher. Could we ignore that reason blindly?

Alignment. Filecoin has an exponential rewards emission schedule, with high rewards that have been useful to bootstrap the network. As the network matures this perspective should be refined to better incentivize storage providers who are aligned with the long-term goals of the network. One year is only six months longer than the current minimum duration, but it shows much stronger commitment to the long-term principles and long-term success of the network.

Again, this is a free world people come and leave, we do not need to decide for people. A simple and flexible base would be fine. And, do not really understand why one-year minimum duration shows much stronger commitment to the long-term principles and long-term success of the network. What we know is that we are setting up a decentralized storage base and also a decentralized CDN. There are looooooots of data will be served in the network for a short period (6 months or less) in this fast-developing world, which we can not just ignore.

[tmellan]

Stability is more depending on decentralization

I commented on why I don't think this proposal implies more centralization here

Do we have concrete proofs for the following reasoning? What's the other impacts? As a layer 1 design, we may want to provide more flexibilities instead of have more limitation.

L2s and HC subnets can be the wild west on top, but as the base L1 we need to maximize long-term stability imho

this is a free world

hmm, debatable

people come and leave, we do not need to decide for people

As a thought experiment, current min is 180d, do you think there should be no minimum at all?

[geoff-vball]

Do we have modelling on the network QAP after this upgrade?

[tmellan]

@jennijuju yep, renewal rate refers to sector extension.

And yes, for sure, if @dkkapur has datacap projections that would be very useful. Thanks!

To clarify, our current assumptions as follows:

• The sector renewal rate is also constant on average over time. Similarly, we use the value currently achieved by the network with a level of variability around the central value.
• We also consider that all new sectors being onboarded or renewed have the same duration. This is a fixed parameter that we vary between 1 and 5 years. Throughout the report, we refer to this parameter as the “average sector duration”.
• The percentage of newly onboarded power that contains FIL+ deals is another parameter we model. To simulate an increase in FIL+ deals, we assume that the rate of FIL+ deals increases linearly at a rate of 50% per year. The initial value at the start of the simulation is the currently observed rate.
• The sector renewal rate is also constant on average over time. Similarly, we use the value currently achieved by the network with a level of variability around the central value.
• We also consider that all new sectors being onboarded or renewed have the same duration. This is a fixed parameter that we vary between 1 and 5 years. Throughout the report, we refer to this parameter as the “average sector duration”.
• The percentage of newly onboarded power that contains FIL+ deals is another parameter we model. To simulate an increase in FIL+ deals, we assume that the rate of FIL+ deals increases linearly at a rate of 50% per year. The initial value at the start of the simulation is the currently observed rate.

These were copied from here.

[jennijuju]

@tmellan appreciate your patient response! will wait for @dkkapur or other folks from DP / fil+ team to confirm wthether the rate of FIL+ deals increases linearly at a rate of 50% per year is reasonable.

[dkkapur]

@jennijuju @tmellan -> DataCap will follow demand, not the other way around. If demand shows up to use the network and its provably useful to the notaries, then DataCap will be provisioned and created. For modeling impact of QAP, we should instead look at the growth rate of deals on chain + anticipated demand over time for Filecoin from the client/data owner side. I think 50% per year is too low for this given how small our current utilization is as a % of the capacity. In the last year alone, we've gone from <20 PiB QAP to <2 EiB QAP. That's 10000% increase.

[dkkapur]

On the data onboarding side, we've been playing around with different growth models and numbers. I'd suggest using something along the lines of 2%-7% growth WoW in data stored per day instead. cc @mr-spaghetti-code do you think this is still reasonable / in line with expectations we should set on demand growth?

[mr-spaghetti-code]

Yes, I believe a target weekly growth rate of 2-7% is appropriate. Our deal pipeline is strong and we are primarily constrained by our onboarding rate (a function of tooling, transfer, and sealing rate).

[herrehesse]

I do not see any new SP's drawn to a place where they need to lock up x50 6FIL per TiB onto a harddrive to then sell that same storage for "free". This is not an healthy place to be in and we will become a staking protocol. Whoever has the most cash, wins.

The reason why FIL+ works is because of a few things:

• It is community monitored (notaries), does not always work but it is way better than unregulated
• It is not easy to receive large amounts of datacap due to the monitoring and process required
• Rules make it fair for all participants, not only the large ones
• Bandwidth and geolocation make it difficult to just easily and quickly grow, there is lots of effort needed to make it work properly

If you make it as easy as "lengthening a lifecycle" to get an x5, everyone who has the funds will do it because it is the easiest way to increase your block reward immediately. You say "great for long term commitment" and I say "everyone not able to compete will cease to exist". Hence no "decentralisation", isn't that why we are all here for?

Do not mistake wishful thinking with simple economics: miners will be FORCED to get the x5 to stay competitive. Because of the decreased FIL/TiB revenue, miners who do not upgrade will go bankrupt. Small to medium sized miners need to lend more than 300%-500% of their initial Filecoin capital to stay competitive. No lender is capable of doing this. The rich will only out-grow all others, keeping equal hardware and not doing any deals.

Please show me the maths and research on the following subjects:

• Increase in collateral costs for SP's ranged between 1-100 PiB in size
• Increase in difficulty to the network with various percentages of empty CC sectors doing x5 (between 10-100%)
• Capabilities of lenders to increase their operations by 500% minimum
• Smaller miners capabilities of surviving with a huge decrease in block rewards (decrease of between 50-80%)
• What percentage of the empty CC will snap deals and store real valuable data (currently >1%)
• What a reasonable amount of collateral value is that is stored on a 18TB harddrive (risk management)

I can keep going since I have another 50+ questions about this. But let's please continue this conversation as we all seek a beter future for the Filecoin network.

[ilovefilecoin]

Hi @vkalghatgi , @tmellan ,

Thank you for the detailed FIP, but impacts are quite difficult to perceive, I’m a small individual miner true Filecoin believer, I quit my job for Filecoin, and I'm very worry about this FIP.

I would like to highlight Two points that doesn't seem to be taken into consideration in the FIP.

Time and effort

• SPs who make the effort until now to onboard data to the network seem to be the big loosers in this FIP. They already spent hundreds/thousands of hours trying to onboard data, testing tools, dealing with clients and they have been rewarded by a x10 for that. That was the deal they have with Filecoin.
  With this FIP, someone sleeping on CC can just get a x5 on its existing power by running a simple command !

I think it’s unfair in regards to the effort these guys have done until now, and I would like to hear just one telling me that dealing with storage deals was not x10 more time consuming than pledging CC all day long; with the great work from the lotus team today even my cat 🐱 can do CC.

Lenders

With this FIP how can we survive without lenders ? If that the new rule, we can probably deal with it, but that should become part of the protocol and give any miners equal access to Fil, otherwise, this is just giving an extreme power to lenders and and unfair competitiveness to large entities.

In a more general manner you present the macroeconomic aspects, I think many of us would like to understand what is going to be the individual impacts.

It would be great if you could forecast the impact on some personas like :

• Persona 1 “Pikachu” :

  • Run on it’s own fund (no lenders)
  • Full on verified deals (>80%)
  • QaP : 2PiB

• Persona 2 “Mewtwo” :

  • Small miner (no lenders)
  • No deals . 100% CC
  • Raw Power : 2PiB

• Persona 3 “Snorlax” :

  • Large mining farm using lenders
  • No deals . 100% CC
  • Raw Power 100PiB

Please ensure that FIP is fair for small players, and respectful for the time and effort they already invest, Long live to Filecoin 🚀

[jennijuju]

.. onboard data... With this FIP, someone sleeping on CC can just get a x5 on its existing power by running a simple command !

@anorth may correct me if I'm wrong, however, if assuming #313 is finalized first or along with this fip, then that means, SPs who have been onboarding verified data, will be able to get both sector duration multiplier * fil+ multiplier upon sector extension. Reason being the fil+ multiplier will be lasting as long as the sector is still live.

[jennijuju]

If a sector with FIL+ is extended to 3y, it gains a further 3x on top of the FIL+ one it already has, so 30x.

Wanna flag this is untrue given the network protocol today, as the QAP will be adjusted according to the deal/sector spacetime upon sector extension.

[s0nik42]

@tmellan I think you have to review one of your assumption :

1/ as @jennijuju mentioned, when you extend a sector with verified deals, the Fil+ x10 is NOT extend. The x10 is related to the deals duration, not the sector duration.

2/ The network doesn’t support Deals extension => Deals have to be fully resealed when the deal expires

3/ SP doesn’t control deal extension, it’s a pure client decision.

This 3 points show how it will be impossible for SPs with deals to keep their mining block rate after this FIP.
If you take the proposed personas from @ilovefilecoin :

Pikachu
- ~85% of it’s power is related to Fil+
- in the hypothesis he get access to Fil loan
To keep is winning block ratio he will have to :
- Wait for current deal expiration (generally 18 months)
- Find new 5 years deals.
- Seal new 5 years sectors.

Mew2
- Extend his CC sectors (5min) , instantaneous, no resealing, no download, no client to find

Pikachu is the big looser here for having accepted deals : This will cost him : Time / Opex / Block reward

You should find some ways to compensate that through the FIP or simply enable this FIP when the last deal is expired.

