Cardano Updates

Upstream cleanup, not Leios-specific: eight unused-import warnings that
upstream's own ghc-options promote to errors via 'nix/ouroboros-network.nix'
('-Werror' on forAllProjectPackages). They've been present at the
'main-peras-5202-merged' SHA that cardano-node and ouroboros-consensus
master both consume; surface when downstream haskell.nix builds also
inherit `-Werror`.

Affected files (one unused qualified import each):
- framework/lib/Ouroboros/Network/Protocol/Handshake/Codec.hs (BL)
- framework/tests-lib/.../ConnectionManager/Experiments.hs (LBS)
- framework/io-tests/.../Driver.hs (List)
- framework/io-tests/.../Socket.hs (IntMap)
- tests/lib/.../TxSubmission/AppV1.hs (BSL)
- tests/lib/.../TxSubmission/AppV2.hs (BSL)
- protocols/tests-lib/.../KeepAlive/Test.hs (BL)
- protocols/tests-lib/.../PeerSharing/Test.hs (BL)

Verified with 'cabal clean && cabal build all --ghc-options=-Werror' →
exit 0.

Co-Authored-By: Claude Opus 4.7 <[email protected]>

Wrap `Channel m a` recv values in a new `Reception a = MkReception
!(IntMap Time) !a`, recording per-chunk arrival times.  In the mux
demuxer the chunk's arrival time is paired with its starting byte
offset; codec/driver layers either strip the time map (the codec sees
only bytes) or propagate it through `wrapMiniProtocolTrailing` at the
mini-protocol callback boundary.  Tests/demos/benches use `IntMap.empty`
when synthesising trailing.

This is the network half of the November Leios demo's arrival-time
plumbing; consumers (e.g. Praos block-arrival tracing on the consensus
side) will be wired up separately.

Co-Authored-By: Claude Opus 4.7 <[email protected]>

Upstream cleanup, not Leios-specific: eight unused-import warnings that
upstream's own ghc-options promote to errors via 'nix/ouroboros-network.nix'
('-Werror' on forAllProjectPackages). They've been present at the
'main-peras-5202-merged' SHA that cardano-node and ouroboros-consensus
master both consume; surface when downstream haskell.nix builds also
inherit `-Werror`.

Affected files (one unused qualified import each):
- framework/lib/Ouroboros/Network/Protocol/Handshake/Codec.hs (BL)
- framework/tests-lib/.../ConnectionManager/Experiments.hs (LBS)
- framework/io-tests/.../Driver.hs (List)
- framework/io-tests/.../Socket.hs (IntMap)
- tests/lib/.../TxSubmission/AppV1.hs (BSL)
- tests/lib/.../TxSubmission/AppV2.hs (BSL)
- protocols/tests-lib/.../KeepAlive/Test.hs (BL)
- protocols/tests-lib/.../PeerSharing/Test.hs (BL)

Verified with 'cabal clean && cabal build all --ghc-options=-Werror' →
exit 0.

Co-Authored-By: Claude Opus 4.7 <[email protected]>

Wrap `Channel m a` recv values in a new `Reception a = MkReception
!(IntMap Time) !a`, recording per-chunk arrival times.  In the mux
demuxer the chunk's arrival time is paired with its starting byte
offset; codec/driver layers either strip the time map (the codec sees
only bytes) or propagate it through `wrapMiniProtocolTrailing` at the
mini-protocol callback boundary.  Tests/demos/benches use `IntMap.empty`
when synthesising trailing.

This is the network half of the November Leios demo's arrival-time
plumbing; consumers (e.g. Praos block-arrival tracing on the consensus
side) will be wired up separately.

Co-Authored-By: Claude Opus 4.7 <[email protected]>

Leftover from previous design. There is no longer a need for
nextWakeDelay to take when the next TX expires from the retained cache.

simplify txIdAckable by removing superflous condition

Cleaning up stale entries doesnt provide peers with new options so we
only need to bump sharedRevision.

with the abstract HasRawTx there is no longer a need to convert the
RawTxId to a ShortByteString.

Differentiate between updates to SharedTxState that needs to be written,
and updates to SharedTxState that needs to be written and which should
wake up other peers.

Allow pipelining of Ack=0 Req=N messages. That is messages that
attempt to fill our window, but don't have any txids to ack.

This matches the way V1 does pipelining, but with the addition of the
limiting of sending new Ack=0 and Req=N messages if the last returned no
new txids.

Replace the central decsision thread by having server threads coordinate
between them by blocking on STM actions.

Fix TxSubmissionProtocolError instance, spotted by claude.

The Eq/Ord instance for txid allocates memory. RawTxId enables the
possibility of more efficient Eq/Ord instances.

Co-authored-by: Javier Sagredo <[email protected]>

Various changes

Allow pipelining of Ack=0 Req=N messages. That is messages that
attempt to fill our window, but don't have any txids to ack.

This matches the way V1 does pipelining, but with the addition of the
limiting of sending new Ack=0 and Req=N messages if the last returned no
new txids.