martino/udpscope
1
0 Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

scope: full-screen figures, second windows, gated measurements, per-configuration events

Browse Source 
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).
This commit is contained in:
Martino Ferrari committed 2026-11-05 17:58:59 +01:00
1 parent bfe13630f9
commit bdf0ff7ddc
39 files changed
+2267 -136
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

docs/claude/probe-readings.md
+155 -23
View File
@@ -305,8 +305,8 @@ Two lines of the self-check output are assertions in disguise:
on screen. Both ranges are on the **stream's** clock, which is the space `measure_window` and
the time marks are stored in -- a live axis is a fixed `[-window, 0]`, so comparing display x
would read no drift whatever the data did. `win` is the window as of the last **tick**
(`App::measure_view_window`) and is deliberately not read back here: the dump runs from the
destroy handler, by which point the exit probe has switched the trigger on, and
(`App::measure_view_window`) and is deliberately not read back here: the dump runs after
`g_application_run` returns, by which point the exit probe has switched the trigger on, and
`stream_at_x0()` then answers for a mode no frame was ever drawn in -- which read as a 45000 %
drift on a run whose every frame was exact. `pin` is the same tick's pin (`App::measure_pin_window`,
`off` when nothing was pinned) and it is the **target** while the view is pinned: a measurement
@@ -1348,6 +1348,125 @@ mock's five signals and dies with `OSError: [Errno 90] Message too long`, so the
two steps to it, which is visible in its `ges/x1`, `ges/x2` lines. A `gave up waiting after N
polls` line means no latch (or no pin) was ever reached, and is a reading about the stream.
## `[gate-probe]`
`UDPSCOPE_PROBE_GATE_MS`. The reading for **what the per-measurement gate does**: the gated
measurement reads the shared band, the ungated one the whole window, and the band is
*relative* — fixed on screen while the data flows under it, live or held. Measured output
(port 8140, `--rate 100000 --batch 100 --freq 1`):
```
[gate-probe] on: gate=[-0.300000..-0.100000] view=[-2.000000..0.000000] s0=2.022990 for=[0.022990..2.022990] band=1 placed=1 plain=1 split=1 gwin=1 ungated=-1.36e-17 V gated=355.31 mV
[gate-probe] still: gate=[-0.300000..-0.100000] s0=[2.022990->2.852990] still=1 s0_moved=1
```
- `gate=` is the band, in **display** x; `view=` the axis; `s0=` the stream time of the
display's x = 0 (`stream_at_x0()`); `for=` the pane's own window (the ungated one), in
stream seconds. The probe measures only once the ring is older than the window (`s0 >= 2`
here): the `plain=` assertion is exact zero because a whole-seconds window of the 1 Hz sine
holds whole periods, and a young ring would clip it to a fraction of a cycle.
- `band=1` — one gated measurement was enough to show the band, because the band is shared by
every gated measurement and shown exactly while one exists. `placed=1` — the row's plot
region holds exactly the bounds the `RowSet` does: the drawing and the state are one pair of
numbers, not two.
- `plain=1` — the ungated average is zero (`-1.36e-17 V`): it read the whole window.
`split=1` — the gated average differs (`355.31 mV`, wherever in its period the stream
happened to be): it read the band. The pair is the per-measurement split on one signal.
- `gwin=1` — the pass booked `App::measure_gate_window` as the band moved onto the stream's
clock, within 5 % of the band (the probe's `s0` read and the pass's can straddle one
production interval). That is the phase-free statement of *which* slice the gated walk
read; the slice's mean value itself depends on where in its period the stream is, which is
why the assertion is on the window and not on the value.
- The second line is the **relative** assertion, one second later: `still=1` — the band
exactly where it was put — and `s0_moved=1` — the stream's clock moved under it.
`still=0` is a gesture echo adopted outside a drag; `s0_moved=0` on a live stream means the
display epoch stopped following, which is the unbound-trace state the probe's assignment
exists to prevent.
