The Send button was enabled whenever the amount alone fit the balance, so
a max-value ETH send passed the confirmation screen and failed at
broadcast, after the user had committed to it.
The arithmetic moves into src/shared/txValidation.js as a pure function
over 18-decimal fixed point: native ETH now requires amount + fee <=
balance, and an ERC-20 transfer requires the ETH balance to cover the fee
on top of the token check, reported as its own error. Validation re-runs
when the async estimate resolves; Send stays disabled while the estimate
is pending and when it fails, so an unknown fee is never treated as zero.
The fee messages are static elements that already reserve their space, so
nothing moves when the estimate lands.
The fee reserved is the one the node will actually require. The send pins
no fee fields, so ethers broadcasts a type-2 transaction and the node
validates it against value + gasLimit * maxFeePerGas; gating on gasPrice
would under-reserve by roughly gasLimit * baseFee and let through exactly
the broadcast failure this change exists to prevent. feeReserveWei()
derives that reserve, falling back to gasPrice only where no type-2
pricing exists. The reserve is read when the screen opens and the
broadcast derives its own, so a base fee that roughly doubles while the
user is at the password field can still outrun it: this turns a
deterministic failure on every max-value send into a rare one, not into
none.
The fee line shows both numbers rather than one - what the transfer is
expected to cost, and below it the larger amount reserved until it
confirms. Quoting only the reserve overstates the typical mainnet cost by
roughly double on every send; quoting only the estimate contradicts the
gate. The second line holds its space from the first paint.
validateTransfer() fails closed: a feeWei that is not a non-negative
bigint under FEE_KNOWN, any unrecognised feeStatus, and a negative amount
all block, rather than counting as a fee of zero or as an amount that
passes every comparison trivially.
