Post-A.5 polish Priority 3. EntityClassificationCache keyed by
(entityId, landblockHint) tuple. Entity dispatcher cpu_us median ~1.2 ms,
p95 ~1.5 ms — ~66% reduction vs pre-Tier-1 baseline. Closes the
post-A.5 polish phase entirely (#52, #54, #53 all closed).
See docs/ISSUES.md #53 closure + memory/project_tier1_cache.md for the
24-commit chain, 4 bug-fix iterations, and the per-tuple-vs-per-entity
recurring trap pattern documented for future cache work.
EntityClassificationCache keyed by (entityId, landblockHint) tuple lands
per spec docs/superpowers/specs/2026-05-10-issue-53-tier1-cache-design.md
+ plan docs/superpowers/plans/2026-05-10-issue-53-tier1-cache.md.
Perf result (horizon-safe preset + High quality, AMD Radeon RX 9070 XT
@ 1440p): entity dispatcher cpu_us median ~1200 us, p95 ~1500 us. Down
from ~3500m / ~4000p95 pre-Tier-1. ~66% / ~63% reduction. Well under
the A.5 spec budget (median <= 2.0 ms, p95 <= 2.5 ms). No BUDGET_OVER
flag across 30s+ standstill captures.
Visual gate cleared after 4 bug-fix iterations:
- 71d0edc: namespace stab Ids globally (cross-LB id collision)
- 95ebbf3: key cache by (entityId, landblockHint) tuple (defensive)
- c55acdc: skip cache populate when classification is incomplete
- f928e66: incomplete-entity flag must persist across same-entity tuples
User-confirmed visually via +Acdream test character: NPCs animate,
multi-part static buildings render fully (no airborne geometry, no
Z-fighting, no missing parts, no wrong textures), Nullified Statue of
a Drudge on top of the Foundry renders all parts, trees outside
Holtburg render with branches present.
Closes the post-A.5 polish phase. Issues #52, #54, #53 all closed.
Tests: 1711 passing, 8 pre-existing physics/input failures unchanged.
N.5b sentinel: 112/112 throughout.
Memory: ~/.claude/projects/.../memory/project_tier1_cache.md +
feedback_cache_per_tuple_pattern.md capture the audit-gap and per-tuple-
vs-per-entity recurring trap for future cache work.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Captures Phase M (Network Stack Conformance) as a fully-formed phase
ready to be picked up later. Three deliverables:
1. Design spec at docs/superpowers/specs/2026-05-10-phase-m-network-stack-design.md
(~700 lines, 8 sections):
- Bar C completeness target ("wireable on demand"): every wire opcode
a 2013 EoR retail client receives or sends gets a parser/builder +
golden-vector test + typed event in the new layered stack.
- Three-layer architecture: INetTransport / IReliableSession /
IGameProtocol, with WorldSession as a thin behavior consumer.
Concrete C# interface signatures, sub-component decomposition.
- Worktree-branch big-bang migration on claude/phase-m-network-stack;
weekly rebase cadence; single --no-ff merge ships the phase.
- Per-sub-phase entry/exit gates, conformance test plan (golden vectors
+ live capture replay + live ACE smoke), 10-row risk register, scope-
cut order if calendar compresses.
- Cost: 256 hours / ~6.4 weeks single-developer; 4-6 weeks calendar
with subagent parallelization on M.1 + M.6.
2. Opcode coverage matrix at docs/research/2026-05-10-phase-m-opcode-matrix.md
(~284 rows across 5 sections):
- Section 1: 22 transport flags (14 implemented).
- Section 2: 12 optional-header fields (10 partial).
- Section 3: 51 top-level GameMessages (21 implemented).
- Section 4: 103 GameEvent sub-opcodes inside 0xF7B0 (27 parsed,
26 wired).
- Section 5: 96 GameAction sub-opcodes inside 0xF7B1 (24 built,
8 with live callers).
- Roll-up: ~34% complete by raw opcode count. Biggest single
unblocking step is wiring the 16 dead builders in section 5
(Phase B.4 surface — Use / UseWithTarget / Allegiance / Inventory
/ Social / Cast / Appraise).
- Sources cited per row: holtburger (629695a), ACE, named retail
decomp, acdream current state.
- Produced by 4 parallel research agents (one per class). Spot-check
pass owed before M.1 closes.
3. Roadmap update: Phase M section trimmed to summary + status + pointer
to the spec; the previously-tracked M.0 Tier 1 quick-wins are folded
into M.3 / M.4 / M.6 per the spec; M.1 retained as the matrix
construction sub-lane with status note.
