`[:xform :anchor]` is deleted. `T(a)·M·T(-a)` is a transform conjugated by a
translation — "do M in a frame shifted by a" — and a parent already IS a shifted
frame, so an anchor was a peg written inline: one that could not be selected,
keyed, shared between nodes, or placed above a measured channel. Same expressive
content, strictly less reach. `node-test` asserts the two produce the same matrix.
Its two jobs split, and neither is a field on a node any more.
A pivot nobody chose is DERIVED PER DRAG and stored nowhere. `gesture/pivot` is
the middle of what the node draws — `pick/bounds-of`, the same call the stage
draws the selection box from, on the same frame — or the node's own origin when it
draws nothing. `gesture/about` solves for the position that holds that point
still, so a turn now writes `pos` as well as `rot`. Nothing is cached, so nothing
goes stale: the stored anchor was that same middle captured once at creation while
the box beside it was recomputed every render, so on anything edited since it was
made the cross and the box disagreed and the pivot was wrong. A symbol with more
than one node diverged on its first edit.
A pivot somebody chose is a PEG — `nest/peg`, an ordinary `:group` parent sitting
on the derived pivot with `:pinv` captured so nothing moves. It is the answer to
the three things a derived pivot cannot do: a pivot that persists (an arm about
its shoulder), a pivot that travels (a keyed `pos`), and a hand transform over a
measured one. The last was impossible before — `local`'s translation is
`pos − M·a`, so under a measured `M` writing an anchor moves the thing it was
meant to leave alone. `flow/freeze`'s `pivoted` pass knew this and skipped every
`node/measured?` node, which is exactly why the traced mouth pivoted about
(-234, -395) on a 320x200 stage: the top-left corner of the footage. That pass is
gone; there is no node a derived pivot can be missing from.
`demo/stage` is the one place the anchor did work a static `pos` cannot: `:scale`
is keyed, and the source's middle has to stay on its authored centre throughout.
It is now seven pegs, identical to the pixel.
Also: `events/ui`'s `fitted` rescales a dropped tracing right after placement, and
the anchor had been silently keeping the picture centred through that; it solves
for the middle explicitly now.
Schema 7. Nothing is converted, as in 6: every project is marked 7 and one still
carrying an anchor is refused by name, with what to do about it. Dropping an
anchor is pixel-exact wherever rotation and scale are the identity — everywhere a
freeze or a drop wrote one — but not on anything since turned by hand, and not at
all where `pos` is dense, so a conversion would be silent and wrong for exactly
the nodes somebody had placed themselves.
601 CLJS tests, 68 Django tests, and the onion and take browser suites pass.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01HkinzDz1VtahZVsujAGRBD
The player's tracker derefed its inputs one by one inside `ratom/run!`. An
edit that also changed the selection ran it from the selection's change: it
read `::render/shown` while still stale, then pulled the dirty
`::render/clip`, which re-ran it nested with the new resolver — and the outer
run, finishing last, wrote the old one back. A deleted layer stayed on the
stage until something else ran the tracker. The snapshot is now one sub, so
the tracker derefs a single clean input and is never re-entered.
Onion skinning only ghosted cels in `:display :lane` symbols, so a drawing
made of keyed shapes showed nothing. A ghost is now just the picture the
resolver draws n frames back and ahead, limited to the selection when there
is one. The scope setting is gone.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The icon run in the top bar read worse than what it replaced. Restore the
earlier top bar — pane words, new, open ▾, undo ▾ redo, the edit menu,
snapshots ▾, title, status, export… — and its CSS. The stage bar's overlay
switches, the palette chooser and `menu/popover` stay.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Top bar: every document command is an icon button in a run — panes, new/open,
undo/history/redo, cut/copy/paste/duplicate, snapshots — with the title and
status in the middle and export, share and account on the right. Pane toggles
are pictures of the window with that pane filled in.
Stage bar: tracing, onion and passepartout are all the same toggle-plus-settings
split button, followed by one zoom group, with fit inside it.
Popovers: one `menu/popover`, measured against its button and dismissed by a
scrim, replaces the <details> popouts that never closed on an outside click.
