Adds the release's line to the sweep notes at the head: FR-CULL-13's
write-path half and NFR-ARCH-1's executors, which now have entries, and
three things that closed without ever having had one — a mask layer's
film settings (FR-DEV-3f), hot and dead photosites repaired before the
demosaic (FR-RAW-3, with the defect-map reader still unwired), and the
tablet's scroll cue already described in §4a.
NFR-ARCH-1's register note says what b1d1c472 and 7be1efff met and what
they left, but outstanding.md, which exists to show the distance between
the register and the binary, had no entry, and catalog.md §6 still said
interactive work runs "on the decode pool with the I/O pool behind it"
when there are no pools.
outstanding.md §4 gains the entry: named and guarded, not bounded — the
counts are a budget, the guard covers block_on only, the two mask
workers and the core crates' threads are outside the module. catalog.md
names the executors the thumbnail and metadata work starts on and says
the counts are not yet enforced.
§2 still said FR-CULL-1 through -5 and -8 through -13 were built, while
the register's own status note (adbb9ac9) says only the write-path half
of -13 is met. The count now says -12, with -13 half built, and §2 gains
an entry for it: what verdicts.rs enumerates and holds to a reviewed
list, what fails the test, and the evidence that is not built — chips
for clipping, focus, burst membership and face counts, shown as absent
rather than zero, and a filter per signal.
2cde2874 made `cargo test -p traceability` fail on any write of a
rating, flag, colour label or trash membership that is not on a
reviewed list, and CONTRIBUTING.md, which lists what CI will stop a
change for, did not mention it. It is now the third invariant beside
the ui-names-no-operation test and the operation schema: what it finds,
the kinds of reason a write can be listed with, that the workspace test
run is what runs it, and `traceability -- verdicts` to print the list.
dr-film's README still described one 32³ lookup that took a negative
through the print and the paper, with the sliders' values baked into
it. Since 6b99f67 nothing a slider moves is baked: the curves are one
row per development time the datasheet measures and push interpolates
between them, a print is two lookups split at the paper's log exposure
with the enlarger's exposure added between them, and exposure, push,
print exposure and format reach the shader as uniforms. That is what
lets a mask layer hold film settings of its own.
The section now says so, and that the film's Exposure on the whole
photograph is the one setting that rebakes, because the enlarger's
filtration is solved against it.
fc54523 (0.18.1) stopped running a layer as a second chain after every
global operation and made its settings offsets applied at each
operation's own place, and nothing outside the code said so. The
manual still read as though a mask's slider were a setting of its own,
frame-budget.md described the masks as "a separate chain per layer",
and no document said how a layer is blended at all.
architecture.md §5.2 now says it: the offset, the blend by the layer's
weighted difference from the global result, what a photograph with no
layers composes to, and the film as the one operation whose settings
are averaged instead (6b99f67). FR-DEV-3 records the change as resolved
beside its local-adjustments bullet, frame-budget.md's note on what it
does not measure describes the cost as it now is, and the manual gives
the arithmetic in one sentence: -20 in a mask over -30 is -50 there.
The bundled manual is regenerated to match.
c50d96e added a repair pass ahead of the demosaic and no document said
so: FR-RAW-3's text names levels, the CFA and the colour matrices, and
architecture.md §5.2's stage diagram went from the upload straight to
black and white levels.
FR-RAW-3 now carries a status note: what counts as hot or dead, what
the photosite becomes, that Bayer and X-Trans share the pass, that
export and every other demosaicing path get it with no setting, the
test that holds it, and that the DNG defect map dr-decode can read is
still not used. §5.2 gains the stage and a paragraph on why it sits
before the demosaic, and its pointer to the raw histogram's tap names
the demosaic box rather than a row count the new stage would have made
wrong.
The single CI runner has one 99 GB disk shared with its container images.
The desktop job reached 92 GB used (2.1 GB free) on v0.18.0's run, and on
v0.18.1's the release build died with "No space left on device", so that
tag has no release page. At rest the runner holds about 23 GB; the
restored target cache, the models and the dependency build bring a job to
about 78 GB before a test runs, and tests plus the release build take the
rest.
The test executables and examples in target/debug are the largest part
of that, are relinked whenever a source changes, and are not what the
cache exists for — the dependency rlibs are. Deleting them after the Test
step, before the release build, frees several GB at the moment the job
is fullest without costing the next run anything the cache would have
saved. The step prints the disk afterwards, beside the existing Disk
before/after lines.
