HHY Language Report

HHY Language Status Report · 2026-09-14

Published status of HHY semantics, Runtime, performance, and engineering quality with reproducible CI measurements.

1.7.1 controlled concurrency and atomicity

This release fixes resource lifetimes and delivers seven experimental APIs, disabled by default. Review state ownership, cooperative file locks, strict transactions and bounded-task guarantees before enabling them.

Read the complete guide and examplesIncludes the guarantee matrix, failure handling and actual validation evidence

HHY 1.7.1 implementation and compatibility

1.7.1 passes local tests, four-platform CI and release validation. Language specification 1.0 remains frozen. Official sample, HTML and macOS/Linux Database 1.0.0 extensions are verified compatible. VS Code 0.2.0 and Sublime 0.1.0 retain independent versions; syntax and editor contract checks pass.

HIR, six optimization passes, integer MIR, type-feedback guards/deopt and local List scalar replacement are implemented. Explicitly enable HHY_COMPILER=ir, HHY_FEEDBACK_SPECIALIZATION=1 and HHY_SCALAR_REPLACEMENT=1; direct Bytecode remains default. Real workloads have not met default-enablement benefit gates. Scalar replacement retains original GC/quota reservations.

Optimization switches, verification and supported scopev1.7.1 implementation 1.7.1 release notesReleased capabilities and limits

1.7.0 measured results and release evidence

One local macOS arm64 run on 2026-09-08 with HHY 1.7.0 and default compiler settings. The raw profiler output follows. Only four CPU samples were collected; short-run timings and hotspot percentages vary and do not establish a speedup.

Actual run

$hhy profile --engine bytecode examples/09-profile-algorithms.hhy -- fibonacci 20

HHY profile: examples/09-profile-algorithms.hhy

Summary
  Engine           bytecode
  Wall time        0.013 s
  CPU time         0.008 s
  CPU utilization  58.8%
  CPU samples      4
  Heap peak        1.8 MiB
  Heap after GC    92.0 KiB
  Allocated        1.7 MiB
  Allocations      33040

CPU hotspots
  CPU%    Samples      Calls  Function
   75.0%        3      21891  fibonacci  examples/09-profile-algorithms.hhy:5:1
   25.0%        1          1  print  examples/09-profile-algorithms.hhy:18:6
    0.0%        0          1  <bytecode-top-level>  examples/09-profile-algorithms.hhy:1:1
    0.0%        0          1  length  examples/09-profile-algorithms.hhy:12:10
    0.0%        0          1  to_int  examples/09-profile-algorithms.hhy:17:19
  Note: fewer than 10 CPU samples; use a larger workload for stable results.

Allocation hotspots
  Bytes          Objects  Function
  1.7 MiB            32856  fibonacci  examples/09-profile-algorithms.hhy:5:1
  11.6 KiB              184  <bytecode-top-level>  examples/09-profile-algorithms.hhy:1:1

fibonacci 6765

Platform	Integer MIR / baseline	Local List / baseline
linux-arm64	0.7624×	0.7495×
linux-x86_64	0.7067×	0.6738×
macos-arm64	0.8392×	0.8553×
windows-x86_64	0.8478×	0.8326×

View 1.7.0 CI and raw performance artifacts8fb0b8f · 2026-09-08

Release summary

Dimension	Question	Current conclusion
Language baseline	Are core semantics stable?	Pipe, Value, Stream, Error, and core callable contracts are frozen
Runtime health	Are resource, memory, and cancellation boundaries reliable?	Resource limits, GC stress, sanitizers, fuzzing, and explicit ownership governance are present
Performance	Is performance measurable and controlled?	Fixed workloads, five-sample medians, machine-readable evidence, and blocking budgets are established
Engineering governance	Are changes auditable?	Four-platform CI, layered gates, version consistency, and release evidence form a closed loop

