Cardano Updates

Signed-off-by: Chris Gianelloni <[email protected]>

- exp/nbf violations throw InvalidCredentialError with descriptive messages
- timestamps normalized: handles both seconds and milliseconds (Date.getTime())
- catch block rethrows InvalidCredentialError, swallows only signature failures
- removed console.log in favor of proper error propagation
- tests updated to assert specific error messages
- all SDJWT tests (11/11) passing locally

Signed-off-by: Abhigyan Singh <[email protected]>

The 96G ulimit -v killed 1500-node sims at slot 313 with RSS only at
58G — Rust+tokio allocator reserves more virtual address space on
larger topologies than physical commit alone implies. The cap was
sized for 750 nodes; 1500 needs more headroom. 256G is the board's
max physical RAM ceiling; actual commit is bounded by RAM + swap.

Reverts pigz -1 to pigz -9 — the faster compressor did not solve the
end-of-sim EventMonitor spike (still ~8 GB from 11M events flushed
at once, regardless of compressor speed). The bottleneck is the
unbounded mpsc channel, not compression.

Co-Authored-By: Claude Opus 4.7 (1M context) <[email protected]>

Signed-off-by: Allain Magyar <[email protected]>

Two fixes from Copilot's review:

  - HTTP probes now accept 307 and 308 alongside 200/301/302. Modern
    HTTP increasingly uses 307 (temporary, method-preserving) and 308
    (permanent, method-preserving) instead of 302/301; treating them
    as failures could misclassify valid upstream sites as dead. Four
    occurrences updated: probe_site, the Phase 3 main validator, the
    Phase 2b reexport probe, and the Phase 3a rescue probe.

  - Add a Herald changelog fragment under .changes/. The repo enforces
    these via .herald.yml and the existing PR template's YAML block
    isn't a substitute for a tracked file.

Two fixes from Copilot's review:

  - HTTP probes now accept 307 and 308 alongside 200/301/302. Modern
    HTTP increasingly uses 307 (temporary, method-preserving) and 308
    (permanent, method-preserving) instead of 302/301; treating them
    as failures could misclassify valid upstream sites as dead. Four
    occurrences updated: probe_site, the Phase 3 main validator, the
    Phase 2b reexport probe, and the Phase 3a rescue probe.

  - Add a Herald changelog fragment under .changes/. The repo enforces
    these via .herald.yml and the existing PR template's YAML block
    isn't a substitute for a tracked file.

* sidecar: classify convergence probe failures

* sidecar: wait for tip-ready producers before setup complete

* testnet: bump sidecar for convergence diagnostics

Fix LSM snapshot restoration by resolving double path nesting

Signed-off-by: Eric Torreborre <[email protected]>

Signed-off-by: Eric Torreborre <[email protected]>

Updates to the latest mithril

Replace the manual segmented-witness serialization with MemoBytes,
which simplifies the code significantly. The block body is now
serialized as a flat 3-element list [invalid_transactions, transactions,
peras_certificate] instead of 5 separate segments.

Also fix PerasCert to encode/decode as bytes (matching CDDL),
reimplement alignedValidFlags using IntSet, and expose
DijkstraBlockBodyRaw/MkDijkstraBlockBody from Internal module.

Co-authored-by: Alexey Kuleshevich <[email protected]>

Change genArrayTerm from Gen to MonadGen so it can be used in
CBORGen contexts (e.g. custom CBOR generators for Huddle specs).