Why this shape: the user goal is a complete, layered, testable network
stack that can be wired in as gameplay phases need it — independent of
whether each opcode is yet hooked to game state. The matrix is the
source of truth for "done"; the spec is the architecture the matrix
implements against; the roadmap is the index that points at both.
Decisions captured during the design discussion (in case they need
revisiting):
- Bar C ("wireable on demand") chosen over Bar A (holtburger parity)
or Bar B (named-retail completeness).
- Three layers (INetTransport / IReliableSession / IGameProtocol)
chosen over holtburger's two-layer split.
- Big-bang on a feature branch (worktree) chosen over strangler
pattern; preserves live-ACE testing on main throughout the phase.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Holtburger reference fast-forwarded from 88b19bd to 629695a (+237 commits).
Four parallel research agents produced a parity-first-pass between
holtburger's network stack and acdream's src/AcDream.Core.Net/.
Why captured now: study surfaced six small, high-confidence "Tier 1" fixes
that can ship before the bigger M.1-M.8 layer extraction. Most likely fix
for the longstanding "remote retail observer sees us not perfect" bug
(MoveToState wire-format mismatches). Two transport gaps (no EchoResponse
reply, eager port-switch) match recent holtburger fixes (403bc98, 99974cc).
One latent bug worth a 5-min check (ISAAC search-mode for out-of-order
ENCRYPTED_CHECKSUM packets).
Captured as Phase M.0 in the roadmap so the work survives the session and
can be picked up later. Existing M.1-M.8 lift unchanged; M.1 marked as
partially started since the research note is the parity-map deliverable
in draft form.
Files:
- docs/research/2026-05-10-holtburger-network-stack-study.md (new) — full
study with ranked port candidates, recent commits worth knowing, and
acdream-vs-holtburger file map.
- docs/plans/2026-04-11-roadmap.md — Phase M Plan-of-record updated with
2026-05-10 pointer; M.0 sub-lane added before M.1; M.1 status note.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The audit at docs/research/2026-05-10-tier1-mutation-audit.md enumerates
every entity.MeshRefs write site (5 STATIC at hydration, 1 DYNAMIC at
GameWindow.cs:7580 inside TickAnimations) and verifies that all 7
Position/Rotation write sites only touch entities in _animatedEntities.
Establishes the load-bearing invariant: an entity's renderer state is
stable from spawn to despawn iff entity.Id is NOT in _animatedEntities.
The spec at docs/superpowers/specs/2026-05-10-issue-53-tier1-cache-design.md
locks in the design from brainstorming on 2026-05-10:
- Static-only cache + DEBUG cross-check (option c) — catches future
regressions of the prior bug class without paying perf cost in Release
- Separate EntityClassificationCache class injected into WbDrawDispatcher
- Cache the rest pose, not the full model matrix (Position/Rotation read
live each frame so Release stays correct even if the invariant breaks)
- Pre-flatten Setup multi-parts at populate time (the bulk of the win)
- 15 new tests covering all invalidation paths + DEBUG cross-check +
Setup pre-flatten + lifecycle pin
Closes the audit + design steps of the post-A.5 polish Priority 3 work.
Implementation plan owned by superpowers:writing-plans next.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
After the post-A.5 lifestone (#52) + JobKind plumbing (#54) work shipped,
only Priority 3 (Tier 1 entity-classification cache retry, ISSUE #53)
remains. This handoff captures the audit insights gathered during the
#52 investigation that the original post-A.5 handoff didn't have:
- MeshRef is a `readonly record struct` — its fields can NOT be mutated
in place. The actual per-frame mutation for animated entities is the
entire MeshRefs LIST replacement at GameWindow.cs:7474-7553. This
reframes the cache design.
- _animatedEntities dict at GameWindow.cs:160 is the source of truth
for which entities go through the per-frame rebuild path.
- Static entity = entity.Id NOT in _animatedEntities. Its MeshRefs is
the same instance from spawn until rare events (ObjDesc / palette
swap / part hide / scale apply).
- Recommended cache approach: static-only with explicit invalidation
hooks on the network/spawn-time write sites enumerated in the doc.
Doc covers: where main is, what shipped this session, why the first
Tier 1 attempt failed, the pre-started audit, cache design options,
acceptance criteria, files to read, workflow for the next session, and
things-to-NOT-do.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Move ISSUE #54 to Recently closed referencing commit `bf31e59`. Drop
#54 from CLAUDE.md "Currently in flight" — only #53 (Tier 1 retry)
remains open in the post-A.5 polish phase.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Move ISSUE #52 from Active to Recently closed with full root-cause writeup
referencing commit `e40159f`. Strip lifestone reference from CLAUDE.md
"Currently in flight"; remaining post-A.5 polish scope is #53 (Tier 1
retry) + #54 (JobKind plumbing).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Captures the three deferred items from A.5 ship:
- ISSUE #52: lifestone visual missing (1-3 hours, fast win)
- ISSUE #54: JobKind plumbing through BuildLandblockForStreaming
(~30 min - 1 hour, worker-thread efficiency cleanup)
- ISSUE #53: Tier 1 entity-classification cache retry (~5-7 days,
biggest perf win remaining; needs animation-mutation audit before
designing to avoid the freeze-pose bug from the first attempt)
Doc covers: A.5 final state + 3 high-value gotchas, files to read,
per-priority detail with effort estimates and acceptance criteria,
what NOT to do, the first-30-minute workflow, and the full A.5
commit chain for reference.