[jennijuju]

You should find some ways to compensate that through the FIP or simply enable this FIP when the last deal is expired.

maybe #313 have to be a prerequisite of this fip

[anorth]

if assuming #313 is finalized first or along with this fip, then that means, SPs who have been onboarding verified data, will be able to get both sector duration multiplier * fil+ multiplier upon sector extension. Reason being the fil+ multiplier will be lasting as long as the sector is still live.

It will be possible, but not a done deal.

• It will only be possible if we migrate all existing deals to the new FIL+ method. It's not decided if we would do that (and it's some real work)
• FIL+ expiration will still be in the hands of clients
• In migration, we would probably set the migrated FIL+ term limits to be only a bit longer than their initial deal, rather than straight to 5y, so extending them would require client co-operation to extend their FIL+ allocation

[Fatman13]

Hello, everyone! I have compiled a list of concerns/questions presented during SPWG-CN meeting on August 3rd. Please note that these are taken on the fly by @Fatman13 which may or may not reflect what the speaker meant as things may get lost in translation or internet connection. Thank you!

• Concerns over general FIP process; old saying of code is law; fip that fundamentally changes tokennomics not anywhere in the initial roadmap but being conceived for just about 1.5 months, feels very rushed; Business planning on previous tokennomics would be completely changed, affecting millions to billions investments;
• Concerns on what if TVL didn't turn out heading toward what is expected, do we have another FIP? This would further lower the confidence people have for this project.
• Negative impact on storage market as when storage client require extension of deal or just simply initiating new deals, SP would now need much larger collateral to fullfil the request
• Negative impact to existing SPs? Stimulus to new SP for on-boarding; In favor of a lower multiplier, 1 year = 1.5 multiplier so as not to add too much cost for on-boarding new sectors; From own experience that SP mostly will go for 1 year, 4/5 year not likely;
• What are the likelyhood of this fip to be included in nv17?
• Concerns over core value of Filecoin is now de-balanced by the Fip; Long-term value of Filecoin come from PoRep, Post and retrieval which is essentially a POW+POS model. Should find the balance of POW and POS as opposed to favor one too heavily. Value of the sector should be determined by storage market and f+, but not protocol level interference. FVM will better serve the innovation and development of new storage markets and leave protocol (layer1) in a relatively easy-to-understand and clean slate;
• Concerns on Filecoin's previous value proposition of being way cheaper than centralized solutions, which may not be the case anymore (if the fip is accepted)

[lyswifter]

When do extension, let’s say for just a short time like a couple of days, will still be allowed after this FIP? And will initial pledge be calculated every time even for a short sector extension?

[FroghubMan]

Hello everyone, for this, I put forward some of my personal views, welcome to discuss:

1. Commitment sectors that can be utilized within a certain period are meaningful to the filecoin network. The CC sector that has not been used for a long time is a consumption of the network, a waste of resources, and also dilutes the income of the SP.

2. The SP is harder to exit and harder to enter, which will not make the network more stable. Pow's network does not have these problems, but with the market economy, it is difficult to automatically maintain the network, and the network is equally stable.

3. The storage capacity of the current network will soon exceed 18EiB, far exceeding the initial network baseline and expected storage capacity, but very little is really used. I think it should be how to make full use of the space, rather than blindly making promises. The area continues to grow.

4. It seems reasonable that the longer commitment of FIL+, the more rewards can be obtained.

[AxCortesCubero]

I see some concerns about this FIP driving up token price (as new players wanting to seal 5y CC try to amass FIL for collateral), to a level where current SP's would not be able to afford collateral.

While it is impossible to predict what will happen to Token price, and it is a fruitless effort to try to predict this accurately, let's follow the assumption from these concerns and see if they are justified.

The implicit (and not unreasonable) assumption is that token price would rise (proportionately?) with network QAP (as more SP's compete to amass FIL). Ok so let's assume this.

In this case the collateral that needs to be paid for a sector is also inversely proportional to network QAP, so the amount in fiat that needs to be paid remains constant.

Also the amount of block reward that the SP receives, is inversely proportional to the network QAP. So while they will receive a smaller piece of the pie as network QAP grows, assuming token price increases also, then their block reward remains unchanged too in terms of fiat.

[AxCortesCubero]

Let's also add another potential gamble. Lets say, as well as FIL price being proportional to network QAP, it is also inversely proportional available circulating supply, because of the concerns here that all the 5y CC sectors will snatch up all the FIL out there.

This case is actually even more favorable to the current SP's (opposed to the previous example which was neutral).

Collateral paid (at least the consensus pledge which is the much larger fraction) is also proportional to circulating supply, so less FIL will be paid, meaning no change to the amount paid in FIAT.

Now, Block reward is not actually proportional to circulating supply, so FIL price is going up, and block reward stream remains unchanged, this means the reward is actually larger in fiat terms.

Again, I am not betting this is what will happen to token price, I have no idea what will happen to token price.

But it has been expressed that there is worry that FIL price will go up and this will be bad for SP's. I am saying IF FIL price goes up because of this FIP, that will be a net good thing for SP's.

[f8-ptrk]

@tmellan

quick question:

slope 2 + 40% vs slope 1 + 50% and maybe lets do slope 0.25 and 80% in addition

how do the pledge amount / 32gib curves look over time if we assume 10% of the non 5year sectors go to 5 years each month until we have the whole network 5year'ed

[herrehesse]

Can someone show me the actual calculations for SP's of various sizes and their expected growth and collateral requirements for 1x 2x 5x and 10x?

I'd like to see the actual numbers in excel/sheets please. Before we can make any real assumptions about this proposal. Until then I think it is a huge gamble to go x5.

[f8-ptrk]

@AxCortesCubero

But it has been expressed that there is worry that FIL price will go up and this will be bad for SP's. I am saying IF FIL price goes up because of this FIP, that will be a net good thing for SP's.

the earlier you do this the more FIL you need to lock up as the pledge will go down with the increased QAP. The later the less pledge you need but at a higher FIL price potentially. The overall rewards do not change - i am not sure if this is a net win for anyone in the end. As this is nto creating any utilization for FIl besides locking them up we will most likely see the same result in a few month after this FIP - price going back to where it is right now and all SPs spend a great amount of capital to go no where

of course thats all just reading clouds, consulting the glass org, shaking the magic 9 ball....no one knows what will happen

[AxCortesCubero]

Just following the guess that "FIL price may go up because of this FIP". My interpretation of that statement as I said could mean two things:

Guess 1: FIL price goes up because it increases with network QAP. In this case this would be neutral to SP's, as their collateral and block reward are not changed in fiat terms.

Guess 2: FIL price goes up because it increases as circulating supply decreases. In this case this would be a positive for SP's, as their collateral would not be changed, but their block reward would be greater than before in fiat terms.

A million more things influence prices of things, so it is not the most useful conversation to have. But if the two above where the concerns, then I am saying those specific concerns should not be a problem for SP's

[tmellan]

Attention!

Hi everyone,

Taking into account many of the points raised regarding SP preferences, we've adapted two key elements of the proposal:

• slope --> 1x
• InitialPledgeLockTarget --> 50%, with gradual ramp-up time.

In addition, several points have come up repeatedly, and we attempt to address why we think these will not cause problems.

The document is here:
https://pl-strflt.notion.site/Duration-FIP-revisions-7426f344685940409ac513a0ffcccc86

Thanks to all who can take the time to read, and especially those who can take the time to comment. This is a critical feedback loop for us.

Pending community response, we'd like to include these changes to the current FIP draft. To help direct, please 👍 or 👎 to this post. Thanks all.

cc @zixuanzh @vkalghatgi @anorth @jbenet @momack2 @Fatman13 @Kubuxu @f8-ptrk @kaitlin-beegle @geoff-vball @AxCortesCubero @misilva73 @jennijuju
(apologies for spamming a random of commenters but let's keep the ball rolling to keep progressing!)

[jennijuju]

Secondly, because the initial draft was already extreme in its effects with x5 it feels to me like an intentional raise to make the argument of lowering accepted by the community.

i think this is a very extreme, defensive speculation. And I hope the community would have some trust in core devs, have the best interests on “what will make network more useful and being more value to the network” in heart.
Honestly I found this to be rude, shows large ignorance and underestimation on the effort folks spent on talking to the community, gathering feedbacks by joining ~5 spwg calls in 3 days, engaging in GitHub discussion and notion comments and etc, so to further analyze and adjust the proposal accordingly.

[herrehesse]

@jennijuju I am not doubting any of the hard work this team has done, and I am also very supportive of the proposal. But some really important calculations and effects have not been thought through properly. And some things feel rushed or odd. This has nothing to do with "being rude". We are all adults and can take a hit or two, don't you think some storage providers feel awful right now because of this proposal?

Continue the conversation, my point remains the same, I believe x5 is a to high factor and the negative network effects will not be pleasant. I believe that x2 is more than high enough. Problem is I have not seen calculations for either of them, only formula's. If someone can concretely give me an example, from a SP perspective, on why x5 is needed I might change my opinion.

