Highlights trimmed to the four genuinely major entries and ordered by
user impact; the markdown-preview comment surface folded into the chat
comment bullet, Cmd+P path matching folded into the unified search
matcher bullet, the instant session-open and uncached fade-in clauses
folded into their owning entries; fixes grouped after features, visual
polish last.
- Markdown DOM cache: key includes a content-length fingerprint so an
edited or reverted part re-materializing under the same id cannot
restore stale DOM, memoization uses scalar identities instead of the
part object (store reducers recreate part objects on unrelated updates,
which re-ran the async render pipeline for identical content), and a
probe with a mismatched locale/directory no longer destroys the entry
it failed to claim.
- Sidebar bootstrap: the layout-level sync owner only knows known
directories, so expanded projects bootstrapped serialized at background
priority. The visible collection now publishes a second, expansion-
aware demand owner, restoring concurrent hydration for expanded
projects and worktree groups.
- Settings: local changes still sitting in the debounce buffer are not
yet tracked as mutations, so a settings GET racing the debounce window
briefly reverted them; reconciled results now reapply the pending
buffer.
Main replaced the chat timeline scroll engine while this branch was in
flight, which obsoletes two of its subareas and reshapes a third:
- Chat timeline: main's LegendList-based MessageList/ChatContainer win;
the activation-overscan staircase targeted the removed tanstack path
(LegendList provides adaptive rendering natively) and is dropped along
with its test.
- Scroll shadows: main's hook-based masks stay (the virtualized list owns
its scroll element — there is no wrapper to hand the styling to); the
viewport-wrapper ScrollShadow rewrite, its index.css replacement, its
test, and the call-site viewportClassName adaptations are reverted to
main. The chat OverlayScrollbar keeps this branch's disableHorizontal.
- OverlayScrollbar: the direct-DOM rewrite lands, but binding now follows
the live container node instead of binding once per ref object — the
chat scroller remounts on every session switch, and a bind-once
contract left the scrollbar attached to a dead element.
- Markdown renderer: the detached-DOM cache and warm-block fast path
merge with main's block-commit reveal (enter cascade), streaming code
highlighting, and gutter reservation; the per-block reconcile keeps
both the decoration-refresh path and the reveal cascade.
The composer rendered a second StatusRow instance carrying the
pending-changes bar and the todos dropdown, so every restyle of the
floating assistant-status chip (glass, placement, sizing) silently
restyled the composer bar and its dropdown too — for the fourth time.
StatusRow is now only the floating chip above the composer; the
composer's own bar is a new ComposerStatusBar with the pre-glass layout,
its own container name, and its own container-query classes (the
anti-overlap rules that hide the long active todo under 38rem and the
changed-files label under 30rem moved with it — losing them during the
split let the two dropdown triggers overlap on mobile). The two
components share no markup and no CSS hooks anymore.
The base rule declared --scroll-shadow-size as var(--scroll-shadow-size,
48px) — a self-referential cycle, which computes to invalid and silently
resolves every mask below to none. The old ScrollShadow component masked
the bug by setting the variable inline; the hook-based path (the chat
scroller) sets only the data attributes, so the chat lost its top and
bottom scroll fades. A concrete 48px default restores them; inline
overrides from the component still win.
The animated maintainScrollAtEnd also ran outside streaming, so opening
a historical session glided visibly through the whole conversation as
late row measurements corrected the end position, and an in-flight glide
could supersede explicit navigation. Corrections are instant unless the
session is actively working.
Returning to the bottom during a stream also lands on the end as of that
moment, and the list's own follow may not have re-armed after the user's
earlier gestures — the queued-send jump then fell behind the growing
reply. goToBottom now re-asserts the edge a few times (150/400/800ms)
until it holds; a new user gesture cancels the window.
During a long stream the follow scroll masks the entrance entirely, so
the blur was invisible where it streamed and only a cost risk on weak
GPUs. Keep the compositor-only fade and rise (and the same-tick cascade)
— those are effectively free and still show on short replies where no
scroll motion is involved.
The 140ms fade read as a pop next to the scroll glide. The reveal now
runs 320ms on an ease-out-quint curve with a 10px rise and a touch of
blur resolving — paint-only properties throughout, so row measurement is
untouched — and blocks committed in the same tick cascade with a 55ms
stagger instead of appearing together. Reduced motion still disables the
whole thing.
