A scope or detached window drew its traces and nothing else: no cursors, no gate band, no trigger
marker, no measurement marks. They are not per-window state to be recreated -- the cursors are one
pair of times, the gate one band, the trigger one instant -- so the main window's row set stays the
**owner** and every window is written from it.
`sync_scope_instruments()` runs in the main tick after the main's own `sync_cursors()`/
`sync_gate()`: a gesture inside a scope window is read back from *that* window's controllers first
(those two calls only adopt while their own plot holds the grab, which is exactly "the drag
happened here"), adopted by the owner, and then every window -- the dragged one included -- is
written from the owner with the same cursor pair, mode and visibility, gate bounds and visibility,
and trigger marker. `RowSet::set_cursors()` is the new pair setter; `trigger_marker_x()`/`_visible()`
are the getters the mirror reads. The measurement marks travel by signal name
(`RowSet::MeasureMark::signal`, filled by the measures pass), so each window finds the row that
plots the signal and draws its ruler or mark there -- or none, when it does not plot it. The search
marks are times on the window and are placed on every window by the same per-frame path, and the
crosshair-axis-labels preference is restated to open windows when it changes.
Probe: `[scope-probe] instruments windows=N cursors=N gate=N trigger=N marks=N (main trigger
on=..)`, after adding a real gated `Top` measurement on `sine` -- the gate is restated visible each
frame by `update_measures()` and marks come only from `show_on_plot` measurements, so the leg uses
the app's own path. Docs: README, the cursors invariant, and the probe readings.
A scope or detached window was never written to the session, and a detached figure vanished from
the layout at the next start -- the figure had been removed from the main rows, and nothing kept
the window that held it.
Each window is now a `[[window]]` block in the session -- `name`, `detached`, and its own
`[[window.plot]]` rows with the same keys the main `[[plot]]` rows carry (signals, row weight,
cell width). `collect_layout()` writes the open windows' present arrangement plus the ones already
closed; loading puts **all** of them in `App::closed_windows` and never opens one: a window is a
view the user closed, and a launch that tiled windows onto the desktop would take that decision
for them (a tiling WM would tile them straight away). The overflow menu gains **Reopen closed
windows (N)**, which opens every remembered window through `open_scope_window_from_state()` -- the
one window constructor, then its rows, weights and signals dealt into its row set.
Closing a window records it, by name (a close/reopen cycle replaces its entry rather than
multiplying them), and schedules the save; **attaching** records nothing, because that window's
figure is on the main window again. A detached figure therefore survives a quit through its
window's entry.
Probe: `[scope-probe] loaded closed=N scopes=0` (remembered, none opened) and `reopen
scopes=..->.. remembered=.. left=0` (close, remember, the menu item's own click reopens).
Docs: README, the Mallard layout page, the probe readings and the session invariant.
A detached window is a view with an arrangement of its own -- and there was no way to change it:
the layout editor was the main window's, and a second window's rows could not be added or removed.
Every scope window now carries its own editor: `L` (or the grid toggle in its corner, revealed by
the same proximity rule as the attach button) shows the main editor's `+` / `×` / Done controls,
and they act on **that window's** row set (`set_scope_layout_editing()`, every button tagged with
its window), so a window holds as many figures as the user gives it and the two editors cannot
move each other's rows. Nothing a scope window arranges is written to the main window's session:
the windows themselves are not persisted.
The keyboard: each scope window gets a `GtkEventControllerKey` running the application's one
accelerator table -- `L` acts on the window the key arrived in, every other action reaches the
application through `run_app_action()`. The plot's own keys ('f', the arrows, Escape) already
worked there, through its plot controller.
The window struct now always builds its overlay (the corner controls' host); the attach button
stays detached-only, and both corner toggles step aside while the editor is up (Done is the way
out).
Probe: `[scope-probe] layout buttons=3 rows=1->2` -- the toggle builds 2*rows+1 buttons and a
boundary `+` adds a row. `L` in a real scope window verified by OCR ("Done" appears, the row's
`×` at its left). Docs: README, the Mallard layout page, the probe readings, and the layout
invariant (a view's arrangement is the view's).
