-
Notifications
You must be signed in to change notification settings - Fork 17
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
poke once a day, use struct, more changes #4
base: sticky
Are you sure you want to change the base?
Conversation
uint32 public age; // time of last poke | ||
struct Accumulator { | ||
uint256 val; | ||
uint32 ts; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm not sure tracking the exact poke ts is worth the added poke cost (vs just using the eod ts). Generally if we expect the window to be much larger than 1 day (e.g. 30 days) the gain in precision is likely to be negligible (I did some quick approximation math to convince myself of it but might be good to test with actual values).
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Well the poke is only done once a day, so even if it's a matter of a few storage operations (probably less) I think it's negligible.
I think storing the timestamps is the more simple and standard way (for example Uniswap have this Observation struct - https://github.com/Uniswap/v2-periphery/blob/master/contracts/examples/ExampleSlidingWindowOracle.sol#L19).
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ok, though note that in the Uniswap example the timestamp is necessary as the accumulators are not indexed by days. That said I don't mind storing the ts here given the frequency of poke
calls. I guess we don't need it to be packed as a uint32
though, uint256
would work fine and save a tiny bit of gas.
Co-authored-by: sunbreak1211 <129470872+sunbreak1211@users.noreply.github.com>
// days_ == N will fill up a window corresponding to [lo == N, hi == 1] along with the current day | ||
// days_ should be selected carefully as too many iterations can cause the transaction to run out of gas | ||
// if the initiated timespan is shorter than the [lo, hi] window the initial cap will just be used for longer | ||
function init(uint256 days_) external auth returns(bool) { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
function init(uint256 days_) external auth returns(bool) { | |
function init(uint256 days_) external auth returns (bool) { |
@@ -17,29 +17,39 @@ | |||
pragma solidity ^0.8.16; | |||
|
|||
interface PipLike { | |||
function read() external view returns (uint128); | |||
function peek() external view returns (uint128, bool); | |||
function read() external view returns (uint128); // TODO: shouldn't this (and our function) return bytes32? https://github.com/makerdao/osm/blob/e36c874b4e14fba860e48c0cf99cd600c0c59efa/src/osm.sol#L150C49-L150C56 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The main reason for using a uint128
type was to be able to pack val
and age
in the same slot (and avoid an explicit cast for a value which under the hood is anyway a uint128
in the pip/osm). But since these variables have been removed, feel free to use a bytes32 (like in the osm) or a uint256 (like in the pip).
uint96 public slope = uint96(RAY); // maximum allowable price growth factor from center of TWAP window to now (in RAY such that slope = (1 + {max growth rate}) * RAY) | ||
uint8 public lo; // how many days ago should the TWAP window start (exclusive) | ||
uint8 public hi; // how many days ago should the TWAP window end (inclusive) | ||
uint96 public slope = uint96(RAY); // maximum allowable price growth factor from the average value of a TWAP window (in RAY such that slope = (1 + {max growth rate}) * RAY) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What I was trying to say here is that, given that the TWAP is an unbiased smoothed estimate of the price at the time in the middle of the TWAP window, the slope
can be interpreted as the allowable growth between that particular time and now. But maybe it's simpler to just describe it as a TWAP multiplier.
uint96 public slope = uint96(RAY); // maximum allowable price growth factor from the average value of a TWAP window (in RAY such that slope = (1 + {max growth rate}) * RAY) | |
uint96 public slope = uint96(RAY); // maximum allowable price growth factor in reference to the TWAP (in RAY such that slope = (1 + {max growth rate}) * RAY) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Ahh ok got it. Although I think it is easier to describe it as the multiplier of the TWAP.
Your correction looks good to me.
No description provided.