With block-level commit the open code fence grows by whole lines at the
throttle cadence, so the old reason to skip highlighting it (re-tokenizing
a growing block ~40x/sec) no longer applies: a partial fence now
highlights too, and streamed code arrives colored instead of colorizing
only when the fence closes. Fences beyond 300 lines fall back to plain
text until closed, keeping the repeated worker re-tokenization bounded.
A freshly committed block also enters with a short fade-and-rise. The
class goes on the block's children (the wrapper is display:contents and
cannot animate) and the transform is paint-only, so virtualized row
measurement is unaffected; reduced motion disables it.
Verified over CDP on a production build: streamed code inside a still-
open fence renders highlighted, and the code text's x-position is
identical during and after the stream.
Token-by-token streaming mutates the trailing paragraph in place on
every tick: words rewrap, the last line jitters, and the whole reply
reads as flicker. Streamed text now commits only up to the last complete
line — prose arrives a paragraph at a time (a markdown paragraph is one
logical line), code fences reveal line by line, tables row by row — and
a shown block never changes again. A paragraph that runs long without a
newline releases at the last sentence (then word) boundary so the stream
never stalls. Applies to assistant text and reasoning; tool output keeps
its raw tail.
With growth arriving in block steps, the end follow switches to the
list's animated mode so each step is a glide — reveal and scroll read as
one continuous motion. The gesture opt-out now measures at-end from the
live list state instead of the cached flag, which the animated glide
deliberately leaves stale while trailing the edge; without that, a drag
during a glide could leave the scroll-to-bottom pill unshown.
Measured: end-following holds at distance 0 for the whole stream, and
the mobile drag opt-out shows the pill in three of three runs. The
continuous glide costs ~15% more main-thread time per streamed character
than the instant follow — the price of the motion.
Three first-render polish issues:
Code blocks jumped at end-of-stream: streaming defers the per-line
line-number markup, so the finished decorate pass inserted the gutter
column and shifted every code line right. The gutter's horizontal
footprint is now reserved with CSS while the markup is deferred, so the
final pass only fills in numbers and colors.
File references were verified twice per render and re-verified with the
wrong directory: the annotation pass's own DOM writes re-triggered its
MutationObserver, and the pass also ran before the effective directory
resolved — issuing stat probes under an empty directory and a second
time under the real one. The observer now ignores the pass's own
mutations and annotation waits for a resolved directory.
Content replacing the hydration skeleton popped in: it now plays a
one-shot 180ms fade (reduced-motion aware); cached session switches
never carry the class and stay instant. Also removed the dead
disableStaging prop and its unused pendingRevealWork threading.
Verified on a production build over CDP: a streamed code block's text
keeps its exact x-position across end-of-stream, and stat probes for a
message with file mentions are unique per path with the directory header
always present. The hydration fade path could not be exercised in the
harness (the live event stream pre-populates parts in a single-project
environment) and needs an eyeball check on a cold multi-project open.
Sending from mid-history positioned the new turn with an animated
scrollToIndex; over a long distance the smooth scroll gets cancelled by
rows mounting and measuring along the way — a weak scroll attempt that
dies partway, leaving the reader where they were. Park instantly when
the viewport was not at the end at send time; the short smooth park
remains for sends from the live edge.
Sending while the agent is working queues the message without creating a
user row, so the anchor-arming send path had nothing to claim and the
viewport stayed parked mid-history. Queued sends now jump straight to
the live edge (instant, so end following re-latches immediately).
Rail selection also switches from smooth to instant scrolling: a long
smooth scroll through the virtualized timeline gets cancelled by row
remounts and lands mid-way or on the wrong message, while a teleport
always arrives.
3.3.x makes maintainScrollAtEnd follow content growth on its own — a
tail row growing in place included — which is exactly what the manual
totalSize correction existed for. Delete that correction (the totalSize
listener now only drives the anchored-turn glide) and pick up 3.3.x's
measurement batching, prepend-flash fixes, and web programmatic-scroll
fixes. Opt the explicit maintainScrollAtEnd config into footerLayout per
the 3.1.1 guidance.
The library's own released-on-user-scroll heuristic proved unreliable
one run in three against synthetic touch, so the gesture state machine
stays authoritative: while a real gesture owns the scroll, the list's
end pinning is switched off through a threaded endPinningReleased prop
and re-engages when the user returns to the end.
