refactor(bootstrap): cap refresh concurrency

[kariy]

Summary

• Cap the Settings-change refresh at 8 concurrent in-flight probes via futures::StreamExt::buffer_unordered. Unbounded parallelism worked against local katana but trips rate limits on hosted Starknet RPC providers for larger manifests. 8 is high enough that the common case (2–10 items) finishes in one batch with no measurable overhead, low enough to stay under typical free-tier 20 req/s caps.
• Delete TODOS.md. Both entries are resolved:
  • Bounded concurrency: addressed here.
  • Account-change refresh trigger: the original premise was wrong. The dirty flag already fires on any Settings field edit, not just RPC URL. compute_deploy_address already uses the current account, so unique=true deploys naturally re-probe at the new address when the signer changes. Locked down with tests.

Details

• New REFRESH_CONCURRENCY: usize = 8 constant in tui.rs, with a const _: () = assert!(REFRESH_CONCURRENCY >= 1) compile-time guard (because buffer_unordered(0) would deadlock).
• Probe futures boxed into Vec<Pin<Box<dyn Future<Output = ()> + Send>>> so declare and deploy probes share one stream. The async blocks have distinct anonymous types otherwise.
• Three new unit tests:
  • settings_dirty_fires_on_account_edit_not_just_rpc — locks down that editing the Account field sets the dirty flag.
  • compute_deploy_address_uses_current_account_for_unique_deploys — confirms unique=true yields different addresses per caller, unique=false is caller-independent. This is the coupling that makes account-change refresh correct-by-construction.
  • needs_refresh_includes_done_and_unknown_but_not_pending_or_failed (pre-existing, unchanged).

Test plan

• cargo nextest run -p katana-bootstrap — 98 passed, 0 failed
• cargo clippy -p katana-bootstrap --all-targets --no-deps — only pre-existing warnings remain
• cargo +nightly-2025-02-20 fmt --all
• Manual sanity: queue a 10+ item manifest, execute, switch RPC. Confirm the Refreshing state completes in one batch and items resolve correctly.

🤖 Generated with Claude Code

[github-actions]