**The blank scope window in trigger mode.** A held event pins the main window to
`[-pre, +post]`, and the tick mirrors that pin onto every scope window. It passed
`pinned_range().min/max` to `pin_event_window()`, which takes **spans** (`pre, post`): `min` is
signed, so `pre` came out negative and the group range became `[+pre, +post]` — an inverted x
window. The plotter then had nothing to draw: no trace, no x-axis labels, an empty plot that
looked like a broken window. `RowSet::pin_display_range(x0, x1)` states the pin as the range a
mirroring view actually has, and the tick uses it; live mode was never affected, which is why the
defect only showed under the trigger.
Also restores the scope-window `rebuild_if_dirty()` in the tick, which a reproduction experiment
had left commented out (`REPRO-TEMP`) and a later commit carried: a window opened empty and filled
by a drop kept the figure it was built with.
Probe: `[scope-probe] pin_mirror=1 main=[..] scope=[..] x=[..]` pins the main window, reads the
scope's pin and plot x range a frame later, and reads `0` for an inverted or mismatched pin.
Docs: probe readings and the trigger/pinning invariant (range vs spans).
A second toplevel's frame clock is not the main window's: the tick binds, queues a render and
mirrors the span, but with nothing asking GTK for frames on *that* window the area can sit with
its queue full and present nothing -- a blank plot. Every scope window now carries a tick
callback on its own area (`gtk_gl_area_queue_render` per frame), which is the request that keeps
its frame clock running.
The "cannot paint" diagnosis now names the window's GSK renderer (`[GSK renderer: GskVulkanRenderer]`)
and says to restart with `GSK_RENDERER=gl` when it is the Vulkan one -- the interaction between
the Vulkan GSK renderer and an embedded `GtkGLArea` in a second toplevel is a known trouble spot,
and a tiling WM has no title bar to carry a title, so the sentence goes to stderr and the status
bar instead.
The scope menu carried a y fit only: with one shared time span, the x fit *is* the main window's
fit, and the next frame mirrors the result back -- so `Fit X axis` and `Fit all` now sit in the
menu beside the y entries and reset the shared range.
And a blank plot becomes a sentence: a scope window whose GL context failed, or whose area has
painted no frame two seconds after opening, writes the reason into its own title and to stderr
(`[scope] window 'plot 1' is not drawing: GL context failed: ...`). A blank rectangle is
indistinguishable from a failed context or an area that never got a frame, and the report "the
plot is not there, no x axis" is exactly what this answers: if the title stays unchanged, the
area is painting and the question moves to the compositor presenting the second window.
A scope window is a view with its own arrangement, but its plot had no menu: every right-click
went nowhere, because the main window's row menu acts on the main window's rows.
`on_scope_right_click()` wires each scope window's GL area (beside its drop target) to a menu
that acts on **that window's** row set: y axis auto scale, y axis fixed range (the main menu's
dialog, aimed at the window's rows through `FixedRangeContext::rows`), fit, and one
*Remove '<signal>'* per trace on the plot. A detached window's menu also carries **Attach back
to the main window**. What it deliberately does not carry is an x fit: the window follows the
main window's time span, so an x fit here would be undone by the next frame's mirror.
`remove_trace()` and `FixedRangeContext` are generalized to take the row set they act on (the
main window passes its own). Probe: `[scope-probe] rclick=1` asserts the click gesture is wired
per window; the menu build is exercised by the real right-clicks the smoke makes.
`[scope-probe]` grows the two readings a "blank scope window" report needs before it can be
blamed: `empty_drop=`/`empty_slot0=` (the same drop onto the empty window, the user's own
path) and `gl copy=/gl empty=/gl detached ok= frames=` -- each window's adapter context
status and staged-frame count. A scope whose context failed draws nothing at all: no axes,
no traces, exactly the "blank plot widget, no x axis" report -- and that is a driver/context
finding, not a layout one. The detached window is left open by the probe, so a run's
screenshots can answer "does the detached figure draw" with pixels.