- The exit dump carries the session leg: `gate: on=1 bounds=[-0.300000..-0.100000]`, and a
second run on the same config with the probe unset restores both measurements with their
gate flags (`measurements: 2 1.sine average=-0.000005 1.sine average=-0.338823` on one
restore run -- numbered with the signal's active-channel number, the same one the pane's
rows carry) and the same band.
## `[crosshair-probe]`
`UDPSCOPE_PROBE_CROSSHAIR_MS`. The reading for **the crosshair's axis badges**: a real pointer
hover — synthesised through the GTK adapter's own controller, a plain move with no button —
must place the crosshair, and with it the two badges the plot writes in the label strips: the
y value in the y strip, the x time in the x strip, each inside its strip and outside the axes.
Measured output (viewport pixels; the absolute numbers are the headless window's own):
```
[crosshair-probe] active=1 y=[74x47@31,588] x=[59x47@749,1529] y_ok=1 x_ok=1 y_centre=1
[crosshair-probe] off: plot_flag=0
```
- `active=1` is the hover having reached the plot (`Crosshair::active`) — the adapter's
viewport conversion is what this probe exercises; the pixels themselves are the plotter
suite's (`crosshair_ink_stays_inside_the_axes`, which asserts the badges' rects against the
frame and pins the off leg too).
- `y_ok=1 x_ok=1` are the two placements: the y badge right-aligned inside the y label strip
(its right edge on the axis line, where the y ticks end), the x badge inside the x label
strip below the axes. Zero-sized badges, or one outside its strip, is the defect.
- `y_centre=1` is the y badge **centred on its line**: its middle row is the pointer's screen
y (the probe injected one). `0` is the top-aligned badge — the line running along the
badge's top edge, reading as if the number belonged to the tick below it — which is half a
badge off, well beyond the one pixel of slack.
- The second line flips the flag the preference drives (`set_crosshair_axis_labels(false)`)
and reads the plot's own flag back: `plot_flag=0` — a preference that reaches the file but
not the plots is the failure this leg names.
- The pixel-level assertions live in the plotter's suite; this probe is the adapter's half.
The preference itself is `[prefs-probe]`'s `xhair=` field, which reads the settings file
back from disk after the dialog toggles it.
## `[maximize-probe]`
`UDPSCOPE_PROBE_MAXIMIZE_MS`. The one-click **full-screen row** toggle, exercised through the
`RowSet` call the button's handler makes (a probe cannot click an overlay button the frame
places), arranged as two single-cell rows with one signal each. The two reads after it are the
**corner reveal**: no pointer over the area, then a pointer position in the first row's corner
(set directly -- the probe owns the state the motion controller fills). Measured output:
```
[maximize-probe] single_buttons=0 before: rows=2 frows=2 row0=sine row1=square buttons=2; max: rows=1 frows=1 flag=1 row0=sine; back: rows=2 row0=sine row1=square flag=0
[maximize-probe] corner: hidden_when_off=1 shown_when_near=1 other_row=0
```
- `single_buttons=` is the toggle count *before* the second row is added, while the layout is
the default one figure: **0** — one figure has nothing to enlarge *to*, so it carries no
button (and `buttons=` is what the same count reads once the second row exists).
- `before:` is the arrangement as set: two plots, two figure rows, the signals on their rows,
and `buttons=` one overlay toggle per figure row (the toggles are on screen whether or not
the layout editor is up).
- `max:` is the toggle: one plot, one figure row, `flag=1`, and the *maximized row's own*
signal (`row0=sine`) — the row that was asked for, not whatever sat first.
- `back:` is the toggle again: the arrangement, the plot count and both rows' signals exactly
as before, `flag=0`. A signal that changed here is the state save/restore (`RowState` in
`plot_host.*`) losing a field; the restore is a view change, never an edit.
