A stay belongs to one of the trip's places, so on a new trip the
accommodation sheet only said "Add places to your trip first", and the
booking could not be saved without leaving the sheet (TestFlight feedback,
5.1.3). The hint now carries the "Add place/activity" button for anyone
allowed to add places: the place form opens over the sheet, and the place
is listed to pick once it is saved.
The app hides the keyboard's bar with its "done" button, and a textarea
keeps its return key for new lines, so after typing a note in the place
editor nothing closed the keyboard and it covered the rest of the sheet,
times and Save included (TestFlight feedback, 5.1.3). A tap away from the
focused field now lets go of the focus. A scroll keeps the keyboard open,
and a tap on another field moves the focus there.
watchOS crops the icon to a circle on a black face, so the black icon did
not read as a circle at all (App Review, guideline 4). The mark now sits
dark on a light grey ground, drawn from the same master silhouette.
While TREK Cloud's orders are closed, the plan picker on iOS offers the
subscriptions only to builds that buy in the App Store sandbox (TestFlight,
App Review, Xcode). That was read from the receipt's file name alone, which
App Review's build did not carry: the reviewer saw "TREK Cloud opens soon"
and no way to buy, and the app was rejected under 2.1(b) because its
in-app purchases could not be found.
The environment now comes from the signed app transaction (StoreKit 2, iOS
16 and later), with the receipt's file name as the fallback on iOS 15 or
when the app transaction cannot be read. A production build still sees the
plans without a buy button until orders open, and the cloud refuses a
production purchase before the opening as before.
Five sorts in the Postgres bootstrap and the move from SQLite relied on the
default sort, which SonarCloud reports as a bug because it compares by code
unit implicitly. They order migration names, table names and JSON keys for a
checksum, where locale-aware ordering would change the result, so each now
names the code-unit compare the default already used. Nothing changes.
The v1 bodies were cast to their types unchecked, so a flight page whose data
was not a list, or a passenger list of another shape, failed deeper down as a
TypeError instead of an AirTrail request error. Each v1 answer (a flight, a
page of flights, a created flight, a reference lookup, the key owner) is now
read through a small schema that keeps the fields TREK does not need lenient.
The legacy save refuses an arrival before the departure; the v1 body passed
it on, so an overnight arrival time without its date would have been stored
before the departure. The v1 write-back now refuses it with the same message.
AirTrail 3.13 adds /api/v1 and marks the /api/flight/* routes as
deprecated. TREK now asks the public discovery document GET /api/v1
and uses v1 when the instance has it. It uses the legacy routes for
an older instance. The client methods keep their signatures, and the
sync hash stays the same, so a linked flight does not sync again
when an instance changes to v1.
The probe cache holds only a clear answer: 200 with the discovery
document gives v1, and 404 gives legacy. The cache is per AirTrail
URL, for 10 minutes. A timeout, a 5xx or a network error fails that
call only. A 401, a 403 or a page that is not the discovery document
uses legacy for that call only. A 404 for one flight makes the client
ask the probe again before it reports the flight as deleted.
A missing scope gives AirtrailScopeError, which names the scope. It
does not unlink a flight. The API key hint and the help step list
the scopes for read and for write-back, in all languages.
Each new request goes through safeFetch, with the SSRF guard, the
timeout and the size cap. The flight list reads pages of 100 and
fails above 50 pages. A failed list unlinks nothing.
airtrail-v1.ts translates between the v1 shapes and the legacy
shapes. A write-back looks up the airport, airline and aircraft ids,
and changes each local time to a UTC instant in the zone of its own
airport. It writes each instant in the +00:00 form of the legacy
save, because AirTrail stores the string as it gets it and the sync
hash reads that string.
The tests cover both APIs over HTTP, the probe cache rules, the sync
hash on a switch and on a save, the page limit, the scope error, and
the UTC conversion across midnight, the date line and daylight-saving
changes.
TestFlight feedback: during a trip the widgets stayed on day 1 until the app
was opened again. The app wrote only today's day into the widget snapshot,
and only while it ran, so every timeline entry after midnight still showed
the day it last ran, with its stops all ahead again.
The snapshot now also carries the trip's days from today on (a month at
most, sixteen stops a day), each with its date. On iOS every timeline entry
takes the day of its own date, so the quarter-hour entry at midnight turns
the widget to the next day, and the bookings list drops the days behind; the
Android widgets do the same on each update. A snapshot from an app without
the days reads as it was written, and an older app binary ignores them.
TestFlight feedback: on an iPhone Duo the status capsule in the top right
corner covered the trip's controls, its day chips, the map's category bar and
the first cards of every trip tab, upright and on its side. The app has
measured the capsule since 0f42b8b8a, but only the shell's own bars and page
used it, and the trip, its map and the sheets place their own controls.
- Upright, a control row in the capsule's height stands beside it and the
content starts below it: the trip's top row and day chips, the map's
category bar, the road trip's stage bar, every trip tab's first row, the
places browser, the plan timeline, docked and card sheets, the assistant,
the account menu, the image viewer's close button, the atlas search,
vacay's header and a journey's content and map controls.
- On a wide window the page gives up the capsule's side over the whole
height instead, as an iPhone on its side does for its camera: lifting each
page below the capsule cost a quarter of the height of an unfolded Duo.
The list tabs, the assistant panel and the form sheets end before it too.
- Screens whose first row is part of the page (settings, administration,
collections) are lifted until that row is below the capsule; screens that
keep clear of it by themselves (atlas, vacay, a journey) are not lifted.
- Right to left the rail stands on the capsule's side and takes it.
All of it reads the corner variables the app publishes, which are 0 on every
other device and before iOS 26, so nothing else moves. The rules for the
screens that place their own controls live in mobile/corner.css.
React Router lands the URL without the link as a transition. Until it does,
every render still read the link, and the tab link took it again on each of
them: selectTab is a new function per render and onLinked sets state, so a
notification's ?tab= link rendered the planner into React's update limit and
the route's error page replaced the trip on an iPad or an unfolded foldable.
Both link hooks now take a link once per URL and wait for the cleaned one;
the same link in a new URL is taken again.
The rail's list ref was an inline function that set state. React 19 detaches
and re-attaches a ref whose function changed, so every commit set the list to
null and back and asked for one more render of the rail. The ref is one
function for the rail's lifetime now.
Adding a day before a trip's start with "Keep bookings on their dates"
moved almost everything a day earlier and emptied the last day. The day
grid was always rebuilt by position: the n-th day row took the n-th date
of the new range. With bookings kept on their dates, only timed bookings
and stays were re-anchored afterwards. Places, day notes, bookings with no
time and a booking's place in a day's order stayed on their re-dated rows
and slid to the day before. Days added at both ends had the same problem.
planDayGrid now takes keepDates. When the new range still holds every
dated day, each of them keeps its date and the new days go around them.
The server passes it unless the traveller chose to shift everything, and
the trip dialog's warning check uses the same rule. A moved or shorter
range still follows the positional rule the dialog describes, and "Shift
everything" still moves the whole plan with the new start.
Two things keyed outside the day rows followed the old rule. Road trip day
boundaries are keyed by day number, so a new first day left each one on
the day before its own. They now follow their days to their new numbers
and go out to every screen. The owner's leave days moved by the change of
start even when days were only added. They now move by as far as the
trip's first day moved, which is nothing when days are only added. The
help text of the option and the MCP description say what happens when
days are added.
Closes#3119
Tapping a transport in the compact Transports or Bookings list on the phone
opens the transport sheet, and its time boxes showed only the clock time. The
sheet named a date only in its subline, and only for a booking with no route
and no airline or number, so a flight or ferry could not be placed on a day
without unfolding the whole list. A transport holding only times (its date
lives on the linked trip day) had no date at all, and the desktop transport
detail dialog had the same gap: it read the date off reservation_time alone
and never named the arrival day of an overnight trip.
The dates now come from one helper, bookingDates(), which reads the booking's
own date, then the stop's local date, then the linked day's date, the rule
bookingFacts already used and now shares. The phone sheet prints the date
above each time (a leg opened from its row uses its own days); when no time
box carries a date, because no time is entered yet or the live flight card
stands in their place, the subline names the day instead. The desktop dialog
names the day, or both days of an overnight trip in reading order, in the
user's locale.
Closes#3122
With "Connect the days" and "A colour per day" on, the drive from one day's
last stop to the next day's first stop was coloured as the day it left. A stop
reached after midnight was handled the same way: the line into it kept the
colour of the day it came from. The road trip panel counts both drives toward
the arriving day (its distance includes them, and the setting's hint says
connecting drives count toward the arrival day). So the map showed the drive
under one day while the panel counted it under the other. This also happened
with daily travel times set to end the day at the last place, whenever the
schedule could not be kept (a drive longer than the window) and the days were
drawn as stored.
assembleRoadtrip now puts every line drawn into a day under that day's
number, the day whose figures count it, and still marks it as a join, so the
one-day stage view still leaves it off. Every map takes its day colours
from this list: desktop Leaflet and Mapbox GL, the phone map and its stage
view. The client tests that pinned the old colouring and the old join order
now expect the new ones.
Closes#3123
Every push to dev started test.yml twice for the same commit: the push
run on dev and the pull_request run of the open release PR into main.
Together about 72 long jobs exceeded the 60 organisation slots, so jobs
queued and ci-ok landed minutes later.
On a pull request from dev into main (head in this repository) the
changes job now sets mirror and turns every area off
(scripts/ci/changed-areas.mjs), so no gate, test, coverage or scan job
runs. ci-ok keeps its needs verdict and then runs
scripts/ci/mirror-push-run.mjs, which waits for the ci-ok of the push
run of test.yml on dev for exactly the PR head commit and passes or
fails with it. A missing push run, one that ended without a verdict,
an API that keeps failing or a wait past 17 minutes fails with a
re-run hint. ci-ok reads with actions, checks and contents only, and
its timeout rises to 20 minutes.
A push to dev now lists its changes against the merge base with main
instead of the previous push, so the run the mirror copies covers
every area the release pull request brings in, also when an earlier
push's run was cancelled or red.
The release workflows keep reading the newest ci-ok on a commit. A
green mirror implies a green push ci-ok before it; a mirror that fails
on its own blocks a release until it is re-run.
The move guide tells the operator to check that the log reports the Postgres engine, but the start banner never said which database the process runs on. It now carries a Database line with the engine's own description: the SQLite file, or the Postgres URL with its password masked. The guide's check and the manual checklist point at that line.
The compose pages name postgres:18, list the volume path for 18 and for 14
to 17, and say that changing a database's major needs pg_upgrade or a
backup and restore. The move steps say to carry the install's own
environment variables, container name and port over, and to append the
passwords and COMPOSE_FILE to the existing .env instead of replacing it,
so later plain docker compose commands, the ones in Updating included, keep
using the Postgres file instead of starting TREK on the old travel.db. The
pages also say which data is lost to an anonymous volume, which compose
service waits for the database, what tables that differ from the Postgres
schema mean, and how to recover from a first start without a password.
The postgres image refuses a volume at /var/lib/postgresql/data from 18 on,
so the db service now mounts /var/lib/postgresql, and the file says that 14
to 17 need the old path and that an existing database keeps its image tag
and path. Compose ignores a config's mode outside Swarm and warned about it
on every command, while the image sourced the init script anyway. The mode
is gone and the script runs its work in a subshell, so set -eu and exit stay
inside it and a failure still fails the first start whether the script is
sourced or executed.
The refusal listed the differences and nothing else. A table or column
only one side has now comes with what that means: the file is at this
version's level, the start that brought it there added every column it
could, the start log names the rest, and the operator keeps running on
SQLite and reports it. A column of TREK's own schema the move cannot copy
is listed apart and named a bug in TREK, since the file is not at fault.
The server guide describes the gap, the repair, the check and the refusal.
After the migrations, both engines compare the declared entities with the
live schema in one catalog statement and log a warning naming each table or
column that is missing, what fails because of it, and that it should be
reported. It never refuses a start, not even when the check itself fails,
and says nothing on a database the migrations built; a Postgres lane suite
holds that on a real server.