Validated with the CDP battery on a production build: stream follow
stays at distance 0, mobile drag releases with the pill shown in three
of three runs, resize oscillation stays at the reduced level, the rail
reaches the last turn, and profiled streaming cost per rendered
character matches the tuned 3.2.0 numbers.
A slow edge drag settles the resize detector repeatedly, and every settle
replayed the instant return to the live edge — reading as exactly the
jumping the resize suspension removes. Stay where the reader is; geometry
changed, and the next explicit action or stream correction re-engages the
edge as usual.
A pinned viewport shook during window resizes while a scrolled-up one
stayed calm: both pinning mechanisms — the growth corrections and the
list's maintain-scroll-at-end — re-assert the end against rows that are
still re-measuring, fighting the size compensation that keeps the free
case stable. While the width is actively changing, stand both down and
hold the reading position the same way the free path does, then re-assert
the live edge once with a single instant write after the resize settles
(only when the viewport was still following).
Both scrollToTurnId and scrollToMessageId hard-returned false when the
target was the trailing (last) turn and its element was not mounted, so
clicking the rail's last item in a long scrolled-up session did nothing.
The trailing entry is a regular list row at the end of the data — scroll
to its index the same way history targets are reached.
A width change re-wraps every row at once, and with size restoration off
the accumulated height delta above the viewport threw the visible content
hundreds of pixels up and down while dragging the window edge. Size
restoration must stay off in steady state — a tool result expanding in
place has to grow downward — so a ResizeObserver on the scroll node
enables maintainVisibleContentPosition size compensation only while the
width is actively changing and releases it 300ms after it settles.
Measured on a continuous 1400-800-1400 drag over a long session pinned to
the live edge: the largest per-step displacement of a visible paragraph
drops from ~680px (median of three runs) to ~200px, and the viewport
never detaches from the end.
On touch surfaces the gesture opt-out waited for the viewport to have
already left the end, but streaming corrections re-pin the viewport every
chunk, so the drag never registered: the user could not scroll away
mid-stream, the scroll-to-bottom pill (which waits for a gesture) never
appeared, and live-follow stayed armed — which also made the mobile
load-older button throw the viewport to the bottom, since the prepend's
content growth ran the end correction.
Track the finger's Y between touch events and treat a downward drag (the
touch mirror of wheel-up) as an immediate opt-out. Loading older history
also releases live follow explicitly before the prepend, covering the
case where the button is reachable without any prior scroll.
The old scroll engine needed message parts to report content growth
(onContentChange) and per-message animation lifecycle callbacks
(AnimationHandlers) so it could re-pin the viewport. The timeline list
measures growth itself now, and the replacement hook had already stubbed
the whole contract with no-ops kept only for source compatibility.
Remove it end to end: the hook exports, the container and list threading,
the ChatMessage/MessageBody signal-only effects, and every part-level
prop and call site. Expand/collapse behavior and reveal animations are
untouched — only the reporting channel goes.
New Streaming section on the Chat settings page with a checkbox that
controls whether the viewport follows new content while a response
streams. Default stays on. With it off, the anchored user message still
parks at the top on send, but no glide or end-follow correction runs and
the list's maintain-scroll-at-end stays disabled; the scroll-to-bottom
pill and session open keep scrolling explicitly.
Persisted through desktop settings like the other chat toggles (auto-save
diff, authoritative apply, sanitize), registered in settings search, and
localized in every locale.
Streaming follow scheduled a two-frame animation-frame chain per content
growth and then went through the list's programmatic scrollToEnd, whose
scroll bookkeeping schedules further frames. Together they roughly
doubled frame production for the whole stream (16 -> 35 frames/s in a
profiled streaming session), with the matching style recalc and layout
cost per frame.
In following-end mode, correct synchronously in the totalSize listener by
writing scrollTop directly — the same write path a user gesture takes —
and keep the animation-frame chain only for the anchored-turn glide.
Profiled streaming frame production returns to the previous engine's
range and peak heap drops from 164 MB to 90 MB.
The streaming tail only overlaid the actively streaming message with live
parts from the sync store. When a turn moved to its next step message, the
previous message fell back to its lagging base record, briefly dropping its
completed tool parts — remounting them and replaying their reveal
animation once the record caught up.
Overlay live parts for every message of the streaming tail via a new
useSessionPartsForMessages hook, guarded so an empty live array never
erases parts the record does have. Also key generate-effect glyphs by
index so appended text does not replay earlier characters, and keep tool
row reveal wrappers mounted unconditionally.