[jennijuju]

We are all adults and can take a hit or two, don't you think some storage providers feel awful right now because of this proposal?

Yes, all feelings are valid and fully ack!

[DecentCr8]

100% no.

100% no to the mechanism, introducing a sector duration multiplier, whatever the slope is.

[songxianghui19890506]

You're right. This proposal is very hasty and does not take into account the actual situation of most storage providers. Therefore, I suggest that the storage providers be cut off as leeks.

[SparkHao]

I have an idea, If you really think circulation is a big problem and have to deal with immediately, instead of trying to find tons of reason to convince the SPs to be the only one to bear the cost, why doesn’t PL or the foundation burn some tokens they control just like CZ does for BNB? It seems like a least destructive nor controversial way to achieve the goal, you don’t even need a FIP and can do it immediately, problem solved!

[f8-ptrk]

or we extend the vesting schedules of ICO tokens by 500-100% - show some long term commitment!!!

edit to add [edit]
to make this clear: that was a joke!

[SparkHao]

or we extend the vesting schedules of ICO tokens by 500-100% - show some long term commitment!!!

Absolutely, that will work too.

[songxianghui19890506]

You're right. This is a good suggestion. You should let Hu Pang know.

[herrehesse]

Can someone explain to me with corresponding calculations why we can not do a 2x for a 5 year sector?

[donkabat]

This is the direction what I can imagine... and initial pladge will not need to change

[herrehesse]

It feels to me like a LOT of worries and arguments are not listened to. Is there a list online where all of the arguments are stated? Love to see this, and if not I will make it myself and share it.

2/3rd of the community vote also sound reasonable for this proposal.

Will this proposal be pushed whatever the feedback from the active community? I do need an answer on this because then I can prepare for it. Right now it seems most active and knowledgeable data onboarders and SPs are against the current form, but with vague and evasive answers it seems to me that it will be implemented regardless. Would be appreciated if we know what our voices are “worth”.

[tmellan]

@herrehesse ppl are listening and giving answers:

• SPs asked for more info on the distribution of deals and sectors and Sentinel produced a dashboard.
• CEL have answered FAQs coming up from in-person and zoom SP meetings here.
• Starboard produced a duration-dependent ROI calculator in response to SP questions about more details on specific scenarios.
• There should be an AMA/panel with FIP authors coming soon.

And for sure, a few more things need more detail --- clarifications on porep security, deal extension and lending --- but ppl are very actively aware of this, and to be clear SPs are absolutely being listened to.

[herrehesse]

So, if everyone is listening to SP concerns, is it now finally clear that we do NOT want this in its current form?

[jennijuju]

Have been getting questions and wanna leave a quick comment
#429 does not implement deal extension or deal renewal for ANY type of deal. There is a possibility that enables existing and still valid data-capped data, to be accounted for an updated QAP to the sector.

If an existing deal renewal or extension is necessary for this FIP, then please add that as a prerequisite to this FIP. And some core devs might prioritize coming up with a proper solution for that first.

[DSS-AL]

We’re a Sydney based SP with 25PiB of proven capacity in the Filecoin network, as such we’re motivated to support activity that will boost the price of FIL and reduce volatility.

Having said this, we are vehemently opposed to the artificial manipulation of FIL supply designed entirely to boost the price of FIL for the benefit of a select few significant holders of supply. The price manipulation is a short term fix to long term problem and we’d much rather PL focus their efforts on increasing demand for FIL by onboarding meaningful customers and improving the customer utility, stability and retrievability.

The proposed FIP does:
• Artificially inflate the price of FIL in the short term with no long term value proposition,
• Result in greater concentration of storage power to the few SPs globally with large balance sheets and access to capital in an all but closed capital market environment particularly in the Crytpo sector,
• Reduce long-term profitability and return on capital for SPs,
• Reduce Filecoin’s current cost advantage of Web2 storage providers such as AWS and Azzure.

The proposed FIP does NOT:
• Solve a real world customer problem,
• Attract more customers to the network,
• Improve product utility,
• Improve stability of Lotus,
• Improve retrieval,
• Attract more SPs to the network,
• Improve competitiveness over Web2 storage providers.

A 5 year FIL+ sector requires 50x the amount of capital to collateralise that sector (compared to a 1 year CC sector), where do you expect SPs to produce that amount of capital if they want to remain competitive?

[mxz9102]

As a small SP with 4PiB adjusted storage power, I don't have much additional capital for 50x collateral. I even couldn't get any FIL loan now.

[jnthnvctr]

As a non-SP, I'm for this FIP in principle but from what I understand / am hearing from folks on this thread I'm concerned about moving forward.

Reading through a lot of other folks' concerns - a few things stick out (maybe these have been addressed, but it's honestly hard to keep track with this thread):

1. It sounds like one part of the push for this FIP is around timing.

"Given the timeline for common holidays + FVM, time needed for final call and implementation, conclusions have to be drawn soon. If not this proposal won’t be considered again until quite far into the future, if at all."

This strikes me as very weird reasoning. It's either the case:
a) This is FIP is critical for the network, and therefore will be continually pushed on even post holidays
b) This FIP is non-critical, and will eventually be forgotten

In both cases, timing of upgrades seems like an explicitly bad reason to try and push this forward - as its the importance of this FIP that will dictate on whether it eventually is accepted by the community. Given the state of the discussion, it seems there are many folks in the community who feel strongly against (not even the idea itself, just the current timing / incarnation) - and its not obvious all the concerns have good answers (see below).

2. Many of the objections are fundamentally hard to answer (e.g. how people will react to some of the changes). However, one crucial concern seems clear and non-theoretical - lending is not well supported by the network (and will not be until post-FVM).

Many of the lending programs today are limited in scale in terms of SPs they can concurrently support, and tend to target SPs who can accept large volumes of FIL (someone correct me if this is wrong!). This means the average smaller SP will not have access to capital to grow at the same rate as the larger SPs.

I know there are a number of teams working on liquid staking like programs when the FVM launches (which hopefully will serve a broader range of SPs) - but until that's a reality, we're exacerbating a divide which will only worsen diversity on the network.

At least for me, it feels like this is a clear issue with the FIP as it stands - and one that requires no modeling or theory. It is practically the case that there are not lending options accessible to all SPs - and until thats remedied making this sort of change will have a perverse effect on SP diversity. (Even if there were a new lender to appear who would work with every SP it also seems concerning that everyone would need to set up a multisig for owner accounts and the potential risk that might incur to the network. EDIT: been poitned to FIP29 as a potential answer on this point, though idk if any loan programs use it)

Is there a reason then to implement this FIP before the FVM / permissionless lending markets are live? Am I missing some context here which should change the reasoning above?

[DecentCr8]

Block rewards are exponentially decreasing with time. If you combine this with the pledge dynamics, the typical ROI will likely halve within around a year. So we've gotta be serious, next year is not just as good as this year when it comes to making these kinds of changes. They only become harder with time.

@tmellan , are you talking about security issue by saying So we've gotta be serious? Is this change trying to solve this "serious" issue?

And, there are two things you may want to think about:

1. if next year is not just as good as this year when it comes to making these kinds of changes, I would expect, it might be worse the 2 years later. It means again it is for short-term purpose. Are you considering removing this change some day in next years if we do it this year?
2. It makes it harder because the typical ROI will likely halve within around a year, do you understand this change makes it harder for the existing participants?

[tmellan]

@DecentCr8 No, I wasn’t talking about security/bug or anything like that

if next year is not just as good as this year when it comes to making these kinds of changes, I would expect, it might be worse the 2 years later.

Yes, agree

Are you considering removing this change some day in next years if we do it this year?

Sorry I don't follow, maybe rephrase?

It makes it harder because the typical ROI will likely halve within around a year, do you understand this change makes it harder for the existing participants?

Yes, for existing and future participants, in FIL-on-FIL ROI terms. The cost of the change is not getting 80% FIL-on-FIL ROI. But there are benefits, which we've tried to set out as clearly as possible.

The way I see it there's 3 choices:

1. Invest in the future of the network now, when times are hard, but will never be better in terms of block rewards.
2. Put it off, maybe try to do it in 9 months or a year. I don't see this happening. This pill becomes much more bitter to swallow in the future.
3. Reject the change. Maybe the technical innovations coming up are enough. From our perspective, we’ve flagged the economic issue. We’ve provided a solution. We can tune it a little further, make sure all bases are covered, but we won't attempt to force people to accept it.

[Crypochen]

However, given we have a number of SPs in this thread - perhaps we can just get direct feedback. Are you able to get capital easily? If not, why not?

There’s very limited corporation( 2-3 ) that can provide loans longer than 360 days and only changeowner to them. But You still need do a lot of KYC work and a long time to complete the loan. Which of course if you're a small SP, and you even don't have a corporation and staff to do that, it will not be a simple task.

With other corporaion, you need to pledge your hardware and even put your hardware in their datacenter. Or the time is too short even to seal 210 days sector.

The interest will be 20%-30% now, and if the FIP is implemented, the Miner annualized income will be 20% as this one pointed：#421 (comment)

In this situation, the loan from corporation interest must be lower than 5% (1/4 of block reward maybe) , which of course is Negative incentive for the corporaion.