The manual said the lists on a tablet scroll by flick alone, and
outstanding.md §4a that Android had no scrollbars by design. Both now
describe the cue from #69: a thin line while the view moves, gone once
it stops, that a flick starting on it passes through; and §4a notes
DR_SCROLL_CUE for looking at it on a desktop, and that it is not yet
checked on the tablet. The bundled manual is regenerated to match.
On Android the desktop scrollbar is off, so every scroller that has one
on a desktop (the develop column, the grid, the collections sidebar,
Settings, the help sheet and the film list) gave no sign of how long it
was or where the view was in it. That is how the black-and-white film
stocks came to look deleted when the list could not scroll.
ScrollBar now has a second mode, chosen in the one Scrolling global:
where bars are off and `cue` is on, it draws the thumb alone, 3 px wide
against the edge, while the viewport moves, and fades it 500 ms after
the last move. It has no TouchArea, so a flick that starts on it
scrolls the content. Rust sets cue on touch-first builds; a desktop
build shows it when DR_SCROLL_CUE is set, to look at it without a
device. Desktop is otherwise unchanged.
Test: on the testing backend's help sheet, a press on the cue moves
nothing, a drag starting on it carries the list with the finger, the
cue is drawn while the list moves (drag, fling, wheel) and not once it
is idle; with bars on there is no cue and the thumb still takes a drag.
apply_judgement and apply_label carry TRACES: R7, the burst
representative callback FR-CULL-13 with a note that choosing it writes
the grouping and no verdict. The register's status for FR-CULL-13 says
the write-path test is met and the evidence chips are outstanding.
Traceability matrix regenerated.
FR-CULL-13 says evidence never writes a rating, flag, label or trash
membership, and nothing enforced it. tools/traceability/src/verdicts.rs
parses the shipped code with syn and enumerates every write: calls to
the catalog setters and trash recorders, SQL that assigns those columns,
sidecar Amendment::Judgement, and fields named rating/flag/label. Each
site must be in ALLOWED with a reason, as Input (inside a Slint on_*
closure, checked structurally), Relay (its callers are checked in
turn), Carried (a verdict made elsewhere: sidecar and XMP pulls, sync
merge, catalog mirrored to file, duplicates consolidation) or
NotAVerdict. Unlisted sites and stale entries both fail
`cargo test -p traceability`; `traces verdicts` prints the list.
syn and proc-macro2 were already in the lockfile as proc-macro
dependencies; this adds the edges, no new crate and no version change.
`a_resaved_file_owes_a_reread_and_loses_its_stale_hash` failed once in a
full dr-catalog run (updated 0, expected 1) and passed three times alone.
The incremental scan tells a changed file by its mtime at whole-second
resolution, and the `resave` helper wrote and renamed the file within
the same second as the scan before it, so on a fast enough pass the
resave looked like no change at all.
The helper now sets the file's and its folder's mtime two seconds ahead,
which is what a real resave some time after a scan looks like. Only the
test helper changes; the scan's rule is right for real files.
§7.1 now says where its table lives in code, how the UI thread is
marked and guarded, and that the counts are a budget rather than a
bound until work is pooled by executor. The register's status note says
what is met — named, spawned through one module, block_on guarded and
tested — and what is not: pooling, a guard for synchronous file reads
and catalog queries, the two mask workers, and threads the core crates
start.
Thirty-nine spawn sites in dr-ui, and one in the Android entry point,
called std::thread::spawn or a Builder of their own, and most of the
threads they started were <unnamed> in a panic message or a profiler.
Each now calls executors::spawn with its executor and a role, so the thread is
named <executor>:<role> — net:sync, decode:thumbs, io:catalog-open —
and knows which executor it is on. The three that already set a name
(automation, import, prefetch) keep their name as the role.
Behaviour is unchanged: each job still gets a thread of its own when it
starts, and spawn panics where std::thread::spawn did.
The module's documentation now says how a job is assigned: by what it
spends its time on, so a sweep that fetches bytes and then decodes them
is Decode, and a sidecar write that touches the catalog is Network.
Left as they were: the segmentation and refine workers in masks_ui.rs,
which another change is reworking, and test-only threads.
architecture.md §7.1 stated five executors and their thread counts, and
no code named them. dr_ui::executors now does: the Executor enum with
each one's thread name and the count §7.1 gives with its reason, and
spawn, which starts a thread named <executor>:<role> and marks it with
the executor it belongs to. The counts are the stated budget, not yet a
bound: a job still gets a thread of its own when it starts.
run marks its own thread as the UI executor before it builds the
window. net_runtime::build now returns a NetRuntime whose block_on
asserts, in debug and test builds, that the caller is not that thread;
everything else derefs to the tokio runtime. The login, folder-list and
remote-folder workers built the same runtime by hand and now take it
from net_runtime, so their block_on is guarded too.