Beside each figure row's full-screen toggle -- same corner, same show-on-approach reveal, same
one-figure/never-maximized rules -- a **detach** button moves the figure into a window of its
own: copied into a new scope window (titled "detached") and removed from the main window's rows,
which then save the change. The detached window is the one scope window with an **attach**
control: an overlay wrapping its GL area, a button in the top-right corner revealed by the same
pointer-proximity rule from the overlay's own motion controller, and one click appends the
figure back to the main window's rows (cells, weights, every slot's signal in its colour) and
closes the window.
`detach_figure()` / `attach_scope_window()`; the reveal and placement share the maximize
buttons' geometry and pointer state, per row. A detached window's rows participate everywhere a
scope window's do (drops, the signal page's *Plot on* groups), which the previous commit gave
them.
Probe: `[maximize-probe]` now counts the detach buttons (`detach_single=0` for one figure,
`detach=2` for two, absent while maximized) and reads their corner reveal
(`detach_near=1 detach_other=0`, the pointer parked on the detach button's own centre -- the
strictest zone of the pair). Real-pointer smoke: hover reveals the pair, clicking detach opens
the window, clicking attach closes it with the figure back. Docs: README, the Mallard layout
page, the probe readings, the layout invariant, and the tour's plot stop.
The second windows were read-only by accident: nothing accepted a drop on them, and the
signal page's *Plot on* buttons knew only the main window's rows.
- **Drop target per scope window**: the same string payload a badge drags (the signal index),
handled by `on_scope_drop_channel()` -- the target carries its own window, the row is that
window's cell under the pointer, and `assign_signal_to_scope()` lands the signal in the
window's own row set (enabling an unacquired signal first, exactly as the main drop does).
The rebuild and the area's re-point are the window's next frame in the tick: a drop arrives
from inside the very widget it rearranges, and a synchronous rebuild would reset the drag
controller under it.
- **Plot on, per window**: the signal page carries one named group per scope window ("In
\"plot 2\""), one button per plot of that window, sensitive on that window's own rows. A
window closing refreshes the pages in its destroy handler, before the deferred delete could
dangle a button's scope pointer.
- Probe: `[scope-probe]` drops through the wired handler (`drop=1 slot1='square'`) and counts
the page's groups (`page_groups=2`); the section documents that it must not run with
`UDPSCOPE_ENABLE_ALL_MS` -- the enable-all probe's second leg clears the front plot, which
is the slot this probe copies. README, help and the layout invariant follow.
The slide-over's header already prints the open page's title beside its Close button, so the
signal page's body carried the name twice and a "Back to signals" button that only repeated
navigation the header owns. Both are gone.
The header's title is now the **numbered** name -- `<n>. <name>`, `measure_channel_number()`'s
number, the same one the signal badges and the measurements pane use -- so the panel says
*which acquired channel* it is and not only what it is called. `slideover_set_title()`
retitles a page in place (the open one included): the number follows the stream's channel
numbering, and the page is refilled whenever the model's signal list changes.
Probe: `[signalpage]` reads `title='1. sine' title_ok=1` (the header text against the numbered
name) and `no_dup=1` (no "Back to signals" button and no `title-3` label in the body); the old
`back_to_list` reading is gone with the button. Help, README and the probe doc follow.
A batch of view and instrument work, with the probes and docs for each.
- **Full screen per figure row**: a toggle on each row's top-right corner, revealed by
pointer proximity (`kMaximizeRevealPx` around its placed rect, per row, read from an
overlay-level motion controller so a revealed button under the pointer cannot flicker);
the plot menu offers the same as *Full screen this figure* / *Restore the layout*. The
arrangement is saved and restored through `RowState`, so nothing plotted moves. A layout
of one figure carries no toggle.
- **Second window**: `open_scope_window()` copies a figure (cells, width shares, signals in
their colours) or opens an empty one, with its own `RowSet` and `PlotArea`; the main
window's tick binds, acquires and repaints it, mirroring the time window and pin. Entries:
the plot menu's *Open this plot in a new window* and the overflow menu's *New scope window*.
