[otherview]

Description

Fixes three defects in the thorclient/bind SDK, all in transaction construction:

• SendBuilder.Submit selected dynamic-fee mode when the best block has a base fee, but built the transaction with new(tx.Builder). The tx type stayed at the legacy zero value, so the fee caps were silently dropped and post-Galactica submissions paid the legacy gas price instead of the selected cap. The builder is now created with tx.NewBuilder(txType).
• DeployContract never set BlockRef (0) with Expiration(10000), so its transaction was expired on arrival on any chain past block 10,000 — it never worked on mainnet or testnet. It also hardcoded the legacy tx type (no dynamic-fee option), left the nonce at 0 (duplicate tx IDs), and hardcoded a flat 10M gas. It now simulates the creation clause up front for strict gas estimation (IntrinsicGas + simulated GasUsed) and a pre-flight revert check, selects the tx type from the best block's base fee, anchors BlockRef to the best block, and uses a random nonce. The public signature is unchanged.
• The ChainTag() doc example passed the chain tag to tx.NewBuilder(), which sets the tx type — it compiles (both are byte) but is wrong.

A shared clause-simulation helper (simulate.go) backs the estimation in both Submit and DeployContract; per-site policy is unchanged (Submit keeps the 15k OP_CALL buffer and no revert check; DeployContract is strict).

Client-side SDK only — no node code paths and no public API changes.

No linked issue.

Type of change

• Bug fix (non-breaking change which fixes an issue)

How Has This Been Tested?

• go build ./... and make lint clean
• Full ./thorclient/... suite passes; the bind tests deploy a contract through the new DeployContract path on a Galactica-enabled chain and exercise Submit end-to-end
• During triage, a regression test asserting tx type and fee caps was verified to fail on the unfixed code (test not included in this PR)

Test Configuration:

• Go Version: go1.26.5 darwin/arm64
• Hardware: Apple Silicon (macOS)
• Docker Version: N/A (in-process test node via testnode)

Checklist:

• My code follows the style guidelines of this project
• I have performed a self-review of my code
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes
• New and existing E2E tests pass locally with my changes
• Any dependent changes have been merged and published in downstream modules
• I have not added any vulnerable dependencies to my code

[codecov-commenter]

Codecov Report

❌ Patch coverage is 71.64179% with 19 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
thorclient/bind/send.go	74.19%	4 Missing and 4 partials ⚠️
thorclient/bind/simulate.go	72.41%	4 Missing and 4 partials ⚠️
thorclient/bind/contract.go	57.14%	2 Missing and 1 partial ⚠️

📢 Thoughts on this report? Let us know!

[libotony]

Direction looks good. A few non-blocking notes:

• DeployContract now re-implements most of Submit's tx assembly: type selection, fee default, blockRef, nonce, gas estimation. The two copies already differ. A shared helper would mean the fee-cap and gas-margin fixes below are made once.
• Pre-existing, not from this PR:
  • Submit writes the defaults back into the caller's TxOptions (opts := b.opts), so reusing a builder reuses the nonce, gas and blockRef.
  • Setting only MaxPriorityFeePerGas gets a zero effective tip, because MaxFeePerGas defaults to the base fee. The fee-cap fix covers this too.

[otherview]

Thanks. All points are in 2d99678: the shared buildTx/estimateGas, Submit copies TxOptions, and the default cap is 2 × baseFee + tip. I added TestContract_SubmitTxConstruction. It fails on the previous code.