Tests: a block_on on a thread marked as the UI executor panics naming
the UI thread; the same call on a worker returns; a spawned thread
carries its name and executor.
A hot photosite went into the demosaic as it was read, and came out as a
coloured cross three pixels wide that nothing later could take back
out. Night and long exposures showed them; the defect-map reader added
for FR-RAW-3 was never wired in, and a CR2 carries no map anyway.
A pass over the mosaic now runs ahead of the demosaic, into a second
buffer. A photosite is hot when it reads more than twice every
same-colour photosite in its 5x5 window plus 2% of the range, and more
than twice each of its eight immediate neighbours of any colour. The
second half keeps stars and glints: real light reaches the sensor
through a lens and an anti-aliasing filter and lights a patch, so the
photosites beside it are lit too, where a hot photosite's are dark. It
is replaced by its brightest same-colour neighbour, which invents
nothing. Dead photosites are the mirror case, judged only where the
neighbourhood is above 5%, so shadow noise clipped at black is left
alone.
The colour of each photosite comes from a 6x6 sensor-anchored tile, so
Bayer and X-Trans share the pass. Export and every other path that
demosaics get it too, and there is no setting: the repair only fires
where a single photosite disagrees with everything around it.
Cost, warm, on a Canon 6D frame (RTX 3050): 91-99 ms to demosaic
before, 94-98 ms after; the extra pass is inside the run-to-run noise.
Tests render a frame with and without the defect and compare the
finished pixels. Without the repair a hot photosite showed by 230 and a
dead one by 168; with it neither shows, and a 3x3 highlight at white
survives.
A layer offered the film's sliders and they moved nothing: its copy of
the node was never given the stock, so it stayed inactive. Film now
works in a layer the way the other adjustments do, as offsets to the
photograph's settings, but blended as settings rather than as results,
since a film is a rendering and cross-fading two developments is not
what a region on a pushed film looks like.
- dr-film bakes no slider. Exposure is a gain in the shader; push
interpolates the stock's measured processes, one curve row each; the
print is split at the paper's log exposure, so print exposure is an
addition between two lookups and exact at any setting. The enlarger
stays balanced at the photograph's exposure.
- film_sim reads all four settings as uniforms, format one-hot over a
grain count per format, so every uniform is linear in what it does.
- Operation::blends_settings lets the composer average each overlapping
layer's uniforms with the global ones by mask weight, the global
setting taking whatever weight the layers leave, and run the fragment
once. Three layers at full weight give the mean of their settings.
- The stock picker is hidden on a layer. Only the photograph's exposure
re-solves the print balance; push, print exposure and format need no
rebake at all now.
A local adjustment ran as a second chain after every global operation,
then blended by the mask. So global contrast -30 with -20 on a face was
contrast -30, the rest of the chain, and contrast -20 again on the
result, rather than -50 where contrast runs. The two edits compounded in
ways neither slider showed; a flattening applied to an already
flattened picture is how the shadows of a night shot went magenta.
A layer's setting is now an offset from its default, added to the
global setting (clamped to the parameter's range; a moved switch or
choice replaces it) and run at that operation's own place in the chain.
At each operation the global fragment and each touching layer's
combined fragment read the same input colour, and the pixel moves by
each layer's weighted difference: c_g + sum w_i (c_i - c_g). At full
weight that is the combined setting exactly, at zero the global result
exactly, and no setting is applied twice. An offset that brings an
operation back to neutral emits an empty version, which undoes the
global setting inside the mask.
Blending the colours rather than the uniforms is deliberate: the tone
curve and colour mixer emit code only for the channels and bands that
are touched, so the global and combined versions of one operation need
not share a uniform set.
A photograph with no masks compiles to the same shader byte for byte.
Test: global -30 with a whole-frame layer at -20 renders within one
count of global -50.
Reducing contrast turned every black in a night photograph pink. The
fragment lifted each pixel's luminance to its target by multiplying the
colour by target/luma. For a pixel at 0.001 on the way to 0.09 that is
a gain of ninety, and in the deepest shadows the channels are sensor
noise: after white balance the red and blue noise sits above the green,
their multipliers being nearly twice its, so ninety times that noise is
magenta.
Flattening now mixes the colour toward middle grey, which gives the
same luminance and adds the lift as a neutral. A black goes to grey and
its noise stays the size it was.