A lot of guys think DIFI will help, i think it well, but we still need to see

1. Are people willing to staking for 5%?
2. How fast will this be online after the FVM？

As i know, the largest DeFi of FIlecoin only get 1.5M staked in ethereum before. That's really not enough.

[jnthnvctr]

To address a few points above:

1. In the short term, @mxz9102 this is a limitation of centralized loans. There is work going on with the existing lending providers to increase capacity (specifically looking at trying to both increase number of providers out there and more capital in these programs)
   With the current key issue, maybe there's a standard we can try and work with lending providers to ship for transferring loans to each other (so a Coinlist lender can manage a loan from Anchorage, or rotate ownership of the existing loan to Anchorage)

Fwiw, I've been advising some teams working on solutions for the FVM in this space, and I think the state of lending will get substantially better once the FVM ships (don't want to out anyone who is in stealth mode - but the technical designs feel substantially friendlier on both sides). Fwiw - I'm generally bearish that we will be able to lean on DeFi in other ecosystems to solve our capital efficiency problems (just as a matter of how complicated the solutions tend to be and the risks involved).

2. @Crypochen - i think this depends on the market. I think we're seeing evidence from other protocols that single digit staking is acceptable, and can get meaningful traction. In most protocols that were in the 5-10% range (SOL, ADA, etc) iirc the stakerate was in the 40-70% range

In general, I think folks will demand a higher ROI (in FIL terms) if they're incurring higher risk - but to @tmellan 's credit, if we can smooth out the waves of expirations (so it isn't an overhang on the FIL economy every couple quarters) this may be offset in general improvement. I guess the impt point here is most folks looking to stake I don't think purely denominate their ROI in FIL on FIL returns - its also a function of FIL/Base currency (whether thats USD, ETH, BTC, you name it). If FIL/Base is steadily chugging upwards, folks are likely to accept a lower yield (ETH as an example is targeting single digit yield, and folks are generally ok with this as theyre not thinking in purely eth on eth terms.).

One slight advantage we have (vs some other protocols) is that there are different profiles of folks who might be staking on FIL - rather than just having folks who are locking FIL who are long the network, you could imagine tokenized representations of debt being used in a number of contexts:

• Held as assets in DataDAOs
• Held as collateral (would need to be pooled) for retrievability oracles

The DataDAO use case is an interesting one - where if the DAO is looking simply to index on FIL inflation, they aren't just looking for the maximum reward, they're looking for stable returns for their endowment to fund ongoing storage (https://twitter.com/jnthnvctr/status/1545401176392437762). I have more thoughts on this (but also don't want to out some early ideas on the DeFi side).

NFT Storage is looking to eventually DAO-ify (and hopefully lay blueprints for other DataDAOs and contracts that can be used) to enable the above use case.

3. @tmellan - I know you guys have been pushing hard on this, so definitely want to appreciate your team's effort. I think the challenge I (and others) have been struggling with us to understand:
   a) What is the motivation for this FIP
   b) Why is timing a specific factor here
   c) Does your team (in this FIP) have full context (e.g. on the loans) - or are there assumptions you're making that feel off

I think for everyone to be able to evaluate things with clear eyes, there is base level of shared context we need to reach. Personally, I was unaware of the upcoming wave of sector expiries - and I'd be curious how others see this problem.

Pulling together many threads, it feels like a few things are true:

• We do not live in a vacuum - we're heading into unprecedented global macro economic times. Not much anyone in this thread can do about it - we may be fortunate and things continue as is, or it could substantially deteriorate in the winter.
• Filecoin has a cyclicality to it due to the nature of sector expiries -> this creates an overhang which seemingly hits us every couple quarters. This cyclicality adds volatility / pressure - and combined with a downswing in global macro, could spell pain for smaller SPs. I'll admit that I did not have an appreciation of the scale of the next wave of expiries and the potential implications that might have for folks (though perhaps SPs keep a closer eye on these things than I do).
• Lending solutions today are inadequate, and while there are solutions in the works they will take time to ship.
• Longer term, I think there's important questions around whether the fixes we're discussing today will need to be adjusted (and if so, how) - which to me is also an important area for discussion.

I think without having a shared understanding of the world, it's hard to draw meaningful conclusions (for me as just a contributor, for SPs knowing their own operations) about what makes sense.

• Personally, I can empathize that we see like a stacking of non-ideal situations (global macro downturn, mass sector expiry) as a good impetus to give an option for miners who are long the network to be compensated higher (for longer term orientation).
• However, I think we can't disregard the reality of existing lending markets - and the implications of such a proposal in the short term (and the reliance that we're putting then on decentralized lending markets showing up as quickly as possible post the FVM. Also the FVM shipping as quickly as it can).
• I think without data on these, its hard for SPs to clearly evaluate how risky they see the next N months for them to be, and whether this proposal (despite the short comings of where we are today) is still worth pursuing. Equally, it's hard to evaluate if the solution proposed is a bandaid fixed (that as @DecentCr8 pointed out may need to be tweaked in the future), or one that we're confident is correct longer term.

One note I'd make is that the proposals you laid out @tmellan are not the only ones that need to be on the table. I think if other SPs appreciate / agree that the macro env is going to be painful, its reasonable to pursue something shorter term to hit the n17 upgrade.

But the alternative doesn't need to be a year from now - it could also be the Feb update no? Imho one of the biggest things we've heard in this thread is trying to tie this with permissionless lending markets (see 1475's post at hte bottom).

[Crypochen]

One slight advantage we have (vs some other protocols) is that there are different profiles of folks who might be staking on FIL - rather than just having folks who are locking FIL who are long the network, you could imagine tokenized representations of debt being used in a number of contexts:

yes, i think defi and other staking stuff shall be better than this FIP. Find the new use case.

[anorth]

I’d like to present my summary evaluation of this proposal as it stands today (partially in response to @Fatman13's request, partly to agitate for action). I am not a member of the CEL team or an author of the proposal, but highly motivated to realise a robust, useful, and growing Filecoin network. This is my opinion, though I have drawn some ideas from the discussion here. I disclose that I am also a FIP editor (an administrative role), but here I am wearing only the hat of community member and core dev.

I think something like this proposal would be good for the network, and I’d like to see it implemented. But the current form has a number of specific issues that, in my opinion, must be addressed before we can be confident it is a safe, beneficial, and equitable change to make.

The right goals & directionally good mechanisms

The proposal is well intentioned. I understand the primary goals to be improving the stability of the storage commitments underlying the network as well as the stability and predictability of storage provider returns. I think these are good goals[1].

It aims to do so by allowing longer-term commitments to storage, allocating a greater share of the block reward stream to those longer commitments, and adjusting the pledge lock target upwards to bring balance to supply flows. There is also an ethos argument that the network should skew rewards towards longer term commitments. I am reasonably convinced that these mechanisms will achieve the goals.

[1] What is the tradeoff for these goals? It’s a bit implicit, but greater stability will come at the expense of absolute commitments. The goals prefer to lower variance and accept a slightly lower total commitment than otherwise. The network will deny SPs optionality for short sector lifetimes, which makes committed storage slightly less valuable to them in aggregate, and so they’ll commit less of it. The most marginal (i.e. least stable) SPs will cease to participate. I think this is a fine tradeoff for the network to make, given the current 18EiB storage commitment. In the long term, stability will lead to greater aggregate utility.

Major issues

The following are major issues with the current proposal. Each one of them is a blocking issue to my support. However, I think all of them could be addressed. If they were, I think this the proposal would present clear benefits with reasonable risks.

Pre-commit deposit risk modelling

The proposal notes that PCD will increase in line with the maximum potential lifetime reward of a sector. Merely noting this is not sufficient. We need to see an analysis of expected returns given assumptions about prove-commit failure rate: perhaps the observed network-wide rate, and some multiple of that for an individual SP’s risk model. The risk is aggravated by batching, which is necessary for high-throughput or high-base-fee onboarding. It is not clear to me that lower-quality sectors (CC or short-duration) are even above FIL-on-FIL water, here, or at what point batching becomes untenable.

Action: SP return projections must take this into account. It’s possible that a compensating mechanism will be required to keep CC and batching viable.

PoRep security impact

This proposal increases the maximum sector commitment to 5 years, which breaks the existing (somewhat implicit) security policy and mechanism for a fault in PoRep. The FIP notes this but does not propose a mitigation or state that its implementation should be conditional one. The FIP authors have attempted to dismiss this issue as being a separate issue and not coupled to their proposal. In my view, a policy and mechanism to address potential PoRep flaws is a pre-requisite to breaking our existing one.

@Kubuxu and I have spent some time trying to design a mechanism that could decouple the validity duration of a proof from the commitment duration of a sector. We hope to contribute this in order to unblock the sector duration multiplier proposal. The FIP should state this as a prerequisite.

See #415

Action: The FIP should state the pre-requisite and authors should contribute to the policy analysis.