The legacy baseline records the positional runner's steps up to
schema_version as executed without running them, so a column that runner
never added stays missing for good. Two things in its history left such
gaps: a step inserted below a version a build had already shipped (step 53,
users.mfa_backup_codes, so turning on MFA failed there) and a column added
to createTables() with no step (users.immich_access_token).
Migration20200101046460 adds, where missing, every column the schema held
at legacy step 258 that head still holds and ALTER TABLE ADD COLUMN can
add, declared as head declares it; on Postgres every one of them is in the
baseline, so nothing is added there. The list is derived again from the
migrations in tests/unit/db/legacy-column-repair.test.ts, which also adds
all of them to a database lacking them and upgrades a v244 install shaped
like the one whose move to Postgres was refused.
better-sqlite3 compiles on install, and the Playwright container has no make or g++, so the first e2e run of a lockfile failed in npm ci. The setup action now installs both when they are missing, before npm ci; a cache hit skips the step.
watch-bridge repeated the client typecheck and three client test files
that test.yml already runs on every change under client/, inside ci-ok.
The shells link only @capacitor/* packages, all declared in
client/package.json, which stays a trigger, so a native dependency bump
still builds iOS and Android while lockfile-only bumps no longer queue a
Mac build.
Rename the workflow to CI and name every job "<Area> <what>[ i/n]" with
ids in kebab-case, in area order. Fold i18n parity, shared coverage and
the plugin SDK into Shared tests, split the client gates into Client
types and Client checks (with the bundle), and merge the image and chart
jobs into Docker and Helm with per-area step conditions. Server and
client tests run in ten slices, the Postgres lane in three per major,
and e2e in two slices inside the digest-pinned Playwright container.
A local setup action restores node_modules and shared/dist from exact
cache keys and falls back to npm ci and the shared build. The
node_modules key covers the lockfile and every workspace package.json,
so npm ci still refuses a package.json out of step with the lockfile.
Only pushes to dev and main save, the npm download cache included, and
only a full install saves that one. The plugin SDK installs and runs
before the root install so it resolves only its own dependencies.
Timeouts drop to 15/10/5 minutes. The workflow tests now check the name
form, the Playwright tag against the lockfile, scan's needs, timeouts,
the cache keys and saves, and that installs go through the setup action;
.github/actions/ counts like test.yml in changed-areas.
Controllers read their path ids through the rowId pipe now, so unit tests that call a handler directly pass the number the pipe hands it. The e2e suites enable addons with true, the type of the entity's field. The realtime gateway test builds its book sockets through one helper instead of a second hand-built TripAccessService.
The union of an integer and a digit string fed z.coerce.number() through pipe(), which Zod 4 types with an unknown input and tsc refuses. The digits are now turned into a number with Number and checked by z.number().int().positive(), the same rule as before.
wiki/Database-Postgres.md, server/CLAUDE.md and the plugin SDK docs now
describe what plugin SQL may name on Postgres, the pinned output
settings, the shared contract (NUL and transaction control refused on
every engine), the move's refusals, the Postgres lane in two slices per
major on 14 and 18, and that TREK runs as one instance on either engine.
server/.env.example names the version 18 client tools the image ships,
and the troubleshooting entry and the Helm values name the full set of
roles a restore refuses. A doc-paths baseline entry that no longer
applies is removed.
trek-plugin dev and createMockHost copied the db:own checks by hand and
let through what TREK refuses (NUL, transaction control outside tx()), so
a plugin worked locally and failed in TREK. Both now read the contract
from src/plugin-data-contract.ts, a byte-identical copy of the server's
file that the parity test compares, and check in the server's order.
Tests compare every refusal of the dev database and the mock host with
the contract's text.
Comments and test titles added on this branch cited review labels and
planning documents that do not exist in the repository, and some used
dashes as separators. They now say what the code does and why in plain
sentences, and the Postgres lane's suites share one header format. Test
ids stay as they were.
Before Postgres 15, every CREATE DATABASE ... TEMPLATE the lane runs per
file is two checkpoints waiting for fsync, and the lane took 20 to 23
minutes on 14 against about 12 on 18. The service now runs with fsync
off, set through ALTER SYSTEM and a reload and checked afterwards. A test
database has nothing to keep through a crash.
Two slices per major replace three, so four long jobs replace six. The
job-count note in test.yml now states the real count: up to 30 jobs, 24
of them long, for 20 organisation-wide slots. At least four long jobs
start only after another has ended. A run with six slices started nine
long jobs four to five minutes late and ended its client coverage eight
minutes after the on-time shards. The job comment says what was measured
on CI and what only locally. The lane docs say two slices, and the root
guide describes the slice gate.
A sliced job states its slice count three times: the matrix shard axis,
the i/n in its name and the --shard=i/n it hands vitest. Changing one and
not the others still passes and leaves part of the suite in no slice or
reports the wrong count. workflow-jobs.test.mjs now treats every job that
names matrix.shard or runs --shard as sliced and fails, naming the exact
fix, until the three agree. A job whose axis is not the one-line
"shard: [1, ..., n]" form fails too, instead of going unchecked, and so
does a fixed --shard. Tests on small inline workflows hold the reader
itself to these cases.
The boot refusal beside a SQLite install told the operator to run the move
with plain `docker compose run`, which uses the SQLite compose file. It now
gives the command the wiki gives, and PGBOOT-037 holds both to the page.
The move read unique indexes by joining indkey to pg_attribute. An expression key such as lower(email) has attnum 0 there, so idx_users_email_lower was missing from the plan. The catalog query now reads each index's key as Postgres prints it, whether any part of it is an expression, and the columns it reads: the plain key columns and the Var nodes of indexprs, without INCLUDE columns or the predicate.
A broken unique key no longer names --drop-nul as its cause. The refusal names the key and the shared value from Postgres's detail. It explains that the SQLite file held the two rows apart, and the hint about SQLite's lower() appears only for an expression key. Values --drop-nul changed in the key's columns are listed as rows to check first, and only when the duplicate value holds their text. An unrelated value that lost a NUL is no longer named. A plain key over a quoted column is named by its column, not as an expression.
A row that breaks a constraint (SQLSTATE class 23) is now a MoveRefusal that says what to correct, in the SQLite file or in a plugin's plugin.db. Before, it was a bare error with a stack. A value Postgres refuses by its type (class 22) stays a failure: it names no row, and values.ts should have refused it first, so the message asks for a bug report. MoveRefusal and the remedy text move to refusal.ts so copy.ts can use them without an import cycle. The usage text of --drop-nul and the wiki's move section describe NUL in JSON, the key listing and these refusals.
A raw U+0000 in TREK's JSON text, which an old or hand-edited travel.db can hold, was refused as "not valid JSON" without naming the NUL or the remedy. It is now refused with the NUL reason, so the refusal names --drop-nul. With --drop-nul, raw and escaped NULs are removed and nothing else, the way TREK's own writes drop them, and each change is listed. Text that is not valid JSON even without its NULs is refused as that, and the checksum's source side follows the same rule.
A \u escape of half a surrogate pair parses in JavaScript but jsonb refuses it. Before, that failed the insert without naming a row. It is now refused by table, row and column.
The plugin SQL guard now refuses what still reached outside a plugin's
schema or depended on the server:
- reads of the system columns (xmin, xmax, cmin, cmax, ctid, tableoid),
which show transaction ids and storage layout and do not exist on
SQLite; a value of the plugin's own may still be named xmin with AS
- the xid and xid8 types wherever the word stands for the type, since
age() of an xid is the cluster's transaction counter
- the text search functions that coerce a text argument to regconfig, a
catalog lookup by name, which leave the allowlist
- collations and access methods outside the set every server has, which
could tell whether a locale or an extension exists
Every plugin transaction pins, with SET LOCAL, the settings that decide
how values print and parse (TimeZone UTC, DateStyle, the C locales, float,
bytea and xml output, the text search configuration), so a plugin's output
neither depends on nor reveals the server's configuration.
Both stores now keep one contract. Its texts and caps live once, in
plugin-data-contract.ts, and both stores throw only those: NUL in SQL,
an argument or a migration id is refused on SQLite as on Postgres,
transaction control is refused in every call on both engines (tx() is
how a batch becomes atomic), a closed store refuses every call, and the
Postgres store answers the cases better-sqlite3 names with its words.
The SQLite store moves beside the Postgres one as SqlitePluginDataStore
in db/engine/sqlite, takes its file from the backend so src/db never
imports src/nest, and no longer needs the driver exemption. A contract
test runs every violation on both stores and fails when they answer
differently. The Postgres lane holds the allowlist, the pinned settings
and the SQL plugins commonly write against a real database.
ROUTEID walks every route the app registers.
- ROUTEID-001 sends 'abc' and the non-canonical spellings of a seeded row's id to each path param named like an id, and requires no 5xx and the answer 'abc' gets. It then sends the spellings of an id no row has and records what the database drivers are handed (tests/helpers/sql-capture.ts, below MikroORM and Kysely). It fails when a spelling, or the number Number() makes of it, reaches SQL, so a route over a table the suite seeds nothing in is held too. Stale NOT_ROW_IDS entries fail, and the walk runs on one listening keep-alive server.
- ROUTEID-002 walks every @trek/shared request, body and query schema and every route's body and query DTO, and fails on a field named like a row id that accepts such a spelling. String ids are listed with their reason; open fields name the handler test that holds them.
Three more suites cover the other entry points:
- MCPROWID does the same for every MCP resource template.
- RPCROWID refuses a plugin RPC id param read past rowIdParam.
- BODYID, JRNID and QUERYID cover the open accommodation update body, the collab reply, the journey fields a handler parses itself and the notifications limit and offset.
All suites that boot the app run in the SQLite suite and in the Postgres lane on 14 and 18. server/CLAUDE.md states the rule and the four gates.
Eleven MCP tool files carried their own parseId, which used Number() and so took '0x1' and '1e20'; they now share uriRowId. The websocket join and book:join read the id with toRowId instead of Number().
Plugin RPC ids go through rowIdParam instead of num(), and a param that may be left out goes through the new optionalRowIdParam. num() let '0x5', 1.5 and 1e20 through, which Postgres refuses or reads its own way while SQLite matches nothing.
accommodations.create read its refs with Math.trunc(Number()), and accommodations.update bound its open input's refs as sent. collab.createMessage took replyTo only when typeof was 'number', which let 1.5 and 1e20 through. All three now read their ids as row ids. A canonical numeric string still passes, as the published contract says, and the refusal keeps num()'s message.
The services and controllers that took an id from a request with Number(), parseInt() or not at all now use toRowId. That covers admin users, notifications (and their limit and offset), trip members and guests, day notes, assignments, memories (photo owner, album link, body photo_id), photos, mail-ingest, the assistant id parsers, dawarich, registration invites, the plugin contribution routes and the collab reply.
The accommodation update body is open, so AccommodationsService reads the place and day refs of every stay write with toRowId and answers a refused one with the create schema's 400.
AdminService.deleteUser checked the caller with parseInt and deleted with Number(), so an admin could delete their own account by spelling its id in hex; both now use one parse.
The OAuth session revoke called Number(id) first and so revoked session 5 for '0x5'; the service now reads the id as sent.
Request fields that hold a row id took z.union([z.number(), z.string()]), a bare z.string() or z.array(z.number()), so '0x5', '5e0', 1.5 and 1e20 reached the handler and from there SQL, where SQLite matches nothing, Postgres 14 and 15 refuse the cast and 16 and later read hex and underscores as a number. Two primitives join idParamSchema: rowIdInputSchema takes a safe integer or its digits and keeps the value as sent, so field types do not change; rowIdStringSchema takes digits only, for multipart fields. Both refuse with 'Invalid id'. The accommodation, assignment, reservation position, admin invite, collab reply, vacay, file and memories photo ids now use them, and every list of ids is z.array(z.number().int()). idParamSchema takes only the decimal spelling or a whole number, and the trip-journal-link and tour library ids use it. A body id spelled '0x5' is now a 400 on SQLite too.