The streaming status and the scroll-to-bottom pill now hand off to each
other in one place above the composer: same anchor, same input-aligned
column, both in the popover glass (ring and soft shadow stay pill-only
as the interactivity cue, with the shadow on a wrapper so the glass
backdrop-filter cannot drop it after hide/show cycles). The pill is a
single full-surface button, moves to the left edge, and while the
session streams it carries a compact one-line status label instead of
the status row's animation machinery, which did not survive a 32px
chip.
Behavioral fixes along the way: a gesture shows the pill immediately
(no debounce) and hides the status in the same frame; gestures register
whenever the viewport can actually move up — an anchored short turn
kept live-follow armed and suppressed the pill entirely; the status
row sheds its in-list morph offsets (mb-6, message column) and its
inline-size container gets the glass on an inner row, since a query
container cannot shrink-wrap; its working text uses full
muted-foreground to match the pill.
The animated glide trails its target between corrections, so the
distance-based at-end signal flapped during streaming and the pill and
the status row crossfaded into overlapping half-transparent mush. The
pill now shows only in free-scrolling (a real gesture), which also
keeps the status row's visibility stable; a gesture taken after our own
movement already left the end offers the pill immediately.
Away from the live edge the floating status row fades out (it was
transparent text over scrolled history) and the scroll-to-bottom pill
takes over the activity signal with a pulsing dot. The pill moves to
the left edge — matching the status row it stands in for — and gets the
same glass treatment as the dropdowns.
The rest gap after completion doubled up: the status overlay reserved
its own end inset on top of the footer's 10vh spacer. The overlay now
floats within the spacer zone — its height feeds only the glide target
(the live line still rests above it), the list gets no extra inset, and
completion cannot shift layout because there is nothing to collapse.
The finished message's metadata lands in the slack the glide left above
the status row, restoring the old morph-in-place feel. The frozen
height hack and the opaque background are gone — the overlay only ever
covers blank spacer, not text.
The barely-visible shift at end-of-stream was the reserved status-row
inset being released when the row unmounted. The measured overlay
height now only grows during a session and resets on session switch, so
the finished message's metadata footer appears exactly where the status
row was, with zero layout movement.
The floating status/working row let streaming text show through it and,
because the scroll math did not know the overlay existed, the reveal
correction targeted the very bottom edge — the live line ended up
hidden beneath the status row ('catches up to almost-the-end'). The
overlay now has a solid background, its height is measured with a
ResizeObserver and fed to the timeline as the occluding overlay height,
so the live streaming line glides to rest just above the status row and
the list reserves matching end inset at rest.
The phase-2 reveal correction now scrolls animated; successive
corrections restart the smooth scroll from the current position, so
streaming follows as one continuous motion. User gestures interrupt the
native smooth scroll, so free-scrolling is unaffected.
Phase 2 of the anchored turn (following the live edge once the reply
outgrows the viewport) never ran: the reveal correction was triggered
only by entries-array changes, and the streaming tail grows inside a
single row without touching the array. The hook now also subscribes to
LegendList's totalSize and re-runs the same guarded correction on every
content growth.
Six reported defects, one root theme — our port diverged from the
reference semantics:
- a user gesture no longer collapses the reserved anchored end space
(that snap-to-bottom was the 'anchoring vanishes on a micro-scroll');
it only disarms the follow/anchor machinery, and anchor remeasures
after the user takes over can no longer re-position the viewport
- the anchor arms synchronously BEFORE the send reaches the store, and
the claim compares against the arm-time baseline — arming a frame
after the optimistic row committed is why anchoring 'rarely worked'
and sending from mid-history did nothing
- gestures are selective (wheel up, touch/pointer away from the end,
PageUp/Home/ArrowUp), so scrolling toward the end or clicking a row
at the live edge no longer kills follow
- 'at end' is a tight 40px band against the full content length instead
of the list's half-viewport isNearEnd, so the scroll-to-bottom pill
appears when the viewport actually leaves the end
- reaching the end re-arms follow without knocking an anchored turn out
of its mode, and the overlay scrollbar suppression follows scroll
ownership rather than the anchor's mere existence
- the status/working row moved out of the list footer to a fixed spot
above the composer: inside the list it walked down with every
streamed line, and its unmount was part of the end-of-stream jerk
Sending a message now parks that message near the top of the viewport
and streams the reply into reserved end space below it, instead of
jumping to the bottom and chasing it.
