feat(x/mint): Implementing inflation in BeginBlocker #311
Conversation
There was a problem hiding this comment.
I would like to add a test case.
Mock StakingKeeper: bonded ratio is constant and within bands (eg: 66%).
Inflation is 10%, adjust context to make it run for a year, with block time always <= params.MaxBlockDuration.
Calculate the expected inflation (note: needs compounding calc).
|
|
||
| // getting last block info | ||
| lbi, found := k.GetLastBlockInfo(ctx) | ||
| if !found { |
There was a problem hiding this comment.
when is the case for which this can be not found?
There was a problem hiding this comment.
I dont think there is a case where this will not be found 🤔
x/mint/keeper/minter.go
Outdated
| // amount of bond tokens to mint in this block | ||
| bondedTokenSupply := k.GetBondedTokenSupply(ctx) | ||
|
|
||
| tokens = blockInflation.MulInt(bondedTokenSupply.Amount).Mul(sdk.NewDecFromBigInt(big.NewInt(int64(elapsed.Seconds()))).QuoInt64(int64(Year.Seconds()))) // amount := (inflation * bondedTokenSupply) * (elapsed/Year) |
There was a problem hiding this comment.
in the getBlockInflation function, we're multiplying the inflation change by NANOSECONDS whilst here we're dividing by SECONDS.
There was a problem hiding this comment.
Also can we split the inflation * bondedTokenSupply and the elapsed/year
There was a problem hiding this comment.
We can also simplify the calculation timeElapsedRatio = elapsed/year without any conversion, just using the raw duration, and then multiply it.
There was a problem hiding this comment.
I am not sure we can do timeElapsedRatio := elapsed/Year as both are internally just int64 values. and smallInt/largeInt will just give 0.
Doing timeElapsedRate := sdk.NewDecFromBigInt(big.NewInt(int64(elapsed))).QuoInt64(int64(Year)) returns the small decimal value. But yea, there could be better ways to do this🤔
| mintedTokens, found := k.GetInflationForRewards(ctx) | ||
| if found { | ||
| // Track inflation rewards | ||
| k.TrackInflationRewards(ctx, mintedTokens) |
There was a problem hiding this comment.
if it's not found I'm not sure if skipping it is the correct way to do it.
Old params would be applied.
There was a problem hiding this comment.
if its not found, that would mean there are no inflation rewards to track right. like it would be 0stake in the current block
…imulation Fdymylja/add inflation simulation
| sdk "github.com/cosmos/cosmos-sdk/types" | ||
| ) | ||
|
|
||
| const Year = 24 * time.Hour * 365 |
There was a problem hiding this comment.
This gets us 31 536 000 seconds per year.
Should we replace this with 31 557 600 seconds to accommodate 4 year avg to handle leap years?
There was a problem hiding this comment.
We should prefer the leap year average... it is better for overall financial math calculations;
|
PR still relevant? |
|
☠️ Closing cuz time based inflation RIP |
Ref: #287
x/mint
x/rewards
note
❗Reviewer pls check 🙏🏻
sdk.Coinfrom asdk.Decvaluearchway/x/mint/abci.go
Line 49 in e8897c4
archway/x/mint/keeper/minter.go
Line 62 in e8897c4