The same fragment clamped luma/0.36 into the curve's 0..1 domain, which
scaled every tone above twice middle grey down to 0.36 in either
direction: contrast +10 took a 230 grey to 162. Those tones are now left
where they are, which is continuous with the curve's top (value 1,
slope 0).
catalog.md §8.2 still said confirmed and rejected faces were matched to
local faces "by box", and that the merge reads only a remote face's box
and model; faces.md said `match_faces` "still matches by overlap alone
across devices". Since #77 the match falls back to embeddings, on the
photographs where a box leaves a remote face over, and since #78
`dedup_people` folds people of one name whose faces agree after every
sync. Both now say so, with the thresholds and the reason a less
decisive pair stays unmatched, taken from the code's own documentation.
`dedup_people COPY.sqlite` prints the listed people and faces before
and after, what the first run merged and kept apart and why, and the
time of three runs. The second and third runs are the cost the job
adds to every sync. `--peer PEER_COPY.sqlite` then plays two sync
round trips. The peer merges with the previous release's code path and
no job, this side merges back through sync::merge_remote, and the named
people each side lists are compared after every step.
On the reference pair: the desktop merges Claudine, Jessie x2, Mathias
and Noemi (80 -> 75 named). The tablet also merges its empty second
Ian (80 -> 74). The first run takes 1.2 s on the desktop, which is
building faces_box; the merge builds it first in practice. Later runs
take 8-15 ms.
A merge is where two devices' people meet: the same name typed on each,
or a redirect one of them made. So dedup_people::run now follows every
successful sync::merge_remote. It runs on the sync worker, never the
UI thread, before the snapshot is pushed, so what it folds reaches the
server on the same pass.
It runs in its own transaction, and a failure is logged, not
returned. What the merge took is committed and valid either way, and
the next pass tries again. Once a catalog is clean it costs 8-15 ms on
the reference library (19k faces, 26k people rows). On copies of the
two real catalogs, a round trip with a peer running the previous merge
converges on 75 listed named people on both sides and stays there
over a second round. merge_remote with the job takes 75-90 ms there.
Seven names are two or three live people on both devices: Ian (756
confirmed faces, and a second Ian with none), Jessie three times,
Claudine, Mathias, Noemi, Pascal and PJ. Each was typed on its own
device and carried across by sync, which keys people on their uuid and
so keeps both. Each half of a person shows half their photographs.
dedup_people::run, in one transaction:
- Same-name people (trimmed, case-folded as the Identity screen folds
them) merge into the one with the most confirmed faces, ties to the
smaller uuid, through faces::merge_people_within, so confirmations,
rejections and the survivor's name are kept. A person holding no
faces at all merges: there is nothing to compare or to carry. Anyone
else needs >= 2 confirmed faces per shared embedder on both sides and
centroids at cosine >= 0.7 in each. A face confirmed as one and
rejected as the other keeps them apart. Unnamed and set-aside people
are never merged by name.
- Faces held twice (one image, one embedder, IoU >= 0.5, cosine >= 0.7)
keep the stronger detector's row (FaceDetector::outranks), then the
confirmed one, then the older. The survivor takes the confirmed
assignment and both rows' rejections. A pair confirmed as two
different people is left and counted.
- Judgements still on a merged-away person move to the person at the
end of its redirects, and a redirect cycle (two devices merging one
pair in opposite directions) is broken at the smaller uuid.
Measured on copies of the desktop catalog and the tablet's server
snapshot, w600k_mbf, confirmed faces only:
- Centroids of differently named people: 2,699 pairs, median 0.02,
99.9th percentile 0.41. One pair reaches 0.70 (0.700 desktop, 0.705
tablet), "Michelle Casanonve" and "Michelle Casanova", one person
typed two ways. Next is 0.62/0.64, "Boris Jost" and "Boris". The
highest pair that is plainly two people is 0.43/0.44.
- One person split in random halves: minimum 0.69, median 0.91 over 72
people. Four faces against twenty-two reach 0.7 in 97% of draws.
One face against twenty of somebody else's reached 0.74 in 3,000
draws, and two faces reached 0.61, hence the two-face minimum.
- Pascal (22 and 4 confirmed) is at 0.57 and PJ (14 and 7) at 0.50,
under 0.7 on both devices, so both pairs stay apart and are logged.
The desktop's second Ian holds 4 suggestions and no confirmations,
at 0.38 against Ian's centroid, and stays apart. On the tablet it
holds nothing and merges.