RowIdPipe (src/nest/common/row-id.pipe.ts) parses a path param with toRowId. It answers a refused spelling with the route's own not-found answer, so '0x5', '5e0' and 'abc' get what an id that names no row gets. ParseRowIdPipe replaces Nest's ParseIntPipe with the same 400 and message: ParseIntPipe let '-1' and numbers past 2^53 through, and Postgres refuses the latter as out of range. Journey, journey recordings, day import, book design, map, public photo, collections, doc-sync, vacay, atlas and the plugin user-data export declare their ids through it. The upload routes parse in the handler with withRowId, so a refused id takes the existing cleanup path, or carry SpoolCleanupInterceptor (collections).
toRowId moves to src/db/row-id.ts so repositories can use it (src/db sits below src/nest), together with a new withRowId helper; src/nest/common/row-id.ts re-exports both. Its docstring is rewritten to state the rule plainly. TripsRepository read its number | string trip ids raw: SQLite matched '0x1' against nothing, Postgres 14 and 15 refused the cast with a 500, and 16 and later read it as trip 1. Its access reads, DaysRepository.existsInTrip and the packing template lookups now parse the id first, so an id toRowId refuses is no row on every engine. TripAccessGuard, TripOwnerGuard and JourneyAssistantGuard used Number(), which read '0x1' and '1e0' as row 1; they now use toRowId.
An attack on the move to Postgres with real SQLite files found five values
that changed without a refusal or a listing, or a valid file that failed:
- Negative zero in a plugin column became 0: pg sends the number -0 as "0",
and the checksum printed both as "0". It is sent as text and kept apart in
the checksum on both sides.
- Text whose bytes are not valid UTF-8 was read with each bad sequence
replaced by U+FFFD, and the checksum was built from the replaced text.
Text is now read as its bytes and decoded strictly in the file's encoding
(UTF-8, or UTF-16 for a file created so); such a value is refused in
TREK's and in plugin data, with --drop-nul too, and listed with its bytes.
- --drop-nul serialized a JSON value again, so a large number lost digits
and 1.0 became 1. Only the \u0000 escapes are removed now.
- A plugin default Postgres refuses for the column (DEFAULT 0 on a BLOB
column) failed the move, and DEFAULT 1.5 on an integer column would have
been rounded. Such a default is left off and listed as not carried over.
- A plugin index or primary key SQLite compared with COLLATE NOCASE or RTRIM
became one that compares exactly. It is still created, and the summary
lists the collation it lost.
Each case has a regression test that fails without its fix, on SQLite and
on the Postgres lane.
Postgres reads x.f as the call f(x) when x has no column f. The guard counted
that spelling only after ")." or "].", so over a function in FROM that
returns a single value, or a parameter, the allowlist was bypassed:
"select g.obj_description from generate_series(...) g" returned the comments
on every object of the database, and g.current_schemas, g.txid_status and
$1.ts_token_type passed as well. A refused catalog function after a dot is
now refused too, except the names in COLUMN_NAMES (money, name, oid, version,
xid, xml), which take no argument or only convert it to their type, so t.name
keeps working. PGPLUG-021 holds that list to pg_proc on the lane's server.
The type aclitem names roles: '=r'::aclitem printed the cluster's first
superuser and 'x=r/x'::aclitem told whether a role exists. It is refused in
every spelling like the reg* types.
The allowlist itself moves to assets/postgres/plugin-functions.txt, beside
the catalog list it is checked against. As TypeScript it spelled 'like' and
'json_each', names of Postgres functions that the SQLite dialect gate reads
as SQLite SQL.
Since the lane runs every suite that boots the app it takes about ten
minutes on Postgres 18 and about 25 on 14, past the job's 20 minute timeout.
Each major now runs it in three slices (--shard=i/3), each on its own runner
and service with its own template; the probe and its artifact stay once per
major, in the first slice. ci-ok still needs server-postgres, whose result is
failure when any slice fails or is cancelled, so every slice is required.
The booking body is not schema-checked, so day_id, end_day_id, place_id,
assignment_id, accommodation_id and the create_accommodation ids can arrive
as any string. The reference guards passed them to SQL as they came. SQLite
compares '0x5', '5e0' or 'abc' against an integer id and matches nothing
(or, for '5e0', the row), so the guard answered "Unknown reference". Postgres
refuses the cast on 14 and 15 and reads hex as a number from 16 on, so the
same body answered a 500 on one major and reached a real row on another;
the Postgres 14 lane caught it in L2.
Both guards now read each id through toRowId, the rule every other untrusted
row id follows: an id it refuses resolves to nothing, so referencesOutsideTrip
does not name it and unresolvedReferences does, and the answer is the same
400 on every engine. create() reads the payload's accommodation_id through
toRowId as update() already did, instead of storing '0x5.0', which the
reads' join could not cast back on Postgres.
ReservationRow and ReservationTraveler move to reservation-rows.ts, the way
reservation-inputs.ts holds the input shapes, so the service stays within
its size baseline; the service exports them on and it is lowered to 1166.
The fakes that answer safeFetch by API took their init as any; safeFetch
declares it as RequestInit, which is what they read (the body as
URLSearchParams or text). The SSRF album case had gained one more of them,
which put the tests' no-explicit-any count one over its baseline. All five
are typed now, and the baseline is lowered to the 1545 the tests hold.
The wiki's move section, the server guide and the nul-free header said
the move drops NUL as every write does. They now say what it does: a
plugin's data moves unchanged or the move is refused with the values
listed, TREK's own NUL is refused unless --drop-nul is given, which lists
every value it changed, and the checksum comes from the SQLite values.
The plugin in the move suite gains a BLOB column holding text and an
untyped column holding numbers, and the plugin store must read both back
as they were stored. A new suite moves a file whose plugin and own rows
hold NUL: the plugin's refuses before anything is written, TREK's
refuses until --drop-nul, a username that dropping the NUL makes
duplicate is named with users_username_unique, and a move without a
collision lists the value it changed and stores it without the NUL.
The move stripped NUL from a plugin's own text, merging keys that were
distinct, and turned a plugin's values into another type (text in a BLOB
column into bytea, numbers in an untyped column into text). Its checksum
was built from the converted values, so both passed as verified.
A plugin's data now moves exactly as the Postgres plugin store would
hand it back, or not at all: a NUL, a value of another storage class
than its column, or an integer past 2^53 refuses the move before
anything is written, listing plugin, table, row and column. A column
left open by SQLite's affinity takes the kind its values need, so text
in a BLOB column stays text.
TREK's own text holding a NUL from an older install is refused by
default too. --drop-nul drops it, lists every changed value in the
summary, and names a unique key that dropping a NUL broke.
Every table's checksum is now built from the values read from SQLite,
so a conversion that changes a value fails the verification.
nulFreeInput left a string as sent under any key containing "password".
That covered the admin settings password_login and password_registration,
which are flags: 'true\u0000' failed the lockout check as a disabled
password login while the database stored 'true'. It also tested the raw
key, so 'pass\u0000word' lost the NUL of the password it carried.
The exemption is now HASHED_PASSWORD_KEYS, the fields TREK only ever hashes
with bcrypt (password, current_password, new_password), matched exactly on
the key without its NUL. A unit test reads every field of the shared
schemas and the instance settings named like a password, passphrase or
secret and fails until each is either in that list or decided against with
its reason.
The plugin SQL guard refused functions by name, so every function that reads
the catalogs or the server and was not listed passed: obj_description,
col_description and shobj_description returned the comments on TREK's
tables, columns, database and roles by OID, format_type named a type by its
OID, and version(), current_database(), inet_server_addr() and the session
keywords described the server.
A call to a pg_catalog function is now refused unless the function is on
PLUGIN_FUNCTIONS (string, number, date and time, JSON, array, aggregate,
window, UUID, range, geometry, network and text search functions), in
either spelling of a call, f(...) and the column notation (x).f. The
catalog's function names of Postgres 14 to 18 are kept in
assets/postgres/catalog-functions.txt, and PGPLUG-021 fails on the lane
when the server has a function that list does not name, so a new major
cannot widen what a plugin may call. The session keywords current_user,
session_user, current_role, user, system_user, current_catalog and
current_schema are refused unquoted.
The plugin-sdk README, the Postgres wiki page and server/CLAUDE.md now name
the layers as the code enforces them: the refusal on a privileged role and
the boot log line, the guard (the reg* types, INTO after the statement's
own INSERT or MERGE only, TEMP only where it makes a temporary object), the
targets Postgres resolves before an ALTER, DROP, REFRESH, TRUNCATE or
CREATE INDEX, one statement per round trip, pg_temp last, and the quota over
every relation with the object cap. They also say that a plugin argument
holding NUL is refused on Postgres and bound exactly on SQLite.
The token guard kept leaking: on a superuser connection an unqualified
`alter type int4 rename to int4x` renamed pg_catalog's type for the whole
server, sequences and catalog-only objects grew the database past the quota,
regtype/regconfig/regdictionary looked TREK's objects up by name, and a
column alias named insert or merge in front of INTO made a table. Instead
of a longer word list, the store now asks the database:
- No plugin SQL while TREK's role is privileged in the sense of the restore
check (PRIVILEGED_ROLE_SQL, imported from server-checks.ts): every db:own
call fails with PRIVILEGED_ROLE_PLUGIN_REFUSAL, read once per pool
connection, and the boot log says why.
- The objects an ALTER, DROP, REFRESH, TRUNCATE or CREATE INDEX ... ON
changes, and the tables a CREATE or ALTER TABLE ties to its own, are
resolved with to_regclass/to_regtype in the transaction right before the
statement, and it is refused unless each one found lies in the plugin's
schema. An unqualified int4, text or box is pg_catalog's and refused even
with the role check off.
- The quota counts every relation of the schema with its TOAST table, and a
schema holds at most PLUGIN_SCHEMA_MAX_OBJECTS catalog objects (relations,
types, constraints, enum labels).
- The guard refuses every reg* type and its array in every spelling, the
has_*_privilege functions, currtid2 and the BRIN/GIN maintenance
functions; INTO only after the INSERT or MERGE that is the statement's
verb or a WITH body's; TEMP only where it makes a temporary object
(CREATE TEMP/TEMPORARY, SELECT ... INTO TEMP), so a column named temp
works again; a base type; and a NUL anywhere in the SQL text.
- A bound value or a migration id holding NUL is refused with the reason
instead of being stripped, so the plugin's keys never change meaning.
Every bypass from the re-review is a test on the real database, on the
role the plugin data is meant to run as and on a superuser connection with
the role check off.
5ff5e7660 stripped U+0000 from every argument a plugin bound to its own
SQLite file, but left the rows it had stored alone. A key written with a
NUL (through a bound arg or char(0) in the plugin's SQL) could no longer be
found by the same key, upserts added duplicates, and keys that differ only
in where a NUL sits ('a\0bc', 'ab\0c') collapsed into one and collided.
A plugin's data is the plugin's, so its file binds the arguments exactly as
sent again, as before that commit. TREK's own data still holds no NUL; the
Postgres store, whose text cannot hold the character, refuses it instead.
The privileged-role check missed three ways a role can reach the database
host from SQL. Before PostgreSQL 16 a member of a CREATEROLE role sets its
role to it and grants itself pg_execute_server_program, since role
attributes are not inherited but the grant checks the role in effect. A role
granted EXECUTE on lo_export, lo_import, pg_read_file, pg_ls_dir and their
siblings reads or writes server files, and lo_export keeps the file after the
transaction rolls back. From PostgreSQL 15 on a role can be granted ALTER
SYSTEM on a setting or SET on a superuser-only one.
PRIVILEGED_ROLE_SQL now counts all three. The parameter grants are read
through query_to_xml inside a CASE taken only on 15 or later, because a
statement naming pg_parameter_acl does not parse on 14. A SET grant on a
setting every role may change anyway, and pg_monitor, stay ordinary. The
lane test creates each kind of role and shows the escalation the check is
for on the server it runs against.
The database stores no NUL on either engine, but the application still held
the value a client sent. A key or a check made from it was then a different
value from the one the database compared: `victim@x.test` followed by NUL
was a fresh key for the per-email password-reset throttle while the lookup
found the victim's row, so every spelling mailed a new link and voided the
last one.