- `corner:` is the proximity reveal: `hidden_when_off=1` (no pointer in the area, so nothing
shows), `shown_when_near=1` (a pointer in the first row's top-right corner brings *that*
row's toggle out), and `other_row=0` -- the second row's toggle stays hidden, because the
zone is per row. A `shown_when_near=0` names a dead reveal path (the motion hook or the zone);
`other_row=1` a zone that leaked across rows.
## `[scope-probe]`
`UDPSCOPE_PROBE_SCOPE_MS`. The **second window**: `open_scope_window()` twice — once copying
the main window's figure 0, once empty — and read back a few frames later (the main tick is the
scope's bind/acquire/repaint pump). Measured output:
```
[scope-probe] count=2 copied='sine' bound=1 empty_ok=1 windows=1
```
- `count=2` is both windows having been created and retained; `windows=1` each being a visible
top-level window.
- `copied='sine'` is the copy's first slot holding the main window's figure-0 signal, and
`bound=1` the tick having resolved it to a live channel — the scope is bound to the same
model, not a snapshot taken at creation.
- `empty_ok=1` is the `-1` call producing an empty row, which is the other entry point the
overflow menu uses.
## `[search-probe]`
`UDPSCOPE_PROBE_SEARCH_MS`. The handler calls `condition_probe()` directly — the condition is
@@ -1432,10 +1551,10 @@ popover row reads `failed: <strerror>` — one stderr line, no retry storm again
```
[defaults] calibrate: 1 0 model=1 0 | trigger: 0.001 0.002 0 0.001 0 50 0.01 0 0 1
[defaults] calibrated: 2.5 -1.25 model=2.5 -1.25
[configs] before=default active=default
[configs] save-as=1 names=probe,default active=probe
[configs] before=default active=default store=default
[configs] save-as=1 names=probe,default active=probe store=probe
[configs] round trip read=1 comment='written by the configs probe' buffer_samples=10000000 relative=0 capture=0 plots=1 signals=4 measures=0 math=0 noise=-1 hf=-1
[configs] after forget names=default
[configs] after forget names=default active=default store=default
```
`noise=` and `hf=` are the first measurement's band and corner as they came back from the file. The probe
sets them to **7.5 and 250** before saving, so a key that never reached the file reads back as the defaults
@@ -1450,16 +1569,17 @@ sets them to **7.5 and 250** before saving, so a key that never reached the file
| `[configs] names` | gains `probe`, and loses it after `forget` | a save that writes somewhere the listing does not read, or a delete that leaves the file |
| `[configs] active` | `probe` after the save-as | the `active` file not written, so the next start opens the wrong configuration |
| `[configs] round trip` | `read=1`, the note verbatim, `buffer_samples` non-zero | the note or the startup settings not persisted — a configuration that carries a layout but not the buffer size it was tuned with |
| `[configs] store` | the **event cache's configuration key**, following the active one: `default` → `probe` → `default` | a cache still keyed on the source alone (the key never moves), or a switch that renames the session but leaves the events list showing the configuration it left: `store=default` on the `save-as` line, or `store=probe` after `forget`, names it |
## `[rowmenu]`
```
[rowmenu] main 4 'Remove signal' h=18 label_x=26 accel_right=-1/174 ok=1
[rowmenu] main rows=5 bad=0 labels=[Y axis: auto scale,Y axis: fixed range...,Fit X axis,Fit all,Remove signal]
[rowmenu] remove 0 'Back' h=18 label_x=52 accel_right=-1/178 ok=1
[rowmenu] remove rows=6 bad=0 labels=[Back,sine,square,ramp,noise,All traces on this plot]