Palette: the strip ends in a palette button whose popover lists every palette
with its swatch strip, renames the open one in place, and marks the ones on
the stage. An info badge says when the palette being edited is not one the
stage draws in at the playhead; both ends of a palette-lane blend count.
`clip/palette-at` is the resolver's root palette lookup, pulled out so the
strip asks the same question the renderer answers.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
One tool at a time, picked from a strip down the left of the stage or by its
letter, as in Photoshop, Illustrator and Flash: V selects and transforms, P is
the pen, B the brush, E the eraser. The options bar above the stage holds the
tool's settings, and the palette is a grid under the tools with the colour a
new shape gets above it, as Deluxe Paint's was.
The pen's draft is filled into the picture as it is drawn, so the edge on
screen is the pixels the shape will be. Click the first point, Enter, Esc or
leaving the pen closes it; the pen stays the tool. Editing a shape's points is
the pen ON that shape — Figma's vector edit mode — so the separate points mode
is gone: click an edge to add a point, ⌥-click one to delete it, on every key
at once so tweens keep meaning something.
The brush paints a mask, and what it will be is shown while painting: the
stroke is traced round its pixel edges, holes and all, and simplified to a
point every so many pixels of outline by the same function the saved shapes
come from. A hole is bridged into the one ring along a whole-pixel row, which
the fill never samples. Blender's Adjust Last Operation re-traces the last
stroke from what it was made of.
A shape coloured CLEAR is a knockout: its symbol is drawn into a layer of its
own and the knockout clears it, every colour or one. The eraser makes one in
the symbol it starts on, of the colour it starts on, or of every colour with
⌥. A click goes through a knockout to what shows. Previews are drawn in the
stacking context of what they will land in.
The polygon's points no longer stay behind when the shape is moved: they were
read off the saved document while the box handles read the one mid-drag.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Opening the 8625 study froze the main thread for 4.7 seconds and settled at
615MB of heap. A profile put three quarters of a project open inside
mix/wav-bytes, which playback had no business calling at all.
Two separate causes. The peak scan built a lazy sequence of one boxed double
per SAMPLE -- ten million of them for a seven-minute mix -- to compute a
single maximum over data already sitting in Float32Arrays; the hand-written
loop is 73x faster and agrees to the bit. The rest was structural: the WAV
existed only because an <audio> element can hold a URL and nothing else, and
the element existed only to be the clock. So a mixdown that was already
rendered got encoded to 73MB of 16-bit PCM, on the main thread, on open, on
every tab switch and on every edit to a track -- and a symbol with no sound
got silence synthesized and encoded full length so the element had a duration
to report.
arthur.clock keeps its interface and all of its arithmetic; the position now
comes from a backend behind a protocol. clock.graph plays the AudioBuffer
through an AudioBufferSourceNode and derives the frame from the context's own
clock, which is the audio device's position in double precision rather than
whatever the media pipeline last published. clock.element is the old path,
kept switchable while the new one earns trust -- BACKEND, or use-backend! --
which is also why every one of the original clock tests passes unchanged: the
derivation they assert is shared, and the backends can only disagree about the
position under it. 6.5s to 1.8s, 4.7s of blocking to 370ms, 615MB to 68MB.
THE POSITION IS COMPENSATED FOR OUTPUT LATENCY, and piecewise because of it.
currentTime is the quantum being rendered, which the speaker is tens of
milliseconds behind; report the renderer and the picture leads the sound,
which in a lip-sync tool is the only artefact that matters. Audio already
rendered cannot be re-rated, though, so reading it back at a new rate jumped
the playhead backwards by three latencies on every press of the rate button.
Each play, pause, seek and rate change now records a segment and a position is
read against whichever segment was in force when that audio was rendered.
One duplicate fell out of this. Opening a project asked for its clock twice --
once from ::opened and once from a ::refresh-clock the shell raised because it
compared symbol ids, and two different documents both open on :main. The
sounds subscription carries the clip id now, so "an edit under the same
symbol" means what it says.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The gesture was handed the pointer's frame and threw it away, creating at
the playhead instead; and the palette row carried :lane? only once its
track existed, so the first double-click on a palette lane -- the one that
has to make the track -- dispatched nothing at all.