- swap @tanstack/react-virtual for @legendapp/list in the chat timeline; the
streaming tail becomes a normal list row rather than a separately rendered
block, so one component owns the scroll position
- add timelineScrollAnchoring: pure anchored-turn geometry plus the three
scroll modes (following-end / anchoring-new-turn / free-scrolling)
- replace useChatAutoFollow with useChatTimelineScroll, which opts out of
automatic movement on real gestures via a generation counter instead of the
timer windows the old implementation needed to recognise its own writes
- move the load-older button, question/permission cards, recap, status row and
bottom spacer into the list header/footer, since the list owns its container
- extract useScrollShadow so the shadows can attach to that container
maintainScrollAtEnd and maintainVisibleContentPosition replace the manual
prepend anchor-hold and the mobile quiet-window prepend deferral.
Validated: workspace type-check, lint, web build, ui tests per file.
Scroll behaviour itself is unverified and needs manual testing on web, desktop
and iOS.
The hover info card (project/branch/PR/time) added noise without pull —
the tab title plus the sidebar already cover it. The tooltip body
component, its per-hover subscriptions, and the native title attribute
are gone; right-click/menu and the status dot are untouched.
The 'Always show editor toolbar' preference is gone — the docked
toolbar under the file tabs is now the only mode, and the floating
hover toolbar branch in the files editor (with its open-state and
outside-click machinery) is deleted. The stored preference is dropped
by a store migration and removed from desktop settings persistence,
settings search and every locale.
Navigation section order now reads: file editor keymap, auto-save,
terminal shell + login shell, Terminal Quick Keys, and the Session
tabs group last.
The General section renders OpenChamberVisualSettings with an explicit
visibleSettings allowlist, so the new control was searchable but never
rendered.
A new Session tabs group (web/desktop only) turns the header session
tabs off; disabled, the header renders the exact pre-tabs view — plain
session title with meta row and the always-visible session menu (the
same block VS Code uses). The Alt+W close-tab shortcut no-ops while
tabs are off. Registered in settings search with a matching anchor;
labels translated in all locales.
The close-with-neighbour-activation logic moved into a shared helper
(the strip and the shortcut use the same path). Alt+W is the default
because the browser owns Cmd/Ctrl+W on web; desktop users can rebind it
— the action is registered as customizable, so it appears in the
Settings shortcuts section and the shortcuts help automatically. VS
Code is excluded (it has no session tabs).
Auto-add means the strip only grows, so past 20 tabs the oldest one
leaves the working set (the newly opened tab is always the survivor).
The branch/worktree/PR/project subscriptions that feed the hover
tooltip moved into the tooltip body component, which mounts only while
the tooltip is open — a resting tab now subscribes only to its status
dot's session status and unread count.
Rename from another tab's menu stored only a boolean that the
session-switch reset effect wiped before the menu finished closing, so
the tab activated but rename never began. The pending rename now
carries the target session id and begins exactly when that session
becomes active (whether the menu closes before or after the switch),
without the reset effect cancelling it — and without an in-flight
rename being cancelled by unrelated re-renders. Tab background/text
state changes animate at 75ms so activation reads immediate.
The title fade no longer sits on controls: it lifts while the tab is
hovered or its controls/menu are open (hard clip instead) and is not
applied at all while the active tab renames. The active tab clips its
title instead of showing an ellipsis, matching inactive tabs. Tab
separators drop the extra transparency that made them nearly
invisible.
Keeps enough space for the chat panel when the sessions sidebar stays visible
Bases the expanded layout breakpoint on the sidebar width instead of a fixed value
Right-click now opens the session menu under the cursor (the sidebar's
context-menu pattern; the same header-supplied items back both the
"..." dropdown and the context menu via injected menu primitives) and
still never changes the active tab. Hovering a tab shows the sidebar's
session tooltip — title, last-activity time, project, branch and PR
status — after a delay, suppressed while a menu is open or a drag is in
flight. Each tab carries the sidebar's status dot at its end (accent
while the session runs, info-blue for unread), hidden while the hover
controls overlay it. Tab titles fade out instead of ending in "...",
and while the active tab is renaming its hover controls stay hidden so
only the rename controls show.
