TL;DR

  • Jackal’s chain software is well behind the wider Cosmos ecosystem. That’s the first real barrier.
  • So we’ve spent the last few weeks properly assessing the upgrade path to SDK v0.47, IBC v7, wasmd v0.46.
  • It went well: Jackal’s custom messages observed in full, the old chain proven deterministic (same inputs, same fingerprint, twice), and the modern protobuf toolchain is already compiling Jackal’s code correctly.
  • Next comes the actual port, replayed against the exact same tests. Results published either way.
  • Nothing touches the live chain until it’s all proven.
  • No reset. Balances, staked JKL, names and history all carry over.

A modern stack removes any barriers; attracting builders, storage providers and activity back is a community and governance conversation that follows although I have a number of exciting ideas I’m itching to share with you once we stabilize things.

Why this work exists.

Jackal’s chain software is built on Cosmos SDK v0.45 and Tendermint v0.34, now several years behind the wider Cosmos ecosystem. Upstream maintenance has moved on. Wallets, bridges, IBC relayers and developer tools assume a newer stack.

This is needed so the chain doesn’t fall further behind with each passing year.

Who we are.

We are Orbital Alchemists - active validators on Jackal for over a year now. Not as long as others, but long enough to realise the importance of this chain.

I’m Si (Sich) and I have been developing software, mostly in corporate and consultancy roles, since the late 1990’s. You can find out more about me in my personal blog “One Alchemist” if you are interested.

We have immense respect for everything the original team built here.

What has actually been done.

Over the last couple of weeks we completed a systematic assessment of the upgrade path from Jackal’s current stack (SDK v0.45, IBC v4, wasmd v0.32) to a more modern one (SDK v0.47, IBC v7, wasmd v0.46)

  • The fee whitelist is fully understood and fully fingerprinted. Jackal’s four free storage messages (proofs, attestations, reports) are “fee-free” thanks to Marston’s modification (“fork”) of the Cosmos SDK. On an isolated testnet we measured that behaviour end to end — balances to the last ujkl, gas to the unit, every event, the fee-granting path included — so the rebuilt, fork-free version can be proven identical before it ever touches the live chain.

  • The old stack is proven a reliable measuring instrument. A chain upgrade’s dangerous moment is the boundary — every validator computes fresh state with new code, and one differing byte halts the chain. That only stays safe if the old chain is a deterministic machine: same inputs, same fingerprint, every time. We tested exactly that on Jackal’s custom stack (forked SDK, six custom modules): the same genesis and same transactions reproduced the identical block hash, twice. Restoring snapshots never proves this — restored nodes borrow history from peers; our rerun recomputed it. Every old-vs-new comparison to come is therefore sound.

  • The modern toolchain works on Jackal’s code today. All 37 protobuf-generated files were regenerated in isolation with the exact builder image modern chains use, and the code compiled unchanged otherwise.

  • Known differences are known, not surprises. Example: the new stack rejects zero-gas transactions earlier and more cleanly than today’s. Ordinary users never send these; the difference is documented and deliberate.

What this means for holders and node operators: no reset by design.

Jackal’s is .. errr … quiet, the storage providers have largely switched off and the community that remains is small. The point of this work is what comes next. An upgrade built this way carries everything over untouched - balances, staked JKL, names, chain history - so revival doesn’t require starting over.

Personally I’d prefer to keep the history and reputation if we can.

What we are deliberately not claiming.

No timeline just yet — the port is only beginning but I think it will be massively expedited by the assessment work done. No code changes exist for the live chain; all work so far is read-only analysis plus tests on isolated hardware, and our production validator runs untouched.

What happens next.

First implementation: a minimal working build of the new stack, then the same test matrix replayed against it, results published. Then a full testnet rehearsal including the state-export boundary. Only after that passes would a governance proposal be drafted — for the community to accept, amend, or reject. Updates will come at phase boundaries, with full technical detail at each step for anyone who wants to verify rather than trust.

— Simon (Sich), Orbital Alchemists