The composer hairline moves one step up from the menus' l1; the seven
menu-fill dropdown cards grow from 16px to 20px corners. Notes and the
token comment record the new layering.
Re-declare the derived elevation tokens on body * so per-surface
--dsw-elevation-stroke-color rebinds reach the consuming shadow (custom
properties inherit with var() already substituted); pin that mechanism
and add synthetic rejection cases to the stylesheet scans; take the
ring-track basenames through node:path so the exemption matches on
Windows; update the ModelsSection row-card spec to the hairline recipe;
align the elevation note with the shipped l1 menu rebind, refresh the
feedback-popover note's surface recipe, and document the soft tier.
Apply global visual polish across the web client: corner-shape:
superellipse(1.5) with corner-shape: round pairing for full circles,
elevation tokens that draw 0.5px stroke outlines inside box-shadow for
floating surfaces, 0.5px hairline borders and divider lines for
neutral-token strokes, and tuned stroke contrast plus larger radii for
menus, settings panels, and cards. Stylesheet-scan specs in ui-theme
reject unpaired circles, border+shadow mixes, and 1px neutral hairlines
repo-wide.
Closes#3287
Review follow-up: the per-step gate recomputed view(previous) on every
changed apply — a property read for identity-stable views, but a fresh
throwaway object per change for computing views. The registry now keeps a
WeakMap from state object to raw view: the previous state's view was
cached when that state was current, so each distinct state's view computes
exactly once (gate and snapshot share the memo) and the quiet path
allocates nothing. Unlike the earlier lastView record, an entry is keyed
by the state itself — the view of that exact state by the pure-view
contract — so no stamping discipline exists to get wrong. Primitive
states bypass the WeakMap and compute directly.
Review consensus: for the only declaring unit the token function was
literally the view function, so the extra wire member bought nothing —
default raw-view comparison produces the identical Object.is on
state.turns at the identical cost (one property read per side). The
per-step gate stays: it holds the no-stored-baseline property, and its
measured overhead (~ns per changed state) is four orders of magnitude
below the push path it guards.
Review follow-up (imccyu): comparing view(previous) recomputes the view on
every quiet change. The wire block now takes an optional viewKey(state)
declaring the cheap comparison token; the drive compares tokens across the
previous and next states and calls view only for an actual push. Default
stays the raw view output. turnOutline declares viewKey: state => state.turns,
making the identity-stable-turns convention an explicit contract; the
registration erasure forwards viewKey (dropping it silently reverted the
gate to the fallback, caught by the new view-call-counting test).
Review suggestion (imccyu): the drive holds both the previous and next
state, so the identity gate can compute view(previous) and view(next) in
the driving step and compare them directly. The stored lastView cell field
and its stamp-on-every-change rule are deleted; with no dedup memory,
nothing can go stale across listener generations by construction, and a
rebuilt cell no longer pushes an unchanged view on its first live event.
Costs one extra pure view() call per changed state.
The change feed stamped lastView only when a listener was subscribed, so a
value change during a listener-free window (HMR swap) froze the baseline
and a later transition back to the old value was silently deduplicated.
The baseline now advances on every changed state, heard or not; broadcast
still only happens with listeners. Docs, catalog, and the feature note
follow the corrected semantics.
Outline entries gain the turn's final text-bearing assistant preview:
each assistant message overwrites a state draft and turn/end commits
the survivor, matching the loaded rail's findLast semantic; the bare-
array wire keeps its identity across draft changes, so pushes stay at
three per turn. Preview budgets shrink to the rail card's clamps — one
50-character prompt line, up to three 120-character response lines,
ellipsis on clip — on loaded and unloaded turns alike (stateVersion 2
discards v1 cache rows).
Clicking an unloaded rail mark from the pinned tail raced the pinned
scroll snap: the first prepend's compensation fires a non-reader scroll
delivery, the snap called toBottom, and toBottom cancels a pending jump
— so the jump silently stayed at the tail while history loaded. The
click now drops atBottom itself (jumping into history is leaving the
live tail), pinned by a jsdom regression and re-verified live: a
118-turn session lands on turn 1 in ~250ms from click.
READMEs record the rail's outline merge and the loadThrough paging
verb in both languages; the feature Agent Note owns the decision,
alternatives (sparse windows, minSeq wire bound, outline RPC, height
estimation), and coverage map. Regenerates the client/API catalogs the
widened faces feed.
render() loaded the workspace manifests once per external dependency
name through workspaceLinkedManifest, an O(names x manifests) file
read on the cold path; on the loaded self-hosted Windows host with
coverage instrumentation the freshness spec crossed Vitest's default
5000ms budget and failed the serial-windows standby gate four times in
a week. Load the manifests once in render() and thread the map through
the collectors instead.
The serial-windows lane also ran the coverage inventory at the strictest
budget of any lane: add DSH_COVERAGE_TEST_TIMEOUT_MS=90000 to match the
pull-request windows-coverage lane, pinned by ci-workflow.spec.ts.