Why a merge made here survives a peer on 0.17.0: the merged-away
person stays as a merged_into redirect with a bumped revision, which
the catalog merge has always taken on revision. The peer hides the
duplicate and never sends it back as a live person. Its own
confirmations of that person stay on the redirect, because a merge
never overwrites a local confirmation. The manual merge has always
left them there too. They follow the redirect when the peer runs this
job. A test syncs two catalog files through the previous merge code
and back, and the people converge and stay converged.
Once a catalog is clean the job reads 80 redirects, the named people,
and the face boxes from the covering faces_box index. That is ~10 ms
on the reference library. There is no schema change. The index is
created IF NOT EXISTS, as the merge already does.
merge_people moved a person's faces onto the target and left the
"not this person" rejections on the redirect. A rejection there binds
nothing: once Annie is Anna, the grouping pass is free to suggest the
face the user pushed away from Annie as Anna, which is the behaviour
rejections exist to prevent. The Identity screen's merge has done this
since it was written, and the deduplication job for #78 merges through
the same function, so it would have done it for every same-name pair.
The source's rejections now move to the target (INSERT OR IGNORE, so
one the target already holds is not doubled). Where the two halves
disagree about one face, confirmed as one and rejected as the other,
the confirmation stands, as `confirm` already rules for one face; and a
moved rejection withdraws a suggestion of the same face, as `reject`
already does. Confirmations and names are unchanged.
The body is split into merge_people_within, taking the caller's
transaction, so a job that merges several pairs commits once
(unchecked_transaction cannot nest). merge_people keeps its signature
and its one transaction.
The grouping pass never puts two faces of one photograph in the same
group (the cannot-link in dr_face::cluster). The merge did not check
this. When the two devices disagree about which face in a frame is a
person, merge_people_within applied the remote's confirmation, or the
anchor of a set-aside group, to face X. This device already held the
same person on face Y of the same photograph, so the person ended up on
both faces.
The reference library has 80 such person/photograph pairs on the
desktop and 89 on the tablet: 79/88 unnamed set-aside groups and one
named person confirmed on two faces. There are no duplicate faces (no
pair of faces in one image and embedder with IoU >= 0.5).
An incoming assignment is now refused when another local face of the
same photograph already holds that person. The one exception is an
incoming confirmation against a local suggestion: the suggestion is
withdrawn and the confirmation is applied. Two confirmations stay as
this device has them, the same rule as a local confirmation outranking
a remote one. A face that already holds the person is not a rival to
itself, so a steady-state pass is unaffected. On the reference pair
this refuses 0 assignments and writes the same 8,954 as before; it only
changes what a future disagreement does. The refusals are counted in
MergeReport::faces_one_per_photograph.
Existing pairs are left alone. They are two different faces (cosine
0.31 for the named one), not one face twice, so there is nothing to
fuse, and which face is the wrong one is not the merge's to guess.
On the reference library, 631 of the faces in the tablet's snapshot
match no desktop face by box (IoU >= 0.5), so a name on them stays on
one device. Twenty of those are the same face with the box drawn
somewhere else. Whole photographs sit at IoU 0-0.48 with cosines of
0.72-0.96 between the two devices' vectors, and ten of them already
carry the same person on both sides. The merge never read the
1 KB embedding every face row carries.
match_faces now runs two passes. The box pass is unchanged except that
a pair must now be unique on both sides: a remote face with two
overlapping local faces, or a local face overlapped by two remote ones,
is no longer settled by whichever overlap is larger. Only for the
photographs where a remote face is left over, and a local face is still
free, does it read vectors: one json_each statement per side, keyed by
row id. That was 357 photographs on the reference library, not all
19 MB of vectors. A left-over face pairs with the local face it
resembles most when:
- the cosine is >= 0.7,
- the two are each other's best,
- each leads its runner-up by >= 0.2, and
- a box has not already claimed the local face.
Anything less decisive stays unmatched, so a new face stays new.
Why the threshold is safe, measured on both catalogs (w600k_mbf):
- Of 169,548 pairs of different faces in one photograph, 4 reach 0.7
(lookalikes in one frame) and the maximum is 0.82.
- At 0.6 the rule would claim two pairs that carry different people on
the two devices. At 0.7 it claims 20, none contradicted and 10
corroborated, each leading its runner-up by more than 0.5.
- It only compares faces the boxes left unmatched on both sides: 737
such pairs, so about 0.02 false pairs expected.
- A low cosine never overrules a box. About 150 box-matched pairs fall
below 0.45, because two detectors cut the same tiny face differently.
73 of them carry the same person on both devices.
- Faces are compared only within one file_id and one embedder, because
the same person in another photograph reaches cosine 1.0.
- `Embedding::cosine` refuses a comparison across models.
