Add a StateExactCirculatingSupply command #4148

arajasek · 2020-10-03T03:32:34Z

At height 136,374, circ supply is 15579755.735329679830248779 FIL

documentation/en/api-methods.md

magik6k

Logic looks good

magik6k · 2020-10-05T23:55:29Z

api/api_full.go

@@ -400,6 +400,9 @@ type FullNode interface {

 	// StateCirculatingSupply returns the circulating supply of Filecoin at the given tipset
 	StateCirculatingSupply(context.Context, types.TipSetKey) (CirculatingSupply, error)
+	// TODO: Remove StateCirculatingSupply maybe?


It's used by chainwatch/visor to put this data into postgres, which the CE team can query, so probably not

We could add a comment to StateCirculatingSupply saying that it's an approximation, and that people should use StateCirculatingSupply in most cases

*StateExactCirculatingSupply

magik6k · 2020-10-06T00:01:58Z

cli/state.go

@@ -1638,6 +1639,34 @@ var stateCircSupplyCmd = &cli.Command{
 	},
 }

+// TODO: Remove stateCircSupplyCmd


I'd

Make this the default output of circulating-supply

Add a --vm-supply flag to get the api.StateCirculatingSupply one

Yeah, sounds like the approx CircSupply isn't going anywhere, so I will do this

whyrusleeping · 2020-10-09T04:59:52Z

chain/stmgr/stmgr.go

+func (sm *StateManager) GetExactCirculatingSupply(ctx context.Context, height abi.ChainEpoch, st *state.StateTree) (abi.TokenAmount, error) {
+	circ := big.Zero()
+	unCirc := big.Zero()
+	err := st.ForEach(func(a address.Address, actor *types.Actor) error {


should skip further processing on actors with zero balance

arajasek · 2020-10-11T05:51:18Z

chain/stmgr/stmgr.go

+			a == builtin.SystemActorAddr ||
+			a == builtin.CronActorAddr ||
+			a == builtin.BurntFundsActorAddr ||
+			a == builtin.SaftAddress ||


We could start including funds in this address post-liftoff, but eh

or just include it now, there's only 4M in it.

raulk

Not being completely in the know about the rationale behind these changes (the PR description doesn't explain much about the motivation of this change, which makes it hard to follow), it looks like a smell that:

we have two circulating supplies,
we're choosing to expose the "exact" one via the API by default, but letting the VM use the non-exact value.

Intuitively, something feels a bit off here, either the divergence in values, or the naming. The circulating supply is a crucial cryptoecon variable, and IMO it should be as minimally ambiguous as possible.

arajasek · 2020-10-11T18:18:37Z

@raulk You're not wrong, as I understand it, we're kinda stuck between two requirements here.

We must have an accurate calculation of the liquid FIL in the network. This is necessarily slow, but is the correct number to report to anyone external who wants to know how much FIL is circulating.
We must have a quick approximation of the liquid FIL in the network. This is needed to be used internally when doing CE math (and so by the VM).

The naming is intended to reflect that external parties probably wanna use the slow, accurate method. Very open to hearing recommendations, though.

raulk · 2020-10-11T18:25:12Z

@arajasek

This is needed to be used internally when doing CE math (and so by the VM).

Apologies for ignorance, what does CE mean here?

arajasek · 2020-10-11T22:17:50Z

@arajasek

This is needed to be used internally when doing CE math (and so by the VM).

Apologies for ignorance, what does CE mean here?

CryptoEcon, sorry

arajasek requested review from Kubuxu, magik6k and whyrusleeping as code owners October 3, 2020 03:32

arajasek commented Oct 3, 2020

View reviewed changes

documentation/en/api-methods.md Outdated Show resolved Hide resolved

arajasek force-pushed the asr/circ-supply branch from 60a9ae9 to bc72681 Compare October 3, 2020 20:49

magik6k reviewed Oct 6, 2020

View reviewed changes

arajasek force-pushed the asr/circ-supply branch from bc72681 to c938424 Compare October 6, 2020 04:26

whyrusleeping reviewed Oct 9, 2020

View reviewed changes

arajasek force-pushed the asr/circ-supply branch from a909a8c to fbcfa7c Compare October 11, 2020 05:49

arajasek requested a review from raulk as a code owner October 11, 2020 05:49

arajasek commented Oct 11, 2020

View reviewed changes

arajasek force-pushed the asr/circ-supply branch 2 times, most recently from 971985a to 4762430 Compare October 11, 2020 06:30

raulk reviewed Oct 11, 2020

View reviewed changes

arajasek force-pushed the asr/circ-supply branch from 4762430 to a56683c Compare October 11, 2020 22:17

arajasek added 2 commits October 12, 2020 16:38

Add an endpoint for precise circulating supply

96e1dfd

Rename StateVMCirculatingSupply to StateVMCirculatingSupplyInternal

83624a8

arajasek force-pushed the asr/circ-supply branch from a56683c to 83624a8 Compare October 12, 2020 20:42

whyrusleeping approved these changes Oct 12, 2020

View reviewed changes

arajasek merged commit 195d96c into master Oct 12, 2020

arajasek deleted the asr/circ-supply branch October 12, 2020 21:02

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add a StateExactCirculatingSupply command #4148

Add a StateExactCirculatingSupply command #4148

arajasek commented Oct 3, 2020 •

edited

magik6k left a comment

magik6k Oct 5, 2020

magik6k Oct 6, 2020

magik6k Oct 6, 2020

arajasek Oct 6, 2020

whyrusleeping Oct 9, 2020

arajasek Oct 9, 2020

arajasek Oct 11, 2020

arajasek Oct 11, 2020

raulk left a comment •

edited

arajasek commented Oct 11, 2020

raulk commented Oct 11, 2020

arajasek commented Oct 11, 2020

Add a StateExactCirculatingSupply command #4148

Add a StateExactCirculatingSupply command #4148

Conversation

arajasek commented Oct 3, 2020 • edited

magik6k left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

raulk left a comment • edited

Choose a reason for hiding this comment

arajasek commented Oct 11, 2020

raulk commented Oct 11, 2020

arajasek commented Oct 11, 2020

arajasek commented Oct 3, 2020 •

edited

raulk left a comment •

edited