[rowmenu] summary main_rows=5 main_bad=0 remove_rows=6 remove_bad=0 after_back_rows=5 after_back_bad=0 second_rows=6 second_bad=0 remove_row_in_main=1 remove_row_gone=1 can_open_again=1
[rowmenu] remove-clicked 'sine' trace_freed=1
[rowmenu] main 6 'Remove signal' h=18 label_x=26 accel_right=-1/248 ok=1
[rowmenu] main rows=7 bad=0 labels=[Y axis: auto scale,Y axis: fixed range...,Fit X axis,Fit all,Open this plot in a new window,Edit the layout...,Remove signal]
[rowmenu] remove 0 'Back' h=18 label_x=52 accel_right=-1/118 ok=1
[rowmenu] remove rows=1 bad=0 labels=[Back]
[rowmenu] summary main_rows=7 main_bad=0 remove_rows=1 remove_bad=0 after_back_rows=7 after_back_bad=0 second_rows=1 second_bad=0 remove_row_in_main=1 remove_row_gone=1 can_open_again=1
[rowmenu] remove-clicked '' trace_freed=1
```
| field | value it must read | the defect a wrong reading names |
@@ -1467,6 +1587,7 @@ sets them to **7.5 and 250** before saving, so a key that never reached the file
| `h` | 18 for every row of both views | the plot menu reverting to Adwaita's 34 px buttons — the row style not applied (`popover.udpscope-menu`) |
| `label_x` | 26 for every row, 52 for **Back** | labels not sharing one column; the Back row's 52 is its leading icon, which every menu indents for |
| `accel_right`/`W` | equal when the row has a shortcut (`Fit all` → `F`), `-1` when it has none | the shortcut not flush right, or shown where the action has none |
| `main_rows` | 6 for a one-figure layout, **7** with `Full screen this figure` (a layout of two or more figures), plus 1 with `Restore the layout` while one is maximized | the full-screen entry missing where it would act (it is deliberately absent for a single figure: there is nothing to enlarge to), or the layout-editor entry missing: the plot menu is the third route into the editor beside `l` and the overflow menu |
| `remove_rows` | 1 per occupied trace slot, plus `All traces on this plot` when there is more than one, plus the **Back** row | a remove view built from the wrong row, or missing its way out |
| `remove_row_gone`, `remove_row_in_main` | 1 | a submenu that floats beside the menu instead of replacing it |
| `can_open_again` | 1 | the second click not reopening: the `GtkMenuButton`-owned submenu that collapsed and left its row dead |
@@ -1620,7 +1741,7 @@ run they report:
[events] after events page opened: available=1 dock_sensitive=1 page_sensitive=1 page=events
[events] open alone: page=events page_box=347
[events] preview redrawn: signals=3 stored=480x132 row_drawn=1
[events] dir=/home/u/.cache/udpscope/events/127.0.0.1-21023 n=3 newest: stem=20261013-184140-0003 dur=2.000000 pre=1.600000 post=0.400000 edge=5.660100 signals=3 samples=20001 rate=10000 seq=3 thumb_bytes=190134 recorded=1 page=1 plot_w_stable=1 plot_w=756/756
[events] dir=/home/u/.cache/udpscope/events/127.0.0.1-21023/default n=3 newest: stem=20261013-184140-0003 dur=2.000000 pre=1.600000 post=0.400000 edge=5.660100 signals=3 samples=20001 rate=10000 seq=3 thumb_bytes=190134 recorded=1 page=1 plot_w_stable=1 plot_w=756/756
[events] activated row0: replaying=1 page=events page_box=347 tick=0 window=2.000 axis=[-1.600,0.400] marker=0.000 thumbnail=1 thumb_size=138x46 thumb_visible=1 rows_kept=1 selected=2/2
[events] dock events: rows=3 pane_rows=3 same_titles=1 sensitive=1 dock_w=268 on_screen=1 marked=0 members=0 max_ok=1 expected_marked=1 at=after pane replay
[events] dock replay: clicked=1 replaying=1 waveform_changed=1
@@ -1857,9 +1978,9 @@ These ride on the `[cursors-probe]` readings (every state line gets a `dock curs
## `strip:` (the readouts strip)
```
[cursors-probe] <what> strip: visible=1 expected=1 visible_ok=1 dock_shown=1 cursors=4 measures=0 same_text=1 h=26 sens_same=1 collapsed=1 window=756 groups=100 groups_ok=1