Benchmark

Details

Benchmark suite	Current: a1b558d	Previous: 3db9b65	Ratio
CompiledClass(fixture)/compress	2759042 ns/iter (± 15607)	2574151 ns/iter (± 8127)	1.07
CompiledClass(fixture)/decompress	2873840 ns/iter (± 13908)	2840607 ns/iter (± 12378)	1.01
ExecutionCheckpoint/compress	32 ns/iter (± 7)	35 ns/iter (± 4)	0.91
ExecutionCheckpoint/decompress	25 ns/iter (± 7)	27 ns/iter (± 9)	0.93
PruningCheckpoint/compress	32 ns/iter (± 8)	35 ns/iter (± 2)	0.91
PruningCheckpoint/decompress	25 ns/iter (± 5)	27 ns/iter (± 5)	0.93
VersionedHeader/compress	646 ns/iter (± 5)	658 ns/iter (± 10)	0.98
VersionedHeader/decompress	815 ns/iter (± 6)	866 ns/iter (± 17)	0.94
StoredBlockBodyIndices/compress	78 ns/iter (± 5)	81 ns/iter (± 1)	0.96
StoredBlockBodyIndices/decompress	36 ns/iter (± 9)	40 ns/iter (± 8)	0.90
StorageEntry/compress	145 ns/iter (± 2)	183 ns/iter (± 2)	0.79
StorageEntry/decompress	145 ns/iter (± 5)	155 ns/iter (± 6)	0.94
ContractNonceChange/compress	146 ns/iter (± 3)	184 ns/iter (± 7)	0.79
ContractNonceChange/decompress	241 ns/iter (± 4)	261 ns/iter (± 5)	0.92
ContractClassChange/compress	216 ns/iter (± 3)	217 ns/iter (± 4)	1.00
ContractClassChange/decompress	250 ns/iter (± 4)	281 ns/iter (± 5)	0.89
ContractStorageEntry/compress	162 ns/iter (± 3)	170 ns/iter (± 2)	0.95
ContractStorageEntry/decompress	311 ns/iter (± 8)	347 ns/iter (± 6)	0.90
GenericContractInfo/compress	136 ns/iter (± 5)	139 ns/iter (± 3)	0.98
GenericContractInfo/decompress	106 ns/iter (± 9)	111 ns/iter (± 4)	0.95
Felt/compress	82 ns/iter (± 3)	91 ns/iter (± 12)	0.90
Felt/decompress	69 ns/iter (± 7)	64 ns/iter (± 10)	1.08
BlockHash/compress	81 ns/iter (± 3)	91 ns/iter (± 18)	0.89
BlockHash/decompress	58 ns/iter (± 5)	61 ns/iter (± 13)	0.95
TxHash/compress	82 ns/iter (± 3)	91 ns/iter (± 3)	0.90
TxHash/decompress	58 ns/iter (± 7)	61 ns/iter (± 6)	0.95
ClassHash/compress	82 ns/iter (± 7)	90 ns/iter (± 11)	0.91
ClassHash/decompress	61 ns/iter (± 5)	61 ns/iter (± 7)	1
CompiledClassHash/compress	81 ns/iter (± 5)	90 ns/iter (± 4)	0.90
CompiledClassHash/decompress	58 ns/iter (± 7)	61 ns/iter (± 4)	0.95
BlockNumber/compress	47 ns/iter (± 3)	51 ns/iter (± 3)	0.92
BlockNumber/decompress	25 ns/iter (± 4)	26 ns/iter (± 1)	0.96
TxNumber/compress	47 ns/iter (± 2)	51 ns/iter (± 2)	0.92
TxNumber/decompress	26 ns/iter (± 0)	26 ns/iter (± 1)	1
FinalityStatus/compress	1 ns/iter (± 0)	0 ns/iter (± 0)	+∞
FinalityStatus/decompress	12 ns/iter (± 2)	13 ns/iter (± 0)	0.92
TypedTransactionExecutionInfo/compress	16247 ns/iter (± 90)	14845 ns/iter (± 45)	1.09
TypedTransactionExecutionInfo/decompress	3612 ns/iter (± 107)	3709 ns/iter (± 70)	0.97
VersionedContractClass/compress	367 ns/iter (± 5)	362 ns/iter (± 6)	1.01
VersionedContractClass/decompress	766 ns/iter (± 20)	814 ns/iter (± 3)	0.94
MigratedCompiledClassHash/compress	159 ns/iter (± 5)	179 ns/iter (± 4)	0.89
MigratedCompiledClassHash/decompress	147 ns/iter (± 4)	162 ns/iter (± 9)	0.91
ContractInfoChangeList/compress	1515 ns/iter (± 45)	1524 ns/iter (± 78)	0.99
ContractInfoChangeList/decompress	2254 ns/iter (± 381)	2395 ns/iter (± 422)	0.94
BlockChangeList/compress	667 ns/iter (± 36)	653 ns/iter (± 101)	1.02
BlockChangeList/decompress	902 ns/iter (± 155)	980 ns/iter (± 164)	0.92
ReceiptEnvelope/compress	30908 ns/iter (± 1922)	28030 ns/iter (± 1145)	1.10
ReceiptEnvelope/decompress	6143 ns/iter (± 240)	6431 ns/iter (± 268)	0.96
TrieDatabaseValue/compress	165 ns/iter (± 2)	159 ns/iter (± 2)	1.04
TrieDatabaseValue/decompress	226 ns/iter (± 4)	242 ns/iter (± 4)	0.93
TrieHistoryEntry/compress	289 ns/iter (± 2)	292 ns/iter (± 2)	0.99
TrieHistoryEntry/decompress	262 ns/iter (± 11)	280 ns/iter (± 12)	0.94

This comment was automatically generated by workflow using github-action-benchmark.

[github-actions]

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.30.

Benchmark suite	Current: a1b558d	Previous: 3db9b65	Ratio
FinalityStatus/compress	1 ns/iter (± 0)	0 ns/iter (± 0)	+∞

This comment was automatically generated by workflow using github-action-benchmark.

CC: @kariy

[github-actions]

Runner: AMD EPYC 7763 64-Core Processor (4 cores) · 15Gi RAM

[codecov]

Codecov Report

❌ Patch coverage is 78.94737% with 8 lines in your changes missing coverage. Please review.
✅ Project coverage is 66.68%. Comparing base (9bde0ae) to head (a1b558d).
⚠️ Report is 367 commits behind head on main.

Files with missing lines	Patch %	Lines
crates/bootstrap/src/tui.rs	78.94%	8 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #534      +/-   ##
==========================================
- Coverage   73.32%   66.68%   -6.64%     
==========================================
  Files         209      308      +99     
  Lines       23132    42486   +19354     
==========================================
+ Hits        16961    28333   +11372     
- Misses       6171    14153    +7982     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.