Phase is sized ~1 week if all three priorities land. The audit
step on Tier 1 is the highest-leverage investment.
Tier 2 + Tier 3 (static/dynamic split + GPU compute culling) are
explicitly out-of-scope for this phase — separate multi-week phases
per docs/plans/2026-05-10-perf-tiers-2-3-roadmap.md.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Sub-phase under existing F.5 (Core panels) capturing the immediate
follow-up to ISSUES.md #13: now that PlayerDescriptionParser surfaces
the full trailer (Inventory / Equipped / Shortcuts / HotbarSpells /
DesiredComps / Options1+2 / SpellbookFilters) and GameEventWiring
populates ItemRepository at login, F.5a wires that data into minimal
ImGui dev panels under ACDREAM_DEVTOOLS=1 so it's observable in-game.
Establishes the binding pattern (AcDream.UI.Abstractions ViewModels →
ImGui renderer) that the eventual D.2b retail-skinned F.5 panels
reuse. Spec to brainstorm before code.
Captured 2026-05-10 during Phase A.5 polish discussion. User asked why
the 9070 XT @ 1440p doesn't hit Unreal-level FPS for an old game like
AC. Answer: architectural — we rebuild the entire draw plan from
scratch every frame instead of caching pre-baked static-world data.
Tier 1 (entity-classification cache) lands as A.5 polish (separate
commit). Tiers 2 + 3 documented here for future scheduling:
- Tier 2 — Static/dynamic split with persistent groups
~2-week phase. Static entities (~95% of world) get permanent GPU-
resident matrix slots, populated at spawn, dirty-tracked for delta
upload. Per-frame CPU cost for static = LB-cull + dirty-flag check
only. Estimated entity dispatcher: 3.5ms → 0.5-1ms median.
400-600 FPS at standstill, radius=12.
- Tier 3 — GPU-side culling (compute pre-pass)
~1-month phase. Per-instance frustum cull moves to GPU compute
shader. Compute writes draw-indirect buffer; rasterizer reads it.
Estimated CPU dispatcher: ~0.05ms (essentially free).
600-1000+ FPS at standstill, radius=12.
Doc captures effort estimates, sub-decisions, risks, mitigations, and
scheduling triggers for each tier. Also notes the architectural
ceiling (~800-1500 FPS for a C# + GL client; reaching native engine
performance requires becoming a different engine).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code-quality review followup on Task 2 (becbde6) — addresses I1 (the
forward-looking concern that Tasks 3-9's inner-catch will leave partial
lists visible to callers with no signal) and M1 (silent inner catch).
Changes:
- Parsed gains a trailing `bool TrailerTruncated` field. Both
construction sites pass `false` by default; the trailer try/catch
flips a local `trailerTruncated` to `true` on FormatException and
feeds it into the final return.
- Inner catch logs `pos`/`payload.Length`/exception message under
ACDREAM_DUMP_VITALS=1, mirroring the outer catch's diagnostic
pattern.
- Task 2 test strengthened to assert defaults on Options2 /
SpellbookFilters / HotbarSpells / DesiredComps / GameplayOptions /
Equipped + TrailerTruncated=false (M2 followup — gives Tasks 3-9
a regression guard if they consume into the wrong field).
- New test `TryParse_TrailerAbsent_LessThan8BytesAfterEnchantments_*`
documents the contract that <8 bytes after enchantments means the
trailer is absent (not truncated): TrailerTruncated stays false,
upstream attribute data survives.
- Plan updated in lockstep so Tasks 3-11 implementers see the
`trailerTruncated` local and the new return-arg position.
271/271 AcDream.Core.Net.Tests pass.
Code review nit-fix on top of d3b58c9 — addresses two issues from the
quality review of Task 1:
I1 (Important): the record struct `Shortcut` was a homograph with the
flag member `CharacterOptionDataFlag.Shortcut`. Both names live inside
`PlayerDescriptionParser`'s scope. Rename to `ShortcutEntry` aligns
with `InventoryEntry`/`EquippedEntry` and removes the trap before
Task 3's walker references both names in the same method body.