The strip now owns a single dropdown per tab, fed by the header with
items bound to that tab's session — rename (activates the tab first),
copy id, share/copy link/unshare, export and move-to-worktree (active
tab only, they need the loaded directory), close other tabs, archive
and delete with the confirm dialog targeting the right session. The
separate inactive-tab menu is gone, and there is no Close item — the
tab's close button covers it, now placed after the menu button.
Right-click opens that menu without activating the tab. The menu's
anchor overlay stays mounted until the close animation finishes, which
removes the popup flashing in the top-left corner on close. The
scroller hides its scrollbar via a dedicated CSS class (the bar was
shifting the header content vertically).
The old flex spacer split the header in half, boxing the strip into the
middle while tabs shrank to slivers; the spacer now renders only for
the VS Code and surface-title layouts, tabs keep a fixed width (w-44)
and the strip scrolls behind the right-side buttons. Every tab gains a
hover-revealed close button next to the menu; the active tab's menu is
hover-revealed too (it was always visible) — the header passes it into
the strip so both overlays behave identically. The active tab's inner
padding matches inactive tabs.
Uniform compact tabs (h-7, 6px radius, 13px medium) that share the strip
width and shrink before scrolling, thin separators between inactive
neighbours that disappear around the active or hovered tab, a
hover-revealed menu button anchored to the tab's end, a soft selection
pill for the active tab, and hidden scrollbars with fade edges. The
active tab is single-line: the project/branch meta row stays only in the
VS Code header, which keeps the plain title.
Web/desktop header replaces the single session title with a strip of
soft pill tabs, one per session the user has opened (sidebar, palette
or deep link — opening anywhere adds a tab once). The active tab is the
familiar title block — rename, meta row and the full session menu —
inside a gently selected pill; a brand-new draft shows as a transient
pill until its session exists. Inactive tabs show the title with a
hover-revealed "..." menu (close tab, close other tabs, copy id) that
nudges the text like sidebar rows, and close by middle-click too.
Tabs drag to reorder, scroll behind the right-side header buttons with
soft fade edges, respect the reserved window-controls inset, and
persist across reloads. Closing the active tab activates its neighbour
(or opens a new draft when it was the last). Tab ids whose session is
not in the loaded list stay stored but hidden, so a partial session
list never destroys the working set. VS Code keeps the plain title;
mobile is untouched.
activeSurface was permanently 'chat' after the legacy mobile layout
removal, so the whole concept is gone: the store field, surfaceGuard,
setActiveSurface/setSurfaceGuard, the per-runtime surface memory in
prepare/restoreForRuntimeSwitch, and WorkspaceSurface itself. All ~30
setActiveSurface('chat') call sites were no-ops and are deleted;
always-true 'is the chat active' checks in keyboard shortcuts, Header
and ChatContainer are unconditional now. FilesView's dirty-file guard
kept its file-switch and close protection but drops the surface-switch
branch nothing could trigger. TerminalView visibility comes only from
its callers. The router keeps parsing legacy ?tab= links (they open the
matching context-panel surface) via its own RouteTab type and no longer
serializes a tab or diff file into URLs — desktop URLs never carried
them anyway.
MainTab/activeMainTab/setActiveMainTab/setMainTabGuard were deprecated
mirrors of the surface names — every call site now uses
activeSurface/setActiveSurface/setSurfaceGuard directly and the aliases
are gone, including the persisted mirror field.
The 'diagram' surface had no way to open it (navigateToDiagram had no
callers except a .drawio attachment click that navigated to a surface
nothing rendered); the surface, DiagramView, and its store plumbing are
removed, and a .drawio attachment now opens in the file panel.
?tab= deep links map to the matching context-panel surface instead of
setting a main-area surface nothing renders, and a persisted non-chat
surface can no longer rehydrate into a blank main area.
Phone viewports run the separate MobileApp shell (and a viewport crossing
now reloads into it), so the desktop layout's mobile branch was
unreachable: the drawer machinery, the full-screen secondaryView surface
switch (including the terminal/diagram desktop carve-out nothing could
trigger), the mobile header with its tab bar, the Cmd+number tab
shortcuts, the mobile quota panel, and the surface guard that reset
non-chat tabs. DrawerContext had no consumers left and is deleted.
Header drops from 2630 to 1853 lines; the desktop render is unchanged.
The mobile-vs-desktop surface is stamped once at boot, so a browser
window narrowed past the phone threshold kept the desktop shell (and
its legacy squeezed layout) until a manual reload. A viewport watcher
now reloads into the other shell once the resize settles — the same
mechanism the old Settings toggle used. Fixed shells (Capacitor,
desktop, VS Code) and ?surface= overrides never switch.