Before -> after on the reference pair: matched by box 18,348 -> 18,348,
by embedding 0 -> 20, unmatched 631 -> 611, ambiguous 0 -> 0. The
report counts the embedding matches. No schema change.
The bench merged the catalog with a copy of itself. Every face in that
merge matches its own box, so the pass never reaches the faces the two
devices disagree about. On the reference library that is 631 of the
tablet's 19,052 faces, and it is the work #77 adds to.
`--remote PEER.sqlite` now also times `merge_remote_catalog` against a
copy of the peer file, and prints the first pass's report so two builds
can be checked for agreement. The first run writes what the peer
brought. The runs after it are the steady state, so compare builds from
two fresh copies of one catalog.
Blue sky, Deep blue sky, Polariser, and Blue sky with golden land, in a
Skies section after Essentials. Each darkens the colour mixer's azure and
blue bands and adds chroma to them — what a polarising filter does to a
clear sky — and brings the highlights down with it, so a white cloud does
not read as a cut-out against the deeper blue. The stronger ones nudge
azure towards blue and add dehaze.
They are looks and work only on the hues a sky occupies, so an overcast
frame is left nearly alone: there is no blue for them to deepen, and
tinting grey cloud blue would be worse than doing nothing. Tuned by eye on
the demo library's alpine and Manhattan frames, with an overcast Étretat
frame as the control.
saf.rs and the export path's SAF branch shipped tagged FR-PLAT-AND-1,
and the matrix counted the requirement as covered. Its subject is the
library — reached through SAF grants — and Android still reaches a
library over a Nextcloud account or a folder path. What the SAF code
does is give an album a folder on the tablet, which is FR-EXP-10.
outstanding.md said the figure overstated it and should be read with
this one subtracted; it now says the tags were narrowed, and coverage
reads 161 of 192.
The sidebar's "All photographs" row was bound to library-total, which
is the scope's count: the header's "412 images" and the scrollbar's
size. Under a collection, and now under an album, the row read as the
album's size — 4 where the library holds 70.
It reads a separate library-whole-total now: the same number as the
scope's when nothing is scoped (no second count), and otherwise the
unscoped count under the same filter, read only when the view's facts
move, so scrolling inside an album does not recount the library.
Every scene is recorded again on the 0.17.0 build, because the header
(Export to Exports), the sidebar (Albums) and the develop column had
all moved. The launch pictures showed the typed folder field, and the
export settings "Export to"; the presets picture predated the sections.
New scenes: presets_film scrolls the sheet down through the shipped
sections and applies Ilford HP5 Plus, a look that changes the film and
nothing else; albums exports four New York frames to an album, selects
it to show the originals behind its files, and opens a new album's
sheet before deleting what it made. record.sh now points the profile's
old export folder at DR_HOME/Exports, which the app turns into the
album "Exports" on first open - the one way to have an album without
the portal's dialogue, which Xvfb cannot show. The launch scene signs
out of a remembered folder instead of typing one, so it shows
"Open folder" beside the folder used last.
Not recorded: the server browser's New folder, which needs a Nextcloud
server, and the download screen, which a folder library never reaches
(the original is a local read, over before the first poll). The manual
says so where it describes each. film_reach and duplicates passed;
inference was pinned to the CPU.
NFR-COMPAT-2's table has CI building every channel, the Flatpak
included. No workflow in .gitea/workflows/ builds it, and the sweep
before this one records that none has been built by hand either. A
status note under the table says so, so the decision and the state are
not read as the same thing.
benchmarks.md and catalog_open.rs both said schema::backfill runs on
every Catalog::open. Since ffdd640 it runs on the first open of a path
in a process and is skipped while the stamp matches, so in dr-bench
catalog_open_ms still includes it and catalog_open_warm_ms, the second
open in the same process, no longer does. That is what a library
reopened in one session costs, and both now say which figure is which.
The comment keeps its line count, so no tag below it moves.
Since 6683c14 every folder on the desktop is chosen through rfd, which
on Windows is the common item dialogue. windows.md's list of what is
already handled now says so, and that nothing has opened one under
Wine or on Windows. "The import flow's path picker" is now the import
page's Browse… button.
The README said export went "to a folder here or back into the
library", which albums replaced and forbid, and called presets "named
presets" beside a shipped collection of film looks and Lightroom
imports. It now says both, and that a photograph only on the server
opens on its thumbnail with its download's progress. The requirement
count comes from traceability.md's summary: 192, 84% claimed; the
version and release count are left to the release commit.
docs/README.md's manual row gains presets and albums. CONTRIBUTING
said 177 requirements and 826 crates, and the Flatpak manifest 826
crates; the lockfile now holds 849 packages, 822 from the registry.
