README
¶
Actors
This package contains shims for abstracting over different actor versions.
Design
Shims in this package follow a few common design principles.
Structure Agnostic
Shims interfaces defined in this package should (ideally) not change even if the structure of the underlying data changes. For example:
- All shims store an internal "store" object. That way, state can be moved into a separate object without needing to add a store to the function signature.
- All functions must return an error, even if unused for now.
Minimal
These interfaces should be expanded only as necessary to reduce maintenance burden.
Queries, not field assessors.
When possible, functions should query the state instead of simply acting as field assessors. These queries are more likely to remain stable across specs-actor upgrades than specific state fields.
Note: there is a trade-off here. Avoid implementing complicated query logic inside these shims, as it will need to be replicated in every shim.
Documentation
¶
Index ¶
Constants ¶
const (
Version0 = iota
)
Variables ¶
This section is empty.
Functions ¶
func QAPowerForWeight ¶
func QAPowerForWeight(size abi.SectorSize, duration abi.ChainEpoch, dealWeight, verifiedWeight abi.DealWeight) abi.StoragePower
Doesn't change between actors v0 and v1
Types ¶
type FilterEstimate ¶
type FilterEstimate = smoothing0.FilterEstimate
func FromV0FilterEstimate ¶
func FromV0FilterEstimate(v0 smoothing0.FilterEstimate) FilterEstimate
type SectorInfo ¶
type SectorInfo = proof0.SectorInfo
TODO: Why does actors have 2 different versions of this?
Source Files
Directories