Every input now loses its NUL where it enters TREK: the nulFreeRequest
middleware after the body parsers (path, query string, body, cookies), the
first global pipe for what a route parses itself (a multipart body), the
websocket adapter for every frame, and the MCP registry through a new
host-supplied input mapper, since /mcp keeps its raw body. Headers need
nothing: Node refuses a NUL in one with a 400. A string under a key naming a
password stays as sent: it is only ever hashed, and stripping it would lock
out an account whose password was set with one. The stripping in the
database layer stays as the backstop for what does not come from a request.
Updating offered a second way to roll back a Postgres update: drop the
database, create it again and pg_restore the pre-migrate copy into it. TREK's
copy is a dump of its schema by name, so pg_restore creates public again,
stops at "schema public already exists" and rolls the load back, leaving an
empty database in the middle of a rollback. The page now says so and keeps the
steps that clear the schema instead.
Environment-Variables listed /usr/lib/postgresql/17 to 14 as the default
search path for the client tools; the code and the image start at 18.
createFromBaseline created the migration history's table through MikroORM's
storage before, and outside, the transaction that runs the baseline. Without a
master transaction MikroORM commits that table on its own, so a first boot that
failed in the baseline (a server without ICU, a timeout, a dropped connection)
left an empty mikro_orm_migrations behind. The next start with
RESTORE_FROM_BACKUP took that for a database in use, skipped the restore with a
warning and brought up an empty instance, after which the restore never ran.
The table is now created on the baseline's own transaction, so a failed first
boot leaves the database empty. The boot restore also treats a lone migration
history that records nothing as no database, which a restore clears anyway; any
other answer keeps it, so the check fails closed. Database-Postgres says both.
The two checks over every file under src/ awaited ESLint's config file by
file, which lint:warnings counts as no-await-in-loop. They read the configs
with Promise.all instead.
The guard's header, server/CLAUDE.md and the Postgres wiki page named
extended-query.ts and a hook on every pool connection. They now say that the
store sends each plugin statement over the extended protocol on a client it
checks out per call, and drops that client's temporary tables and session
state before handing it back.
Plugin statements on Postgres went over the extended query protocol through a
marker parameter and a hook that replaced query() on every client of TREK's
pool. That kept a plugin to one statement per round trip, but it changed
TREK's own hot path for a concern only plugin data has.
The plugin data store now checks a pg client out of the ORM's pool for each
transaction, through MikroORM's getNativeClient, and runs the transaction on
it directly. Every plugin write goes as a query with queryMode 'extended' and
every read through a pg-cursor on that client, so Postgres parses exactly one
statement either way. The settings the transaction makes are local, and
before the client goes back to the pool the store drops temporary tables,
cursors, listeners, session advisory locks and sequence state, or closes a
client it cannot bring back to that state. The row cap, quota, time budget,
untimed migrate, per-plugin queue, pool share and script commits are as
before, and plugin args are still sent without NUL.
The marker, the hook and its install in the Postgres driver are gone.
The ESLint walls against raw(), the sql tag and SQL text handed to
execute/run/all/get applied to src/db/repositories and src/db/seeders only. A
service, job, guard or MCP tool could send its own SQLite-only SQL through
em.getConnection().execute(...), which the Postgres probe never runs, since it
only calls the repositories. The walls now cover all of src/. The dialect layer,
the engines and the migrations stay the homes of raw SQL, and the four boot-time
lifecycle files that hold SQL text today are named and pinned to their count.
The wrapper took the chunk size of a stream as optional, which lint:strict refuses, and dropped the abort options Kysely hands executeQuery and streamQuery, so a cancelled query kept running.
resetDatabase empties tables and puts back the default categories one after another through the repo's oneByOne helper, where the order can matter, and the default addons and photo providers together; the MCP suites switch their addons on together and create ordered stops through oneByOne. The warnings this brings below the baseline, and the no-explicit-any and no-unused-vars counts earlier commits lowered, are lowered in scripts/eslint-baseline.json.
gen:pg-baseline rewrote the baseline with its cut at the newest migration on
disk, so after a migration landed after the cut, running it moved the cut over
that migration while Postgres installs created from the old baseline would
still run it through the migrator. It now refuses while a migration follows the
committed cut, names those migrations and the routine, and moves the cut only
with --move-cut (lint:dialect keeps such a migration under the post-cut rules
either way).
check:pg-baseline stopped comparing SQLite triggers with POSTGRES_TRIGGERS once
the baseline was frozen, so a trigger a later migration created on SQLite went
unnoticed. The comparison now runs in both states, with advice that points at
the migration when the baseline is frozen.
lint:dialect missed several classes of SQL that SQLite reads and Postgres does
not, or reads differently:
- group_concat, ifnull, instr, iif, printf, typeof, total(), changes() and
json_object/json_array/json_valid/json_type/json_quote, as calls in SQL text
and as the bare name handed to eb.fn.
- LIKE in SQL text, Kysely's 'like' operator and MikroORM's $like, which ignore
case on SQLite only. A LIKE that is exact on purpose now says so in a
like-exact comment; the six existing ones (setting keys, storage paths, the
lowercased reference search, sqlite_master) carry it.
- The sqlite member of engineDdl was not read at all, so a trigger or a data
backfill put there never reached Postgres. It may now hold only the schema
change and a rebuild's copy and sqlite_sequence bookkeeping, and SQLite
statements with an empty postgres member need a sqlite-only reason.
- execBestEffort, bestEffort and execUnlessTableMissing in a post-cut
migration: on Postgres a swallowed failure aborts the migration transaction.
- The post-cut rules follow POSTGRES_FIRST_CUT as well as the manifest, so
regenerating the baseline over a new migration no longer releases it.
src/db/engine/sqlite/ is SQLite's own code and is no longer scanned.
The gate matched only isSqlite(/isPostgres(, instanceof on a platform class
and a comparison on a property named engine or id. A destructured engine name,
the literal on the left, a switch over the engine, instanceof on a driver class
and a comparison with a driver class all passed with an empty baseline, and
standalone-orm.ts already held one such check. The regex now matches every one
of them, the guard in standalone-orm.ts moves into the engine port as
assertSqliteOrmConfig, and the advice names where a check belongs.
server/CLAUDE.md names where NUL is dropped, INTCOL and the .int() rule
for integer columns, and the calendar-date check for a bound date, and no
longer lists the places NUL suite as SQLite-only. The Postgres wiki says
the move drops a NUL instead of refusing it.
Postgres text cannot hold U+0000, so any write or filter carrying one (a
request body, a query string, an import, a feed, a forwarded mail, a
plugin's own SQL) answered a 500 on Postgres while SQLite stored it. Both
engines now drop the character where every value leaves for the database
(src/db/nul-free.ts): the MikroORM platforms when they inline a string or
serialize a JSON value, the Kysely dialects for every parameter, a
plugin's own SQLite file for its arguments, and the move from SQLite,
which used to refuse such text. The same request now answers the same
status and leaves the same row on both engines, so the places NUL suite
runs on Postgres as well and leaves SQLITE_ONLY.
The release queue narrows each owner's rows with LIKE '%run%' before it
parses them, and relied on SQLite's LIKE ignoring ASCII case. On Postgres
LIKE is exact, so a row naming an object in uppercase was not found and
the release deleted an object that row still names. likeAny now matches
LOWER(column) against the lowercase run, which reads alike on both
engines. RELI-010 fails on Postgres without it.
The nearest-day fallback of a booking's time cast the bound date to a
timestamp on Postgres, which refuses 2024-02-30 with a 500, while SQLite's
JULIANDAY rolled it over into March and placed the booking there.
dayDistance now refuses anything but a calendar date on both engines
(isCalendarDate), so a caller that forgets the check fails on SQLite too,
and findNearestDayId answers no day for such a date. The size baseline of
Reservations.repository.ts drops by one line.
A vacation day can be taken as a half day, so what a year carries into
the next can end in a half. vacay_user_years.carried_over was an INTEGER
column: SQLite stored 29.5 as a REAL, Postgres (bigint) rounded it to 30,
and the move refused such a SQLite value. The column becomes REAL on
SQLite (a table rebuild that keeps every id and the AUTOINCREMENT counter)
and double precision on Postgres, through the first migration after the
Postgres baseline's cut. VACAY-046 books a half day and reads the
carry-over on both engines.
The album list asks three sources at once; which call reaches the fake first
depends on how database reads interleave, which differs between the engines.
The fake now answers by API and category and the cases assert the retry with
a fresh session itself.
Every integer column is bigint on Postgres. SQLite stored a fraction sent
to one (a packing quantity of 1.5, a budget item for 2.5 days, an id of
3.5) as a REAL, while Postgres rounded it in a statement MikroORM renders
and refused it with a 500 in a Kysely statement, so the engines held
different data for the same request.
The request schemas of those fields now take whole numbers only, and
INTCOL (tests/unit/contracts/integer-columns.test.ts) holds the class: it
reads the integer columns from the entities and fails on any DTO field or
inline MCP tool field of that name that takes any number.
The Postgres wiki page and the plugin author guide list what the plugin
SQL guard now refuses and the two layers behind it, and the wiki and
server/CLAUDE.md name every kind of role a restore refuses, including
CREATEROLE on PostgreSQL 14 and 15, which the managed postgres role on
such a server holds.
The privileged-role check looked only at the role's own rolsuper and at
pg_execute_server_program and pg_write_server_files. A role that is no
superuser itself but a member of one passed it and could SET ROLE to the
superuser from a crafted archive; on PostgreSQL 14 and 15 a role with
CREATEROLE could grant itself pg_execute_server_program; and
pg_read_server_files let COPY read the server's files into TREK's tables.
The check now counts membership in any superuser role, the three
predefined server roles, and CREATEROLE before 16. The boot warning and
the restore refusal describe the same list from one constant. A Postgres
lane suite creates each kind of role and evaluates the check as it, and
shows on the server it runs against whether CREATEROLE can grant itself
pg_execute_server_program.
The guard ended a -- comment at LF only, while Postgres ends it at CR as
well, so "select 1 --\rfrom public.users" hid a qualified name from the
guard that Postgres then read, and with no args a second statement after
the CR ran as TREK's role. It also only looked for INTO at the top level
of a SELECT, so "(select ... into temp users)" made a temporary table that
shadowed TREK's own users table on the pooled connection.
The tokenizer now skips exactly what Postgres skips as whitespace, ends a
line comment at CR or LF, and refuses any character the two read
differently (other control characters, non-ASCII whitespace such as
U+00A0, U+FEFF and U+2028, and the non-decimal and underscored numbers
Postgres before 16 reads another way) instead of guessing. INTO is refused
anywhere but after INSERT or MERGE, at any depth, and the words TEMP and
TEMPORARY anywhere outside a string or quoted name.
The guard tests carry both payloads and their variants; the Postgres lane
runs them against the real database, proves a second statement in one
text is refused by Postgres over the extended protocol, and that a
temporary table never shadows TREK's or a plugin's tables.
The plugin SQL guard was the only thing between a plugin and TREK's tables
on Postgres, and two gaps in it were enough to leave the plugin schema.
Two layers now hold even when the guard is fooled.
A plugin write with no bound args went over the simple query protocol,
which runs every statement in its text. Plugin statements now go through
extendedQuery, which pg sends over the extended protocol whatever the
args, so a second statement is a syntax error instead of a statement that
runs. A marker parameter carries the request to the pool's clients, and a
client without the hook binds one value too many, so it fails closed.
pg_temp is searched before the search_path unless the path names it, and
a temporary table lives as long as its pooled connection, so one left
behind would stand in for users, sessions and every other table for the
next caller. Every connection of TREK's pool now puts pg_temp last on the
search_path the role configured, and a plugin transaction sets its schema
with pg_temp last.
They seed through the booted container's ORM instead of the SQLite snapshot
mock and reset through resetDatabase, so they run in the Postgres lane as
well. The tag notification case waits for the fire-and-forget row with
vi.waitFor instead of one turn of the event loop: on SQLite the insert
lands within that turn, on Postgres it needs a few round trips more.
