Loop playback — the transport repeating the open symbol — is editor state and
the document has never heard of it. A looping INSTANCE is a node repeating the
symbol it places, which is the four-frame tire turning for the hundred and
twenty frames it is on screen, and it lives in the document as
`:playback {:end :loop}`. Same word, two scopes; named apart here before
anything is built on either.
The status of the second one is the surprise: it already works and cannot be
asked for. `node/placed-frame` does the modulo, `node/problems` already admits
`:end :loop`, `audio-tracks` already expands the periods — and no control sets
it anywhere in the UI. So the tire is three pieces of work, not one: somewhere
to edit an instance's playback, a block that draws its repeats rather than
only its first pass as the timeline's own docstring admits it does, and the
audio period guard.
This also pins down when the held cel can be torn out. Turning every drawing
into a one-frame looping instance is only safe once a looping instance can be
seen and edited; otherwise every drawing in the document quietly acquires a
property with no control on it.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Holding a drawing and looping a one-frame symbol are the same picture, and the
second is a case the model already carries, so the first is a special case
kept for nothing. Speed 0 goes; a cel becomes a one-frame symbol with
`:end :loop` and a span. Looping and span are already properties of the
instance and `:frames` already belongs to the symbol, so nothing moves — a
mode is deleted. Two spellings of looping collapse to one, `:time :loop?`
giving way to `:playback :end :loop`, and `extend-hold` collapses into
`resize-out`, since it exists only to refuse anything that is not frozen
before editing a span.
What it buys is one rule where there were three refusals. `nest/inside` has no
invertible clock for a hold, for `:end :hold`, or for a loop, so
shift-to-reparent refuses all three — which is most of what anybody would drag
onto. Resolving the move with the destination's map at the CURRENT frame
covers every one: a loop is affine within the period the frame falls in, a
one-frame loop is that rule with a period of one — which lands exactly where
the nesting notes argued it should from first principles — and `:end :hold` is
affine in the played part and frozen in the tail.
The trap is written down beside it, because it would be found the hard way:
`audio-tracks` expands a loop into one walk per period, and is saved today
only by the `(pos? speed)` guard that a held cel fails. Make every drawing a
loop and a drawing held for 120 frames becomes 120 walks emitting any nested
sound 120 times. An audibility precheck has to land in the same change.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The plan asked what to draw when a symbol in lane mode holds overlapping
children, and offered to report it as a decision waiting for a person. Wrong
question: they never overlap. Placement claims time — anything placed, moved
or grown over occupied time trims, removes or splits what it lands on — so the
operation that could have made an overlap did not, and `span/finish`, which is
already the single commit path and already refuses rather than half-applying,
is where that is enforced. An overlap is then a bug in a command and not a
state to design around.
The check stays, named `symbol/overlaps` and used three ways: the commit path
refuses one, a property test asserts no command can produce one, and a
document that somehow holds one still LOADS and is drawn visibly wrong with
the status saying so. Not `problems`, which stops a document loading, and not
`conflicts`, which means somebody has a decision to make — a display hint must
never be able to keep a document from opening.
The one place a person can ask for the impossible is toggling lane mode on
over children that already overlap. That refuses and offers to trim them into
a sequence, through the `:required-frames` retry the model already uses.
Also written down, because it is the pair the modifier exists to separate:
a plain body drag is temporal and replaces, trimming extents as needed; shift
is structural and goes through `nest/move-node` into the symbol under the
pointer, which has to keep working for symbols held in a lane.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The lane model put a second container in the node map — a group with
`:layout :sequence`, with its own membership, its own validation and its own
fourteen commands — and every part of the editor then had to ask which kind of
container it was looking at. The decision recorded here is to delete all of
it: a lane becomes a way of DRAWING a symbol whose children are sequential and
non-overlapping, the display goes back to a row per symbol, and the word
survives only in the timeline and in the drag handling that re-spans a
symbol's children while it is drawn that way.
The commands are not the part being thrown away. `extend-hold`, `resize-out`,
`roll`, `blank` and the rest are what endpoint dragging IS, and their
arithmetic is right; what changes is their subject, from "the children of lane
L in symbol S" to "the children of symbol S". They belong in `span.cljs`,
which already owns re-spanning and `finish`.
The plan takes a position on the one question that decides whether this is a
simplification or a circle: lane mode is a saved hint on the symbol rather
than unsaved view state, because the drag rules follow the mode, and a toggle
the document does not record would make one gesture do two different things
to it. Nothing outside the timeline may read the hint.
It also lists what must not be lost on the way, all of which broke at least
once today: a held clip's contents reachable with no keys and no draggable
edges, double-click to open surviving the selection it leaves behind,
selection waiting for pointer-up, no drop silently deleting what it lands on,
and a sound drawn once.