Review findings on the CPython backend: a child that never reads fd 3 leaves
the reply pipe full forever, so the drain loop waits on 'drain' while every
call frame it keeps sending resolves a binding and queues another reply —
the backlog (and the binding results it pins) would grow until the wall
clock. sendReply now caps the pending backlog at MAX_PENDING_REPLIES and
settles the run as worker-exit past it, mirroring the frame cap; a child
flooding calls against a binding that never settles would otherwise bypass
that cap (pendingReplies grows only after the await), so the dispatcher
counts in-flight binding calls before dispatch and releases the slot in the
async body's finally, capping outstanding closures at the same bound. The
drain also compacts its consumed prefix (replyQueue.splice(0, head)) once
head reaches the bound, so a drain that stays alive without emptying cannot
grow the backing store linearly with cumulative throughput.
The completion-value meter counted lone surrogates with
_SURROGATE.findall(folded), materializing one single-character string per
surrogate: a surrogate-dense value near the budget (millions of surrogates,
each serializing to six bytes) allocated millions of objects before the meter
returned, defeating the meter's counting-without-building contract. The count
is now the length difference between folded and the without string the meter
already computes; a standalone equivalence check confirms it matches findall
across lone-high, lone-low, paired, astral, and mixed cases.
validateBindings read namespace.global/errorClass.name/memberNameProperty
several times and retained the original errorClass object for the boot
frame, whose JSON.stringify re-read it after validation: a stateful getter
could throw or change between the two stages, turning the seam-misuse
rejection into a worker-exit or injecting an unvalidated name. Each field is
now read once into a plain value and the bindings map stores a plain
{ name, memberNameProperty } copy, so validation and the boot frame see
identical values.
Regression tests: a hostile child floods 5000 sequential valid calls without
reading fd 3 and the run settles worker-exit with the reply-queue message
before maxWallMs; a 3,000,000-surrogate completion succeeds at an
18,000,002-byte budget and reports output-limit one byte under; a 5000-call
flood against a never-settling binding settles worker-exit with the
call-backlog message; getter-backed namespace metadata that throws or
changes on a second read boots and runs with each field read exactly once; a
two-wave flood whose replies exceed the writable high-water mark drives the
drain past the compaction bound mid-delivery and verifies all 1524 replies
arrive. README Known Limitations gains the reply-backlog and call-backlog
bounds (en/zh, pairing re-recorded); a new Agent Note registers the findings.
Review findings on the CPython backend: an explicit pythonBin path bypassed
the load-time checks (missing/non-executable/directory paths surfaced only
as a run-time worker-exit); a throwing binding member accessor escaped the
fd-3 data callback and terminated the host; the reply drain waited on
'drain' alone, so a pipe destroyed under the wait hung forever; and two
staging-leak assertions diffed a global tmpdir that parallel workers can
perturb.
resolvePythonBin now applies the same accessSync(X_OK) + isFile check to
explicit paths (resolved against the host CWD), and the load error message
distinguishes 'is not an executable regular file' from 'does not resolve on
PATH'. validateBindings snapshots callables into a plain record during run()'s
synchronous validation, turning an accessor throw into the seam-misuse
rejection and fixing the key set the boot frame and dispatch share. The reply
drain waits on drain/close/error together and short-circuits on
proto.destroyed. The staging-leak assertions check the exact paths this test
file staged (recorded by the mocked mkdtempSync) instead of a tmpdir diff.
docs(code-runtime-python): add the alternatives section to the hardening note
docs(config-catalog): refresh the code-runtime-python Config source line
test(code-runtime-python): cover the async spawn-error worker-exit path
The review's items: the experimental group README's Packages table and Summary
now list code-runtime-python (CPython subprocess backend, ctx.codeRuntime),
paired; the portable-identifier note's Scope drops the 'has since shipped …
now lists' change narration in favor of the current state, removing the
apparent contradiction with 'the worker is the only shipped backend'.
The fd-3 note's Status line moved off line 3 when the experimental-location
fact was added; it is back as the sole line-3 status. The zh portable-identifier
note's merged Scope paragraph lost its blank-line separator, which the
md-wrap gate read as one hard-wrapped paragraph — the blank line is restored.
The review's move-follow-ups: the Windows test exclude now points at
packages/experimental/code-runtime-python (the constructor throws by design on
Windows, so the suite must stay excluded); the invariant companion and
@module annotations use the new npm name; the truncation marker text and
tmpdir prefix stay as-is (tests anchor them); the package JSDoc and READMEs no
longer call the private experimental backend 'published'/'shipped'; the
code-runtime README row describes the package as protocol AND runtime; the
fd-3 note records the package's experimental location.
The CPython code runtime's complete public contract is experimental, so it
moves to packages/experimental per the experimental-packages rules: npm name
@deepseek-ai/dsh-experimental-code-runtime-python, private: true, no
publishConfig. All references updated (code-runtime READMEs, config-catalog
and module-graph regenerated with zh alignment, tsconfig paths, doc-standard
and workspace-constraints scripts, the fd-3 and settlement Agent Notes, and
the package README links); md-links and translation pairing pass, and the
suite still runs green.