It takes the ORM from the app buildApp boots, seeds through the factories
and resets through resetDatabase instead of the SQLite snapshot mock, so
it runs in the Postgres lane as well.
The MCP suites seed through the booted container's ORM instead of the
SQLite snapshot mock and reset through resetDatabase. The route inventory
never touched the database and only mocked it; without the mock it
compiles AppModule over the real module, so it runs in the Postgres lane
as well.
They seed through the booted container's ORM instead of the SQLite snapshot
mock and reset through resetDatabase, so they run in the Postgres lane as
well. The trip warnings suite seeded through the factories but only closed
the snapshot handle; it now boots the container to reach the ORM the reset
needs.
They seed through the booted container's ORM instead of the SQLite snapshot
mock and reset through resetDatabase, so they run in the Postgres lane as
well. The transit suite saves its permission levels through the
container's PermissionsService in a request context instead of one built
by hand over the SQLite handle.
They seed through the booted container's ORM instead of the SQLite snapshot
mock and reset through resetDatabase, so they run in the Postgres lane as
well. The road trip suite saves its permission levels through the
container's PermissionsService in a request context instead of one built
by hand over the SQLite handle.
They seed through the booted container's ORM instead of the SQLite snapshot
mock and reset through resetDatabase, so they run in the Postgres lane as
well. The plugin search suite seeded nothing through the snapshot handle
and only closed it; it now boots the container to reach the ORM the reset
needs.
They seed through the booted container's ORM instead of the SQLite snapshot
mock and reset through resetDatabase, so they run in the Postgres lane as
well.
They seed through the booted container's ORM instead of the SQLite snapshot
mock and reset through resetDatabase, so they run in the Postgres lane as
well. The collab sub-feature case in the help suite toggles through the
container's AddonsService in a request context instead of one built by
hand over the SQLite handle.
They seed through the booted container's ORM instead of a second ORM over
the SQLite snapshot mock, and reset through resetDatabase, so they run in
the Postgres lane as well. The settlement parity cases call the
container's BudgetController in a request context instead of one built by
hand over the SQLite handle, and the collections suite drops its raw
per-table DELETE for the same reset.
They seed through the booted container's ORM instead of a second ORM over
the SQLite snapshot mock, and reset through resetDatabase, so they run in
the Postgres lane as well. The PermissionsService the AirTrail cases save
through and the CategoriesMcp and DemoService the categories write
harness attaches are the container's own, called in a request context,
instead of instances built by hand over the SQLite handle.
include_coords compares the coordinates as a set: getTravelStats reads
them through a SELECT DISTINCT without an order, and Postgres returns
them in hash order.
They seed through the booted container's ORM instead of a second ORM over
the SQLite snapshot mock, and reset through resetDatabase, so they run in
the Postgres lane as well. The tour the assignment cases put on a day is
made with makeTour instead of a raw insert.
They reset through resetDatabase and seed through the ORM factories on the
booted app's ORM instead of the SQLite snapshot mock, so they run in the
Postgres lane as well.
They reset through resetDatabase and seed through the ORM factories on the
booted app's ORM instead of the SQLite snapshot mock, so they run in the
Postgres lane as well.
They reset through resetDatabase and seed through the ORM factories on the
booted app's ORM instead of the SQLite snapshot mock, so they run in the
Postgres lane as well.
They drop the SQLite snapshot mock and, where they seed, reset through
resetDatabase and seed through the ORM factories, so they run in the
Postgres lane as well.
They boot on a SQLite file the retired positional runner built, which no
Postgres database ever came from, so they belong in SQLITE_ONLY rather
than in the snapshot baseline. The lane scan now tells such a suite apart
from the snapshot mock: it mocks src/db/database onto a database of its
own, never names createSnapshotTestDb, and is named in SQLITE_ONLY with
its reason. A snapshot suite added to SQLITE_ONLY stays in the baseline.
They reset through resetDatabase and seed through the ORM factories
instead of the SQLite snapshot mock, so they run in the Postgres lane as
well.
IMMICH-057 and SYNO-035 need a shared photo row whose trip is gone, which
the foreign key's cascade rules out on a database that enforces it; they
write it with SQLite's foreign keys off as before and are skipped on
Postgres, which has no such switch.
They reset through resetDatabase and seed through the ORM factories
instead of the SQLite snapshot mock, so they run in the Postgres lane as
well. The OAuth service they drove by hand over a second ORM on the
SQLite handle is now the container's own, called in a request context.
The MCP audit cases wait for the mcp.tool_call row: the transport writes
it without awaiting it, so on Postgres it can land after the response.
They reset through resetDatabase and seed through the ORM factories
instead of the SQLite snapshot mock, so they run in the Postgres lane as
well.
The two cases about a NUL byte in a place name move to
places-nul-sqlite.test.ts, listed in SQLITE_ONLY: they pin SQLite's value
quoting, and Postgres text cannot hold U+0000. The control characters
0x01 to 0x1F stay in places.test.ts and run on both engines.
They reset through resetDatabase (the seam suite needs no reset) and seed
through the ORM factories instead of the SQLite snapshot mock, so they run
in the Postgres lane as well.
They reset through resetDatabase and seed through the ORM factories
instead of the SQLite snapshot mock, so they run in the Postgres lane as
well, the user erasure cases included.
They boot the app without the SQLite snapshot mock, so they run in the
Postgres lane as well. The demo seed suite spies on the real runDemoSeed
export instead of replacing the database module to wrap it.
They boot the app without the SQLite snapshot mock, so they run in the
Postgres lane as well. sessionRowsOf reads a user's session rows through
a booted app's ORM, next to sessionRows for the suites that still hand in
a better-sqlite3 handle.
Seeds through the ORM factories, resets through resetDatabase and mints
ws-tokens through the container's TokenService instead of one built by hand
over the SQLite snapshot mock, so it runs in the Postgres lane as well.
Seeds through the booted container's ORM instead of a second ORM over the
SQLite snapshot mock, and resets through resetDatabase, so it runs in the
Postgres lane as well.
src/db may not import src/nest, so the '<id>.0' shape is spelled where the
filter is. The flag type's undefined read had no test, which left the types
folder below its branch threshold.
resetTestDb needs the SQLite snapshot mock's better-sqlite3 handle, so a
suite resetting with it can only run on SQLite. resetDatabase takes the
booted app's ORM and leaves the same state on SQLite and Postgres: every
table but the kept ones emptied (read from the live schema), sequences left
counting, seedDefaults' rows back through the ORM, and on the first call the
users sequence restarted as the snapshot mock normalises it.
Reading the TEXT column as an integer also matched '14', ' 14' and '14abc',
which the legacy REAL-bound compare never did (AC37/AC40). Comparing with the
'<id>.0' text the services store keeps that contract on both engines.
Nothing is inserted after the first refused value, so the self-references held back had no rows to land on, and the move failed on them instead of listing the values (MOVE-006).
ESLint refused both: better-sqlite3 is imported only in src/db/connection.ts and its named neighbours, and an error thrown in a catch carries what that catch caught.
On Postgres a fire-and-forget notification still ran when the case read its recording, so it landed in the next case. The harness now waits for NotificationsService.settled() after each write and each exemption; a late send of any other kind still fails the next case.
Most writes start their notification without waiting for it. NotificationsService now tracks each send until it settles, and settled() resolves once every send started before it has finished. The channel check moves to channel-gate.ts, which keeps the service under the size limit.
The CI runner gives a child process a socket for stdout, which /dev/stdout cannot open, so every case whose pg_restore streams its script failed there. The COPY terminators in PGTOOLS-020 are spelled as a dump spells them.
trek/transactions-through-unit-of-work reads the receiver's type and reports
transactional() on anything but a UnitOfWork (an untyped receiver
included), and begin() or transaction() on an EntityManager, a connection
or Kysely. src/db/engine/ and the two boot-time schema files are left out.
It runs with the rest of the server's ESLint, so lint:warnings fails on a
new one, and its message names UnitOfWork, writeTransaction and the rule in
server/CLAUDE.md.
On Postgres a check read before a transaction, or outside one, no longer
holds at the write that follows: the pool runs other requests' writes in
between. Moved into the unit of work that writes:
- the last-admin guard of an admin's role edit and of an account deletion
(deleteUserCompletely takes a precondition run in its transaction)
- the login lockout guards of the auth settings save and the SSO settings
save
- the first-account check of an SSO sign-up, which starts over as an
ordinary sign-up when another account won
- the OAuth client caps, the token cap and the mail intake limits
- the year check before a vacay year is opened, and the label name check of
a rename
A password sign-up that loses the last use of an invite is now refused, as
the SSO sign-up already was, and a reset link is burned by one conditional
update, so two resets with one link cannot both change the password.
UnitOfWork took the engine's write lock, but four repositories opened
transactions of their own: the demo addon set, the plugin user erasure's
savepoints, and the plugin metadata and scheduler cap checks. On Postgres
those ran beside the others on the pool.
writeTransaction (src/db/engine/write-transaction.ts) is now the one
transaction that writes: UnitOfWork wraps it, and the demo seed and reset,
which run outside the container, call it with the engine the environment
names. The capped plugin writes run in their caller's unit of work instead
of a transaction of their own, and the plugin erasure runs in the erasure's
transaction without savepoints, since its three tables come with the
migrations on both engines. lint:tx counts writeTransaction as a
transaction.
The rotation ran on the SQLite file only and refused when TREK_DB_URL was
set. Its statements now go through a small store per engine: better-sqlite3
for the file, unchanged, and on Postgres the server's own MikroORM
connection, in one transaction that takes the engine's write lock first. The
Postgres path checks that it reaches a database holding TREK's tables and
that the data directory is writable before it asks for a key, writes a
pg_dump copy before it changes anything, refuses DB_PATH or TREK_DB_FILE
beside the URL, and reads the jsonb addon config the driver hands back
parsed. The SQLite transaction is now BEGIN IMMEDIATE, since better-sqlite3
takes no async function.
npm run db:move-to-postgres (dist/move-to-postgres.js in the image) copies
a SQLite file at this build's migration level into an empty Postgres
database. It refuses a file at another level, rows whose references point
at nothing and a target holding any table or plugin schema, creates the
schema with the engine's own bootstrap, then in one transaction copies every
table in foreign key order (setting a self-reference once its rows are in),
converts each value to its column's type and refuses one Postgres cannot
hold exactly, copies each plugin's plugin.db into its schema, moves the
sequences past the copied ids and compares every table's row count and an
order-free checksum over all rows before it commits. A failure drops the
schema again. The SQLite files and the uploads are never written.
Setting TREK_DB_URL on an existing SQLite install booted a new, empty
instance beside its data. The Postgres bootstrap now refuses an empty
database, or a migrated one without a single account, while the data
directory's travel.db holds accounts, and names the move command. A file
that cannot be read counts as holding them. TREK_DB_IGNORE_SQLITE_FILE=true
starts anyway, with a warning; the file is opened read-only and only when
the database holds no TREK data.
reservations.accommodation_id is TEXT and holds '14.0'. Compared with a bare
number, Postgres binds the number as the text '14' and finds nothing, so
deleting a stay, a day or a place left its bookings behind with no error. The
three filters read the column as an integer, as the joins already do.
The server compiles without strictNullChecks, where a narrowed value returned
as is keeps its full type, and two literal arms of one UNION column must share
a type.
Covers what every call sets up, the schema check before plugin SQL runs, the
row cap, the time budget between chunks and at statement_timeout, the quota,
a script committed statement by statement, the tx and migrate shapes and one
call at a time per plugin, without a database. The Postgres lane runs the
same store on a real one.
The plugin-sdk README says what dialect a plugin writes when TREK runs on
Postgres, what the guard refuses and how the limits hold there. The wiki's
Postgres page describes the plugin schemas, their backup and restore and
the one-instance-per-database limit, and its key rotation path now names
the file where it is.
pg_dump now names the trek_plugin_ schemas TREK's role owns beside its own
schema. A restore reads the dump's plugin schemas from its table of
contents, drops the plugin schemas the role owns, creates the dump's and
loads them with TREK's schema, all in the one psql transaction. Schemas of
another role and every other schema of the database are never read or
dropped.