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>
Now that a lane is generic, the obvious next move is to read a clip dropped on
another clip as "put it inside that symbol". It cannot mean that: dropping a
clip on occupied lane time already means it claims that time and trims the
incumbent. Structural nesting therefore needs an explicit affordance -- a grab
handle with `grab`/`grabbing` cursors, distinct from the body's temporal move
and the edges' trims -- and its drop must route through `nest/move-node`, which
preserves world transform and root timing, rather than through a weaker
`:parent` assignment that would make a drawing jump when it is rehoused.
The second half of the note is what expansion should be. One permanently
expanded row per lane clip is the vertical growth the one-row lane exists to
avoid, so an expanded lane shows exactly one portal: the currently selected
clip, swapped in place when the selection changes, not following the playhead.
The portal header is the structural drop target, sub-expanding it walks the
source symbol's lanes through the existing recursive root-time mapping, and
collapsed clips carry their instance-level keys as ticks. The cost of
inspection stays constant.
No lane-as-symbol type and no second ownership edge: the hierarchy is still
symbol, lane, clip, source symbol, its nodes. Lane membership owns time; a
symbol instance owns composition. Written before the interaction is built,
because the capability it protects is easy to lose by accident.
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>
Pulls the preserve-snap forward to step 2: it is the half with no UI and nothing
proposing today, and it needs nothing from the plate side but a fold that can
land last.
Makes the rule exact, because the direction is easy to get backwards and the
draft was vague about it. Slot k reads the latest preserved frame in
(d(k-1), d(k)], else d(k). The closure at native 13 in a 12-from-30 output is
recovered by slot 6, whose default is 15, reading 13 — NOT by slot 5 reaching
forward from 12, which would show it 17ms before the mouth shut and break the
invariant cadence_test already asserts.
Records where it goes, which is the part that was understated as "about thirty
lines". The slot interval exists only at clip.cljs:261 and group identity exists
only at symbol.cljs:399, so the interval has to be threaded down: two internal
signatures, not a drop-in. The rule's seat is the `(js/Math.floor lf)` default
`base-channel-frame` hands `pose/source-frame`, which leaves an explicit hand cut
beating a snap, as manual precedence requires.
And records the shortcut not to take: snapping at clip.cljs:261 needs no
threading and is wrong, because one native frame per output frame means the whole
picture reads 13 instead of 15 — a 67ms stale head to fix the mouth, fighting the
trace selection's own opinion about which head frame to show.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Corrects this plan's central claim. It said one component used twice, because
the prototype's `suggestPlateFrames` and timing-handoff's performance poses
looked like the same function. They are not.
A plate selection has to be NON-UNIFORM — that is the entire reason it exists. A
head still for sixty frames and then whipping across in ten wants two drawings
and then eight, and no nudging of a uniform grid produces that distribution,
because the spacing itself is the answer. A tolerance belongs here: the artist is
buying drawings.
A performance selection is not choosing sparseness at all; the output rate or the
exposure setting already did. What is left is which native frame each decided
slot reads, so nudging the grid is the right size of answer and a tolerance would
be a knob with nothing to control.
And the harder reason, which settles it: a selection cannot put a frame on screen
that the output grid never samples. At 12fps out of 30 a closure at native 13
falls between output frames 5 and 6, so keeping 13 in a set makes it available
and never shows it — exactly what time.md says about an event between output
frames. Only moving what output frame 6 reads can show it.
So the performance half is a backward-only snap of the grid's own pick toward a
preserved frame, about thirty lines plus a prepare step, and its marks come from
cuts `flow/freeze` already stores: the mouth's aperture threshold and the eyes'
`resolve-blink` with its hysteresis. No aperture signal, no new dense track, no
threshold decided twice. The nesting survives unchanged, because a kept plate
frame is just another mark in the same pile.
Records why the aperture signal is impossible as drafted — positions 5 and 15 of
LIPS-INNER survive `freeze/rings->flat` only at verts divisible by four — and
names commit 02069e8 as the one holding the now-uncalled extrema detection, with
the two conditions under which it would be wanted again.
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>
Constant, ramp, and return offsets now append ordinary channel layers to a lane or cel in explicit owner frames. The inspector exposes the commands as one undoable transaction, shows conflicts, and offers removal or retry while preserving generated bases through regeneration.
Ordered-stack compatibility is shared by validation, conflict reporting, and regeneration, including adjacent replacement coverage. Cel-sheet gaps and headers select their lane, so commands cannot fall through to another column's stale selection.
437 tests, 5,804 assertions; both browser flows; 56 Django tests; optimized frontend build.
The cel sheet is a second projection of the same lane rows: frames run down, lanes run across, and each occupied cell carries the timeline cel's exact selection address. The shared action strip proves the point in the browser test by selecting a cell and issuing the existing hold command.
Before exposing that second entrance, fix the boundary mistakes it revealed. Nested commands now convert the open playhead through their enclosing instance path. Overwrite composes blanking with non-rippling placement as one transaction. Picture-rate and pose sampling select only the generated base frame while hand corrections retain the node's authored frame. Stack validation follows covering replacement layers so a document accepted by the validator cannot throw solely because a later offset sees a different shape.
429 tests, 5,767 assertions; both browser flows; 56 Django tests; optimized frontend build.