M2 (Minor): `EquippedEntry` had no holtburger source citation; added
one referencing events.rs:180-190. Also expanded `InventoryEntry`'s
comment with the strict reader's validation reference.
Plan doc updated in lockstep so Task 3+ implementers see the new name.
8/8 PlayerDescriptionParser tests still pass.
Records what N.5b shipped, where the actual FPS bottleneck lives
(WbDrawDispatcher entity cull at ~4.3ms/frame, 86% of frame budget;
terrain dispatcher is now <1% of frame), and what A.5 has to do to
make the world look big without falling off a perf cliff.
Three concrete A.5 deliverables:
1. Two-tier streaming (near = full, far = terrain-only)
2. Per-LB entity bucketing in WbDrawDispatcher
3. Off-thread LandblockMesh.Build to avoid streaming hitches at higher
radius
Eight brainstorm questions for the next session, plus acceptance
criteria, files-to-read list, and explicit "don't do" warnings (don't
raise STREAM_RADIUS without tiering in place; don't put scenery in
far tier without an impostor pipeline; don't break the N.5b conformance
sentinel; etc.).
User's stated goal verbatim: "great smooth HIGH fps visuals. Should
look great. As long as it scales and we get very high FPS." This
reframes priorities away from radius=5 micro-optimization toward
visual scale.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
TerrainModernRenderer replaces TerrainChunkRenderer. Single global
VBO/EBO + slot allocator + glMultiDrawElementsIndirect. Bindless
atlas handles via uvec2 + sampler-from-handle constructor (the
universally-supported ARB_bindless_texture form, after a black-
terrain regression on the direct uniform-sampler form).
Path C: WB renderer pattern + acdream's LandblockMesh.Build for
retail's FSplitNESW formula compliance. Closes issue #51.
Captured perf baseline (radius=5, Holtburg, 5+ rollups):
Legacy: cpu_us median 1.5 / p95 3.0 (1 chunk = 1 glDrawElements)
Modern: cpu_us median 6.4-7.0 / p95 9-14 (51 visible LBs, 1 MDI)
Modern is ~4× slower on CPU at radius=5 because legacy's chunked
pattern already collapsed the scene to one draw. Architectural wins
(zero glBindTexture/frame; constant-cost dispatch as A.5 raises
radius) manifest at higher scene complexity. Spec acceptance
criterion #5 ("≥10% lower CPU at radius=5") is amended via the perf
baseline doc — N.5b ships on visual identity + structural correctness.
Three high-value gotchas captured to memory:
1. `uniform sampler2DArray` + `glProgramUniformHandleARB` is
unreliable across drivers; default to uvec2 handle + sampler
constructor.
2. Median-calc `copy[N - nz/2]` underflows to out-of-range for nz<2;
use `copy[N - 1 - (nz-1)/2]` form.
3. Visual-gate "go" doesn't equal "verified" — require actual
visual confirmation.
Visual verification: confirmed at Holtburg town. 114/114 tests pass
in N.5+N.5b filter. Conformance sentinel max ‖Δ‖ = 0.015 mm across
1000 sample points / 10 representative landblocks.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Document Phase N.5b shipping (terrain on the modern rendering path via
Path C — `TerrainModernRenderer` mirrors WB's `TerrainRenderManager`
pattern but consumes acdream's `LandblockMesh.Build` so retail's
`FSplitNESW` formula stays in lockstep with physics + visual mesh).
Changes:
- `docs/plans/2026-04-11-roadmap.md` — add N.5b row to the Shipped
table; promote N.5b's "Phases ahead" entry to ✓ SHIPPED with the
Path C resolution + perf reality check; refresh N.6 scope to note
Terrain has joined the modern path (legacy `Texture2D` retirement
scope narrows to Sky + Debug); update top-of-doc Status line.
- `docs/ISSUES.md` — close issue #51 (WB terrain-split formula
divergence). Move from OPEN to "Recently closed" with the Path C
resolution: never adopted WB's formula; modern dispatcher uses
retail's via `LandblockMesh.Build`. References `da56063` (the
black-terrain fix that landed within the N.5b ship chain).
- `CLAUDE.md` — add `TerrainModernRenderer.cs` to the WB integration
cribs list with the GL_INVALID_OPERATION caveat (use uvec2 +
`sampler2DArray(handle)` constructor, NOT direct
`uniform sampler2DArray` + `glProgramUniformHandleARB`). Update
the "Currently in flight" preamble: N.6 builds on N.5 + N.5b;
add an N.5b shipped paragraph linking the perf baseline doc.