The dump's migration history is read from TREK's mikro_orm_migrations only,
so a plugin table of that name cannot pass for it.
PLUGIN_DATA binds the plugin data backend to the engine (the same
{ sqlite, postgres }[engine.id] binding the backup port uses): the SQLite
file per plugin, unchanged, or the Postgres schema per plugin. The host
opens a plugin's store through it and an uninstall that deletes data erases
through it, which drops the schema on Postgres. Nothing outside db/engine
asks which engine runs.
A plugin's SQL is passed through as written, so a plugin needs to know the
dialect: db.engine (db:own) answers 'sqlite' or 'postgres', and the SDK
exposes it as ctx.db.engine(), in the mock host (dbEngine option) and in
dev. The plugin facts are regenerated and the SDK goes to 1.10.0.
The Postgres lane's PGPLUG suite installs, uses and erases plugin data on
Postgres and shows that a plugin reaches neither TREK's users table nor
another plugin's schema.
A plugin with db:own got a SQLite WAL file in the data volume whatever the
engine, so a Postgres install still needed a local disk (discussion #386).
This adds the plugin data port (db/engine/plugin-data.ts) and its Postgres
side: one schema per plugin, trek_plugin_<id>, reached through TREK's pool.
Every call runs in a transaction that creates the schema if missing, puts
only it on the search_path and sets statement_timeout to the call's budget.
Reads stream through a cursor and stop at the row cap, a write that grows
the schema past the quota rolls back, and the calls of one plugin run one at
a time on at most half of the pool.
TREK's role may read every table, and a role per plugin would need
CREATEROLE, so plugin-sql-guard.ts checks each statement first: it lexes the
SQL as Postgres does and refuses another schema as a qualifier, the pg_
catalogs and functions, functions that run SQL from a string or reach an
object by a name in one, temporary tables, and every statement but DML and
DDL on tables, indexes, views, sequences and types.
The users identity guard moves to repositories/_shared/user-identity.ts
(Users.repository.ts re-exports UserIdentityTakenError), the flag text pass
of the entity generator to scripts/generate-entities-flags.ts, and the
Wallet pass query takes the repository's EntityManager. The size baseline
of Users.repository.ts is lowered to what it holds now (1445 to 1421); a
few exports only their own file used are no longer exported.
The baseline was travel-baseline.db on both engines and went straight into
the swap, so a SQLite baseline left in the data directory made every hourly
reset of a Postgres demo fail inside pg_restore, and never got replaced. It
is now named after the engine's archive entry (travel-baseline.db on
SQLite, travel-baseline.pgdump on Postgres), and the backup port's new
checkCopy refuses one the engine cannot restore before the swap; the reset
then saves a fresh baseline and logs the reason instead of failing hourly.
reset-admin.js opened data/travel.db with better-sqlite3 whatever the
engine, so on a Postgres install it created a stray SQLite file or reset
the admin in a leftover one. With TREK_DB_URL set it now resets the account
in Postgres through the built server's engine (its pool, SSL and write
lock, in one transaction), and refuses a TREK_DB_FILE beside the URL; the
SQLite branch is unchanged. migrate-encryption.ts cannot re-encrypt the
secrets in Postgres yet, so it now stops before it opens or writes any file
when TREK_DB_URL is set. The wiki says what each does on Postgres.
SQLite takes NULL as the smallest value and Postgres as the largest, so any
list ordered by a nullable column (bookings without a time, trips without
dates) came back in another order on Postgres. Instead of every ORDER BY
spelling its NULL position, the Postgres engine renders SQLite's wherever a
statement names none: a Kysely query compiler for this.kysely() statements
and a MikroORM platform for find() and the query builder. An explicit
position is kept.
A new Postgres-lane suite checks the answers the probe cannot see: the NULL
order, a stay linked as '14.0' in the public reads, the traveller owner
flag, the identity errors for a lost race, and the plugin caps under a race
on the pool.
meta.set and scheduler.set checked their caps in a transaction of the
repository's own, which keeps check and write together on SQLite's single
writer but not on Postgres, where two pool connections both read the same
count. Both RPCs now run the capped write inside UnitOfWork.transactional,
which takes the engine's write lock; the repository's transaction becomes a
savepoint of it.
addons.enabled was the one flag MikroORM's boolean type declared as a
Postgres boolean, while SQLite and every other flag hold 0/1, so an addon
migration's usual SET enabled = 1 would fail there and a Kysely read would
answer true instead of 1. The column now uses IntegerFlagType: the entity
still reads a boolean, both engines store 0/1, and the regenerated Postgres
baseline declares it bigint default 0. The entity generator renders the
type for the columns in BOOLEAN_COLUMNS.
- Database-Postgres: supported versions, the version 18 client and the tool
search order, sslrootcert with a client before 16, how to create a role
that is not a superuser and move an older compose install to one, a
current schema other than public, and the space a restore needs.
- Updating: the manual rollback clears the whole schema and loads the copy
in one transaction, as TREK's restore does; dropping only the tables left
the collation, the trigger functions and the sequences behind. A unit
test keeps the printed statement equal to CLEAR_SCHEMA_SQL, and the boot's
rollback hint points at that section.
- Troubleshooting, Install-Docker-Compose, Environment-Variables, README
and server/CLAUDE.md follow.
The compose file let the postgres image create TREK's role as the
superuser, so a crafted backup restored by TREK could run commands in the
database container. The superuser is now postgres, with a password of its
own, and an init script creates the role trek that owns TREK's database and
nothing more. The Helm values say the same about the role in the URL.
A client handles every server up to its own major and refuses a newer one,
so the image needs the newest supported major, 18, which Debian 13 does not
ship. The PGDG repository is added with its signing key kept in the
repository and its component pinned to 18.
- The client tools are looked for newest first, from 18 down: a client
handles every server up to its own major and refuses a newer one. A dump
or a restore now refuses a client older than the server up front, naming
both versions, and the boot refuses a server outside PostgreSQL 14 to 18.
- sslrootcert=system is used only by a client from libpq 16 on; an older
client is refused with a message that names sslrootcert and the bin dir.
- The restore's CVE-2025-8714 table covers 18.
- A restore refuses a role that is a superuser or may run programs or write
files on the database host, since the archive's SQL runs as that role;
the boot warns about such a role.
- pg_restore's script is streamed to its work file and stopped once it
passes BACKUP_MAX_DECOMPRESSED_MB, so a small dump can no longer expand
without limit on the data volume.
- A failed tool's error keeps only the summarized stderr; its cause names
the exit code and signal instead of carrying the process error, whose
message and stderr quoted row data into the log.
- The migration history is looked for in the current schema, where the
tables and the backups are, so a role whose search_path is not public
boots a second time.
guardIdentity matched only SQLite's message text, so on Postgres a lost
race for an email or a username escaped as a 500 instead of a 409. The new
dialect module driver-errors.ts reads a unique violation by Postgres's
SQLSTATE 23505 and constraint name, or by SQLite's message, and the users
repository maps the key names of both engines. isMissingTableError moves
there too and also knows Postgres's 42P01 for an unconverted error.
The owner and member branches of listTravellerUsernames selected bound
values, which Postgres resolves to text in a UNION column, so the public
API reported every traveller as owner: false. The literals are written into
the statement as integers now.
publicStayExists cast reservations.accommodation_id straight to integer,
which Postgres refuses for the stored '14.0', so the calendar feed and the
share links failed for every trip once any booking was linked to a stay.
The correlation now goes through accommodationIdMatches, which compares
through the new dialect helper integerTextEqualsKysely, and the Wallet pass
context uses the same helper instead of its own numeric cast.
TREK supports Postgres 14 to 18, and CI tested 17 only. One server-postgres
job per major now runs the probe and then the lane against the same service,
so the change keeps the job count of the two jobs it replaces. The lane's
global setup refuses a service whose major is not TREK_TEST_PG_MAJOR, and
the image's Postgres boot moves to 18, the major its client tools target.
scripts/ci/workflow-jobs.test.mjs fails when a job other than the scan is
missing from ci-ok's needs, and changed-areas pins that migrations,
repositories, the engine and the lane's files always run the job.
The lane ran a hand-kept list of 26 suites that had already missed
trip-tags-mcp and the broadcast-parity suites, and listed two that never
touch a database. tests/postgres/lane.ts now takes every test file that
reaches src/bootstrap.ts or src/nest/app.module.ts through its imports and
does not vi.mock src/db/database, minus a named SQLITE_ONLY map with the
reason for each.
tests/unit/postgres-lane.test.ts (PGLANE, in the SQLite lane) fails on a
suite that names a boot entry point the discovery does not see, on a stale
exclusion, and on a new suite that boots the app over the SQLite snapshot
mock; the 105 that still do are held in snapshot-suites-baseline.json,
which only shrinks.
A frozen baseline still has its manifest verified, and the check now lists the
migrations that run on Postgres after the cut and what holds them. Every
problem it reports says what to do about it.
A migration newer than the baseline's through (read from
assets/postgres/baseline.json) fails on a schema statement outside engineDdl,
a plain string column for addColumnIfMissing or a SQLite-only helper, and on a
SQLite spelling in an engineDdl postgres member. The advice names the two
steps, and the baseline cannot hold an allowance for these rules.
The script builds a scratch database the way a boot does (the baseline, then
the migrations after its cut) up to the target migration, compares it with the
entities by PARITY-PG's rules and prints the remaining statements as the
postgres member of engineDdl. The scratch database is always dropped.
Which comparator statements change nothing (drop not null on a live primary
key, spelling-only defaults and checks proven by a rolled-back fingerprint, the
hand-ported triggers) is now one module the parity test and the migration
generator share, with unit tests for the sorting and the rollback.
A migration after the Postgres baseline's cut runs on both engines, but its DDL
is not one language. engineDdl(this, { sqlite, postgres }) queues the statements
of the engine the migration runs on, like addSql, and addColumnIfMissing takes a
column definition per engine, both picked in the dialect layer so a migration
never branches on the engine itself.
The merge with dev brought eight migrations written for SQLite (DATETIME,
INTEGER PRIMARY KEY without identity); run verbatim on Postgres they stop the
first boot. No Postgres install has been released, so the baseline moves to
the newest migration and those tables come from the entities.
TREK runs on PostgreSQL when TREK_DB_URL is set, and on SQLite as before
otherwise: an engine port under src/db/engine, the dialect helpers under
src/db/dialect, a Postgres baseline for the migrations up to the cut,
pg_dump/pg_restore backups, a probe of every repository method against a
real server and a Postgres test lane in CI.
The texts card explains each text in a line of its own and the phone sheet
follows it, so the description, notes and day note placeholders are no
longer read anywhere.
The text tabs name their field type for the switch, so it type-checks. The
colour field's sliders run left to right in every language and say so with
dir=ltr and a logical inset. The new phone fields take their sizes from the
type tiers or fs() instead of rem literals, the category field builds its
icon as an element, and the ESLint and theme baselines come down to what the
tree now holds. The tests follow the tabs (only the open tab's field and
formatting bar are shown), the category list and the files row, whose link
button keeps its place before attach.
jsdom 31 failed 56 client test files: it simplifies calc() and reads an
unset border as none in computed styles, and its File no longer carries
what the upload tests hand it. It moves back to 29 with isomorphic-dompurify
3, which needs it; the lockfile is resolved from the one before the bump, so
both sit at the root as they did. jsdom stays a root devDependency, so npm
keeps the copy vitest resolves at the root.
knip 6.45 reports tsconfig-paths, which test.yml loads from the server, as
unlisted rather than unresolved; it is the server's dependency. The removed
RTL text plugin took one swallowed catch with it, so that ESLint count is
lowered to what the tree holds.
Tags on the place form are pills in their own colour, with a palette under
the dot that recolours a tag for the whole trip (PUT /trips/:id/tags/:id was
there already, the client now reaches it through repo and store) and a
custom colour field of its own instead of the system picker. A new tag is
made in a colour picked the same way.