`docs/lane-handoff.md`, after `timing-handoff.md`'s shape, because the next
thread starts cold and the expensive part of that is not the code — it is the
decisions that were argued out and would otherwise be argued again.
So the section that matters most is the one listing what NOT to re-litigate:
the shot length is authored, placing ripples and overwrite is blank-then-place,
a position inside a cel refuses and names split, a correction has no time space
of its own, a layer's values are a channel, a conflict is not a problem, and a
command refuses rather than guesses. Each of those is a paragraph here and a
commit message in full.
Then the vocabulary, since it was settled one commit ago and the old words are
still in this repository's history: instance, cel, lane, drawing, and placement
for where a node sits only. With the warning that `exposure` still means the
`:expose` grid and always did.
Then the mechanisms that keep paying out — `:span` in the node's own frames
above all, which is why split, trim and blank cost almost nothing — the known
gaps, and how to run the suites, including that a release build clobbers the dev
bundle the browser tests need.
Recommended next piece is the cel sheet: it needs no new model, and it is the
first real evidence the document is not shaped by the timeline that grew up
with it.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Four words had accumulated for a node that puts a symbol inside another symbol.
`instance` was the document's, from the model. `placement` was the stage and
export work's. `occurrence` came in with the lane model. `exposure` came in with
me, because it is what an animator would say. Three bodies of work each brought
a word and none of them retired anybody else's, which is how you get a codebase
that reads like three people describing the same object over each other.
It is a CEL. One drawing, held for some duration. `cel` was already the view's
word — `.tl-cel`, the cel strip — so choosing it was also the smallest change,
and the app already says "drawing" for the content, which is what frees the word
up: historically a cel IS the celluloid with the drawing on it, and that sense
has somewhere else to live here.
instance the `:kind`. The general thing, anywhere in a document.
cel an instance in a lane. UI labels, command names, prose.
lane the group with `:layout :sequence`.
drawing the content a cel names.
placement kept ONLY for where a node sits — `nest/placement` and the
transform that puts a face on the stage. Retired as a noun for the
node itself.
occurrence gone.
AND IT SETTLES A COLLISION I SHOULD HAVE SEEN EARLIER. `:time :expose` already
existed and means something else entirely: how many frames each step of a
subtree lasts, which is what shooting on twos is. Had the block been called an
exposure too, `node/expose`, `clock/exposed-frame` and `subs/render ::exposure`
would have been permanently confusable with it. Choosing `cel` lets the word
`exposure` keep the thing it actually names, and every remaining use of it in
`src` is now that one.
`:layout :sequence` stays as the field, and it is the one place two words are
kept deliberately: the layout names the RULE — children follow one another and
may not overlap — and a group carrying it is called a lane. `node/lane?` says so
where the two meet.
The second view is traditionally the exposure sheet. It will be the CEL SHEET,
for one vocabulary.
Renamed with a script and then read, because a blind pass does real damage: it
produced "an cel" thirty times, renamed the `::exposure` sub that is about the
`:expose` grid, and turned an "exposure grid" into a "cel grid" in two
docstrings. All three classes are fixed. `arthur.domain.sequence` is now
`arthur.domain.lane`, which is what its test file was already called.
424 tests, 5,749 assertions, and both browser flows — `test/browser/lane.mjs`,
renamed too.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Trim, move and blank, and the decision they all three walked into: is the shot's
length authored, or derived from what is in it?
AUTHORED. `:frames` is the symbol's window — how long the shot IS — and the
occupied extent of its lanes is a different fact, read off the occurrences. A
command grows the window when the caller says `:grow-symbol` and NEVER shrinks
it, so blanking the end of a shot leaves a shot with empty frames at the end.
That is a true statement about what somebody authored, and the alternative is
deleting the last drawing and quietly shortening the film. `finish` had the
right behaviour by accident — `(apply max (:frames sym) ...)` — and now says
which number is which: `needed` is where the occurrences reach, `:frames` is
what was authored, and the only thing that makes the second follow the first is
a caller asking.
The three commands turned out to be one piece of geometry, which is `split`'s.
A `:span` is in the occurrence's OWN frames and `:time` says where those land in
the lane, so moving an edge of an exposure is ONE WRITE to `:span` and `:time`
and `:playback` are never touched. `local` and `edged` are the whole of it, and
split now goes through them too.
trim narrows one edge and moves nothing else. Lengthening is `extend-hold`,
which carries a ripple policy and a shot-length policy because it needs
them; letting trim grow as well would give one gesture two sets of
rules and a way to overlap its neighbour.
move one write to `:time :at`, and a destination that would overlap is
REFUSED rather than rippled. Moving a drawing and re-timing the ones
around it are different intentions, and a move that pushed the rest
would be the second wearing the first one's name. Clear the room first.
blank leaves a gap and does not close it. Wholly inside the range goes,
overlapping an end is trimmed to it, spanning the range is split — the
one case that needs an ID, and it asks for one instead of inventing it.
