math.scalar |
in part |
19 |
36 exact typed scalar scenarios through the production native MATH provider. |
Full numeric boundaries and all applicable failures remain pending. |
math.arrays |
in part |
4 |
Capabilities, typed evaluation, array outputs, status and alias-preserving mutation through immediate hosted memory. |
Deferred-memory, stale-handle and broader failure/cancellation evidence remain pending. |
math.random |
in part |
2 |
Exact random calls and results, repeat/advance/reseed and zero/negative seeds against existing D/OS normative values. |
Broader argument boundaries and all applicable failures remain pending. |
gfx.pixels |
in part |
13 |
Exact typed calls, buffer contents including viewport output/sentinel, 32x32 independent pixels including rectangle outlines, and publication with the native frame journal enabled and disabled. Text here covers NativeMono space cells and decorations; visible glyph rendering also has the existing paired window scenario. |
Other registered GFX rows, broader dimensions/fonts/flags, deferred memory and physical targets remain pending. |
ui.scene |
in part |
22 |
63 exact UI/GFX calls cover 22 UI rows: node ownership, geometry, Unicode text/caret/hit/selection, idempotent cleanup, unreused identities and retained canvas commit/cancel/move with five independent pixel frames through the native provider, journal on and off. Ten invalid mutation/cycle/stale-handle cases preserve the shared refusal; removing the exact CREATE route refuses before any provider submission. |
Broader widget kinds/geometry, additional malformed state, deferred memory and physical targets remain pending. These cases do not establish full UI conformance. |
ui.snapshots |
in part |
2 |
Exact authorized string and integer array contents and copied grid/table state after source mutation, including styles, authored bits and captions. The production shared UI dispatcher renders through a native GFX sink under a deterministic test routing adapter. An invalid replacement style preserves prior owned text/style atomically; node-quota and zero-byte grants enforce capacity before creating an extra node. |
The native composite port is exercised for scalar UI; these array observations use the same production UI implementation in a test adapter to inspect retained snapshots. Exact independent raster goldens for all grid/table glyphs, broader malformed snapshots/allocation failures, deferred memory/cancellation and physical targets remain pending. |
file.records |
in part |
8 |
22 paired transactions through the shared headless native FILE provider compare every request/reply byte, authorized array shape and unchanged input, plus independent filesystem snapshots after every call. Covers binary bytes, offset gaps, simultaneous readers, staged replacement, cancel/commit, generations, rename, mkdir and delete. |
Other FILE resource/store rows, broader paths/list paging/operation boundaries, deferred transport and physical targets remain pending. Immediate host filesystem checks are not power-loss durability or hard deadline evidence. |
compute.hash |
in part |
1 |
Paired Python/BASIC hashes compare complete typed calls, unchanged input and independently pinned empty/abc/padding/chunk/maximum-prefix digests through the production console provider. Pending polls preserve all caller bytes until terminal completion; input/output aliasing is checked after the provider snapshots its input. Invalid lengths/types/shapes preserve every buffer byte. |
Deferred array memory, caller cancellation and injected read/write/allocation failures remain pending for this paired suite. VECTOR_I16 is not served by the native array adapter. No physical or hard-deadline claim. |
audio.cached |
in part |
6 |
Paired Python/BASIC console programs compare complete typed calls, authorized PCM and handle buffers, independently calculated stereo samples, copied clip ownership after source mutation, attack/release envelopes, pan, volume, tone duration, stop/replay and teardown through the shared native mixer with an explicit offline capture. Invalid carriers, cache exhaustion, busy release, stale generations and removal of SOUND or any exact cached row preserve shared refusals. |
Capture is explicit test/export evidence, not a production-device substitute. Other AUDIO rows, broader sample rates/resampling/waveforms, all cache dimensions, injected allocation/memory failures, deferred array memory, device latency and physical targets remain pending. No hard-deadline claim. |
net.status_resolve |
in part |
2 |
Paired NET.QUERY_LINK and NET.RESOLVE cases compare complete typed records through the production native provider: 24 status snapshots over six endpoint classes and absent/mounted/withdrawn/recovered states; numeric IPv4/IPv6 resolution and family misses; explicit unsupported SRV/mDNS answers; invalid class, length, UTF-8 and array shapes; exact capability/route refusals; unchanged caller requests and replies while pending; and withdrawal/recovery without replay. No external DNS is contacted. |
Full NET stream/grant semantics, NET.POLL/PING/CTL paired success, arbitrary DNS servers and resolver timing, SRV/mDNS implementation, injected allocation/deferred-memory failures, complete cancellation surfaces and physical targets remain pending. The fixed host worker pool may return Busy; OS DNS cannot be forcibly canceled. No hard deadline claim. |
net.streams |
in part |
6 |
Executing paired Python/BASIC TCP scenarios through the production native provider for OPEN, SEND, RECV, CLOSE, LISTEN and ACCEPT: exact typed requests and bounded payload/status replies; real one-byte binary transfers and EOF; stale/recycled generation isolation; listener replay restricted to listener kinds; accepted children surviving listener closure; explicit unsupported HTTP and port-in-use answers; all six capability, route and pre-submit withdrawal refusals; overlong OPEN/SEND refusal with unchanged caller buffers and no transmitted prefix; a caller SEND request deliberately mutated after submission while its original byte reaches the peer; and pending ACCEPT/RECV withdrawal and recovery, including peer closure before teardown. Grant bytes are current adapter observations, not full canonical grant conformance. OS-assigned LISTEN ports are observed separately and independently verified by successful loopback peer connection. The scenarios run inside the existing sole NET aggregate before cancelled resolver cases. Central debug execution passes. |
The canonical NET receipt requires OPEN max_span, latency_class, tls_terminator and host_key_state, LISTEN max_span/latency_class, and ACCEPT peer address/latency_class, but shared net.rs has no authoritative offsets for those grant fields; they remain deferred and no private layout is added. RECV sequence validation, ACCEPT nonblocking flags, WebSocket and other protocol semantics, larger/partial payloads, handle exhaustion, deferred array memory, injected allocation failures, general cancellation and physical targets remain outside this slice. NET.POLL/PING/CTL paired success remains deferred. Only numeric/local loopback peers are used; no external DNS or network timing claim. |
ui.native_events |
in part |
1 |
Four paired native tests execute BASIC and Python with both compiler optimizer settings through NativeEventSession and the shared VM. Complete UIE2 records, bounded repeated chunk reads, immutable copies retained after callback expiry, strict lease/range failures, missing route and foreign surface credentials preserve exact results, unchanged pixels and zero guest leaks. Native timestamps are observed and independently bounded rather than predetermined. The UI census and coverage-ledger gate execute the same success scenario and derive this group's rows from its validated completed traces, requiring identical coverage in all four language/options runs. |
Physical OS/IME delivery, all event kinds and pressure interleavings, injected result-allocation failure and native presentation latency remain pending. Numeric event leases are scoped to the authenticated session; no global uniqueness or complete conformance claim. |