Data at a glance

Signal	Result	Evidence basis
Current formal release	v1.7.1	Four platform archives, per-asset SHA-256, and SHA256SUMS
Execution engines	Bytecode default / AST fallback	Full dual-engine suite and machine-readable decision
Core callables	128	Runtime Callable Contract Registry
Continuous-verification platforms	4	macOS arm64, Linux arm64, Linux x86_64, Windows x86_64
v1.3.10 CI engine baseline	Historical measurement passed	1M CPU 0.3695; short task 1.0088; sustained JSON 1.0207
v1.3.10 Profiler overhead	1.0269× / +2.786 ms	Historical nine samples; limits 1.35× and 12 ms
Bytecode cache	Not admitted	Five workloads × 21 samples; compile+verify is not a major cost
Complete practical projects	6	AST/Bytecode end-to-end acceptance with stable exit status

v1.5.0 / Database 1.0.0 release acceptance

Released 2026-09-07. Dual-database/engine, TLS, scope cleanup, cancellation, CMS database lifecycle and cross-platform release gates pass. Local million-row streaming and five-minute-per-backend sustained reports are available. RDS and 24-hour soak validation remain pending.

Read DB 1.0.0 acceptanceHHY 1.5.0 · Database 1.0.0 Read HHY 1.5.0 release notesHHY 1.5.0 · Database 1.0.0

v1.4.3 Web Runtime and release verification

Check	Result	Measurement scope
Embedded Runtime	100,000 repeated calls	One loaded context with AST/Bytecode differential output checks
HTTP stress verification	1,000,000 requests · zero failures	Loopback on the release-validation host, 16 concurrent clients
HTTP duration and throughput	131.439 seconds · 7,608.1 requests/second	Host-specific short-request workload, not a public-network or business-service throughput claim
HTTP regression	End-to-end coverage	Router, Query, 404/405/413/500, recovery, middleware, static caching, upload cleanup, CORS, signed cookies, and gzip
Streaming and operations	End-to-end coverage	Stream, SSE, Range, metrics, health, multiple workers, and development reload

Read the v1.4.3 release-validation recordSite-hosted release notes with test scale, results, and capability boundaries.

Current capability boundaries

Area	Current boundary
HTTP serving	Runtime provides HTTP/1.1; Caddy, Nginx, or a load balancer handles TLS and HTTP/2
WebSocket	Outside the v1.4 capability boundary
Execution engines	Bytecode default with AST fallback; no JIT, process/disk Bytecode cache, or unverified external Bytecode loading
Native integration	C embedding uses opaque handles and a JSON ABI; it does not expose Runtime internals or a third-party Native Extension ABI

Overall baseline and compatibility

Baseline	Stable commitment	Verification
Language semantics	No second Pipe, Stream, or Error model	Specification examples, Parser/Checker fixtures, and valid-program regression
Callable contracts	Names, arity, effect, lazy, cancellable, and threading metadata are machine-readable	Contract Registry JSON and current callable-contract consistency checks
Diagnostics	CLI text and JSON/LSP share the Core checking path	Diagnostic schema and LSP protocol tests
Extension boundary	Third-party capabilities prefer the Process Extension Protocol	Manifest integrity, Protocol 1, and official-extension acceptance
C embedding	Public opaque Application/Context handles and JSON ABI; Runtime internals remain private	One-time loading, repeated calls, request isolation, and AST/Bytecode differential checks

The current formal baseline is v1.7.1. Bytecode is default; the AST evaluator remains the semantic oracle and is selectable through --engine ast or HHY_ENGINE=ast. Compiler-produced Stream Kernels must pass their independent Verifier; dynamic and unknown shapes fall back losslessly to general Bytecode.

v1.2.2 release and extension status

Capability	Current state	Acceptance result
Extension distribution	Ed25519-signed Registry and deterministic resolution	Tampering, unknown sources, and dependency conflicts fail closed
Reproducible environment	Lockfile and content-addressed offline cache	The same lock yields the same graph and rebuilds offline
Safe change	Transactional install, upgrade, and explicit rollback	Failed upgrades preserve the old environment
HTML 0.2.0	Lexbor, CSS selectors, and single-parse multi-field projection	Malformed HTML, hard limits, truncation, and structured errors pass on four platforms
Protocol decision	Retain the bounded synchronous batch API	HTML retains synchronous batches; DB 1.0 uses scoped handles, cancellation and host cursor Streams; generic Stream credit remains unavailable

Open the HHY Language v1.2.2 releaseDownload all four platform archives and checksums and read the release notes.

v1.3.7–v1.3.10 Bytecode hardening

Version	Core delivery	Verified conclusion
v1.3.7	Named specialization metadata, unified stack/error rules, fallback reasons	No magic kinds; three-path differential and metamorphic gates pass
v1.3.8	Compiler-produced versioned Stream Kernel IR	Runtime does not inspect AST shapes; independent verification; safe dynamic fallback
v1.3.9	Shared optimization decisions for normal and profiled execution	Kernel/opcode, cancellation, CPU/Heap attribution, and machine reports remain consistent
v1.3.10	Performance-triggered cache governance	Evidence did not admit a cache; no process/disk cache; unverified external Bytecode rejected

Open the HHY Language v1.7.1 releaseFour platform archives, per-asset SHA-256, SHA256SUMS, and Web Runtime release notes.