The figures are bee5c58's, quoted as that commit measured them: five
passes to two, 22.9 ms to 9.1 ms at 2560 x 1600 fit and 54.1 ms to
28.1 ms at 4K, the output bit-identical. Under its own heading and
status line, like the 2026-09-25 section, since this file was not
re-run for them.
#75 and the develop-landing work changed how the catalog is opened and
queried without touching its design document. §2 now lists the two
indexes made on first use, keywords_term_version and faces_box, beside
the tables made that way, and says why each exists. A paragraph states
what the backfill stamp in backfilled.rs holds, why it records the
newest rows by content rather than by id, and when the backfill still
runs.
§6.2's example of coalescing was a thumbnail job, a kind nothing
enqueues since #73.
storage.md's trait listing stopped at get; it now has get_reporting,
what the default and the Nextcloud override do, and where develop reads
the figures. A new §5.3 says how folders are chosen — the portal or
Windows dialogue, the server browser whose New folder is create_dir,
SAF on Android — and where an album's files go: a server folder relative
to the account root, with the outbox's third .dest line, or a device
folder that never syncs.
catalog.md §8.2 said only collections merge, which had not been true
since keywords, people and capture metadata joined them, and is less
true with albums; it now lists what merges and why album_folders does
not. §2 records that the album tables, like dedup_probes, are made on
first use rather than by a migration.
outstanding.md said there was no SAF code on Android. There is now,
for album folders only, and it carries TRACES: FR-PLAT-AND-1, which
the entry says overstates a requirement about the library; FR-PLAT-AND-2
and S10's row follow from that.
distribution.md §4, outstanding.md's FR-PLAT-LIN-3 entry, the Flatpak
manifest's comment and the README all said a library was chosen by
typing a path and that nothing in the tree called the FileChooser
portal. Since 6683c14 every folder the desktop asks for is chosen
through rfd's xdg-portal backend, so those sentences were false.
What they now say instead is narrower than "it works in the sandbox":
no Flatpak has been built here, so whether the portal's path opens a
library, holds across a restart and takes a sidecar is unobserved, and
volumes() still cannot see a host card. The chooser also landed in ui/
rather than behind the dr-plat seam distribution.md had proposed, and
both documents say so. FR-PLAT-LIN-3 gets a status note to the same
effect.
The backfill stamp read max(id) of images and versions and max(rowid)
of keywords. None of those tables is AUTOINCREMENT, so SQLite hands a
freed newest id out again: empty the trash of the newest photograph and
scan a new one, or let a local folder's walk delete a renamed file's row
and insert the new name in the same pass, and the new image takes the
old id. max(id) does not move, nor does count(*), and when a newer
version elsewhere keeps max(versions.id) still too, the stamp matched
and the open skipped the backfill.
That row is exactly one that needs it. Neither scan path creates the
default version: scan::persist and walk insert the image and leave the
version, the RAW/JPEG pairing and the keyword terms to the next open.
Skipped, the image went without them until the app restarted, so a
rating or a pulled sidecar judgement had no version to land on and a
JPEG beside its RAW showed twice.
The stamp now carries the newest row's content: the newest image's id,
path, added time and whether it has a version; the newest version's id
and image; the newest assignment's rowid, version and word. Whether the
newest image has a version is the part that cannot be fooled - after a
backfill every image has one, and a row that has just taken a freed id
has none - so the two stamps differ even when the same file comes back
at the same id in the same second. Still one statement: three reverse
rowid scans that stop at the first row, and one probe of versions_image.
An open that skips still costs ~1 ms on the reference catalog copy.
This closes the hole in the stamp itself rather than by a forget() at
each delete site, so a delete path added later, or one in another
process, cannot reopen it. Two tests delete the newest image and insert
another at the freed id on a separate connection, with a newer version
elsewhere holding max(versions.id); both fail against the old stamp.
holds_original, the prefetch worker's check that a neighbour's original
is already cached, opened the catalog for every neighbour it asked about.
Its own comment called it a row check; the open around it was four of
the five opens a develop landing made.
The worker now keeps one Catalog for the batch it is serving, opened at
the first check and reopened only if the batch names another catalog
file. The connection runs in autocommit, so each check still sees what
fetch_original committed in between. fetch_original is unchanged.
With the backfill no longer run on every open, a landing whose
neighbours are all cached goes from five opens to two, and from ~80 ms
of CPU to ~1-2 ms on a copy of the reference catalog.