Description, notes and the note for one day's visit sit in one card under a
switch, each with a line on where the text shows up; the day tab is named
after its day and is only there with one visit in context. Nothing about the
three fields changes underneath.
Opening hours get a switch for open around the clock next to closed. Google
writes a place open all week as one period on Sunday that never closes; it
read as Sunday open and every other day closed, and now opens every day.
The phone form follows the desktop: category as one field with a list, tags
in one field, the texts switch, and phone, e-mail and opening hours, which it
did not have. The hours and the text tabs are shared hooks (usePlaceHours,
usePlaceTextTabs). Files and costs on the phone sheets are drawn as fields
like the rest, and the paste hint, a keyboard shortcut a phone has no use
for, is gone with its key.
A place keeps its one category, which decides its marker, and can now carry
any number of tags as extra labels to filter by.
- Tags belong to the trip (tags.trip_id): every member sees the same list and
whoever may edit the trip's places adds, renames and deletes. Personal tags
stay what they were.
- The place form picks them on desktop and mobile through one shared hook,
usePlaceTags, and can add a new tag to the trip on the spot.
- The existing category filter of the places list lists the tags as well.
- Tags show on the place (inspector pills, phone place sheet); the marker is
untouched.
- MCP: create_place, create_and_assign_place and update_place take `tags`,
and list/create/update/delete_trip_tag manage a trip's tags.
- /api/trips/:tripId/tags and the tripTags:changed event keep every client
of the trip in step.
The minor and patch group raised the Alpine pins to 4.68.3 while rollup
itself stayed on 4.62.0, so npm nested a second 4.62.0 copy under rollup
and the hoisted pins went unused. The pins follow rollup again.
The minor and patch group of #3108 without MikroORM: 7.2.9 refuses to set
the OAuth client reference from its client_id (the Postgres probe fails on
OauthTokens.insertToken and OauthConsents.upsertGrant), so MikroORM stays on
7.2.8 and with it better-sqlite3 and kysely, which it pins exactly.
mapbox-gl 3.33 shapes right-to-left text itself and removed setRTLTextPlugin,
so the engine registers nothing any more and @mapbox/mapbox-gl-rtl-text, its
Vite alias and the registration guard go. The WebAssembly module it loads
instead comes from api.mapbox.com, which the CSP already allows.
jsdom 31 and isomorphic-dompurify 4 move together, since the sanitizer needs
jsdom 31. jsdom is also a root devDependency now: listed only in the client,
npm nests it under client/ where vitest's built-in jsdom environment does not
find it.
dotenv 18 is loaded through `import 'dotenv/config'` only, which stays quiet.
Closes#3108, closes#3109, closes#3110, closes#3112
Since 1.22 Dawarich sends a visit's display name, which is "Unknown Location"
for a visit without a place. TREK showed that as the name, although the points
recorded during the stay are reverse geocoded. A stay without a real name now
takes its place's name or address, and otherwise the city most of its points
fall in. A visit without a place also takes the middle of its points as its
coordinate.
Closes#3105
The print-mode photos held their source back for the print view, but the
assistant also draws pages in print mode and needs its photos loaded. Only
the export's sheets now ask for it, the loader reads a minimal print
document like whenReady, and the new test helpers stay within the client's
TypeScript target and lint baseline.
After switching the display currency the flows stayed in the old one until
the new read landed, or for good when it failed offline, and a settle-up
booked those amounts in the new currency. The ledger now remembers the
currency it read, drops answers to older reads and only settles flows read
in the one shown.
A move keeps the entry's resolution and changes only its trip and created
ids, so an undo that read them before a move landed still claimed the entry,
removed nothing and left the moved booking behind. The claim is now bound to
the trip and ids it was read with.
The booking dialog and the phone sheet hid their departure and arrival tiles
as soon as the store held any leg of the flight, but a leg long over or never
seen by the feed draws no live card, so such a flight showed no times at
all. Both views now hide the tiles only for legs that get a card.
With two booked sectors under one flight number, the second is in its watch
window for the whole first one, and the aircraft flying the first sector was
read as the second having departed, with its departure notice hours early.
An aircraft that fits a sibling sector better is now left to it.
The CHECK on mail_ingest_cursor.transport did not allow 'notice', so the
hourly job that tells address holders the import is paused failed on its
lease insert every run. A new migration rebuilds the table with 'notice'
allowed, keeping every reader's position and lease.
The offline flag lived in the process, so after picking another server the
recreated activity still answered with the previous server's start page. The
copy now stands in only for the server it was activated for, and an online
start clears it.
A refresh stored the new start page before the build files it names, so one
cut off halfway left a page whose scripts were missing. The page now goes in
last, once every file is stored.
isBareHeaderBlock rescanned the header run below every sender line, so a
plain-text mail of nothing but "From:" lines without a subject took
quadratic time and held the event loop for seconds before the address was
even checked. The block starts are now worked out for all lines in one pass
from the bottom up, with the same rule.
Every printed photo got its src at once, off screen in the editor and again
in the print frame, so a journey with a few hundred photos asked for every
original in parallel and the browser failed with insufficient resources.
The print markup now holds the source back, and the print view hands it over
four at a time, retries a failed photo once, gives up on one that never
answers and shows the progress on the Save button.
Closes#3102
The desktop day list hides the stop a lodging booking puts on the day, but
the places are stored across the full day. Bookings and notes were positioned
by counting only the visible places, so every hidden row ahead of a stop
pushed it one slot behind the booking it was dropped above, and it fell
below the evening flight on every drag or arrow move.
Position bookings and notes in the same full-day count the places are
stored in, and share that count with the page's reorder.
Closes#3100
A tester reported that their own plugin, fine on the web, slowed the app down
badly on a phone. A sandboxed plugin iframe keeps running its timers, polling and
animation even off screen, and on a phone it shares the WebView's main thread with
the whole app, so a heavy one janks everything.
The frame now watches whether it is on screen and the app in the foreground
(IntersectionObserver + page visibility) and forwards it as trek:visibility, so a
plugin can pause its work while hidden; the design kit exposes trek.onVisibility /
trek.isVisible and pauses its own CSS animations via data-hidden. On a phone a
dashboard-widget plugin scrolled well out of view is suspended outright: its iframe
is unmounted and a placeholder holds the widget's height, so an offscreen widget
costs nothing and remounts when it scrolls back.
The picked visibility is a raised light segment on the grey track, like the
other pills in the window, instead of a black block, and the +N of further
bringers sits beside the first picture rather than over its initial.
The item window sits on the layer of the other floating menus and the Lists
menu just above the bars, through the z-index tokens instead of numbers, and
the theme baseline follows the files the redesign left with fewer literals.
Clearing a common item's owner and dropping the bringer row now happen inside
one unit of work, so a failure cannot leave one without the other. The lint
and RTL baselines follow the rows that lost their last physical sides.
Lower the lint and strict baselines the redesign made smaller, keep the
detail window and the todo tab under the complexity limit by moving their
choices into helpers, read the toast in the detail test through its module
rather than a window global, keep the new dialog focus case in the module's
existing test file, show the panel's progress once in the standalone header,
and point the planner wiring tests at the Lists menu.
The badge holds its text in an element of its own, which the ellipsis needs
inside a flex badge, and the row's middle column may shrink below its content.
The bringer tests read their items through a small type instead of any, the
panel hook names its two ceilings after the panels rather than the sides, and
the stored-width test runs on a window wide enough for the map's share.
The PENDING pill beside a title broke long names over several lines; a small
dot on the type icon's corner now carries the status's first letter (A for
Ausstehend, P for Pending), with the word as its tooltip and accessible name.
The time and the trip or place below the name are soft badges instead of a
line strung together with dots. Desktop and phone alike.
The search runs from a small button inside the field, dark once there is
something to look for, instead of a greyed-out button beside it, and the row
keeps its distance from the address box below. A dialog can now mark the field
it starts in, so this one opens in the search rather than in the name above it.
The collections map gets the trip map's padlock (#2010): locked, a click on a
place leaves the camera where it is, so places can be clicked through without
the map zooming in and out. One setting for both maps. On the full map a
picked place no longer brings the list back: its details dock on the map's
edge, or open as a sheet on a narrow screen.
The progress card says how many tasks are done with a slim bar, a round "+"
for a new task and a "..." for the order and the lists; the filters sit in
one tinted track, with "Mine" kept short there; and the tasks are rows of one
card, with priority and due date under the name and the assignee's face.
A row now shows its box, its name and only what is set: the amount from two
up, the pieces packed while packing, and who brings it. A click on the name
opens a small window at the row with the list, amount, pieces packed, weight,
bag, who brings it and who sees it, each as one pill, plus copy and delete,
instead of fields, a bin and menus in menus on every row. Above the lists the
Lists bar keeps Add list and gathers templates, import, export, the order,
hiding packed items, the bags on a narrow screen and the clean-up in one menu;
the progress moves next to the Shared/My list switch. The bag overview stays.
The phone follows the same shape: one "..." sheet for the list's actions, a
sheet per list for its own, a slim row and an item sheet that saves as you go.
Who may change what about an item lives in one module both views read.
A common item only let you pledge yourself. Nobody else could be named,
and a bringer could never be removed, so an item that went personal and back
kept its owner's avatar for good. The item menu now lists every member to
tick on or off, a click on the avatar opens it there, and the add route takes
an optional user_id limited to the trip's members. Taking the owner off
clears the owner. The remove route now checks visibility for the caller, not
for the user named in the path.
The side panels stopped at a fixed 520 px, which left a wide screen with a
narrow day plan. Each panel may now take whatever the window leaves once the
map keeps at least 35% of it (never under 360 px), the day plan keeping its
width first when the window shrinks. While a handle is held the width
transition is off, so the panel no longer trails behind the map and its
buttons.
The packing list keeps its entry field open after Enter, and each added item
lands above it. With the keyboard already up no keyboard event fires, so the
field slid under the keyboard after the first entry. While the keyboard is up,
a change to the page now checks the focused field again, once a frame at most.
A nullable day_assignments.stay_minutes holds the stay of one visit; a visit without one keeps reading the place's duration, so existing trips drive as before. REST PUT /trips/:tripId/assignments/:id/stay and the MCP tool set_assignment_stay write it with day_edit.
Follow-up messages failed with HTTP 400: a whitespace-only text block, two model messages in a row where the turn had pushed for an answer, or a tool call renamed after its provider state was recorded. The rebuilt history now matches what the provider saw during the first answer.
Keeping the whole right inset clear gave up a strip of the full height for a
capsule that only sits in the corner. The iOS app now measures an element in
a top corner (iOS 26 safe area with corner adaptation, vertical against
horizontal) and hands its reach and width to the page: the floating bars
stand beside it, the page keeps the full width and starts below it. Without
such an element, on every other device and before iOS 26, nothing changes.
TestFlight feedback from an iPhone Duo:
- Folded, the status capsule stands in the right inset and ran over the hero
card's actions; the page now keeps clear of the side insets as iOS asks
(.m-page), and a floating bar keeps a little air next to an inset.
- Unfolded, the top bar ignored the trailing inset and the capsule sat on the
avatar; on a wide window the bars now end before it, by their box, so a
bar that centres itself still does.
- Dragging the dashboard's second column back up refreshed the page: a pull
that starts inside a box scrolled on its own now belongs to that box.
A tester with a server on his home network could not open the Android app
away from it: after closing the app and turning Wi-Fi off it stayed on the
address screen ("No server answered at this address").
- Android keeps the server's UI itself (TrekOfflineShell): the start page
and every file of the build, listed by native/offlineShell.ts from the
server's manifest and downloaded natively. A web view runs no service
worker on a plain-http server, the usual address of a server at home, so
the old "Android has the service worker" assumption did not hold.
- Without a network, or once the server failed to load twice, the web view
serves the copy under the server's origin with Capacitor's bridge; the UI
runs on its own offline data and the API answers again once the server
is back. The address screen opens the copy too (openOfflineCopy).
- The check an update asks for no longer locks the app out offline: on both
platforms a server out of reach is put back on its copy and checked on
the next launch that reaches it. iOS no longer throws its copy away when
the update check runs.