Historical performance baseline · v1.3.10

Final v1.3.10 CI evidence comes from commit 4ddc8c3 on GitHub Actions Ubuntu 24.04: schema-2 paired/interleaved engine benchmarks plus independent Profiler and cache-decision artifacts. Ratios are Bytecode/AST wall time; lower than one means Bytecode is faster.

Gate	Measured	Limit	Result
1M CPU	0.3695×	0.90×	Pass
Short task	1.0088×	1.25×	Pass
Sustained JSON/I/O	1.0207×	1.10×	Pass
Profiler overhead	1.0269× / +2.786 ms	1.35× / +12 ms	Pass

Cache-admission workload	Compile+verify median	Cold-run median	Share
Hello	0.0097 ms	5.860 ms	0.1649%
Advanced Flow	0.0269 ms	5.916 ms	0.4546%
Stdlib	0.0286 ms	7.003 ms	0.4086%
Sustained JSON	0.0089 ms	33.518 ms	0.0266%
Core Flow 1M	0.0105 ms	166.285 ms	0.0063%

v1.3.10 six-runtime same-machine rerun

Rerun on 2026-09-01 on macOS 26.6.2 arm64 with HHY 1.3.10, PHP 8.5.10, Go 1.27.0, Python 3.14.7, Lua 5.5.1, and OpenJDK 26.0.2.1. The fixed task maps, filters, stable-distincts, materializes, and counts one million integers; all six implementations validate output 333334. After two warmups, two independent rounds use seven deterministically shuffled and interleaved fresh processes per runtime. Wall time includes process startup; Go and Java are precompiled and compile time is excluded.

Implementation	Version	Round 1 median	Round 2 median	Combined range
Go	1.27.0	7.995 ms	7.969 ms	7.622–15.325 ms
Lua	5.5.1	18.368 ms	17.957 ms	17.582–19.822 ms
PHP	8.5.10	43.174 ms	43.549 ms	42.588–49.033 ms
Java	OpenJDK 26.0.2.1	49.394 ms	48.153 ms	46.995–53.066 ms
HHY Bytecode	1.3.10	55.297 ms	53.404 ms	51.027–78.702 ms
Python	3.14.7	81.747 ms	86.459 ms	79.993–87.774 ms

Comparison	Round 1	Round 2	Interpretation
HHY / PHP	1.28×	1.23×	HHY uses about 23%–28% more wall time on this task
HHY / Java	1.12×	1.11×	Java is slightly faster including fresh-JVM startup
HHY / Python	0.68×	0.62×	HHY takes less wall time on this task
HHY / Lua	3.01×	2.97×	Lua is faster on this integer loop
HHY / Go	6.92×	6.70×	Precompiled Go remains substantially faster

Governance conclusion and watch list

Overall status: v1.7.0 / DB 1.0.0 passed local, CI and cross-platform release gates; real RDS and 24-hour soak validation remain pending.
Web conclusion: embedding, HTTP, Router, middleware, uploads, Stream/SSE, multi-worker serving, and observability ship together.
Engine policy: Bytecode is default; AST remains permanently available as the semantic oracle, differential-test engine, and --engine ast emergency fallback.
Performance conclusion: one million Web short requests completed with zero failures; existing CPU, short-task, sustained-JSON, and Profiler gates continue to pass.
Cache conclusion: compile+verify is not a major cold-run cost; no cache is introduced until new data triggers review.
Update rule: synchronize this report whenever the release baseline, measurement method, engine/cache decision, or overall risk conclusion changes.

Open v1.3.11 continuous-verification evidenceFour-platform builds, sanitizers, fuzzing, performance gates, and practical-project acceptance.