Both are one rule now. ::new-symbol-at uses the frame it is given and
resolves every row through drop-destination, with the destination deciding
what is created: a clip of a palette track is a palette symbol, a clip of
any other lane is blank. A row with no path of its own resolves to the
symbol it names, which is what lets a palette track -- hanging off its
owner by :palette-track rather than placed in it -- be reached without a
special case; it also stops a palette cel's slide resolving against the
open symbol and looking like a transfer out of the track. The one thing
left that knows about palettes is materializing the lane a palette row
names before anything asks where the row leads.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
A paste put an instance of "bg" inside "bg" itself. It saved, loaded, and then
threw "symbol cycle in audio" out of nest/audio-tracks, because a container that
contains itself has no finite expansion.
place-symbol and ui/drag already refused that, each by asking
clip/contains-symbol?. Paste asks clip/problems instead, and problems did not
encode the invariant at all -- it checked missing symbols, pose tracks and audio
links, but never the placement graph. So the rule goes where every command is
already checked: paste, cut, duplicate, correction and the span ops all gate on
problems, so one rule covers them all.
Why it looked like a reasonable thing to do is the other half. place-symbol
copied the symbol's name onto the instance it made, and that copy went stale on
the next rename: the symbol read "bg" in its tab while an instance of it still
read "symbol-18", which is its id from before it was named. One object under two
names, with nothing on screen to connect them.
So instances are no longer given a name at creation, and clip/node-label reads
the symbol's name through on every render. :name on an instance now means only
what a person typed, which is what tells two instances of one symbol apart --
"8625 left" and "8625 right" of one "face" -- so an authored name still wins and
read-through is the fallback. The label logic was duplicated across four call
sites with three different fallback orders; location.cljs already read through
and the others did not. They now share one function.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Making a lane one row cost the thing a row was for. A clip stopped being a row,
so there was no longer any way to open a clip and see what was inside it, and
the inside of a drawing — the most ordinary thing in the document — became
reachable only by opening it as its own tab. This is that capability back,
from the root timeline, down as far as it goes.
An expanded lane opens exactly ONE clip: the selected one. Its own keys, then
the lanes and nodes of the symbol it places, then theirs, each mapped into this
ruler by the recursive walk that was already there. Twelve clips in a lane
still cost one row, and inspection costs one branch rather than twelve.
Two things that only showed up once it ran. The portal is chosen by the whole
LINEAGE of the selection and not by the selected id: selecting a shape inside
the clip — or the end of its span — is still working inside that clip, and
matching the id alone shut the portal the instant anything under it was
touched. And selecting now waits for the pointer to come UP, because selecting
on the way down re-drew the timeline before the gesture had said anything: it
shut the portal holding the lane being dragged INTO, out from under the
pointer.
A HELD clip opens too, which the old row walk never did either. `source-time`
is nil for a hold, so the walk stopped there and the contents of every drawing
were invisible from here. Its rows are shown across the hold — which is when
the node is on screen — and marked `:unmapped?`: no keys, and no draggable
edges, because a frozen clock gives no frame inside it a place on this ruler.
Refusing to place the keys is the honest half; refusing to show the rows was
not.
Double-clicking a clip opens the symbol it places as a tab, as double-clicking
the same symbol in the pool does. That was already written and had never once
run: the track captures the pointer for a slide, so the click and double-click
that follow are delivered to the track and never to the block. The track now
resolves them itself. Fixing the delivery exposed two more: `symbol/lineage`
reported a `parent cycle` for any id in a symbol with NO nodes, because a
one-element chain is longer than zero nodes — and opening a symbol left the
selection pointing into the symbol being left, which the breadcrumb and the
inspector then tried to resolve. The editor unmounted. Both are fixed where
they were wrong, and the browser test asserts the editor is still standing
afterwards.
Audio is a clip in a lane like everything else. A dropped sound lands in one
and is trimmed and moved by the same commands; a lane holds picture or sound
and not both, which is the explicit capability the model asked for rather than
a guess per frame. The refusal lives in the commands and not only in
validation, because placement claims time: `blank` would have deleted the
sound to make room for the picture and left a perfectly valid document behind.
What is in a lane of the open symbol is drawn as a lane; what is nested inside
a placed symbol is still flattened by `audio-tracks`, so no sound is on two
rows.