[measures-probe] <what> strip: visible=1 expected=1 visible_ok=1 dock_shown=1 cursors=0 measures=12 same_text=1 h=26 sens_same=1 collapsed=1 window=756 groups=010 groups_ok=1
[measures-probe] start strip: visible=0 expected=0 visible_ok=1 dock_shown=1 cursors=0 measures=0 same_text=1 h=0 sens_same=1 collapsed=1 window=756 groups=000 groups_ok=1
[cursors-probe] <what> strip: visible=1 expected=1 visible_ok=1 dock_shown=1 cursors=4 measures=0 same_text=1 h=26 sens_same=1 collapsed=1 window=756 groups=100 groups_ok=1 add=1
[measures-probe] <what> strip: visible=1 expected=1 visible_ok=1 dock_shown=1 cursors=0 measures=12 same_text=1 h=26 sens_same=1 collapsed=1 window=756 groups=010 groups_ok=1 add=1
[measures-probe] start strip: visible=0 expected=0 visible_ok=1 dock_shown=1 cursors=0 measures=0 same_text=1 h=0 sens_same=1 collapsed=1 window=756 groups=000 groups_ok=1 add=1
[cursors-probe] dock hidden strip: visible=1 expected=1 visible_ok=1 dock_shown=0 cursors=4 measures=0 same_text=1 h=26 sens_same=1 ...
[cursors-probe] dock restored strip: visible=1 expected=1 visible_ok=1 dock_shown=1 ... collapsed=1 window=756
```
@@ -1868,7 +1989,7 @@ One `strip:` line follows every `dock cursors:` / `dock measures:` line, and eac
- **`visible` equals `expected`, and `expected = (!dock_shown || collapsed) && (cursors on || measures non-empty)`** (`dock_shown` is `show-sidebar`, `collapsed` the split view's, both read in the same frame as `visible`; "cursors on" is the cursors toggle, not the `cursors=` chip count, which is a rebuild's and is 1 even with them off). `visible_ok=1` is that equality; `visible_ok=0` is `sync_readouts_strip()` not run, or run before the reconciliation. A single such line right after the probe's own `update_side_visibility()` or on the very first line (the tick has not set `collapsed` yet) is the sub-frame before the next tick and not a defect.
- **What each leg reads.** `[measures-probe] start` (before the probe adds anything, from a **fresh** config: a stored session brings its measurements back) is the "neither" state: `visible=0 expected=0 groups=000 h=0`. The measures probe's later lines have measurements and the cursors off: `groups=010`. The cursors probe alone has the cursors on and no measurements: `groups=100`; run after the measures probe in the same session it has both: `groups=111`. A line with the strip visible and `groups=000`, or hidden with a group on, is `sync_readouts_strip()` not called from that path.
- **`groups=<cursors><measures><separator>`** is which of `strip_cursor_group`, `strip_measure_group` and `strip_sep` are visible; **`groups_ok=1`** is that they equal (cursors on, a measurement, both). `0` is a group left over from the previous state.
- **`groups=<cursors><measures><separator>`** is which of `strip_cursor_group`, `strip_measure_group` and `strip_sep` are visible; **`groups_ok=1`** is that they equal (cursors on, a measurement, both). `0` is a group left over from the previous state. **`add=1`** is the strip's `[+]` sitting after the last measurement chip (a button, not the chip box) -- the strip adds a measurement itself rather than only opening the pane; it is structural because a probe that clicked it would open a dialog on every report.
- **`h=0` with `visible=1` on the `cursors-probe ... on` and `measures-probe ... added` lines** is the strip shown in the same frame the probe reads it, before its first allocation; the next line reads 26. It is not a `visible_ok` miss: the toggle handler and `rebuild_measures_pane()` call `sync_readouts_strip()` themselves, so `visible` is already right.