Because the source clock is untouched, trimming the front of a playing insert
starts it LATER INTO its animation rather than restarting it, which is the
difference between trimming and slipping and the reason they stay two commands.
The test samples the frames it kept and asserts they show what they showed.
Blanking leaves the drawings in the library. A lane does not own its content,
and a drawing whose last exposure is gone is still a drawing somebody made.
Overwrite is now `blank` then `place` and needs no policy argument of its own,
which is why it still is not one.
424 tests, 5,749 assertions. The browser flow trims an exposure at the playhead,
moves it into the gap that made, blanks it, and checks the shot is still as long
as it was authored.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The loop the layer design exists for, tested for the first time: correct a
generated channel by hand, turn the generator's knob, and get the new base with
the correction still on it. `replace-feature` already carried `:over` across —
somebody anticipated this — so the feature path needed a test and not a fix.
The head path needed a fix, and there was a second fault of my own making.
`regenerate-head` leaves the head's authored channels alone once somebody has
placed it by hand, and decided that by `(= (:channels old) (:measured old))`.
Sound, until a correction exists: an `:over` layer makes those unequal, so the
FIRST correction anyone made would have stopped the head following
re-measurement for good — the exact opposite of what a layer is for. It compares
the channels without their layers now. The test fails against the old guard,
which is how I know the bug was real and not a story about one.
The other fault was mine, from the commit before this one. An `:offset` whose
shape does not match its base threw, which is right for authored data — the
validator catches it — but WRONG for the case the model actually names: turn the
mouth's `:verts` knob and the re-freeze gives it a different number of points,
so a correction that was correct when it was made stops fitting through nobody's
error, and a throw in the read path takes the stage down.
So a base that has outgrown a correction is a CONFLICT, and a conflict is the
third thing beside applied and discarded. The regeneration records `:conflict`
on the layer; the layer stays exactly where it is; `over-at` skips it, so the
picture is the base meanwhile; and `clip/conflicts` lists them for a view to
offer. A later regeneration that restores the shape clears the mark, so
resolving one can be as simple as putting the knob back.
Deliberately NOT `problems`. A document with a conflict loads, evaluates and
saves — it contains a decision nobody has made yet, and refusing to open it
would be the persistence layer taking a side in an editing question. The
distinction in the validator is one line: a shape mismatch nobody has recorded
is an authoring bug, and one a regeneration recorded is a conflict.
`channel/conflict-with` is the single rule for "can this layer apply to this
base", used by the validator, by `conflicts`, and by the regeneration that marks
them. Only `:offset` can conflict, since `:replace` states a whole value and has
nothing to agree with; a shape that cannot be read yet — an empty key map — is
not a disagreement. `value-shape` answers it without sampling anything.
414 tests, 5,696 assertions, and `:verts` in the test is a real topology change
rather than a synthetic one.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`:over` was specified in animation-model.md, refused in two places, and
produced by nothing: `check-unimplemented!` threw on read and `channel/problems`
reported it. It reads now. This is the part of the model the rotoscoping half
depends on — generate motion, correct it by hand, turn the knob, keep the
correction — and it was the last thing in the design that had never been tried.
The shape that made it small: A LAYER'S VALUES ARE A CHANNEL.
{:id :nudge :support [88 98] :op :offset
:values {:animated? true :interp :linear :keys {88 [2 0], 96 [0 0]}}}
So the three commands the lane model asks for over a selected range — a
constant adjustment, a ramp, a return motion — are one mechanism and not three:
framed values say the same thing on every frame they cover, keyed values move,
and neither needs a new way to say what a value is over time. A layer reads
through `value-at` and `cursor` like any channel, which is also what stopped
blending from becoming two implementations: `over-at` is shared, and the
specification and the playback path differ only in how they READ a layer —
recursively through `value-at`, or through a reading head of its own. One level
deep; a layer's values may not carry layers, which the stack already orders.
That was the risk worth spiking for. A cursor that drifts produces the wrong
pose rather than an error, and a stack means several reading heads per channel
where there was one. The agreement test that holds the cursor to the
specification in forward, backward and random frame order now covers stacked
channels too — including a layer whose head is asked for nothing across the long
stretches outside its support and then asked again, which is where drift would
hide.
`:support` is half-open and explicit. Outside it the base evaluates exactly as
it did before, which is the whole difference between a bounded correction and
inserting boundary keys: the latter alters the neighbouring segments, and the
lane model says so.
A LAYER HAS NO TIME SPACE OF ITS OWN, and this is the design question the doc
left open. Its support and its values' keys are in the frames the base channel's
keys are in — the node's. A correction on a lane is therefore in lane frames and
reaches across the drawings exposed beneath it; one on a single occurrence is in
that occurrence's frames and travels with it when the exposure moves. Ownership
had already answered it, so there is no field to disagree with, and both halves
are under test at lane level.
Two things cost nothing, which is worth recording. A channel is ONE LEAF, so a
correction persists inside it with no codec change at all. And `node/problems`
already reports every channel's problems, so a malformed layer surfaces at the
document level and in the sequence commands' post-check without plumbing.