Catalog::open ran schema::backfill every time, and every worker thread
opens its own connection. A develop landing made five opens, and each
paid the RAW/JPEG pairing, the default-version anti-join over every
image, the uuid pass over every default version and the keyword check:
17 ms of CPU an open on a copy of the reference catalog, ~80 ms a
landing, to confirm that nothing had changed since the open before.
Everything the backfill repairs is a row some write added: an image a
scan inserted, a version or keyword assignment a merge brought in. So
the open now reads a stamp - user_version, max(id) of images and
versions, max(rowid) of keywords, and the file's device and inode - and
skips the backfill when the stamp matches the one recorded at this
path's last backfill in this process. The maxima are each the last page
of a b-tree; an open that skips costs ~1 ms.
The backfill still runs:
- on the first open in a process (nothing recorded yet);
- on any open that migrated the schema, unconditionally;
- after a pull: merge_remote forgets the path, so the next open
backfills even when every incoming row collided and nothing moved;
- when the file is replaced under its name: the inode is in the stamp,
and recovery::set_aside, the first step of a restore and a rebuild,
forgets the path;
- when another process or thread adds rows, because the stamp is read
from the file, not from anything this process did.
The stamp is taken before the backfill, not after. Read after, it would
describe the backfill's own inserts, and could record an image another
connection inserted in between as covered when it was not. Read before,
the worst case is one redundant pass after a backfill that did real work.
Kept in memory rather than in the catalog: a stamp row would need a
table an older build does not have and would travel in the sync
snapshot, where a flag from another device's catalog says nothing about
this one. No schema version bump, so the tablet on 0.16.0 still reads
the snapshot. Tests cover the skip, a scan's new image, a migration, a
pull and a replaced file.
Landing on a photograph in develop opens the catalog once to fetch the
original and once more per prefetched neighbour to ask whether the cache
already holds it: five opens, each running the whole backfill. The bench
timed one open but not the landing, so the cost of the shape was not
visible and a fix to it could not be measured.
Two figures now, both against an empty cache so the question is asked the
same way whatever the answer: the five-open shape the app had, and the
two-open shape where the prefetch worker keeps one connection for its
batch. On a copy of the reference catalog (23,582 images) under load, the
five-open landing costs ~80 ms of CPU.
Dehaze cost 22.9 ms of a 2560x1600 frame on the reference laptop RTX 3050,
and 54.1 ms at 3840x2160, with the memory clock held at 810 MHz by the power
cap (graphics 1762 MHz). It ran five passes: a run and a span erosion along
x, the same along y, and the recovery. At those clocks a detail pass costs
what it reads and writes, not what it taps: a pass with an empty body -
one render-sized rgba16float read and write - measured 4.0 ms, and each
dehaze pass 4.4-4.6 ms, so the taps were about 2 ms of the 22 and the four
hand-offs between passes were the rest.
Each axis is now one pass that takes the minimum over the whole window
directly, and the recovery rides in the y pass, which already holds the
veil and the pixel's own colour. That is 36 texture reads per pixel at
2560x1600 in place of 12, nearly all of them cache hits, and two passes in
place of five.
The picture is the same bits. A minimum is exact in any order, and the
window is the one Split always covered, the surplus pixel on the far side
included (Split::first and Split::width). The veil crossing the removed
hand-offs was already exactly representable in rgba16float - a minimum of
channels read from rgba16float, floored at zero - so storing it between
passes never rounded anything that the fused form now keeps unrounded.
Measured with a scratch probe that renders the synthetic 60 MP frame from
examples/frame_budget.rs, only a detail parameter moving so the fused pass
is reused, 30 frames per scene after six of warm-up, five runs of each
binary alternated, median of the per-run p50:
scene before after
dehaze 2560 fit 22.88 ms 9.06 ms
dehaze 2560 1:1 23.41 ms 9.52 ms
dehaze 3840 fit 54.09 ms 28.12 ms
all detail 2560 fit 53.11 ms 39.97 ms (NR, sharpen, clarity,
all detail 2560 1:1 67.48 ms 56.42 ms texture, dehaze)
every op 2560 fit 57.59 ms 44.19 ms (with film)
every op 2560 1:1 71.83 ms 57.93 ms
controls without dehaze (NR, sharpen, clarity, texture): within +-2%
The rgba8 output hashed identically before and after for every scene -
dehaze alone, all five detail operations, every operation with film, and
each other detail operation alone - at fit and 1:1, at 2560x1600,
3840x2160, 1917x1203 and 333x211: 64 of 64.
dtourolle