Everything that enters the timeline now enters a lane: a converted take, a
symbol brought in from another project, a sound. One rule answers where —
`lane-destination` — and every symbol is born with a lane for it to answer
with. An unaimed drop fills an EMPTY lane rather than taking an occupied one
nobody pointed at, because the alternative is trimming away what was there to
make room for what was dropped.
Shift during a clip-body drag means the other intention: put this node INSIDE
the symbol the clip under the pointer places, through `nest/move-node`, which
is what keeps the world transform and the root timing. Overlap cannot say
which of the two is meant — dropping on occupied time already means claiming
it — so the person says, and a label by the pointer says it back. The label
asks `nest/move-refusal`, the same check the command makes, so it cannot
promise what the drop would refuse. Today it refuses more than it allows:
both clips have to be on screen at one frame, which two clips in one lane
never are, and a held destination has no clock to move through at all.
`docs/lane-nesting-notes.md` argues that the second refusal is stronger than
the facts require and says what would settle it.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
A lane was a drawing lane: the only thing that could go in one was a one-frame
held cel, and every other symbol instance stayed a permanent root row of its
own. Those are not two kinds of timing, they are one kind with two creation
policies. `lane/place-symbol` drops any library symbol in as a clip that plays
naturally at speed one, `lane/adopt` moves an instance that is already in the
document into a lane keeping its source, span, playback and corrections, and
`append-drawing`/`overwrite-drawing` keep being the policy that makes a new
empty symbol a one-frame hold. The child shape they produce is the same.
Both new commands claim their interval through `blank` before they write, so
the partition rule is unchanged and unduplicated: placing into occupied lane
time trims, removes or splits the incumbents, and a lane still never stores an
overlap. Real compositing overlap is another lane, where the order is explicit.
Creating a symbol with nothing aimed now makes a lane and a clip in it instead
of a loose root instance, and a pool drop prefers an explicitly targeted lane,
then the selected one, and makes a lane only when there is neither. That is
what stops the row-per-symbol growth coming back in through the drop path, and
it is why `add-lane` now takes a z in front of the existing root nodes and
calls what it makes a "lane" rather than "drawings".
The timeline learned the two gestures that a generic lane needs. A clip body
dragged over another lane's track previews there as a dashed block and lands
through `::adopt-in-lane`; the track is found with `elementsFromPoint` and its
selection read back off the element, because a pointer capture does not
retarget. A pool drop over an existing lane previews as a dashed clip inside
that lane instead of a temporary new row that appears and then vanishes --
which also needed the drag-leave check to be geometric, since inserting the
preview changes the element under the pointer and Chromium then reports a leave
with no related target. Lanes are renameable from their label, by double-click,
F2, or the pencil, through `::rename-node`.
`symbol/lane-cels` is `symbol/lane-clips`, and the vocabulary table in the
handoff now separates the two words it had merged: a clip is an instance in a
lane, and a cel is specifically the one-frame held source that drawing creation
makes. Keeping `cel` for the policy is what lets the lane stop being about
drawings at all.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Step 2 of docs/frame-selection.md, which f7e16e5 planned and left unbuilt: the
preserve-snap, and nothing of the plate side. `pose/snapped-frame` is the whole
rule — the latest mark in `(lo, hi]`, and `hi` when there is none — where `hi`
is the native frame the slot defaults to and `lo` is the one the slot before it
defaulted to. So the interval is exactly the frames this slot is the first to
cover, which are exactly the ones the grid shows to nobody: it recovers a
dropped frame out of its own gap, can never read a frame another slot already
showed, and cannot reach past `hi`.
Backward only. A closure at native 13 in a 12-from-30 output is recovered by the
slot whose default is 15, reading 13 — not by the slot at 12 reaching forward,
which would show the mouth shut 17ms before it did and break the no-lead
invariant `cadence_test` asserts over every grid and native pair. That test now
covers the snap too.
Seated as the DEFAULT pose that `pose/source-frame` reaches, so an explicit hand
cut beats a snap with nothing having to say so, and per pose group rather than
at the slot: snapping where the grid becomes native is one frame for the whole
picture, so a head would go two frames stale to fix one mouth. Groups exist only
in `symbol/base-channel-frame`, which is why the interval is threaded that far
down — `(:pre parent)` carried beside `(:f parent)` through the same time maps,
so an ancestor's exposure fold or retime is already in it.