Inequality of access for existing sectors with FIL+ deals

The proposed mechanism is agnostic to deals, but current limits on capability of the built-in deal market mean that, for existing sectors, those already hosting verified deals can't be extended immediately to fully exploit the multiplier, while CC sectors can. The raw byte power attributable to the sector could be multiplied, but the FIL+ boosted power would be countered by the QA power’s “spreading out” effective. This means existing CC sectors can increase their proportion of total power and hence rewards by some multiple. Assuming any significant uptake of extension of CC sectors, the share of per-sector reward to existing FIL+ sectors could fall significantly faster than without this proposal.

This issue applies only to existing sectors, not new ones. But of course existing sectors account for 100% of committed storage provider participants. The problem is the inflexibility of existing deals, so the proposal being agnostic to deals has real implications. This problem goes away in 1.5y when all such deals have expired.

I am working on some improvements to the flexibility of FIL+ datacap and deals, but even if they land simultaneously, I am not confident they adequately address the issue. Most likely, extension of existing FIL+ deals would be at client discretion and require action and (small) cost on their part. SPs would experience divergent outcomes depending on their luck of their specific clients. I could be convinced otherwise by evidence that clients representing, say, ¾ of current verified deal bytes stand ready and willing to extend their deals when possible, or SPs clearly confident of that. Or a FIL+ notary vote that they will mint new datacap to extend any and all existing deals.

I think the option of awarding duration multipliers only to newly-sealed sectors should be investigated more thoroughly. Even with that restriction, the future rewards of existing FIL+ sectors would fall, but at least their providers would be on equal footing to seal more, high-multiple sectors. Ramping the maximum duration very slowly (18m) could also reduce the impact (because the old deals will expire). I would love to see other ideas to restore equality.

I’m aware the CEL team’s position is to advocate for applying the multiplier at extension because to apply it only to new sectors “risks us seeing much smaller positive effects” and “will be an incentive to let sectors expire” (… in order to seal and commit new ones). I don’t find this convincing, and certainly not more powerful than the inequality against the network’s most utility-aligned SPs. Especially since more muted effects would reduce the risk of many uncertainties and shortcomings, and I’ve seen no modelling. In short, I don’t think they tried to make the idea work because they haven’t been forced to.

Action: Authors should model and publish the best possible version of a change that applies only to new sectors. Show analysis for early-termination incentives, and propose mechanisms to mitigate them. Pretend it were against the law to change the policy for existing sectors: what’s the best you can do?

Support for FIP-0045 ([draft](#432)) is also welcome.

The FIP also needs to fix simple policy numbers like max deal duration

Shortcomings

The following are significant shortcomings or weaknesses that I would like to see addressed, but don’t quite rise to the level of a blocking issue to me. They are risks that I am unhappy about.

Uncertain short-term dynamics

The proposal appears to prompt an immediate drop in per-sector profitability-on-pledge (according to published charts), and there’s also some incentive to delay extending sectors until the pledge-per-QAP requirement falls. There may also be an incentive to terminate the oldest existing sectors to re-use the pledge for new commitments that gain a greater return on hardware cost (but lower return-on-pledge?). These all point towards a short-term drop in onboarding rate or even total power. (My guess is that this is more likely than a rapid increase in QA power, but see “Rollout shock” for that one).

Some of these dynamics might counter each other, and other system properties may also act as negative feedback, but I haven’t seen any attempt at dynamical system modelling to understand the instabilities, feedback loops etc, that might be in play. I’m willing to take the long term projections as given, but in a system like this with reflexive dynamics on token price, the long term could be seriously undermined by a short term feedback loop.

A 3-6 month ramp on the maximum sector duration/multiplier could significantly mute any unpredicted dynamics. Denying multiplier on extension of existing sectors would also constrain any such effects. Or I’d welcome other solutions.

Action: Analyse and show best version of a ramp on the multiplier. Analyse and show best version of applying only to new sectors. Show dynamic modelling to demonstrate self-healing behaviour without other mitigations.

Rollout shock

If CC sector extensions can enjoy full multiplier, there is a positive feedback loop for one or more large SPs attempting to gain a dramatic increase in power. A provider/consortium that did have the capital could pack blocks with extensions and censor extensions of other providers. If they gain any momentum, their increased power would increase the rate at which they could pack blocks and the effectiveness of censoring others: a positive feedback loop that increases the effectiveness of their attack. This is independent of the consensus risk analysed and noted below.

I think this scenario is less likely than the one above in which onboarding decreases, but “less likely” isn’t an appropriate level of assurance for a billion-dollar blockchain.

Rollout shock would be effectively prevented by a ramp on the maximum sector duration/multiplier. Or I’d welcome other solutions.

Action: Analyse and show best version of a ramp on multiplier. Develop alternative solutions to consensus risk due to rollout shock.

Moving towards proof-of-stake security

This proposal shifts the basis for consensus power sharply towards money rather than physical storage. I am not comfortable that PoS networks are sufficiently validated as a secure foundation for trustless collaboration (although eagerly await learnings over the next few years as Ethereum and other networks mature). This seems to be an unintentional side-effect of the proposal; it’s certainly not stated as a goal nor invited as a topic for discussion. I wish there were proposals for ways to achieve the goals of stability and predictability without relying so much on stake.

“Useful consensus is one of Filecoin network's contributions to the blockchain world.” [DecentCr8](#421 (comment))

Underestimating a complex adaptive system

Filecoin is a complex adaptive system. It is not generally possible to predict the effect of some changes to the system, because second- or third-order effects might dominate and agents will change their behaviour in response to a changing environment. There is non-trivial risk that the system has been mis-characterised. Changes aimed at stability could end up reducing network stability, or increasing fragility. Perhaps the storage commitments will be more stable, but the system more vulnerable to external shocks? Perhaps stable profits will deceive SPs, and they’ll take less risk-mitigation measures as a result? How would a major SP failure affect the network perception? How does this compare with a simple POW system, which is less stable but more anti-fragile?

We can’t predict these, and the authors have stayed clear of predicting any outcomes (see uncertain system dynamics and rollout shock, above). But working with such a system requires a “probe, sense, respond” engagement rather than a “sense, analyze, respond” approach that would suit a complicated but predictable system. This proposal seems like a significant step change based on a belief that the system is understandable.

There is some risk of an adverse outcome. The authors have justified the large step change with (paraphrasing) “SPs need stability, we don’t want to change often”, but also downplayed with risk with “we can change it later if it doesn’t work out”. Pick one.

A more incremental approach, or at least a ramped rollout, could help buffer us from getting it wrong. Nevertheless, I don’t consider this a blocking issue, just a pre-mortem.

Buying an expensive option

This proposal achieves stability by inducing SPs to make longer commitments. The network essentially pays the SP a premium to give up the option of dropping a sector. (The SP can pay termination fee to buy the option back).

5x reward seems like a very expensive option.

We don’t know whether this scale of incentive is necessary to shift the average commitment duration higher (i.e what the right price for the option is). The authors have offered no analysis of multiplier slopes below 1.0 so the community have no ability to compare the projected outcomes. The one-shot nature leaves little scope for “price” discovery. The size and step-nature of the change introduce many of the risks above.

I don’t consider this blocking, just an unfortunate over-payment.

Weak justification for raising minimum sector duration

The proposal raise the minimum sector commitment from 6 months to 1 year. While it does align with the goals, there’s no evidence it’s necessary after introducing a massive incentive to longer durations. The only argument for this that I think holds any water is the ethos one: “because we think it’s right”. Arguments that sectors expiring every six months need to find new sealing power to compensate are flawed: a 6-month sector can be extended without resealing. Arguments about client demand are similarly orthogonal to the issue.

I don’t consider this blocking, just unnecessary rules that exclude a market segment.

Missing concrete numbers for pledge, reward, etc

SPs have requested concrete numbers for the pledge, day reward etc derived from the network conditions today. The proposal should show the current values, and the values immediately after the proposal was adopted, and, if relevant, the range of values as they would change in the future. This is a reasonable request from SPs which should be answered.

The PCD values are already given.

Action: Give actual numbers in the FIP

Closed-source modelling and simulation

The FIP authors have presented some analysis documents, including charts of projections and simulations. They have not provided the code for these models so no-one in the community can check their correctness, or use them to answer questions other than the ones that the authors wish to publish. The authors are understandably reluctant to present projections that have anything to do with token price, but others can use their code to make their own.

The source code for and data inputs to analysis should be provided to network participants to verify. Preferably checked into the FIPs repo.

Resolved or non-issues

The following are issues that have been brought up, but I think are adequately addressed already.

• Lesser incentive for short deals, interaction with Decoupling FIL+ term from storage marketplaces (FIP-0045) #313
  • I don’t think it’s a practical problem, and there’s still a market for shorter deals (tho probably not below the minimum sector lifespan).
• Consensus power attack
  • Seems reasonably analysed, and became even harder with slope = 1.0
  • I would like to see further mitigation with a ramp on max commitment duration, if possible
• Duration multiplier applies to CC sectors too
  • Storage capacity is the basis for consensus security, and a network good
  • Applying only to deals would make it even more proof-of-stake
  • But inequality for existing sectors is an issue
• Bias against small miners
  • Proposal shifts the focus more to financial capital than hardware, which is more widely accessible
  • To the extent borrowing capital is difficult, it’s already difficult and I don’t see this change as making the matter practically worse. SPs need to borrow for hardware ops too, changing the balance of what they spend funds on doesn’t make a huge difference.
  • De-emphasises economies of scale in operations
• Timing
  • We need to take time to do proper analysis, but beyond that we should make improvements as fast as reasonable. I don’t support generic arguments to slow down, but calls for specific research or analysis to be performed.