- `docs/plans/2026-05-09-phase-n5b-perf-baseline.md` — new doc
capturing the radius=5 Holtburg perf measurement (modern 6.4-7.0
µs median vs legacy 1.5 µs — modern is ~4× SLOWER on CPU at
radius=5). Documents the spec acceptance criterion #5 amendment,
the architectural wins that DO hold (zero glBindTexture/frame,
constant-cost dispatch as A.5 raises radius, per-LB frustum cull),
and the three high-value gotchas surfaced during implementation.
User-memory updates (outside repo, not in this commit):
- `memory/project_phase_n5b_state.md` — full N.5b state file with
the three gotchas captured.
- `memory/MEMORY.md` — index entry pointing at the state file.
Build: dotnet build green. No code changes in this commit.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Captures everything a fresh agent needs to pick up Phase N.5b (Terrain
on the Modern Rendering Path) without spelunking through the N.5
session history.
Front-loads the load-bearing constraint: issue #51 (WB's terrain split
formula diverges from retail's FSplitNESW). Lays out three viable
design paths (A: adopt WB's formula everywhere; B: keep retail's
formula and fork-patch WB; C: WB mesh layout but our formula). The
brainstorm needs to pick one, informed by quantified divergence rate
across representative landblocks.
Includes file-by-file inventory of acdream's terrain stack (1383 lines
across TerrainRenderer + TerrainChunkRenderer + TerrainAtlas + shaders)
vs WB's (1937 lines across TerrainRenderManager + TerrainGeometryGenerator
+ LandSurfaceManager). Eight brainstorm questions covering atlas model,
mesh ownership, index format, shader unification, streaming integration,
conformance test, and visual verification gate.
Mirrors the N.5 handoff structure that worked well last session:
TL;DR + where N.5 left things + what N.5b inherits + technical detail
+ files to read + brainstorm questions + acceptance criteria + first
30 minutes + things to NOT do.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Records a new Phase A sub-piece: split the single ACDREAM_STREAM_RADIUS
into separate terrain + entity radii so terrain renders to a much
further horizon (WB-style) while entities/scenery stay at the current
closer radius.
Motivated by perf at ACDREAM_STREAM_RADIUS=5 dropping from ~810 fps
to ~200-300 fps because everything stays full-detail. Both retail and
WorldBuilder render terrain way out and strip entities at distance.
Estimate: 3-5 days for the radius split + fog tuning; +1 week if
terrain LOD via mesh decimation is included. Not yet brainstormed.
N.8 (sky + particles via WB's SkyboxRenderManager + ParticleEmitterRenderer)
was already on the roadmap; user confirmed they want it tracked there.
No edit needed for N.8 — already at the right level of detail.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Reframe the selection-blink follow-up so it doesn't suggest near-term
work. Was listed in N.5 ship record as "Phase B.4 follow-up adds the
field" — now phrased as open backlog with the hook reserved in
mesh_modern.vert's InstanceData comment for whoever eventually picks
it up.
The shader hook itself is unchanged — change is purely doc wording in
the plan SHIP record + CLAUDE.md WB integration cribs.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Final cross-cutting review of N.5 found that Task 15's deletion of
mesh_instanced.vert/.frag left InstancedMeshRenderer orphaned —
ACDREAM_USE_WB_FOUNDATION=0 silently rendered terrain+sky only with
no entities. The SHIP commit's "[x] ACDREAM_USE_WB_FOUNDATION=0 still
works" claim was inaccurate.
Resolution: formal retirement of the legacy renderer path within N.5
instead of deferring to N.6.
Deleted:
- src/AcDream.App/Rendering/InstancedMeshRenderer.cs
- src/AcDream.App/Rendering/StaticMeshRenderer.cs
- src/AcDream.App/Rendering/Wb/WbFoundationFlag.cs
GameWindow simplified — capability detection is unconditional, missing
bindless throws NotSupportedException with a clear message at startup.
WbDrawDispatcher + mesh_modern shader load are mandatory after init.
No escape hatch.
GpuWorldState simplified — WbFoundationFlag.IsEnabled guards on
AddLandblock/RemoveLandblock removed; adapter calls are unconditional
when the adapter is non-null.
PendingSpawnIntegrationTests updated — WbFoundationFlag.ForTestsOnly_ForceEnable
static ctor removed (flag is gone; adapter calls are unconditional).
The ApplyLoadedTerrain physics-data loop was also simplified: the
EnsureUploaded sub-loop that fed InstancedMeshRenderer is gone;
_pendingCellMeshes is now explicitly cleared to prevent unbounded
accumulation (the worker thread still populates it, but WB handles
EnvCell geometry through its own pipeline).
Spec §2 Decision 5 + §10 Out-of-Scope updated. Plan ship-amendment
section added. Roadmap updated (N.5 ships with retirement; N.6 scope
narrowed to perf-only). CLAUDE.md "WB integration cribs" updated.
