x/incentives: index synthetic lockups by denom

[czarcas7ic]

Background

Synthetic lockups aren't indexed by denom as expected

osmosis/x/incentives/keeper/distribute.go

Lines 91 to 109 in fb0de8c

	// getLocksToDistributionWithMaxDuration returns locks that match the provided lockuptypes QueryCondition,
	// are greater than the provided minDuration, AND have yet to be distributed to.
	func (k Keeper) getLocksToDistributionWithMaxDuration(ctx sdk.Context, distrTo lockuptypes.QueryCondition, minDuration time.Duration) []lockuptypes.PeriodLock {
	switch distrTo.LockQueryType {
	case lockuptypes.ByDuration:
	// TODO: Create issue for this
	// Confusingly, synthetic lockups aren't indexed by denom as expected.
	// Thus you have to do this as a hack
	denom := lockuptypes.NativeDenom(distrTo.Denom)
	if distrTo.Duration > minDuration {
	return k.lk.GetLocksLongerThanDurationDenom(ctx, denom, minDuration)
	}
	return k.lk.GetLocksLongerThanDurationDenom(ctx, distrTo.Denom, distrTo.Duration)
	case lockuptypes.ByTime:
	panic("Gauge by time is present, however is no longer supported. This should have been blocked in ValidateBasic")
	default:
	}
	return []lockuptypes.PeriodLock{}
	}

Suggested Design

Will need to come back to this to figure out a suggested design.

Acceptance Criteria

• Synthetic lockups are indexed by denom
• Hack no longer required in getLocksToDistributionWithMaxDuration

[stackman27]

Hi @czarcas7ic is this issue being worked on? can i take a shot at it? Will follow up with @ValarDragon regarding the design

[czarcas7ic]

@stackman27 please do, I will assign it to you!

[ValarDragon]

Hey, just saw this issue. I think we should not index these, and instead be planning to delete the synthetic locks entirely.

Rather than maintain the current state, would rather put such energy into getting it deleted 💪

[stackman27]

Are we trying to make it a part of period locks?

[ValarDragon]

This is the design we should go towards imo #942

[ValarDragon]

Were not going to pursue this.