[jennijuju]

Thank you for the clear write up! 💙

[herrehesse]

@anorth. This is extremely helpful. Thank you so much.

[DSS-AL]

How is "financial capital" easier to come by than any other capital?

Equipment finance is a far more liquid market with significantly lower cost of capital, hardware can be financed over 5 years at 5%... this is the easy part.

Borrowing FIL requires collateralisation and greater than 20% cost of capital with earnings locked up for 6 months.

This is not a "Resolved or non-issue" the fact that this concern has been dismissed means that you have absolutely no idea how financial markets or SP businesses operate and I have zero confidence in the assumptions behind this FIP.

Raising capital for the purpose of buying FIL on balance sheet in the current climate is extremely difficult, the last time we did that 95% of this value got wiped out, try explaining to an investor why this is a good idea to do it again.

[tanukiroku89]

Arock supports the FIP. We believe that it benifits the Filecoin ecosystem. The FIP is not forcing Fil miners to stake/pledge more FIL, just an option to do so, and it is rewarding miners that are locked in for a longer period. But we should adjust the rewarding system to scale according to amount of stake and duration of the sector, with a proper formula, rather than a fixed multiplier.

FIL+ locking period is not decided by miners, data providers will define how long it should be stored. With the multiplier, there will be more long term real data storage demand, more LDN application, less notary collusion.

From my personal point of view, this proposal should be approved as early as possible but implemented after the FVM upgrading. We expect that there will be some DeFI dapps on chain once it is supported, for FIL holders to lend, and for miners to borrow FIL to pledge, with an open interest rate for everyone to calculate and make decision. It could also benifit the actual network utilization. FVM upgrading, retrieval, CDN market and other development in near future need a stable network and FIL price. Less token circulation in current market situation will be more friendly for user adoption.

[cbtan21]

Sharing a conversation I had with SPs. I'm not a SP. I am just a community member, who is 100% devoted to the growth of Filecoin.

My view is any time there is a change, there are always tradeoffs. So I understand the concerns about existing deals being dead space.

My overarching views are:

• a lot of SPs will not survive this period if nothing changes soon - this is the most important point.

• about deals being done and are now "dead space" - that's likely a problem one will face regardless when it is implemented. if it is implemented 18 months from now, what happens if one does deals at the 15-month mark for example. he/she will still face the same issue unless the underlying protocol changes (another FIP). or does he/she not do deals at all until the 18-month mark?

• for any good protocol with decelerating rewards, miners (for a lack of a better word) will always have to grow to compete. if one stagnates, he/she will probably end up losing, albeit slowly. doing deals yesterday is great, but he/she will have to do deals tomorrow too. if one has to do more deals and the rules are the same tomorrow for everyone, but the new change enhances the long term prospect for filecoin and the entire ecosystem, then what really is the collective downside?

[1475Notary]

We open sourced our model here (https://github.com/lyswifter/FIP36_simulation-mild_growth_assumption) to simulate Filecoin network dynamics post FIP36, specificly what’s the SP’s ROI looks like at any time of points in next 6 years by assuming
1, this FIP will go live in nv17 (early Nov)
2, RB power add/expire and QAP add/expire on epoch basis after this FIP, which can be modified by yourself to check how SP’s ROI will change accordingly

The ROI doesn’t include operation and hardware costs, nor gas fees because that’s variant SP by SP, it’s just fil based return, concretely, 365 days of block reward divides pledge, you need to dedicate the gas fee, collateral and fiat cost from the aggregated block reward by yourself based on your own cost structure. In short, you might see much lower ROI in your real business operation than the model tells you.

This model is written by python Jupiter notebook, if you don’t have experience but still want to try it, you can download Anaconda and use “jupyter notebook” command to start it, and you will find tons of online guides to make the learning curve not intimidate at all.

Caveat! This notebook should not be viewed as financial decisions by any means, there might be defects of simulating Filecoin consensus or false assumptions of network parameters and etc. Although it’s worth to mention that with two months of emulation from Jun 17 to Aug 22, it’s pretty close between the model output and network data (block rewards and circulation).

Again, you are free to modify any part of the model, especially the part in “## Assumptions we can modify” code block, and re-run the notebook to check how that impact SP’s ROI. By changing these assumptions and compare the results, we have observations as below:

1, SP ROI will continue to drop and will be in below 20% fil based APY range after FIP36
2, Larger total daily QAP commit to the network will make the SP ROI dip faster,
3, Larger total daily QAP commit to the network will make SP ROI dip deeper
4, There will be a turning point where SP will see ROI raise again, this point mark that baseline power overtakes QAP, which is determined by consensus pledge using max(QAP, baseline) as denominator, and again larger daily QAP, later we will see.that time point.

Since this APY is only about (block reward)/pledge, so with 20% to 30% fil-based APY SP will enter “bleeding zone”), assuming in real operation, SP still need to deduce the cost of collateral(30%ish loaning rate at this moment) cost of gas (0.6% of pledge for 50x sector to 30% of pledge for 1X sector), hardware and datacenter cost, fixed cost (team, network, etc)

Based on above 4 observations, we infer that
1, After a relatively short period of roll out shock (1 to 3 month), we will see new committed QAP quickly shrinked, as larger daily total QAP means SPs collectively push their ROI further down in quicker way, down to the road, more and more SPs will stop committing new QAP until no one can be profitable for doing so,
2, The network will be more unstable than stable (short term growth by roll out shock and long term shrink for negative incentive for SPs)
2, Therefore there will be no long term benefit in terms of locking more tokes in TVL.
3, When SPs stop committing new QAP, other tracks of Filecoin ecosystems like Fil+ might see negative impact as well

Advise:
Push this FIP to further later, until like decentralized loaning platform is ready, which give SPs have chance to access some low cost collateral and earn positive profit. But whatever it is, looking at other POS staking, the ROI should not be lower than 30% at any time of point. (In Polkadot staking 26 lock days make 14% with very little slashing risk APY just for reference)

If you see any discrepancies between this model and the Filecoin consensus, or suggestions to improve the model, please let me know.

[Crypochen]

When invest 100T , just get 10000T QAP.

i don't know. The whole FIP is not decided now. And it seems the key point is whether the previous sector will be multiplier.

[AxCortesCubero]

@1475Notary

I am trying to go through your code to understand your calculation and how it differs from other models.

I have one significant objection to this model, which is this doesn’t really seem to be simulating anything about the duration multiplier itself. It seems just a simulation based on different raw and QA power growth rates assumptions, that seems mostly unrelated to this FIP.

In particular it seems the only way you are modeling the effects of this FIP is by assuming that after the duration multiplier comes online, you assume there will be a significant drop in network raw power growth rate. And then you use as input; some decreasing growth rates at different points in time after the multiplier comes online, and make some also pessimistic assumptions about the network QAP growth.

The current simulation seems to be just answering the question “what happens to SP ROI if there is a drop in network power growth?”.

I think this would be even a fine assumption if you have some explainable argument or mechanism that shows implementation of duration multiplier leads to decrease in network drop.

Instead this pessimistic outcome which is presented as a conclusion which emerges from some simulation, in fact is not a conclusion, but an assumption built into the simulation.

So to be clear what I think you are presenting is “If there is a significant slow down of network power growth, this would be bad for ROI”, what you left out is a justification of how this relates to FIP 0036, unless there is some other simulation that shows that the effects of FIP 0036 would lead to such a power slow down.

I would recommend others to also consider the ROI calculator built by starboard (https://observablehq.com/@starboard/sproi-fip-duration-v2) which is also open sourced and you are free to explore the code. I believe here the effects of a duration multiplier are actually examined.

First this model needs an assumption of raw power growth (which you can tune, optimistically, pessimistically or neutral). Then the duration multiplier comes in in two ways: user can input a prediction of what they think the average sector duration would be (if you think everyone will go ahead and get the maximum multiplier, or not). This average sector duration controls in the simulation what the total network QAP will be given the total network raw power assumption, and also average sector duration affects the expiration rates used to calculate circulating supply. Is it a perfect simulation with all possible effects taken into account? probably not, but it is one with clearly stated assumptions and limitations.

[1475Notary]

Thanks for raising the questions @AxCortesCubero. About the duration multiplier from the FIP, my view is this is just a lever for SPs to choose how they want to allocate their capitals between hardware investment (represented by raw bytes power) and token investment. Therefore

1, My ROI is calculated by (365 days aggregated block reward) divides pledge, so duration multiplier has no impact in this calculation what so ever. For example, if one SP sealed 3 CC sectors for 1 year each, and another sealed 1 CC sector for 3 years, they have the same fil-based ROI as pledge and block rewards will be all the same, given they sealed at the same time of point.