The marks are the `[:vis]` cuts `flow/freeze` already stores, with their
thresholds and hysteresis already decided: no new signal, no new stored field.
A whitelist of `:roto/mouth-aperture` and `:roto/blink` and not a test for
`:generated`, because a skipped frame, a hidden feature and an absent
measurement are three different facts — snapping onto the frames a teeth contour
happened to be missing on is the cadence being dragged about by an absence.
Off is the default and needs no second code path: an opts map that says nothing
gets the behaviour it got before the snap existed. `:snap` is asked about the
SYMBOL, because the cuts are the face's own nodes' and every placement of one
face has the same ones.
The switch is `performance · <face>` in the inspector, and the readout says it
is the stage only. The document setting docs/frame-selection.md specifies wants
a leaf and a round trip of its own; until then this is `[:ui :smart]`, not
undoable, not synced, and unable to reach an export.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`domain/select`: the choosing half of a selection, pure and handed a plain
vector by each site. `pose/held-frame` was already the shared reading half;
nothing chose automatically at all.
The greedy walk rather than the DP, deliberately — it is what the prototype
shipped, and having two implementations is how the DP gets tested. Several tests
pin its exact output for that comparison and will need rewriting when it lands.
Protection is settled before the walk and is not unioned onto its result,
because a protected frame anchors the walk like any other kept frame. The anchor
is what is on screen, so measuring the next frame's drift from a frame that is
no longer displayed holds a one-frame closure across two frames and shows it
twice as long as it was measured. The test that says so is about the picture
rather than the algorithm: the frames the mouth was measured shut on and the
frames the selection shows it shut on are the same list.
A strict local extremum compares against the nearest DIFFERING samples, not the
immediate neighbours, and a run of equal samples is one extremum at the frame it
begins. Immediate neighbours find nothing at all on a closure lasting more than
one frame, which is most of them.
One width for the whole signal, so `distance` is never comparing the prefix two
samples happen to share; a non-finite tolerance falls back to nought, since NaN
compares false against everything and would quietly collapse a performance to a
single pose.
THE EXTREMA HALF HAS NO CALLER under the plan as it now stands, and this is the
commit to revert if it stays that way: `segments`, `turns`, the `:extrema`
branch of `propose` and the multi-component refusal that only guards it, plus
the three tests over them. Plate selections take no extrema by design, and the
performance half gets its closures from the `[:vis]` cut `flow/freeze` already
stores, so nothing asks this code for anything. It is correct and tested and
speculative; docs/frame-selection.md says so beside the build order.
Also corrects `ring/subsample-slots`, which claimed every even budget lands on
the cardinal positions. The corners do; the lip centres survive only multiples
of 4, so the aperture pair cannot be read off a subsampled ring at verts 6, 10,
14 or 18. That is why the performance signal reads a stored cut instead.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The source-to-stage mapping moves off :main's :face group and onto each face's
own :place, above its head. `face-placement` computes exactly what it computed
before, over every subject together, so two faces filmed side by side keep
their filmed relation — it is written into each face instead of onto a group
above them all. Same transform, same subtree, one level lower, and the
composite is identical to the pixel: a digest over every op :main emits across
the whole take is unchanged either way.
THE OWNER IS THE POINT. A face carrying its own mapping is the right size
wherever it is put — dropped into another symbol, or opened in its own tab to
be drawn over — and the take that holds it needs to know nothing. On a group
above the instances the scale belonged to the take, so a face taken out of it
had no size at all and drew at a fraction of a pixel.
The pool's thumbnails drop the workaround that knew about this: a symbol is
rendered rooted at itself again, because a face now carries the placement that
makes that honest, so the pool needs to know nothing about where a symbol
happens to be used. `domain/node` and `arthur.export` leave its requires with it.
The tests here were reading the placement off :main. The photo registration
test changes shape rather than location: its premise was that face-1's head is
its own root, so a photo sitting where it was filmed was image pixels over
image height and nothing else. The head still cancels — that is what the test
is about — but it now cancels against the face's own placement, which is why
the photo comes with the face into its own tab instead of sitting at a
fraction of a pixel beside it.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