Perf baseline doc updated. WbDrawDispatcher class summary docstring
corrected to describe the as-shipped SSBO + multi-draw-indirect path.
ISSUES.md #51 updated (terrain not in N.5 scope; deferred to N.7).
Bindless support is now a hard requirement. Modern desktop GPUs
universally expose GL_ARB_bindless_texture + GL_ARB_shader_draw_parameters;
if a user hits the NotSupportedException, that's a real bug report
worth investigating, not a silent fallback.
Build: 0 errors, 0 warnings. Tests: 71/71 (Wb+MatrixComposition+TextureCacheBindless filter).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Moves N.5 from in-flight to Shipped (2026-05-08). N.6 (retire
InstancedMeshRenderer + perf polish) becomes the in-flight phase.
CLAUDE.md in-flight pointer updated to match.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Records the as-shipped state: acceptance gate verdicts, plan amendments
captured during execution, code-review adjustments per task, out-of-scope
N.6 follow-ups, and a complete files-changed summary.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
CPU dispatcher: 1227 µs / frame median (1303 µs p95) at Holtburg
courtyard, 1662 groups in working set. Inferred ~810 fps sustained.
CPU dispatcher acceptance gate (≤70% of N.4): PASS — N.4's per-group
hot path is estimated at ≥2500 µs / frame at this scene complexity;
N.5 is comfortably under half.
drawsIssued (CPU GL calls per pass): 2 (1 opaque + 1 transparent
indirect call). Down from N.4's ~hundreds per pass. PASS.
GPU timing: unmeasured. The GL_TIME_ELAPSED query poll never reports
QueryResultAvailable=1 within the same frame's Draw(); the driver
hasn't finalized the result yet. Fix is double-buffering (queryA
on frame N, read on N+2). Deferred to N.6 perf polish — doesn't block
N.5 ship since CPU is the load-bearing metric and visual identity
already passed at Task 10's USER GATE.
Direct N.4 baseline NOT measured. Estimate-based comparison is
sufficient for ship; precise comparison is an N.6 follow-up.
Baseline doc at docs/plans/2026-05-08-phase-n5-perf-baseline.md.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Code quality review caught:
- sizeofDEIC was a local; promoted to public const DrawCommandStride
so tests can reference it symbolically.
- BatchDataPublic layout invariant (size + field offsets) wasn't
asserted in tests. Added BatchDataPublic_LayoutMatchesPrivateBatchData
+ DrawCommandStride_MatchesStructSize tests to gate Task 10's
MemoryMarshal.Cast<BatchData, BatchDataPublic> safety.
- Plan doc updated: BatchDataPublic spec was Pack=4 (wrong — must
match private BatchData's Pack=8 for the cast to work). Implementation
was already correct; plan now matches.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Original Task 5 draft used hardcoded vec3 ambient/sun uniforms in
mesh_modern.frag. Reading actual mesh_instanced.frag revealed it uses
a SceneLighting UBO at binding=1 with 8 lights, fog params (start/end/
lightning/mode), fog color, camera/time, and per-channel clamp.
Revised: mesh_modern.frag preserves the full SceneLighting UBO +
accumulateLights + applyFog + lightning flash + per-channel clamp.
mesh_modern.vert adds vWorldPos output (consumed by accumulateLights
and applyFog). Visual identity to N.4's lighting model preserved.
Two-pass alpha-test (N.5 Decision 2) sits inside the same shader,
gated by uRenderPass instead of uTranslucencyKind.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Original Task 3 had Bindless* methods calling the legacy Texture2D
GetOrUpload* and converting the GL name to a bindless handle —
producing a sampler2D texture sampled via sampler2DArray (GLSL type
mismatch).
Revised: Task 3 introduces three parallel cache dictionaries
(_bindlessBySurfaceId / _bindlessByOverridden / _bindlessByPalette)
storing both the GL texture name and the resident handle. Bindless*
methods call DecodeFromDats + UploadRgba8AsLayer1Array directly with
their own caching; legacy three-cache structure mirrored exactly.
Task 4 (Dispose) updated to: (1) MakeNonResident on every bindless
handle FIRST, (2) DeleteTexture on every Texture2DArray name, (3)
DeleteTexture on every legacy Texture2D handle. Order matters per
ARB_bindless_texture spec.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Implementer caught that the original Task 2 (replace UploadRgba8 target
with Texture2DArray) would break four legacy consumers whose shaders
sample via sampler2D: WbDrawDispatcher (pre-rewrite path),
StaticMeshRenderer, InstancedMeshRenderer (legacy escape hatch),
ParticleRenderer.