2, About raw bytes growth after the FIP, my assumption is, all business operators want to optimize their cost structure, given my example above, 3 CC sectors 1 year each definitely requires higher hardware cost than 1 three years CC sector, though the token return is the same, so a nature choice from SP will be committing as little raw bytes as possible will be a far more efficient way of spending capital , but again, that’s only represent my view, and maybe some other community members who view this change as a “POW to POS change”, if you disagree with this assumption, you can state why and tuned the raw bytes growth parameters

[AxCortesCubero]

I think it is fine for everyone to have their own opinions about what they think the impact of this FIP would be on storage power growth.

The important thing I think is to make sure it is clear that this is an assumption you built into your simulation, stemming from your heuristic analysis, and not an outcome or conclusion that emerged from your simulation.

That said we could also now discuss your assumption instead, which I think has some issues.

First you state the duration multiplier has no effect on your definition of ROI. This is strictly true in a far future equilibrium sense, when Duration multiplier has already been established and network steadies around a sable average quality multiplier, then it is true duration multiplier just cancels between numerator and denominator. 

There are however more subtle non-equilibrium effects, where the change in the duration multiplier will lead to different circulating supply dynamics that affect ROI, as network reaches equilibrium. The simulations CEL and Starboard are working on examine these effects.

On the assumptions of raw power growth mainly let’s agree to disagree. Let’s examine and follow your scenario.

I know it has become popular to say this FIP will turn Filecoin from PoW to PoS, but it Is not true in the sense that you don’t get higher multiplier simply because you locked higher collateral, you get it because you are on the hook for a longer time commitment, which involves more risk as well as more value to the network.

Sealing 3 sectors for 1 year vs sealing 1 sector for 3 years: the first option has the benefit that you are free to do this for one year and then abandon the network, while the second option keeps you on the hook for longer, which the network values therefore rewards.

But say following your assumption that all the SP’s decide it makes more sense to seal their sectors for 3 years. If that is the case, then 3 years simply becomes the new normal, as the average network quality multiplier is now 3. So the SP will get the same return on their one sector as before, just that the network standard has now become 3 years long. So if everyone had chosen to do 3 years, they still need to seal the three sectors if they want to have the previous ROI.

What I’m saying is, if everyone decides to to seal a certain amount of time, that just becomes the new normal, if you want to keep your 3 sector’s worth of reward, you need to seal your 3 sectors, so I disagree with you on there existing this incentive for everyone to seal fewer sectors. If anything there exists an incentive to raise the average duration of a sector in the network.

[1475Notary]

@AxCortesCubero I don't want to be rude but I might need to ask you to stop putting words in my mouth, as I never concluded anything by misleading the community to believe my assumptions are facts, as you repeatedly implied. All I wrote in my opening and I quote here -
"specificly what’s the SP’s ROI looks like at any time of points in next 6 years by assuming
1, this FIP will go live in nv17 (early Nov)
2, RB power add/expire and QAP add/expire on epoch basis after this FIP, which can be modified by yourself to check how SP’s ROI will change accordingly"

Now back to the model,

• Duration multiplier, my ROI calculation is simple, reward divides pledge, you can just explain to me what I missed here how duration multiplier plays a role in either reward or pledge mathematically. Again, direct impact to both terms are QAP, while the QAP is from CC, or fil+ or duration multiplier doesn't really matter to me. I don't know what CEL is simulating now but I guess it will be a very different model for different purpose, you can't ask everyone to strict follow your logic to doing things, unless there is outstanding discrepancy with how the network parameters are calculated in consensus level

• Let's say my assumption about RBP growth is a false, but the debate itself about RB is pointlless within the context of my model, cuz I still get the very similar ROI while I increase the RBP growth ten times.

[zixuanzh]

Filecoin Network Macro Commentary

• While everyone has commented from their individual perspective, I would like to shed some light on the network macro to provide context based on public facts and numbers. Emotions have run high during this FIP process, but the discussion might not all be working from the same shared context.

• This post intends to be less technical and focuses more on the big picture since many analyses delved deep into the technical aspects and how it affects individuals. Thank you Starboard for extracting these charts and numbers from the Filecoin blockchain.

• Disclaimer: Numbers can be wrong. Not advice of any kind. Please do your own research.

• Just a reminder, as all FIPs in Filecoin, it should never be us vs them. We are all part of the Filecoin community. It should be about how to make Filecoin better as a whole. This is the principle of CryptoEconLab and its network of teams and collaborators.

• Storage onboarding - a key health metric for the Filecoin network - has slowed down significantly in recent months. There could be multiple factors here:

  • Broad crackdown on cryptocurrencies of any kind in China
  • Crypto market winter
  • Global macroeconomy downturn

• The Filecoin network was hit as one can see from the slower rate of onboarding.

• Worse, sectors that were added have an expiration. Not only does the network get a slower onboarding rate, the network is also facing threats from sector expiration. This directly impacts the circulating supply. In recent months, there is a net inflow of ~500k FIL/day (or 15M FIL/month) to the circulating supply (this number was ~120k FIL/day in April 2021 at its lowest).

• Bars above the x-axis mean that it’s positive inflow to the circulating supply and bars below the x-axis mean that it’s negative outflow. The red line is the net change on a daily basis.

• As one can see in the above chart, as onboarding slows down, the amount of FIL Locked has significantly dwindled. There was even an upshoot around Feb 2022 in FIL Locked when the first large wave of sector expiration started and releasing locked initial pledge into the circulating supply.

• Strong Fil+ growth since Feb 2022 (see below) has helped with locking more tokens on the macro level but it is not enough with sluggish sector onboarding growth in recent months (~15PiB/day).

• More bad news, the network also expects another large wave of upcoming scheduled expiration in the coming months (potentially on the order of 50PiB/day for Oct 2022 to April 2023, yes per day). And no, these are no longer SpaceRace sectors since the maximum sector duration is 540 days. Most of these sectors are likely added around March 2021 (we can drill into these sectors for sure but here is the big picture).

• If we do nothing as a community and the status quo of sluggish growth persists, the network may see sector expiration exceed new sector onboarding soon and that in and of itself creates other complex issues.

• If the protocol does not increase the average sector duration, the network will have to deal with the short sector expiration cycles every 6-12 months. These short-term cycles of sector expiration create a constant drag on the overall growth of the Filecoin network.

• For everyone’s information, the upcoming scheduled expiration also means an additional of 300k FIL/day of initial pledge released. If nothing changes, there might be a net inflow of 800k FIL/day (or 24M FIL/month) to the Filecoin economy. Coupled with broader global macroeconomic trends (rising prices, higher interest rates, and slower growth), this winter can be a long and painful one and some SPs might not survive if the global economy takes another downturn.

• Going back to this FIP, this is not about locking Filecoin for the sake of locking. This is to give participants who are committed to the network long term to express such a view and be rewarded for it. It also alleviates the short sector expiration cycle issue for the network.

• With or without the network macro condition, this optionality is net good for the ecosystem and should be here to stay.

• There are many other projects that teams in the ecosystem are working on in parallel, from FVM, to retrieval, to compute-over-data, to data-as-an-asset protocols, and to creative financial applications. All these take time, talent, and capital. Filecoin has a long roadmap ahead (both for the protocol and layer-2 applications). The protocol should align incentives and dynamics to ensure the network survives to the distant enough future to see the fruition and reward mission-aligned and long-term-oriented participants.

• Decision on this FIP still lies in the hands of the community but some meta points:

  • We should not take things for granted as a community. Researchers can work on many things and we chose to work on Filecoin and on this. Please be kind. People are pushing on this because we genuinely believe this FIP will help the network achieve its mission. It is okay to disagree and teams are here to provide analysis and heed your feedback. Similarly for builders, lenders, token holders, entrepreneurs, clients, and of course Storage Providers. We should treat them all like gold. The network cannot thrive without any one of these groups.

  • We should be able to react to changing circumstances and make collective decisions for the good of the ecosystem. While CryptoEconLab fights hard to keep a high bar for any economic policy changes, the world is rapidly changing and sometimes the network must adapt. There will be tradeoffs and some might win more than others but without the common good and a healthy network, there is no one left.

  • We should always grow the Filecoin network and ecosystem. How can we bring in more new talent, new capital, and new ideas to our ecosystem and network? An economy that does not grow is bound to become obsolete and we should all continue to push for the growth of the Filecoin network.

• Lastly, all charts can be found on https://dashboard.starboard.ventures/ and data available on https://observablehq.com/@starboard .

[DecentCr8]

Did you see the initial pledge per TiB is increasing? As a provider, one would prefer sector extension to on-boarding new sectors from the cost perspective. In other words, the provider who does not want sector extension, will not onboard new sectors. In this case, what do we lose?

[TippyFlitsUK]

Thank you for the less technical writeup @zixuanzh!! Sincerely appreciated!!

If we do nothing as a community and the status quo of sluggish growth persists, the network may see sector expiration exceed new sector onboarding soon and that in and of itself creates other complex issues.

Would you be able to expand on this point, please? This implies that the SPs who own these expiring sectors will not be using the released collateral to re-pledge new sectors. Are you able to share any modelling or data that supports this?

