`Series::gather()` stages x as f32 relative to `x0()` -- the start of the
*retained window* -- so the value's ULP was a fraction of data the view need
not even hold: 1.9 us over a 20 s ring. A readout quotes `x_at()` in exact
f64, so at a 10 us view across 700 px that put more than a hundred pixels
between where a sample is drawn and the x reported for it.
The origin is now a parameter, placed before `out_xy` so every call site is a
type error rather than a silent change of behaviour. `Renderer::draw_traces`
passes the visible range's minimum, which makes the error a fixed 2^-24 of the
span on screen -- 4e-5 of a pixel over a 700 px axis -- at every zoom.
`stage_xy()` loses its second origin as a consequence: the gather origin and
the draw origin are now always the same value, so the linear case hands the
gathered buffer straight to `add_polyline_xy()` with no per-point copy.
`Plot::crosshair_point()` names the query itself, the one origin that cannot
lose precision for a caller that might.
Guarded by `a_gather_origin_at_the_view_keeps_a_deep_zoom_on_the_sample_it_quotes`,
which measures the error in pixels both ways: under 0.05 px with the view's
minimum, over 10 px with `x0()`.
Breaking: Series::gather emits x relative to x0() and takes a GatherMode;
consumers reconstruct absolute x in f64 before mapping (series.hpp documents
the contract).
Bugs, each pinned by a fail-before regression test:
- render: f32 staging quantised epoch-time x into 128 px columns; every
data-space draw now stages to_space(x) - to_space(origin) with the origin
folded into the offset uniform in f64 (epoch/zero-origin twins rasterise
identically)
- ticks: major test was ULP-fragile and blanked epoch axes; majors are an
integer-index test and are never truncated by the minor cap
- gather: LOD envelope emitted buckets straddling [begin,end); ceil/floor
bucket edges + raw lead/trail spans
- live follow: log x axes held a linear span and zoomed every frame; span and
offset now live in axis space (plot + both figure link-group paths)
- input: a zero-notch wheel disarmed auto_scale on every axis; dy==0 pans x,
factor == 1 / inf refused in wheel and drag zoom
- render: the single text flush painted legend names over the crosshair
readout panel; two batches, under flushed before the panel
- data: NaN dropped from min/max unless it seeded the fold; non-finite
samples are now excluded from every aggregate and from cached bounds
Robustness: viewport normalisation (no clamp UB), non-finite range refusal,
pyramid level cap at 57 (no shift/div UB through public gather), rounded ring
capacity documented + pre-sized, GL blend func and UNPACK_ALIGNMENT
save/restore, Renderer::release_gl() teardown, u64 buffer epochs, annotation
captions clamped, button-aware GLFW row capture, cached figure row rects (no
per-event layout), ruler-pair measurement refused, span-nonfinite guards in
axis_ends.
Features: XY series (sorted x, binary-search neighbours/snap, LOD over sample
index); StreamSeries::freeze(); ArraySeries::push(); per-trace dash and
markers (scatter at width 0) with auto or literal stride; GatherMode
Envelope/Nearest/Mean via bucket_sum/sumsq; Style::line_aa_px; per-glyph text
colour (readout values in trace colour); wheel-dx x pan; arrow-key pan;
Escape cancels a drag and restores captured ranges; modifier-masked box zoom
(Shift both, Ctrl x, Alt y); double-click reset via PointerEvent::time_s;
ImGui key repeat unified; pinch gestures on Qt and GTK; pixelDelta wheel path
and DevicePixelRatioChange handling; GNUInstallDirs, SOVERSION,
SameMinorVersion, generated plot/version.hpp, relocatable plot.pc, licence
install, FILES_MATCHING, GitHub Actions CI (debug/coverage/asan).
Suite 283 -> 350 cases, 0 failed; coverage 93.9% (3730/3971) over 24 files,
floor 80/90 held; ASAN preset suite green; all four demos run headless (the
GTK scripted-resize check needs a window manager and fails identically on the
parent commit); readme_main.cpp committed and compiles with the README
command; pkg-config and find_package consumption verified from a relocated
install prefix.
The repository had no licence at all, so every source was all-rights-reserved
by default. Add the verbatim MIT text at LICENSE (Copyright (c) 2026 Martino
Ferrari) and a two-line SPDX header to each of the 113 first-party sources --
48 include/plot, 26 src, 10 adapters, 5 examples, 21 tests, 3 tools -- so the
grant travels with each file rather than relying on the root document alone.
Scripts keep their shebang on line 1; the SPDX lines follow it. Genuinely
vendored code is untouched: third_party/stb_truetype.h carries no header of
ours. src/render/embedded_mono_font.cpp is generated, so tools/fonts/gen_atlas.py
emit() now writes the two lines into its header comment and the file was
regenerated -- gen_atlas.py --check passes and the regeneration adds only those
comment lines to the previously committed file.
README's provenance section now states MIT and links LICENSE.
No code, test or public API change.
Findings from the review of the sliding-window change:
- A level >= 1 gather emitted only complete, wholly-retained buckets, so the
geometry stopped at the last bucket boundary inside the range: the newest
samples of a live window -- up to `span - 1` of them, about a pixel at the
level level_for() picks -- were not drawn, on the right edge, which is the
part of a scrolling trace that gets watched, and the oldest retained samples
fell outside the geometry too (the level-1 gather of a sliding window started
51 samples in). The two partial edge buckets are now gathered as raw samples
through copy_raw() and bracketed around the envelope -- the same exactness
split reduce/sum/sumsq already use -- so the emitted geometry spans the
requested range exactly. The pyramid still stores no partial bucket, so append
stays amortised O(1); at 1 MS/s the demo stages 511 extra sub-pixel segments
and still draws at 1.05 ms/frame, and the screenshot's trace now reaches one
column further at each end.