- **Layout editor from the plot**: the plot menu gains *Edit the layout...* / *Done editing
the layout*, flipping the app's one toggle.
- **Measurements gate**: `MeasureOptions::gate` restricts a measurement to the held/pause
snapshot (live falls back to the live ring), persisted with the session and shown in the
measure dialog.
- **Crosshair axis badges**: a preference (`crosshair_axis_labels`, on by default) and the
plotter's badges where the lines meet the axes, in the axes' own formats; the plot menu
and the probe read them.
- **Event cache per configuration**: `EventStore` is keyed `<source>/<configuration>/`,
`retarget_event_store()` follows every configuration switch, and a pre-configuration cache
(loose files) is adopted into the configuration's folder on first start.
- **GTK4 lifecycle**: the main window's handlers move from the non-existent `"destroy"`
signal to `close-request`; scope windows are closed with it and the application quits
explicitly (the exit run's hang). The exit probe closes with `gtk_window_close()`.
- Probes: `[maximize-probe]` (single-figure rule and the corner reveal), `[scope-probe]`,
`[split-probe]` cell crosshair leg, `[prefs-probe]` store keying, `[configs]` store key,
`[rowmenu]` layout-editor entry; docs, help pages, README and AGENT.md updated.
Also fixes a probe-only use-after-free: `toggle_maximize_figure_row` replaces the figure
object, so the area must be re-pointed (the button's path does; the probe's raw call did not).
Add cell `+` buttons (n+1 per figure row), column resize handles with
drag, and update the `x` handler to call remove_cell instead of
remove_row. Fix row_weights_ misalignment in insert_row/remove_row.
Split probe steps 9-15 exercise content movement, column drag clamp,
row-weight tracking, and sensitivity at caps.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The dock button's tooltip, the guided-tour menu row's tooltip and the m / g shortcut labels still described a side panel and tabs.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The dock's Events block, the events slide-over page, the availability-rule deviation, the probe lines and the removal of the left pane are described; the false remembered-width claim is gone.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The dock tour stop names the Events block (one click replays, Manage opens the list) and the event-history and record hints point at the page and the dock instead of a left pane.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
The dock tour stop describes the read-only readouts and their Edit buttons, and the cursors and measures hints point to the dock copy and the pages.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Both views are written from the same variable at the same call site, so the dock cannot disagree with the full panels.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Converts every --buffer-mb / buffer_mb mention to --buffer-samples / buffer_samples,
records the unit change as a deviation (README, SPECS, CLAUDE.md), documents the
buffer: and planner: dump lines and the new Preferences row, and updates the help
pages and the help-shot script.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
--buffer-samples (k/M/G) replaces --buffer-mb: the request is rounded down to a
power of two for every signal and the shared time ring. A separate memory limit
(--memory-limit-mb, a Preferences row, default 8 GiB) halves the plan until the
real ring bytes fit and says so on stderr and in the status line. The per-source
key buffer_mb becomes buffer_samples; an old buffer_mb is ignored.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
A Source pill that opens nothing (UDPS, Flat UDP, files) was noise. It is now shown only while the slide-over has a source page; the [badges] source line gains visible=.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The link state is the status bar's, so the Source badge loses its state text. The signals have no lock
(the badges and signal pages are direct manipulation), and the signals page lists only unacquired
signals and uncomputed math channels (plus the acquired ones while the badge bar has overflowed).
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Also wrap the signals page's hint line, which at 375 px minimum widened the whole slide-over panel to 376.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The per-signal page bypassed the signals lock for the acquire tick and the math
Edit/Delete buttons, and offered "Plot on r" for a plot the signal was already on.
A removed page also let another signal's page show during the close animation.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The session old-file probe and save_scheduled now discriminate, Esc is read through the real key path, the dock-section probe proves the Measure section starts out of view, and the tour's spotlight is measured independently of the helper it is drawn from. A target scrolled out of its scroller no longer leaves the previous hole standing.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The two drawn tab glyphs were judged not understandable enough; the
switcher goes to the wide policy with titles only, and the glyphs,
their gresource entries and the icon-naming code are removed.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