- **`visible=1 expected=1` on every line that has something to read, under `xvfb-run`**, the `dock restored` line and the lines with `dock_shown=1` included: the window is 756 px, below the 900 px `kSidebarRoom`, so the split view is collapsed, the dock is an overlay on the plot, and the strip stays. `visible=0` there is the old rule (hidden whenever the dock is shown) or `collapsed` not being read. The wide case (`collapsed=0`, `dock_shown=1`, `visible=0 expected=0`) is **not reachable from a probe**, because a window with no manager cannot be resized by `gtk_window_set_default_size()`.
- **`h` is 26 with `visible=1`, 0 with `visible=0`.** Any other visible height is the chips setting the strip's height (the button trim in `ui_widgets.cpp` or the `height-request` lost).
@@ -2118,15 +2239,22 @@ same reason.
## `[prefs-probe]`
```
[prefs-probe] opened=1 rows=1 dir=/tmp/k5/cache/udpscope/events store=/tmp/k5/cache/udpscope/events/127.0.0.1-8000
[prefs-probe] dir=/tmp/udpscope-prefs-events keep=7 rel=1 theme=dark live=1
[prefs-probe] opened=1 rows=1 dir=/tmp/k5/cache/udpscope/events store=/tmp/k5/cache/udpscope/events/127.0.0.1-8000/default cfg=1 cfg_switch=1
[prefs-probe] dir=/tmp/udpscope-prefs-events keep=7 rel=1 theme=dark xhair=0 live=1
[prefs-probe] default restored: file=''
[prefs-probe] browse: cancelled=1 icon_ok=1 field='/tmp/k5/cache/udpscope/events'
```
`UDPSCOPE_PROBE_PREFS_MS` opens the preferences, changes every row, and reads the settings file *back
from disk* after each change -- not from `App::settings`, because a dialog that edits its own copy and
never writes is exactly the failure this has to catch. **A change applies as it is made**: the rows are
never writes is exactly the failure this has to catch.
`store=` is the directory the cache is **really** using, and the `cfg=`/`cfg_switch=` pair is that
directory's key: `cfg=1` reads `store=` ends `/<source>/<active configuration>/` (the cache is per
source *and* configuration, so two configurations of one source never list each other's events), and
`cfg_switch=1` reads `retarget_event_store()` -- the one helper every configuration switch calls --
moving the key to another configuration and back. A `cfg=0` names a store still keyed on the source
alone; a `cfg_switch=0` a switch that leaves the cache pointing at the configuration it left. **A change applies as it is made**: the rows are
libadwaita's (`AdwEntryRow`, `AdwSpinRow`, `AdwSwitchRow`, `AdwComboRow`), and the `preferences.ui`
dialog has no Apply button for the probe to press.
@@ -2138,8 +2266,10 @@ dialog has no Apply button for the probe to press.
plus the source's folder name**, which is the reading that says the row shows the path in force rather
than a plausible string.
- The second line must echo the probe's own input: directory `/tmp/udpscope-prefs-events`, keep `7`,
relative time `1`, theme `dark`. Any field that differs means the row, the write or the parse dropped
it. `live=1` is the theme having reached the running style manager rather than only the file.
relative time `1`, theme `dark`, and `xhair=0` — the crosshair's axis-badge switch, which the
probe drives off exactly as it drives the rest, and which reaches the running plots at once
(`RowSet::set_crosshair_axis_labels()`). Any field that differs means the row, the write or the
parse dropped it. `live=1` is the theme having reached the running style manager rather than only the file.
- `default restored: file=''` closes the loop from the other side: the probe puts the default path back
into the row, and the file must end up with **no** `event_cache_dir` at all. A path in the file is a
user's choice, and leaving the default is not one -- pinning it there would stop the setting following
@@ -2152,7 +2282,7 @@ dialog has no Apply button for the probe to press.
