Skip to content

Latest commit

 

History

History
270 lines (210 loc) · 10.4 KB

ROADMAP.md

File metadata and controls

270 lines (210 loc) · 10.4 KB
Raw

Roadmap 2023

Welcome to the Cosmos SDK's team roadmap.

This document is meant to help the team get feedback on the proposed work and for others to follow where we stand in our process. This will be a living document updated on a regular basis. If you'd like to participate in any workscope or would like to suggest another feature please reach out to Marko or Sam and we will schedule a call to discuss the feature request.

Q1

Storage

  • Produce a spec for the new store design
    • Research a new design for store. This could entail writing some POC's in order to identify design patterns
  • Store as its own go module
    • Store module should be its own go.mod without a dependency on the Cosmos SDK
  • Begin implementation of store v2
    • Identify the migration path from store v1 -> store v2
  • Parallel execution of state
    • RFC/ADR is merged into the main on the sdk
  • Optimistic execution
    • RFC/ADR is merged into main on the sdk

Client UX

  • Release v1 of query support (auto-cli)
    • A version of query support has been merged, documentation is missing
  • Dynamic metadata support
    • Dynamic support allows the Cosmos SDK to release a cmd line tool that could work with any chain.
    • Add metadata support to latest version of Cosmos SDK and, if possible, backport to older versions
  • Multi-chain command (Done)
    • Release a cmd line tool that can be pointed a grpc endpoint which then can produce cmd lines to interact with the chain
  • Auto-cli tx support
    • Tx support for auto-cli/hubl
      • This would fully remove the need for application developers to write cli commands for their modules
  • Consensus Key Rotation

Dev UX

Testing

  • integration testing framework
    • design and merge a integration testing framework.
    • The goals of the framework would that a module only needs to depend on modules that it depends on outside of testing, not all modules in the sdk like today.

ABCI 2.0

Issue: cosmos/cosmos-sdk#12272

  • ADR ABCI 2.0
    • Write ADR for integration of vote extensions & finalize block

Blocked:

once cometBFT has a release candidate of ABCI 2.0 (cmt 0.38)

  • Integrate ABCI 2.0

Security

  • Circuit breaker
    • Implement the circuit breaker module and prepare releases for Cosmos SDK versions 0.45, 0.46 and 0.47

IAVL

  • ADR 001
    • Implementation has been completed, waiting on reviewers
  • Prepare the migration path
    • Migration path has not been finalized
  • Test on mainnets
    • If possible we would like to test the new IAVL design on mainnets to observer behavior changes.

Issue: cosmos/iavl#548

Q2

Storage

  • Storage v2
    • Objective:
      • Goal is to get 60% of the way through the rewrite of storage
      • Begin testing the rewrite on mainnets if possible
      • External Audit
    • Progress:
      • On pause till ABCI 2.0 ships
  • Optimistic Execution
    • Objective:
      • Users should be able to pick between delayed execution and optimistic
      • RFC/ADR is merged
      • Implementation started.
    • Progess:
      • On pause till ABCI 2.0 is merged

Client UX

  • Hubl/AutoCLI
    • Objective:
      • Allow users to sign and submit transactions using hubl
      • Add module support for autocli
      • Deprecate/remove legacy cli (optional)
    • Progress:
      • Signing support is being refactored and is near completion
      • Adapting modules to use autocli instead of manually written cli
  • Consensus Key Rotation
    • Objective:
      • Allow users to rotate consensus keys for their validators
    • Progress
    • Merge pull request and updated ADR into main
  • Operator key rotation
    • Objective:
      • Allow users to rotate operator keys for their validators
    • Progress:
      • secondary val index was added to support operator key rotation
      • Pr is open and is being worked on

Dev UX

  • Toolkit/SDK ADR.
    • Objective:
      • Produce a RFC/ADR on how to make core composable
      • Merge RFC/ADR into main
    • Progress:
      • on pause until abci 2.0 integration is completed
  • Adopt core api fully in modules
    • Objective:
      • Remove the Cosmos SDK and Comet as a dependency from all modules
      • Release v1 of modules that have their dependency graph cleaned up
      • Spin out 4 more modules into their own go.mods
    • Progress:
      • Core api has been integrated into all modules
      • KvStoreService is being added to all modules
  • Remove global bech32
    • Objective:
      • Depreacte global bech32 setting
    • Progress:
      • All modules except bank use the address.Codec
  • Make sdk.Msg only be proto.message
  • Collections
    • Objectives
      • Migrate all modules
      • Add query support
      • Add schema support
    • Progress:
      • We have migrated three modules

ABCI 2.0

  • ABCI 2.0
    • Objectives:
      • Integrate comet 0.38
      • QA
    • Progress:
      • Integration has started

Testing

Modules

  • Invariant Checking
    • Objective:
      • Design a new system for checking invairants
      • Implement changes
      • Audit current invariants
    • Progress:
      • NA
  • Accounts
    • Objective:
      • Allow users to use account abstractions
      • Implementation is completed
      • External Audit is scheduled (if needed)
    • Progress:
      • RFC/ADR is in review

Research

  • Commitment Structure
    • Objective:
      • Identify different commitment structures that could be used in the sdk and how they would be plugged (highlevel)
    • Progress:
      • Conversations in slack have started
  • Cross lang
    • Objective:
      • Answer, what is needed to support many languages
      • Answer, what sort of encoding of passing between the cgo/ffi boundary
    • Progress:
      • working group is meeting bi weekly
      • ADR/RFC is in review

Q3

Storage

  • Audit & release of storage refactor
  • Identify further optimizations for storage
    • Goal is to identify the next bottlenecks in storage or the state machine

Dev UX

  • Complete Toolkit/SDK implementation refactor
    • Goal is to release the new version of the sdk allowing for further composability
  • Implement fee market abstractions
    • Goal is to release an alpha version of fee market abstractions

Modules

  • Governance
    • Make gov and groups composable with each other, not duplicate
  • Staking
    • Research a new staking design
    • Begin Implementation

Research

  • Nonce Lanes
    • Goal is to produce a spec and/or viability of using lanes for nonces instead of a single sequence number.

Q4

Research

  • Stateless clients
    • research how stateless clients could evolve in cosmos

This document will be updated at the end of the quarter on what was achieved and what was not. Shortly before the quarter concludes a new section will be added for the next quarter. We are working on updating the complete one year roadmap and will be posting it here as well.