What is still missing is a command that MAKES one, and with it the question of
how a view offers a constant, a ramp and a return over a selected range. The
evaluator no longer has an opinion about that, which was the point.
`offset` adds component-wise and never writes into a dense value, which is a
view onto the block itself; a shape mismatch throws rather than being dropped,
since a correction that silently does not take is the failure this design exists
to prevent. `replace` can supply a value over an absent base and `offset`
cannot, as animation-model.md required.
408 tests, 5,655 assertions.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Everything could only be added to the end, because `append` computed its own
position — the max end of the lane — and so had no opinion to state. Insert is
not a new command; it is the argument that function was missing. `:at` takes a
lane frame or `:end`, `:end` is the position where nothing has to move, and
appending stops being a separate operation from inserting. New, reused and
duplicated drawings all take it, because there was only ever one placement rule.
Placing ripples: occurrences at or after the position move later by the new
exposure's duration, and `:keep` against `:grow-symbol` still decides what
happens at the shot's end. OVERWRITE is deliberately not a policy argument yet.
Taking frames away from the occurrence already there is TRIMMING, and an
argument whose second value is unimplemented is worse than an argument that is
not there. A position strictly inside an existing exposure refuses and names
`split`, rather than splitting on the quiet: one command performing two is how
a command stops being predictable.
Then split, which turned out to cost almost nothing, and that is the
interesting part. The two pieces keep ONE `:time` and differ only in `:span`.
The right piece's own frames therefore carry on exactly where the left's
stopped, so its source clock, its keys and its corrections go on meaning what
they meant: a held drawing holds the same frame either side of the cut, and a
playing insert plays through it without a seam. There is no arithmetic on
in-points to get wrong, and no shot-length question, since the pieces occupy
the frames the one exposure occupied. The test samples every frame before and
after and asserts the picture is identical — for a hold, for an exposure with a
correction of its own, and for a playing insert.
That is not a clever split. It is `:span` being in the node's OWN coordinates,
which was decided long before there were lanes, paying for something it was not
designed for. The same property is why extending a hold leaves lane keys alone.
Both new commands act at the playhead, which needed `lane-frame` — the symbol's
frame as a frame of the lane's own time, nil through a stepped or looping lane
where one is not the other. Nil refuses; it does not snap to a nearby frame.
Two smaller things found while doing it. `placeable` promised "a whole lane
frame" in its refusal and then accepted 2.5, so both it and `split` now require
an integer, as `extend-hold` already did for its delta. And `lane-end` is
private: `:end` is the only way to ask for it.
401 tests, 5,612 assertions. The browser flow now splits an exposure at the
playhead and puts a drawing in the gap, and checks that six exposures are still
one row.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The model's whole claim is that content and its occurrences are different
things, and until now nothing in the editor could tell them apart: you could
make a drawing and time it, but not expose one drawing twice, and so never find
out whether an edit arrives in two places. That is the first proof obligation in
the lane model and it was the one the commands could not reach.
Three commands, and the distinctions between them are the point:
reuse another occurrence of the same drawing. A decision to share,
made on purpose, because sharing discovered later — when an edit
turns up somewhere you did not expect — is the bad version.
duplicate a copy of the drawing, appended, for when what is on screen is
the starting point for the next one.
make unique this occurrence gets a private copy; the others keep sharing.
The undo of reuse, and refused when nothing else uses the
drawing: a copy nobody asked for is a second identical symbol in
the library for no reason a person could see.
Duplicate copies the CONTENT and not the exposure. Its new occurrence is a plain
one-frame hold, not a copy of the source occurrence's transform or corrections,
because those belong to that use of the drawing — carrying them over would make
duplicating a drawing quietly duplicate the treatment of one exposure of it.
A copy is SHALLOW by default and keeps its references to other symbols, so a
head built out of reusable eyes still uses those eyes. `:deep? true` copies
everything it places with new ids throughout. The lane model asks for both and
says why: never promise decoupling while leaving the edited object shared, and
only the deep copy can keep that promise. `bring/symbols` already did the
reachability walk and the id remapping, so the deep copy is that function
pointed at its own clip.
`node/sources` was still being read as a SET at five call sites, each with a
comment about a lane that cuts between several drawings — the keyed source that
no longer exists. An occurrence names one symbol, so they now ask `node/source`,
and `placed-frame` answers with `:symbol` rather than `:of`, which was the last
echo of the retired field name.
To let the commands use `clip/free-id` and the copy machinery, the lane's own
validation moved from `domain/sequence` to `domain/symbol`, which is where it
belonged anyway: a sequence is the one composition rule a node map carries, and
it now sits beside the parent and stencil checks rather than in the namespace
that happens to build lanes. That also breaks the cycle — sequence can require
clip and bring, and nothing below it requires sequence. Preconditions still
check only the LANE's shape: refusing an exposure edit over an unrelated defect
elsewhere in the symbol would be this command answering for a part of the
document it never touches.