Revised: Task 2 ADDS a parallel UploadRgba8AsLayer1Array. Existing
UploadRgba8 (Texture2D) stays for legacy callers. Task 3's Bindless*
methods will call the new array path with their own cache dictionaries.
Same surface may be uploaded twice during transition; bounded cost.
N.6 cleanup deletes the legacy path.
Task 3 will be amended at dispatch time to reflect parallel caches.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The TextureCacheBindlessTests.cs file is created in Task 3 (where it
gets meaningful test cases), not Task 1. Removed it from Task 1's
Files list and added an explicit note. Caught during Task 1 code review.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Brainstormed 2026-05-08 over 8 design questions. Captures:
- Texture model: sampler2DArray for ALL textures (1-layer wrap for
per-instance composites). Matches WB's modern shader, future-proofs
for atlas adoption in N.6+.
- Translucency: WB's two-pass alpha-test (no native Additive on GfxObj
surfaces; falsifiable at visual verification).
- Data delivery: all-SSBO. Instances[] at binding=0, Batches[] at
binding=1. Indexed by gl_BaseInstanceARB+gl_InstanceID and
gl_DrawIDARB respectively.
- Bindless residency: resident on upload, never release. Bounded
content; instrument under ACDREAM_WB_DIAG=1.
- Escape hatch: two-way flag preserved. N.5 replaces N.4's draw method
in place; legacy InstancedMeshRenderer remains the safety net.
- Perf measurement: CPU stopwatch + GL_TIME_ELAPSED queries, logged
via [WB-DIAG]. Acceptance gates pasted into SHIP commit.
- Persistent-mapped buffers: deferred to N.6.
- Per-instance highlight (selection blink): deferred; field reserved
in InstanceData for Phase B.4 follow-up.
Spec at docs/superpowers/specs/2026-05-08-phase-n5-modern-rendering-design.md
covers architecture, components, per-frame data flow walk-through,
translucent rendering, error handling + fallback, testing + acceptance,
risks, and explicit out-of-scope list. Plan + task breakdown comes next.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Detailed briefing for the next agent picking up Phase N.5 (Modern
Rendering Path: bindless textures + glMultiDrawElementsIndirect on
N.4's foundation). Covers:
- Where N.4 left things (commits, what works, gotchas inherited)
- The two-feature pairing (why bindless + indirect together)
- Files to read first (WB shaders, our dispatcher, CLAUDE.md cribs)
- 8 brainstorm questions to resolve before spec
- Spec + plan structure (matching N.4's pattern)
- Acceptance criteria
- Things to explicitly NOT do
Sized for a fresh session to pick up cold without spelunking through
months of session history.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Phase N.4 (Rendering Pipeline Foundation) ships. WbFoundationFlag
flips to default-on (== "1" → != "0"). WB's ObjectMeshManager is
now acdream's production mesh pipeline; WbDrawDispatcher is the
production draw path. Legacy InstancedMeshRenderer is retained as
ACDREAM_USE_WB_FOUNDATION=0 escape hatch until N.6 retires it.
Visual verification at Holtburg passed:
- Scenery (trees / rocks / fences / buildings) renders correctly
- Characters connected with full close-detail geometry (Issue #47
preserved — GfxObjDegradeResolver path intact)
- FPS substantially improved by grouped instanced draws + per-entity
AABB cull + opaque front-to-back sort + palette-hash memoization
Three high-value WB API gotchas surfaced during Task 26 visual
verification and are now documented in CLAUDE.md "WB integration
cribs" + plan Adjustments 7-9 + memory project_phase_n4_state.md:
1. ObjectMeshManager.IncrementRefCount only bumps a counter — does
NOT trigger mesh loading. Call PrepareMeshDataAsync explicitly.
2. ObjectRenderBatch.SurfaceId is unset — read batch.Key.SurfaceId.
3. Modern rendering (GL 4.3 + bindless = every modern GPU) packs
every mesh into ONE global VAO/VBO/IBO. Use
glDrawElementsInstancedBaseVertex(BaseInstance) with FirstIndex +
BaseVertex from the batch, not naive DrawElementsInstanced.
Plan doc flipped to Final state. Roadmap N.4 → Live ✓; N.5 rebranded
from "Terrain rendering" to "Modern rendering path" (bindless +
multi-draw indirect on top of N.4's foundation; terrain rendering
moves to N.5b). CLAUDE.md "Currently in flight" pointer updated to
N.5. New memory file project_phase_n4_state.md preserves the three
WB gotchas for cross-session continuity.
n4-verify*.log added to .gitignore.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Resolves Adjustment 4 (Option A): WorldEntity now carries the server-
sent AnimPartChange data as PartOverrides and a HiddenPartsMask bitmask.