- `drop_before`'s doc claimed a no-op "for an empty pyramid": false, and
backwards -- the restart branch is exactly what an empty pyramid hits, which is
what StreamSeries::acquire() depends on when the ring saturated before the
first acquire. The precondition is `absolute_sample <= window_base()`, and the
restart is documented as what it is.
- The pyramid's memory figure was 3x optimistic: an entry is 24 bytes (MinMax
plus the exact sum/sumsq doubles), so level 0 costs 0.375n bytes and the whole
pyramid ~0.75n -- about 19% of 4-byte raw samples, not the 6% a min/max-only
index would cost. Corrected in the header, spec 4.2 and 4.5, with the trade
stated (the doubles buy O(log n) sum/mean/RMS/integral from whole buckets),
and the plan's Task 3 sketch no longer describes per-level prefix sums, a
level() span or an unaligned-end widening that does not exist.
Tests: `gather_at_a_lod_level_reaches_both_ends_of_the_window` pins that the
first and last emitted points are the oldest and newest retained samples -- with
spike values no bucket envelope can produce -- at level 1, at the level a dense
view picks, and for a static series whose end is unaligned. It fails against the
previous commit (first point at the first whole bucket), and so does the sliding
gather mapping check, which now accounts for the raw edge pairs.
Frame cost was O(capacity x levels) once the ring saturated: acquire() saw the
window base move (it always moves once the ring wraps), cleared the pyramid and
re-appended all 1,048,576 retained samples, and append() walked every live level
with no early stop -- 13,647,936 bucket merges and 73.8 ms per frame at a
1M-sample window, scaling with the window, which is what capped the 1 MS/s demo
at 32-58 ms/frame while it was still filling.
Index the pyramid in absolute bucket space instead. Level l bucket k covers the
samples [k * (kBucket << l), (k + 1) * (kBucket << l)), so a bucket's span never
depends on where the window starts and the window can slide over the aggregate
instead of forcing it to be rebuilt. Each level keeps its complete buckets in a
deque plus the absolute base of its first entry; append() folds a sample into the
single in-progress level-0 bucket and, only when that bucket completes, cascades
upward, stopping at the first level that gains no new entry; a new drop_before()
slides the window by discarding leading buckets and remembers where the window
begins. acquire() is now: drop to the new base, then append exactly what was
produced -- 256 merges and 0.034 ms per frame for 8192 new samples, independent
of the window, and memory bounded by the window (32,767 entries after 50M samples
through a 1M ring, the same as after 10M). A producer that laps the consumer by
more than capacity() between two acquires restarts the window rather than
splicing a gap.
Interface: `level(l) -> std::span<const MinMax>` cannot survive a ring and is
replaced by level_base/level_size/bucket in absolute bucket coordinates, plus
window_base/window_end/drop_before. The *_buckets queries stay in window
coordinates (index 0 = oldest retained sample), so Series keeps its window
semantics for callers; its exactness split now aligns in absolute space
(window_base() + the ceil/floor) because the bucket grid is absolute. The sums
move from per-level prefix sums to per-bucket ones -- the only representation a
sliding front can keep exact -- and StreamSeries::y_bounds() becomes
reduce(0, size()), exact in O(log n + kBucket) per frame instead of a scan of the
window. The pyramid owns the published window now, so there is no second copy of
base/end to keep in step, and gather() emits only buckets that are both complete
and wholly retained.
Tests: test-only merge counters (mirroring PLOT_PYRAMID_COUNT_COVERS, absent from
every non-test build) pin an acquire proportional to production and independent
of capacity, an append cost amortised over 16 live levels, and bounded memory;
brute-force checks pin absolute bucket aggregates across slides, sliding-window
reduce/sum/sumsq over unaligned ranges, and the level-1 gather mapping. The demo
now draws 946 samples/px at 1.03 ms/frame (967 fps) where the saturated window
cost 54.96 ms/frame (18 fps).
The pyramid's public query surface is the bucket-grained trio
(reduce_buckets/sum_buckets/sumsq_buckets); reduce() was the one
non-*_buckets query and silently over-covered non-aligned ranges, which
is exactly the misuse the exactness split exists to prevent. Exact range
queries stay on Series.
- delete the declaration, its doc paragraph and the definition; no
helper became dead (fold_range/cover_buckets/ceil_to_bucket are all
still used by the *_buckets queries)
- pyramid_clear_and_rebuild: complete bucket 0 and assert its extrema via
reduce_buckets(0, kBucket) -- a pyramid has no exact sub-bucket query
- stream_series_acquire_publishes_snapshot: s.pyramid().reduce(0, 100)
-> s.reduce(0, 100), keeping the == 99.0 assertion
cover_buckets chose one level per pass, so an odd first whole bucket (half of all
measurement ranges) never reached a level above 0 and folded every remaining
bucket at level 0 -- 15624 entries for a 1e6-sample pyramid. Pick the largest
aligned, in-range, materialised level at each cursor instead; the new cost test
counts folded entries through a PLOT_PYRAMID_COUNT_COVERS-only hook (absent from
normal builds) and fails on the old body.
Ring slots are now relaxed std::atomic: overwrite-on-full means the producer
reuses the slot of a snapshot's oldest sample while the consumer reads it, which
was a data race, not just a stale value. copy_from loads explicitly instead of
std::copy; push stays lock-free. A new producer/consumer case is the one TSan
flags without this.
Also extends the wrapped-window test with x0/bounds/pyramid/sum/sumsq/gather
checks, and corrects the StreamSeries threading note to the sample-count bound.
Martino Ferrari