Additionally, this FIP would potentially lead to a requirement of up to 5X the collateral pledge for the same 1X quantity of onboarded data. Won't this only serve to exacerbate the issue?

Many thanks for your time! 🙏

[brendalee]

thanks for the filecoin macro overview, @zixuanzh - it's helpful to have the broader perspective in mind wrt timing of this particular FIP.

@TippyFlitsUK to answer the question you have - I think for some SPs that I have talked to, there are harder aspects of the around working with investors / lenders (where the return needs to come faster, or the loan duration is only for 6 months to a year). so in that case the released collateral may have to go to pay back loans / investors etc. If the SP does not have enough capital/FIL themselves to continue operations, then that scenario would happen. I personally don't have a very deep understanding of the breakdown of SPs, but I do think a nontrivial percentage of them fall into this category.

[TippyFlitsUK]

Many thanks, @brendalee! It's great to hear feedback from as many SPs as possible and to learn about the various different challenges that we are all facing. This information has been rather lacking until this point so I really appreciate you sharing your findings!!

It does sound like this regrettable outcome is already rather inevitable though. It's tough to see how this FIP will improve their circumstances within the time frames of concerning expiry that ZX has highlighted in his post.

[stuberman]

@zixuanzh I very much appreciate your write up and elevating the conversation to the good of the whole ecosystem.

I do want to dig a bit deeper into your statement Storage onboarding - a key health metric for the Filecoin network - has slowed down significantly in recent months.

I think there is a very important distinction to be made regarding growth of sectors and growth of storage.
In the charts that are provided, these are sector growth charts which include both CC and stored data deals.
Have the growth of data deals been analyzed? I doubt they are shrinking as SPs are adding as much as possible data to the network.
I presume the decline of sectors is due to the decline of CC sectors, which makes very good economic sense given the state of crypto markets. I infer this because deal data makes up around 0.07% of the entire network.

The key metric for our health needs to be growth of deal data onboarded (and yes longer deal durations are critical).
It would be worrisome to see CC growth continue as I understand the majority of those miner operations do not support data onboarding and are not contributing to the purpose of the network, which is to store humanity's most important information.

It would make more sense for this FIP to encourage data onboarding and reduce incentives for CC growth.

[DSS-AL]

If PL and FF had generated meaningful demand for the product through functionality, stability and customer engagement then we would not have this problem. Fixing circulating supply issue artificially through market manipulation at the expense of SP's is not the solution.

The only winners here are those holding massive amounts of FIL, the Filecoin Foundation, Protocol Labs and their direct investors. If PL was not selling down / distributing their massive interest in FIL then perhaps the price would have held up a little better before today.

[brendalee]

There's a ton of work that's being done around functionality / stability / customer engagement as @jnthnvctr has mentioned in the second half of his comment: #421 (reply in thread) . I think theres the reality of the time it takes to build a protocol / tooling / products within a short period of time (and just like any other engineering org, shipping faster also means that we have to resolve tech debt later on) and other factors outside our control (like the market environment).

I think mentioned multiple times though, one of the goals of this FIP is to award the general Filecoin ecosystem -
"We introduce the idea that longer term sectors represent a long-term investment and commitment to the Filecoin ecosystem, and therefore should be rewarded proportionally with greater block reward shares."

I do acknowledge there are realities that makes it difficult to realize these benefits depending on the situation though, many of which are detailed in the comments in this discussion. I think it would be helpful if you can share from your perspective how you wouldn't be able to benefit from different components of this FIP (which might not have already been pointed out), since the biggest motivation of the FIP was to reward participants in the network who have a longer term commitment.

[MegTei]

The support for this FIP will come down to how it will be deployed day 1 and the clarity of goals/ outcomes.
The short term goal is to extend the life of verified deals and the long term goal is to generate meaningful demand for the Network, where incentives go to those SPs who generate it. How do you bridge the short and long term so that SPs with capital constraints arent disincentivised to exit?
@dkkapur From a Governance perspective Im unclear what this FIP requires. Can you clarify?

[psh0691]

It's been a long time since we had such a lively debate.
What I think is that wealthy SPs may survive and small SPs may not be able to keep up.
There is also a problem with the circulation of funds due to too much pledge money.
I think the issue of pledge money should be reduced from various angles.
You can also consider differentiating pledges, such as gradually decreasing pledges by 10% over time (e.g., 6 months), or reducing pledges for effective data or transaction data storage.

[btcioner]

After the great infrastructural leap forward, the secondary market has seen a large number of equipment sold off.

The current market conditions are such that every de-encapsulation of physical data is essentially an increase in fixed expenditure.

There is a situation where the more you encapsulate, the more you lose. For every 1TB of data encapsulated, there are electricity costs, maintenance costs and software maintenance costs.

This is a calculation that is better than not expanding in size.

However, by retaining the current scale, we are unable to expand our profitability fundamentals.

It is a game dilemma.

The launch of FIP-421 breaks the current dilemma.

Effective combination of capital flexibility and stockpiling of assets is utilised.

Realised performance of SP on ecologically bullish long-term options while adjusting for secondary market idle capital.

Lay the foundation for long term development of FIL Eco.

We are in favor of this FIP

[mogli0528]

It's a very weak move, trying to use an economic method to cover the failure of building ecology/applications.

Filecoin fails to attract clients/users to onboard their data onto mainnet. That's the reason why sectors growth is sluggish. The only way to solve this problem is to give more effort/reward to build the ecology, a popular application will be much more impactful to FIL price than a Ponzi FIP.

TBH, even at this price, SP can still make a profit if sealing 100% FIL+ deals. SPs who spend too much on sealing machines and can not seal FIL+ deals are suffering. Those SPs are not here to provide storage service but make quick money, and they are leaving. That's not a bad thing for the community in the long run.

Also, a 5-year pledge from now ensures Filecoin foundation and PL buyers of unfinished products. It won't be good stimulation for these 15% FIL holders who are also the main contributor.

Onboard FVM, build applications and attract more users, that's what we should do.

[dmc314]

I think @jnthnvctr said it well above:

imho for Filecoin its not an either or (re tokenomics and functionality), but an and.

Thousands of developers and hundreds of projects are, in fact, racing to build out valuable, real-world business models on top of Filecoin.

This takes time, however, and requires trust in the stability of resources and infrastructure, which is what this FIP aims to accomplish.

By comparison, AWS, Google Cloud, Azure, and Ethereum have had 16, 14, 12, and 7 years respectively to grow their ecosystems to where they are today. Filecoin has been working at it for 2, and has made pretty incredible progress in that short a time from my perspective.

And regarding the inflammatory use of the term ponzi scheme, that is patently false. The tokens that Protocol Labs and Filecoin Foundation have are not some get-rich-quick scheme, but rather are being judiciously and amply put to work as salaries, grants, prizes, and startup investments towards building the ecosystem exactly as you suggest.

[mogli0528]

I'm pretty impressed by what PL and Fil Foundation have done, and I'm not intending to push them to do it faster. However, this FIP is not like what they used to do. Trying to use economic methods to limit the circulating supply is like a ponzi scheme for me. We have a better way, and we are working on it.

And another reason why the QAP is not growing fast right now is that CC sealing overheated last year. A low-profit reality will force SPs to choose more cost-efficient solutions and participant in high-value projects (like Fil+, slingshot). As you can see, the proportion of verified data grows faster and faster. I think we are heading in the right direction without this Fip.

[jennijuju]

@mogli0528
Disclaimer - I work on lotus @ PL, tho opinions are my own.
I think @zixuanzh 's post has explained what motivated this FIP and the good intention behind it. We don’t all agree the proposed solution is the best, including myself, which is fine. The team is taking all the valid feedback and updating the proposal accordingly. What can be helpful is to brainstorm with us (the community), and provide feedback to the authors - so we can develop a better solution for the network and move forward with it.

I consider a lot of the feedback raised in this discussion to be constructive and points out the other pain points & things that should be considered to be prioritized by the team working on making Filecoin better (including your comments). Many of them should be broken into their own FIP discussion and be in community & core dev's backlogs.

However, this FIP is not like what they used to do.

Now, putting huge and skeptical labels won’t help filecoin to grow. And we need teams to work on different streams in parallel to make sure filecoin improves as a whole: that’s why we have the FVM team that’s head down working on FVM, CrytonetLab continues to research on protocol opportunities, boost working on better deal making software, lotus working on network protocol changes and growth, FF & outercore and many other community/ecosystem teams are raising awareness and onboarding use cases to filecoin daily, and etc. That being said, CEL is doing its part to contribute to the Filecoin network. During their stewardship of Filecoin cryptoecon and network health, they proposed a solution that might benefit the Filecoin network & its participants in the long term. What I'm trying to say is that PL isn’t a monolith - there’s a bunch of different separate orgs working independently of each other, and one cannot make Filecoin successful and better on its own. We should show some appreciation for each other's work, always 💙

[kaitlin-beegle]

The FIP Authors for FIP0036 have submitted a new PR, based on ongoing community feedback. In order to support discussion around this draft, and for ease of access, this thread will be locked.

For updates and continued community conversation, please feel free to weigh in on the updated thread.