The cel strip gains reuse, duplicate and make unique, the last shown only where
the selected exposure actually shares its drawing. Drawing on twos is also now
under test: exposure length is the cadence, the lane's transform has its own
clock, and it still moves on every frame — stepping it would be the cel cadence
leaking into continuous motion.
397 tests, 5,561 assertions. `test/browser/sequence.mjs` drives the three new
commands through the real editor and checks that three exposures are still one
row.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
A lane's drawings were going to be one instance whose source was a KEYED
channel: frame 0 says `:drawing-a`, frame 4 says `:drawing-b`, and the cels of
a row are that channel's keys. Two things followed from it, and both were
wrong.
The first is that playback meant whichever shape the channel happened to have.
A framed source played its symbol; a keyed source froze the selected frame.
So `node/placed-at` read animation out of storage, and adding an ordinary key
to a still turned it into an animation — the last-key bug, which was not a bug
in the code so much as the rule working as written. But WHICH drawing is used
and HOW time runs inside it are independent questions, and all four combinations
are ordinary: hold one drawing, play one animation, cut between held drawings,
cut between playing ones.
So an occurrence names one symbol in `:source {:symbol ...}` and says how its
source time advances in `:playback {:in :speed :end}` — `source = in + speed *
f`, a hold being speed 0, with `:stop`, `:hold` or `:loop` at the end named
rather than guessed. `node/placed-frame` samples it forwards, which works for
holds too, and `node/source-time` is the separate, invertible edit map, nil
where inversion is meaningless. The two were one function before, and a hold
had to lie about one of them.
The second is that a keyed source only looked necessary because an occurrence
was assumed to need a ROW. It does not. A lane is a group with `:layout
:sequence`, its occurrences are ordinary instances in the same flat node map,
and `timeline/rows` draws them as cel blocks on the lane's own row: twelve
exposures, one row, each cel still separately selectable and addressable. The
vertical growth that justified the keyed source is a presentation question, and
it is answered in the view.
`arthur.domain.sequence` holds the first commands over that shape — add lane,
append drawing, extend hold — each one history step, each refusing rather than
half-applying. Extending a hold leaves the lane's keys at their authored times,
because you are adjusting drawings underneath timed motion; a correction owned
by an occurrence travels with it. Ownership does that work, so no key needs a
flag saying what it follows. Ripple past the symbol's end is refused with the
frame count it would need, and `:extent :grow-symbol` is the caller saying yes.
`clip/blank` no longer carries `:subjects {} :features {} :groups {}`. Empty
maps write no leaf, so a blank document could not survive its own round trip —
`leaf/leaves` promises exactness and was the only honest side of that.
Documents are schema 3. A version 2 document is not read; nothing here converts
one. `docs/lane-model.md` is the design, and says which of its parts are built.
392 tests, 5,525 assertions, and `test/browser/sequence.mjs` drives the editor
through create, hold, explicit overflow and undo.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
A face's :head carries :trace {:frames :origin}: the frames its photo holds
on, and whether the head reads every frame, jumps to the trace frames, or
holds frame 0. It replaces :anchors, so which measured frame a head reads is
one stored fact. An instance's :underlay shows the tracing stills over every
face at or below it, registered through each face's own head, at an opacity,
unkeyed. The clip resolver answers where a row path went on its last frame,
so the paint loop reads the photo's matrix instead of resolving again.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A project is only ever at /p/<id>/<slug>; / is the index of the projects
you own or edit. Every project has an owner, who can name editors;
anyone with the link can view. Every edit saves itself, one request in
flight at a time, as a patch of the leaves that changed, and a websocket
(channels + daphne) carries presence and each committed write to
everyone else in the project. The first write to a leaf wins, and the
loser is told.
Undo is per person: a step undoes only if the leaves it touched still
hold what it left, so it never takes a collaborator's work with it.
Named snapshots replace saving, and restore as an ordinary write.
An empty symbol now survives the leaf round trip with `:nodes {}`.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Every node now has the same time map into its parent, local = rate·(parent
− at), with its span and keys in its own frames: what instances had, made the
rule. A shape without a time map reads as it always did, so no data changes.
The per-kind branches, the span-start term and the rate refusal are gone;
node/time-of, then-time and invert-time compose it like the matrix.
clip/move-node puts a node into another symbol without changing the picture
or the timing — its matrix becomes a :pinv, its time a new :at and :rate, and
its channels, keys and span are untouched — and clip/group makes a new symbol
around side-by-side nodes. Generated parts, split stencils, cycles and
looping instances are refused with the reason. Nested sounds use the same
map.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
A shape's :span stays in its parent's frames, but an instance's or a sound's
is now in its own: dropping a symbol at frame 97 gives it span 0 … length and
:at 97, so moving it along its parent is one write to :at. :time :in is gone
(it was the span's start written twice); node/placed-span maps an own-time
span out to the parent for playback, the mixer and the timeline rows, and
node/problems reports a stale :in.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Everything that holds nodes is a symbol (domain/timeline -> domain/symbol,
:timelines -> :symbols) and a node that places one is :kind :instance. The
reserved :main root is gone: which symbol is on screen is editor state
([:ui :open]), every domain function that needs a symbol is told which, and
a document opens on the longest symbol nothing else places.