```
[prefs-probe] opened=1 apply=1 dir=/tmp/k5/cache/udpscope/events store=/tmp/k5/cache/udpscope/events/127.0.0.1-8000
[prefs-probe] opened=1 apply=1 dir=/tmp/k5/cache/udpscope/events store=/tmp/k5/cache/udpscope/events/127.0.0.1-8000/default cfg=1 cfg_switch=1
[prefs-probe] dir=/tmp/udpscope-prefs-events keep=7 rel=1 theme=dark live=1
[prefs-probe] default restored: file=''
[prefs-probe] browse: toplevels=3 extra=1 icon_ok=1 field='/tmp/k5/cache/udpscope/events'
@@ -2862,9 +2992,10 @@ inside the panel does not close it), `click_outside_closes=1` (a click on the pl
[split-probe] ruler-labels split=[row 1,row 2] mixed=[row 0,row 2]
[split-probe] shared-y y_shared=1 y_gutter=1 y_fit=1 ruler_mirror=1 ranges=[A=-3.000000..7.000000 B=-3.000000..7.000000 | fit A=-1.04..1.04 B=-1.04..1.04 data=-1..1] axes_w=[671.599976,671.599976] auto=1
[split-probe] x-axes row0_labelled=[1,0,1] bottom_labelled=[1,1,1] gap_ok=[1,1,1] gap_px=[3.0/3.0,3.0/3.0,3.0/3.0] col_gap_px=3.0 caption_ok=[1,1,1] last_plot_zoom=1
[split-probe] cell-xhair left=1 right=1 centre_dy=0.0 sibling=0 (the horizontal line spans a split row's shared y, and no other row)
```
`UDPSCOPE_PROBE_SPLIT_MS` (stderr; see [verification.md](verification.md)). 28 steps (0..27), 250 ms apart;
`UDPSCOPE_PROBE_SPLIT_MS` (stderr; see [verification.md](verification.md)). 31 steps (0..30), 250 ms apart;
each restructures and the *next* step reports the frame that was laid out with it, so a line's label
names the arrangement the line **reads**, not the call that made it. The state lines share one format:
`plots` (`row_count()`), `frows` (`figure_row_count()`), `cells=[..]` (per figure row), `widths=[..]`
@@ -2887,6 +3018,7 @@ The other lines, in order:
| `row+cell` | `cells=[2,2]`, `weights=[3,1\|-]` | row 0's cell weights lost or moved when a row and a cell were inserted below it (they belong to the figure row) |
| `caps-calls` | `insert_cell=[1,1,0] refused_cell=1 insert_row=[1,1,0] refused_row=1 plots=8` | a cap that is not reached or not held: row 0 starts with two cells and takes two more, refusing the third call at 4; two rows then fit before the plot total reaches 8 and the third is refused; `plots` must stop at **8** |
| `caps` | `plots=8 frows=4 cells=[4,2,1,1]`, `tile=1` | a ninth plot, or a fifth cell in a row |
| `cell-xhair` | `left=1 right=1 centre_dy=0.0 sibling=0` | the crosshair's horizontal line, placed through `Figure::set_crosshair()` on the **second** cell of a two-cell row: it must also be drawn in the first cell (`left=1`) at the same screen row (`centre_dy`) — the cells share one y, so the line is one value — and must not appear in the figure's other (unsplit) row (`sibling=0`). A `left=0` names a row whose cells each keep their own line; a `sibling=1` a y spread across rows that do not share one |
| `remove` | `plots=7 cells=[3,2,1,1]` | `remove_cell()` on a multi-cell row that removed the row, or that dropped a neighbour's traces |
| `roundtrip-io` | `wrote=1 read=1 plots_in_file=3` | a save that was refused, or a file that lost a `[[plot]]` table |
| `roundtrip` (state line) | `cells=[1,2] widths=[756\|524,229] weights=[-\|3,1]` and the three traces on `1`, `2a`, `2b` | the arrangement did not survive `collect_layout()` -> file -> `apply_layout()` |