With the new mobile app reachable this way, the old/new mobile layout
preference is gone: phones always get the mobile app.
Sidebar session/folder/group search and the Todos, Memory, Plans and
Notes filters required the whole query as one literal substring; they
now match tokens in any order and ignore punctuation via
matchesRankQuery, keeping their own list order and tree structure.
Branch, project, agent, model, provider, stash, SSH-host, skill-catalog
and archive filters each had their own toLowerCase().includes (or no
ordering at all); the git branch and gitmoji pickers also let cmdk
re-filter and reorder on top of the manual filter, silently dropping
rows. All of them now go through rankByQuery/matchesRankQuery: results
are relevance-ordered, multi-word queries match in any order, matching
ignores punctuation, and cmdk filtering is disabled where the ranked
list is already final. rankBranchesForQuery keeps relevance order
instead of re-sorting matches alphabetically; the model picker now also
matches model ids.
One matcher for every dropdown filter: multi-token queries match in any
order, matching is punctuation-insensitive, single-token queries
tolerate typos, and results come back relevance-ordered (prefix, then
word-boundary, then substring, original order on ties). matchesRankQuery
is the boolean companion for lists that keep their own grouping.
ghostty-web marks the terminal container contenteditable but only its
hidden textarea opts out of IME text mangling, so iOS and Android
keyboards uppercased the first letter of every command. Set
autocapitalize/autocorrect/spellcheck off on the container after open.
The tab list lived only in per-tab sessionStorage, so a new browser
tab, another device, or cleared storage showed an empty terminal
sidebar while PTYs kept running server-side, and orphans leaked until
the idle sweep. Add GET /api/terminal/sessions and adopt unknown
server sessions into the local tab projection (additive only; a failed
listing changes nothing).
The idle sweep also reaped terminals in background tabs because only
the active tab holds a WebSocket attachment. Add POST
/api/terminal/touch and have open clients periodically refresh
activity for every session their tabs reference.
Directories and files were rendered as fixed category blocks, so an
exact file match sat below unrelated directories. Merge both result
sets and rank them with the shared fuzzy scorer against the full
relative path. Multi-word queries now match tokens in any order
(longest token queries the server, the rest filter client-side), and
path truncation keeps the parent segments next to the file name so
index.md-heavy trees stay distinguishable.
Cmd+P scored server file hits by basename only, so a query like
"solo-is-a" returned nothing for solo-is-a-team-size/index.md.
Score by relativePath and widen the server candidate limit to 40
so client-side reranking has enough to work with.
Two mitigations for local multi-instance contention over the shared relay
identity:
- A standby instance now waits a 2-minute grace period after the host claim
frees before taking over, so a cleanly restarting host (app update or
relaunch) — which reclaims at boot with no wait — always wins the restart
window instead of stranding paired devices on another process.
- Dev instances never host the relay passively: dev scripts set
OPENCHAMBER_RELAY_HOST=off and the Electron dev shell is detected via
OPENCHAMBER_ELECTRON_DEV. Explicit enable/pairing on such an instance still
force-claims; OPENCHAMBER_RELAY_HOST=on overrides.
A dedupe-keyed client re-mint (QR rescan, password/passkey re-login) replaces
the stored record, which also reset the operator-visible name to the app's
hardcoded default ('OpenChamber Mobile'). The app-reported name is now only a
fallback: an explicit pairing label wins, otherwise the replaced record's
label is kept, and the default applies only to a first-ever pairing.
The chat selection menu (desktop pill and mobile sheet), the chat
comment input, the file-preview Comment pill, the shared inline comment
editor, and the composer's context chip preview popover now use the
same translucent blurred glass surface as the btw panel, keeping their
borders and soft shadow.
The project knowledge panel's section sidebar becomes a horizontal pill
strip on mobile, where a half-width side column left no room for
content. The autocomplete popups get their 120px minimum height back —
browser keyboard panning could put the anchor above the measured
boundary and collapse them to nothing. The status row's 24px footer-swap
reservation now applies only to the assistant-status instance; the
accessory-only row above the composer (pending changes) takes the
normal 8px instead of floating a stray gap.