Saved projects move to schema 2 through migration 0007, which rewrites leaf
paths, instance kinds and the feature :symbol key; the client refuses a
schema it does not read.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Detection now walks a browser-seekable H.264 proxy in MediaPipe's VIDEO
running mode. The PNG sequence it replaces was 112MB for 7.6 seconds at
1440x1920 and 1.1GB at the 900-frame limit; the proxy is 6MB, and landmarks
detected off decoded H.264 rather than off the PNGs moved at most 0.0033 of
frame width.
Three things had to be true for video mode to work, and each was measured
against the same footage decoded to PNGs:
/blob/<digest> answers byte ranges. Django's FileResponse does no Range
handling, and a media element handed 200 with no Accept-Ranges reports an
empty `seekable`, no-ops every currentTime write, and detects frame one
ninety times without raising.
A seek aims at the MIDDLE of its frame. Aiming at i/fps sits on a frame
boundary and landed one frame early 31 times in 91; (i + 0.5)/fps was exact
on all 91.
Timestamps are strictly increasing footage milliseconds. Video mode is a
tracker: a repeat leaves the graph in an error state every later call
re-throws, so the landmarker is discarded on failure, and passing the frame
index instead of i*1000/fps moved landmarks six times further from the
per-frame answer.
Frames are verified rather than trusted. requestVideoFrameCallback states
which frame it handed over, the walker discards any other and fails loudly
if the one it asked for never arrives — a stale presentation from the tail
of a previous seek is what produced "asked for frame 1 and it presented
frame 2" on a video whose seeks were in fact exact.
The proxy is re-encoded even when the upload is already H.264: HEVC is not
decodable everywhere, and footage identity is the proxy's digest. The JPEG
stills beside it are tracing references, outside the footage digest because
re-rendering them at another size is not different footage.
Verified end to end in a real browser against real footage: 228/228 frames
detected, a drawn roto face, 37 backend and 234 frontend tests green.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
There were two answers in the tree to "which stored bytes stop being valid when
this knob moves", and only one of them was checked.
`flow/address/block-knobs` is per block and asserted by biconditional —
`address-test` re-freezes the take once per knob and requires that the bytes
changed if and only if the key did. `domain/params`'s `:affects` was per area,
had no caller but a test asserting it returned what it was written as, and was
already wrong in both directions on the one entry where the two granularities
disagree: `:aperture-cut` claimed `#{:mouth}`, where it reaches no block, and
omitted the teeth, whose contour bytes it genuinely moves by gating
`condition/interior`'s smoothing. `:blink-cut` claimed `#{:eye}` and reaches no
block either, because a blink is `[:vis]` keys in tier 1.
So `:affects` and `affected-areas` are gone, and `address/knob-roles` is the
derived inverse of the table that is asserted — which is what a parameter panel
actually wants to ask. A knob absent from it invalidates no block, and that is
an answer rather than a gap.
Two new assertions keep the derivation from rotting at either edge: every role
in the table is reachable from some knob, and every knob a block declares is one
the registry defines. The second closes a real hole — `block-descriptor` checks
only that a knob was PASSED, and the freeze's `merge take/knobs` makes that true
of anything spelled like a keyword, so a typo in `block-knobs` would have named a
setting no slider can move.
228 CLJS tests, green.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Step 9. The tier split was the work; Django was the easy half.
Tier 1 — the authored scene — is the document, and it is addressed as
independently versioned leaves rather than saved whole, so one vertex drag
cannot clobber a collaborator's keying. `domain/leaf` is the document as
path -> value; `domain/wire` puts it on the wire as transit, because JSON
has neither integer map keys nor keywords and a save would quietly turn
`{0 v}` into `{"0" v}`.
Tier 2 — the dense channel blocks — is content-addressed by a hash over
every input, with the detector version inside every key through the
analysis the block descriptor names. `flow/address`'s `block-knobs` is the
invalidation table, and `address-test` does not trust it: it re-freezes the
take once per knob and asserts the biconditional, that a block's bytes
changed if and only if its key changed. That found `brow-pos` not depending
on `contour-avg` — the brow ring is smoothed, the raise is not.
Tier 3 — frames and audio — is served by the hash of its bytes out of the
same store. A manifest now names frames and carries a URL for each, so the
frame layout stopped being a shared secret between a shell script and a
ClojureScript namespace, and the `?v=` cache-buster went with it: a blob's
name is the hash of its contents, so a stale copy is not a thing that can
happen. The synthetic take's `audio.wav` moved to `static/arthur/` — an
asset the project owns, not an extraction that churns.
The server verifies rather than trusting a name it was handed: it
recomputes every key from the descriptor stored beside it, refuses an
analysis that declares no detector version, and refuses a document naming
blocks it does not hold. It hashes the descriptor TEXT, because JS prints
an integral double as `1` and Python as `1.0`, and a scheme where both ends
re-render the numbers disagrees on the first parameter that happens to be
whole.
