Merge branch 'main' into lowhung/rollback-sync-validation-tests
# Conflicts: # common/src/caryatid.rs
# Conflicts: # common/src/caryatid.rs
Signed-off-by: Chris Gianelloni <[email protected]>
Signed-off-by: William Hankins <[email protected]>
The deploy-haddocks job downloaded the haddock artifact but never called deploy-pages, so haddocks were never published. Meanwhile deploy-website correctly deployed the Docusaurus site but without haddocks. Since GitHub Pages only supports a single deployment per repo, merge both into one job: deploy-website now downloads the haddock artifact into static/haddocks/ before building, so Docusaurus bundles everything into a single deployment.
Signed-off-by: Chris Gianelloni <[email protected]>
Signed-off-by: William Hankins <[email protected]>
From https://github.com/intersectmbo/dmq-node at 70763daf0b8c86b74a3ea8cac8d31e7be594d15e
Empty stub pages erode reader trust by promising content that doesn't
exist. Delete 18 stubs across all four Diátaxis sections and update
the sidebar accordingly. Topics that need documentation are tracked
for later phases.
Deleted stubs: consensus_protocol, mempool, network_interaction,
starting_up_consensus_layer, genesis_design, managing_updates, ebbs,
storage_layer/{index,immutable_db,volatile_db,ledger_db},
reporting_a_bug, working_with_stored_data/{analysing_a_chain,
serving_a_chain,manipulating_data}, tutorials/{sop,hfc},
references/serialisation.
Also: add one-line descriptions to section index pages, clean up
haddocks.md TODO, fix broken EBB link in ledger_interaction.md.
replicated content from a word doc Signed-off-by: musikc (Christina) <[email protected]>
Signed-off-by: Chris Gianelloni <[email protected]>
Signed-off-by: William Hankins <[email protected]>
Signed-off-by: Chris Gianelloni <[email protected]>
New explanation page documenting how data flows through the consensus layer, based on code investigation of the actual component interactions. The page covers: - Overview: NTN (untrusted) vs NTC (trusted) connections, header-body split motivation, mini-protocols (ChainSync, BlockFetch, TxSubmission, LocalTxSubmission, LocalStateQuery, LocalTxMonitor), and internal components (ChainDB, Mempool, Block Forging) - Block flow (NTN upstream): ChainSync validates headers using chain state and ledger views from ChainDB, BlockFetch downloads blocks for validated chains, ChainDB performs chain selection - Block diffusion (NTN downstream): ChainSync server announces headers, BlockFetch server serves blocks, diffusion pipelining optimization - Transaction flow: NTN TxSubmission (bidirectional) and NTC LocalTxSubmission, both go directly to Mempool - Client queries (NTC): LocalStateQuery reads from ChainDB, LocalTxMonitor reads from Mempool - Internal flows: chain selection queue, Mempool revalidation on ledger state change, block forging cycle - Passive node: simplified diagram Each section has a focused Mermaid diagram showing the relevant components and data flows. The original reference data flow diagram is preserved in references/ for comparison. Also updates System Overview links to point to the new location.