Selecting text in a rendered markdown preview shows a Comment pill;
attaching stores a file-quote context draft carrying the file path, the
selected fragment (not whole lines), the user's comment, and a
best-effort source line range resolved by anchoring the fragment's
first and last lines in the raw content — a partially located fragment
gets no range rather than a misleading one. The fragment stays
highlighted while the comment input is open, using the selection
overlay rects shared with chat quote comments, and the preview's
native selection color now matches chat messages. file-quote flows
through the same context contract: composer chip previews, the message
context card, and the metadata round-trip.
The diff comment editor is restyled as the shared comment pill (context
line inside the box, auto-growing textarea, round attach button;
Cmd/Ctrl+Enter attaches, Enter breaks the line) and the saved-comment
card matches the composer preview entries. Pierre's gutter utility is
enabled: hovering a line shows a small primary-colored plus that opens
a comment for the line, dragging from it selects a range, and it rides
the bottom of an active selection. Clicking a diff line toggles a
single-line comment, and dragging over content maps to the same line
selection the number column produces — the native text selection is
suppressed once a drag crosses a line boundary — opening the editor for
the range on release. Placeholders and meta lines are dimmed relative
to typed text.
Hovering (or tapping) a context chip opens a stacked preview of its
pending items above the composer: numbered entries with a muted header
band, the captured selection, and the user's comment, which can be
edited in place (save/cancel) or removed before sending. The chips
component now subscribes to the draft store itself; the per-kind count
plumbing in ChatInput is gone.
The openchamber-ui:// handler returned index.html without cache
headers, so the renderer disk cache could keep a freshly installed
build loading the previous version's hashed bundles. Hashed assets
stay cacheable; only the document that names them must not be.
Every user-attached context item (diff/file/plan comments, terminal
selections, browser annotations, PR comments and failed checks, linked
issues/PRs, and new chat-quote comments from the selection menu) is now
sent as its own synthetic text part carrying an openchamberContext
metadata payload. The model-facing text keeps the previous wording; the
timeline reads the metadata back and renders each item as a context card
instead of raw prompt text. Legacy messages still render via the old
text sniffing.
The selection menu gains a Comment option with an inline multiline
input, the quoted fragment stays highlighted while commenting, and on
mobile the input overlays the composer pill by rendering inside the
composer form. Add to chat is renamed Add to input; the menu is
restyled and the mobile Copy tile removed. Terminal drafts move their
terminal id out of the language field (persisted-draft migration v3),
and the dead preview-console source is deleted.
The streaming status line and the finished turn footer are the same visual
line, but the swap used to jump: different font/color, a 2px left inset, a
20px row vs the footer's 32px (its h-8 action buttons set the height), and
the bottom-anchored chat pulling the line up because the finished message
carries more structure below its footer than the status row had.
Match the footer exactly: text-sm at muted-foreground/60, no left inset,
h-8 row, and mb-6 reserving the missing space below. Verified against the
live DOM: the footer appears at the exact pixel position the status row
occupied.
Neither failure came from the change that surfaced them; both suites were
asserting behavior the product had already moved.
issue-2039 mocks `session-actions`, and session creation's authoritative
steps live there: the created session becomes current under the directory
the server confirmed, and that directory enters sync's routing index. The
mock did neither, so every assertion about current session and directory
routing read an empty store. It now performs both, which is what makes
those assertions test the store's real resolution order again.
issue-2903 pinned `<ChatEmptyState />` inside ChatContainer's empty-and-idle
branch. The draft transition animation moved that surface to the draft, and
the branch now returns a neutral placeholder. What the test is actually for
— the idle branch not rendering a status row of its own — is kept.
A project could pin the model new chats start on, but not the level to
run it at: the default cascade dropped any variant as soon as a project
model won, and only ever considered the global one — which belongs to
the global model.
Projects now carry `defaultVariant` alongside `defaultModel`, stored and
sanitized only next to that model, and the cascade passes it through.
Both controls sit in one "Defaults for new chats" group laid out like the
Sessions defaults, and the level appears only for models that offer them.
Changing the default model, variant or agent wrote the chat's live
selection unconditionally. For a session carrying its own choice the
effect was mostly transient — ModelControls restores that choice on the
next hydration — but a message sent right after the change went out on
the new model.
The guard is the same pair of signals ModelControls restores from
(shouldPreserveManualModelOverride): a manual selection plus a saved
selection for the current session. Chats following the default still
switch immediately.
Bohdan Triapitsyn