Two loose ends from step 8 closed on the way. `pack` no longer takes a
`(track, frame)` predicate whose call sites each re-derived a feature from
an index — every track names the feature it follows, which deleted five
hand-maintained mappings. And `:dev-http` is gone: Django serves the page,
shadow-cljs only builds into the staticfiles tree.
227 CLJS tests, 31 Django tests, green.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Step 8's data model, ahead of its controls. Nothing here is a UI.
domain/params holds every knob's definition once — default, applicable area,
value constraints and the areas a change would force to regenerate. flow/take's
literal knob map becomes a view of it, so the take's defaults and the future
parameter panel cannot drift apart.
domain/feature adds subjects, features and groups as document data the renderer
never reads. A feature ID is stable for the whole clip, across occlusion: a run
of visible frames is not a new identity. An eye pair is an explicit group of one
or two eyes of the same subject, so a profile view with one identified eye needs
no invented partner. Settings resolve area -> subject -> group -> feature, and
dropping an eye from a pair materialises its effective values first so playback
does not jump. scene/problems now validates all of it.
Presence becomes per-feature rather than per-subject. freeze's :absent predicate
takes a track as well as a frame, so one occluded eye can be absent while its
partner still has a value; a full-face miss still marks everything absent. A
manifest may annotate known gaps as one-based inclusive intervals, which ingest
expands into observation tracks before measurement. An unobserved eye then gets
no vote in the iris pairing and cannot steer the shared gaze — gaze falls back to
whichever eye is visible. Temporal filters still see a sample on every frame,
held from the last observed one, because the numbers are a rectangular buffer;
the state mask, not the buffer, is what says the frame has no value.
js/app.js gets the same occlusion lesson: leading nulls from a face that starts
occluded used to throw away the whole take, and the neutral frame could be chosen
from a held duplicate pose.
Parameter editing, scoped regeneration and a feature-level detector remain. Until
one exists, footage without annotations falls back to the full-face mask rather
than claiming occlusions it cannot see.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01B87NVmiU36qQmN9gmFYnJ9
`stabilize` is three things wearing one name, and it is now three functions in two
stages: `flow/measure/anchor` fits the rigid transform, `flow/condition` smooths
its parameters, `flow/measure/mouth` measures the lip rings through the result.
Parity is on the COMPOSITION and not on the pieces -- a split that agreed
function by function and not end to end would be a split rather than a port.
The oracle now drives `stabilize` at three configurations and the port agrees to
1e-9 on ref, rigid, transforms, outer, inner and aperture, plus `smoothContours`
at three radii. Two of the three configurations are at aspect 0.5625, a 1080x1920
phone clip, because at aspect 1 `pick` is the identity: a port that dropped the
anisotropy correction outright would pass every other assertion in the suite.
148 tests, up from 134.
Three decisions worth the reading time.
`makeXform` is not ported, and its absence takes the face oval with it. It
centres on the oval's bounding box and zooms until the face is 80% of the raster
height, so every vertex it touched carried a cropping decision made once, at
analysis time, from one frame's landmarks. Geometry belongs in the node's own
local space with the framing as a transform on a node, so this is a deletion. The
oval's only other consumer was the placeholder plate outline, which is painting.
The residual is taken against the RAW fit, and the prototype took it against the
smoothed one. That is the only deliberate numeric divergence here, and parity is
kept by asserting `anchor/residuals` on exactly what the prototype handed it. The
number's job is to say whether a section is stabilisable at all; folding the
smoothing error into it makes a slider look like a property of the footage, and
docs/architecture.md lists the residual under stage 3, which requires it to be
knob-free. `condition/anchor` therefore replaces `:transforms` and leaves
`:residual` alone.
The stage order is not the strict chain the table in docs/architecture.md looks
like, and that document now says so. The fit is knob-free, conditioning smooths
it, and the rings are measured *through* the conditioned transform -- so
`anchor avg` does re-run the ring mapping, which is a few hundred frames of twenty
points. The guarantee was only ever about the part that reads a source pixel, and
that part never sees a transform.
Two things fall out and are asserted rather than assumed. Smoothing and
subsampling commute, because both are per-slot, which is what lets `vertices`
stay a stage-5 knob downstream of a stage-4 one -- and it is also why the port can
smooth the full twenty slots where the prototype smooths eight and still match.
And `condition/contours` is `geom/moving-average` per vertex per axis rather than
its own clamped window, so "radius 2" cannot come to mean two different things at
the two knobs.
One dead end recorded so nobody walks it twice: the synth's head is perfectly
rigid -- its jitter is a whole-head translation, which a similarity absorbs
exactly -- so every frame's rigid configuration is congruent with frame zero's and
the Procrustes mean IS frame zero to 1e-15, jitter or none. "The reference is the
mean and not frame zero" cannot be asserted on this track and is asserted in
geom-test, where the two can differ.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