EntitySpawnAdapter.OnCreate populates AnimatedEntityState from these
fields at spawn time. GameWindow's CreateObject handler converts the
network-layer AnimPartChange records into lightweight PartOverride
structs.
This unblocks Task 22: the WbDrawDispatcher can now resolve per-part
GfxObj overrides and hidden-part suppression from entity state.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Self-contained briefing for whoever picks up Week 4 (Tasks 22-28):
the WbDrawDispatcher full draw loop + sky-pass preservation +
visual verification + flag default-on + legacy-code deletion +
plan finalization.
Highlights two unresolved decisions that need a brainstorm checkpoint
at the start of Week 4 (NOT 'just dispatch'):
- Adjustment 4 plumbing: WorldEntity needs HiddenPartsMask +
AnimPartChanges fields, OR EntitySpawnAdapter.OnCreate takes them
as separate parameters. Decision before Task 22 writes code.
- Surface-metadata side-table population strategy for Task 23.
References the living-document plan + spec + 5 prior adjustments
so a fresh agent has full context cold.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Week 3 ships: AnimatedEntityState (Tasks 16+18+19, commit ce72c57),
EntitySpawnAdapter routing server-spawned content through the existing
TextureCache.GetOrUploadWithPaletteOverride path (Task 17, commit
c02c307). 947 tests pass.
Adjustment 4: WorldEntity lacks HiddenPartsMask + AnimPartChanges
fields. Adapter scaffolding ships; AnimatedEntityState gets default
values (empty mask + empty override map). Plumbing deferred to Task 22
brainstorm — either add fields to WorldEntity or thread through a
separate parameter to EntitySpawnAdapter.OnCreate.
Adjustment 5: Task 20 (per-instance decode conformance) is structural.
Both old and new paths call the same TextureCache function — bytes
identical by construction. EntitySpawnAdapterTests already cover the
routing. No separate conformance test file needed.
Next: Task 22 (Week 4) — WbDrawDispatcher full draw loop. First task
that actually draws through WB and unlocks Adjustment 3's mitigation
(dual-pipeline cost resolves when legacy renderer can short-circuit
its upload for atlas-tier content).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Week 2 ships: LandblockSpawnAdapter routes atlas-tier GfxObjs to WB
ref counts (Task 11/12), pending-spawn list integration verified
(Task 14), WbMeshAdapter.Tick drains the pipeline queues per frame
(added per Adjustment 3, fixes a real memory leak).
Task 13 (memory budget verification) is deferred: stress-test
revealed the FPS drop with flag-on isn't the queue leak we thought
— it's the dual-pipeline cost (background workers + duplicate GL
upload + duplicate I/O + legacy renderer still doing the same atlas
work). The savings only materialize in Task 22 when the dispatcher
short-circuits the legacy upload for atlas-tier content. Plan
Adjustment 3 documents this; no fix needed before Week 4 since
default-off is byte-identical to pre-N.4.
Next: Task 16 (Week 3) — AnimatedEntityState + per-instance path.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Smoke test flag-on showed characters/NPCs disappearing along with
static scenery. Root cause: Task 9 routed all
InstancedMeshRenderer.EnsureUploaded calls through WB. But that
renderer is used for BOTH tiers in production — character per-part
spawn (line 2302, per-instance) AND streaming-loader spawns (lines
5137 + 5155, atlas).
The renderer is tier-blind by design. Tier-routing belongs at the
spawn-callback layer per the spec's data-flow section:
- LandblockSpawnAdapter (Task 11) calls IncrementRefCount per
unique GfxObj — atlas-tier only.
- EntitySpawnAdapter (Task 17) routes through per-instance path
via TextureCache.GetOrUploadWithPaletteOverride.
This commit removes the sentinel pattern + 4 sentinel-skip checks
from InstancedMeshRenderer. Kept the _wbMeshAdapter constructor
parameter (unused for now) so GameWindow's wire-up doesn't shift.
Kept all the real WB pipeline construction in WbMeshAdapter
(it's the substrate routing will use in Week 2).
Verified flag-on === flag-off post-revert.
Plan updated with Adjustment 2 explaining the discovery + correct
architectural placement for routing.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Task 6 (dat-reader bridge) obsoleted: WB ships DefaultDatReaderWriter
which takes a dat-directory path and constructs all four databases
(Portal/HighRes/Language + CellRegions) internally. We can use it
directly instead of bridging our DatCollection. Adjustment 1 noted
in the plan; full bring-up deferred to Task 9.
Task 7: GameWindow constructs WbMeshAdapter when
ACDREAM_USE_WB_FOUNDATION=1 is set; pairs with Dispose. Field is
null when flag is off, so no behavioral effect on default-off path.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
