This is instead of encode::Errors because the encoders should
not be allowed to return errors that don't originate in the writer
they are writing into.
This is a part of the method definition that has been relied upon for a
while already.
Currently whenever an unrecognized network message is received, it is never
flushed from the read buffer, meaning that unless the stream is closed and
recreated it will keep returning the same error every time `read_next()` is
called.
This commit adds the length of the message to `UnrecognizedNetworkCommand`,
so that the `StreamReader` can flush those bytes before returning the error
to the caller.
`cloudhead` on IRC was asking how to properly use
`BlockHeader::validate_pow()` on genesis (or similar) when the
pow_limit field isn't expressible as a compact target (and, thus,
does not actually represent the PoW limit/genesis target). We
swap it for the actual PoW limit by truncating the way a compact
encoding round-trip would.
Note that, in Bitcoin Core, the original value is only ever used
once in its original form:
```
if (bnNew > bnPowLimit) // Note: bnPowLimit is params.powLimit
bnNew = bnPowLimit;
return bnNew.GetCompact();
```
Thus, even if Core adopted our change, as long as there exist no
256-bit integer x which satisfies
`x <= powLimit && x > encoding_roundtrip_truncated(powLimit)` and
`enoding_roundtrip_truncated(x) != powLimit`, the change would have
no impact on consensus.
It is trivial to show that there are no values which are between
the new value
(0x00000000ffff0000000000000000000000000000000000000000000000000000)
and the original value
(0x00000000ffffffffffffffffffffffffffffffffffffffffffffffffffffffff)
which can be encoded in compact form, but it is also critically, no
such values will encode to a compact form of anything different than
the new value as the encoding always truncates the low bits, never
rounding up.
Taking an external dependency just to convert ints to byte arrays
is somewhat of a waste, especially when Rust isn't very aggressive
about doing cross-crate LTO.
Note that the latest LLVM pattern-matches this, and while I haven't
tested it, that should mean this means no loss of optimization.
It has been a long time since Bitcoin's maximum network message
length was 32MB, so we should follow suit. This is also an important
DoS limit, so we should limit it as much as possible.
- Implement psbt::Map trait for psbt::Output
- Add (en)decoding logic for psbt::Output
- Implement PSBT (de)serialization trait for relevant psbt::Output types
- Add macro for merging fields for PSBT key-value maps
- Add macro for implementing decoding logic for PSBT key-value maps
- Add convenience macro for implementing both encoding and decoding
logic for PSBT key-value maps
- Add macro for inserting raw PSBT key-value pairs into PSBT key-value
maps
- Add macro for getting raw PSBT key-value pairs from PSBT key-value
maps