The user sheet's actions on another account (passkeys, two-factor, SSO
link, delete) are MUserDangerActions with their own confirms, the place
names picker is MPlaceLanguage, and useApiKeys takes its copied flag from
a small useCopiedFlag.
- MApiKeysSection: the keys with what each may read, creating one in a sheet
(name, the areas as chips, the key shown once with a copy button), deleting
behind a confirm sheet, and the endpoint to copy.
- The list and every action move from ApiKeysSection into useApiKeys, shared
by both views; the calls go through a small online-only apiKeysRepo, so no
view reaches the API (offline baseline loses ApiKeysSection and
AdminSettingsTab).
- The phone's Integrations tab is always there now, as on the desktop: the
keys in it need no addon.
- A test title no longer breaks on its apostrophe.
- Admin notifications: the defaults every user's cells start as, one row per
event with a pill per channel cycling On, Off and Blocked
(MAdminNotificationDefaults); the grid's constants are shared with the
desktop panel in notificationDefaultsModel.
- Editing a user offers "Remove SSO link" behind a confirm sheet, as on the
desktop, once SSO is set up and never for yourself.
- Settings: the wanderer connection (MWandererConnectionSection, on
useWandererConnection) and the language of place names under General.
- MSetCard takes a brand mark as its icon, not only a lucide icon.
- Place search and routing as one card, as on the desktop: TREK's own
services, today's allowance with a new identity behind a confirm sheet
(MAdminServiceQuota), sharing place details (MAdminPlaceContrib), the transit
backend and the place search provider as TREK selects, and the search log.
- The key card holds Google (with its options and the daily request cap,
MAdminGoogleDailyLimit), Unsplash and Amap, in the desktop order.
- The logic both views need moves into one place: the transit switch into
useAdminSettingsActions, the allowance figures into serviceIdentityModel,
and the share-place-details panel into usePlaceContribPanel.
- MAdminSettingsSection renders each section in a function of its own.
- Tests follow the split admin sections (Save order, 2FA label, field marker).
- The phone's admin sections follow the desktop sidebar: the same groups
(Users, Configuration, Maintenance) in its dropdown, Security and Services
instead of one Settings section, and live flights and booking mails beside
the addons they configure. An old ?tab=settings link still opens Security.
- Passkey login and the 2FA rule sit with the other ways in; the passkey card
keeps the domain, and a managed install no longer shows the danger zone.
- MAdminSelect replaces every native select on the admin screens: TREK's own
field or row pill that opens the picker sheet (permissions, backup schedule,
invite trip, plugin settings).
- Explanations wait behind an info mark (MHintLabel), as on the desktop: admin
rows and fields, the settings rows' new `hint`, addon and provider rows.
- Permissions are one line each with the level as a pill; addon rows lose the
type badge their heading already names; settings buttons no longer wrap.
- "Require 2FA" in every language; admin.tabs.settings is unused and gone.
The settings redesign folded the integration sidebar group into config,
but ManagedAdminTab still accepts group 'integration'. Indexing the label
map with it was an implicit any under the strict check and would have put
such a tab under an empty heading. It now lands under config.
The tests now look for what the pages show after the redesign: the admin
Settings tab is Security and Services, MCP Access and User Defaults are MCP
and Defaults, notification channels and the 2FA rule are rows with a named
switch, hints sit behind the info mark, each addon type card counts its own
addons, OAuth scopes and the AI sharing note are shown in full, and the 2FA
disable password is the field beside the code. The phone's MCP admin header
keeps its own title, MCP Access. The mail pill test renders with
translations, so it finds From mail rather than the key.
The phone's MCP card still read the config fold state that the shared hook
no longer holds, so the client failed to typecheck. The phone keeps its
folded config, so it now holds that state itself.
SettingsColumns takes start and end instead of left and right: the prop
types counted as two physical sides for lint:rtl, and in a right-to-left
language the first column is the right one anyway.
Also: an unused icon import, seven files outside the format baseline run
through Prettier, Routing (de) and Services (fr) confirmed as the same word
as in English, and the size and warning baselines lowered to what the
redesign left (AdminPluginsPanel one wrapped line shorter, two promise
warnings, one long function and two non-null assertions in tests fewer).
The settings pages had grown into a mix of styles: folds, dialogs behind
rows, full-width rows next to boxed ones, pills for every figure and buttons
in four sizes. Every page now uses the same kit.
- Cards keep their tinted head band; every part of a card is a rounded block
with its own small head (SettingsBlock), and rows sit in rounded boxes.
- Nothing folds and nothing hides behind a dialog: the MCP config, OAuth
scopes, mail how-to, AI privacy note and addon sub-options are all shown.
- Each page is split into two columns balanced by length (SettingsColumns).
- Long explanations moved into an info mark next to the label (HintMark).
- One button size everywhere (text-control); tab names are one word.
- Admin: the old Settings tab is split into Security and Services
(?tab=settings still opens Security, and the phone keeps its single
Settings section for both); the Integration group is gone, live flights
and mail import sit next to the addons they configure.
- Addons: one card per type with rows, no search, filter or fold.
- Quota, TREK API facts, mail import status and Google's daily limit are
compact lines and strips instead of tiles and pills.
- Language is a select; routing has its own card under the defaults;
storage saves from the foot of its categories card.
- Unused keys, FoldSummary and the page-width wrapper are removed.
The settings pages list everything at once, so they read as a wall of
forms. SettingsCard and Section take 'collapsible': the card starts shut
and its badge says its state ('Connected', 'Not set up'); its title is the
button, stretched over the head, the head's action stays its own button.
The body stays mounted while shut. A link to #id opens the card and
scrolls to it; the browser remembers a card left open.
Every integrations card is collapsible now; AI parsing shows its saved
model and MCP the apps signed in. The assistant's settings link opens the
AI connection card. Tests that read into a card open it first
(tests/helpers/settingsCards.ts).
From xl up, as on the admin settings: what connects your trips with other
services (booking mails, photo servers, AirTrail, Dawarich, Wanderer) on
the left, AI and developer access (AI connection, AI parsing, API keys,
MCP) on the right. A column with nothing to show drops out and the other
takes the whole width.
Behind the title stood the review mark, the 'From mail' pill and the two
actions, and the title was cut short. Where a booking came from is looked
up in its detail dialog, so the pill stays there only. The AirTrail note
sat above the mail pill; it is back above AirTrailPill.
- The mail import's status tiles (which, labels, tones) were written out
in the desktop and the mobile card alike; statTiles in the shared model
holds them, each card keeps its own look.
- The journey share link set 'copied' back two seconds after a copy even
when the dialog had closed, which also broke a client test run (window
gone at teardown). useCopied, the trip share's hook, moves to hooks/ and
serves both; it clears its timer on unmount.
- PlanLlmConfig is not exported: only the union is used outside.
Only an upload started a processing run; a forwarded mail waited for the
next two-minute tick, and one that came in right after a run waited for two.
The IMAP and relay runs start one run once they took anything in, the
webhook once per mail it accepted.
Stalwart writes Delivered-To as the mailbox itself, its sub-address dropped,
so a forwarded mail to trek+<token>@ was refused as unknown_address: the
first delivery header won even though it carried no token. The headers are
now read in their order of trust, and the first that names a live address
decides.
A ChatGPT sign-in only ever ran the assistant: parsing's clients call an
endpoint with a key, and a plan has none. AI connections now lends the plan
as a reader. Each document is one Responses call with the schema as the one
tool the model has to call (tool_choice required). The token is read and
refreshed per call by ChatgptAuthService, and a sign-in that has to be
renewed lends nothing.
- ResolvedLlmConfig is a union: ApiLlmConfig (what an admin configures) or
PlanLlmConfig (provider chatgpt, a read function, never stored).
- ConnectionReadClient sends the text plus photo or scanned pages, and reads
the filled shape, or the shape written as text; prose is unreadable.
- AiChatClientService.chatgptPlan is the one place that builds the plan's
client, for the assistant and for parsing.
- Wiki: a ChatGPT sign-in reads bookings on its plan.
- The upload posted its FormData through the client's JSON default, so the
form went as JSON and the server saw no file (400). It goes through
postMultipart now, like every other upload.
- reservation-origin.ts becomes reservation-origin.contract.ts: the booking
import sets an origin, and a contract is what another domain may import.
- CI: the client's strict type check holds the whole program with strict
checks and outgrew the default ~4 GB heap; it gets 8 GB like the server's.
The own address was made only when asked for, and the card had no way to ask
before one existed, so it showed none. Once a template is set and the
person's import is on, the first look provisions it (idempotent; read after).
- auth.service: how an admin form value is stored (validated, encrypted, a
masked echo kept) moves to the app settings registry as storedAdminValue
and isMaskEcho.
- reservations.service: a booking's origin (withoutOrigin, originColumns)
moves to reservation-origin.ts, its input shapes to reservation-inputs.ts,
exported on from the service so importers stay as they were.
- Reservations.repository: the joined projection, the origin lookup and the
origin columns move to _shared/, like the Wallet pass context before.
- test-sql: MailIngestApplyService.prepare is marked as no SQL statement.
- The pending bookings dialog opens centred, and says 'nothing waiting' once.
The last read sits in the head line with a green dot, amber while the last
run failed; the day's numbers are tiles with their icon in a soft tone
(mails, added on their own, waiting, failed turning red above zero); the
last error is its own amber note. Desktop and phone.
Upload only, IMAP and webhook each explain themselves in the admin card,
desktop and phone: what the way means, what it needs (a mailbox of its own
taking plus addresses, a signing secret), with its icon centred beside the
title and the sentence. All 28 locales.
add, move, file, discard and undo of a mail booking and the import
settings each get a case with their REST twin. A trip expectation may
name its own room, for the move that announces in two trips.
- mime-walker: trim each line of a stripped HTML body, so the space a tag
leaves is not read as a folded header line that swallows the booking
under an Outlook forward's header block.
- settings: an unset daily AI limit is the default of ten, not zero
(Number(null) is 0), which kept the AI from ever reading a mail.
- env: MAIL_INGEST_IMAP_PASS and MAIL_INGEST_WEBHOOK_SECRET are secrets
and never printed in the boot report.
- pipeline test: the subject is kept without its forward prefixes, the
demo move adds its booking before the second trip exists, the typed
notification mock, and AI reading stays offered while the person
switched it off.
- e2e: the pipeline tick runs in a request context of its own, as the
cron registrar runs it, and the 401 checks run without awaiting in a loop.
ai_available folded the person's own switch in, so switching AI reading off
greyed the switch out and it never came back. It now says only whether AI
reading works on the instance (the AI Parsing addon and a connection). When
it does not, the settings card says it needs the AI Parsing addon, by its
localized name, in all 28 locales; the wiki says the same.
- mailInboxModel.test: actionsOf takes the entry alone since the AI reread left.
- booking-origin.test: its endpoint helper gives the sequence an endpoint has.
- test-new-service baseline follows booking-import's single service factory.
- lint:tx: the mail transports and the intake document why their writes stand
alone (@txIndependent, network I/O between them, one copy per address); the
JSDoc of NotificationsService.respond is back above it after the response
actions were registered in between.
- lint:boundaries: booking-import asks McpToolGuardsService whether the person
may edit the trip's bookings instead of reading the Users table; its unit
test builds the service through one named factory.
- lint:query-api: findIdByExternal reads through Kysely.
- lint:test-mocks: the mail e2e hands its own snapshot db to the ORM instead
of mocking the database module.
- Entities regenerated (index order only).
- Client: the review row reads the live booking, the shells share
useMailInboxShell, and the admin card's buttons are one size down.
- Admin settings (desktop and phone, self-hosted only): the instance switch,
how mail arrives (IMAP or a signed webhook) with its fields, the address
template, the AI allowance per person and day, a connection test and the
status in counts. Changed fields only are saved; secrets stay masked.
- The review mark of an imported booking takes its mark off: the pill in the
desktop dialog and a row in the phone's transport sheet ('Looks right'),
through the ordinary booking update. Its hint no longer speaks of airports.
- The inbox follows mailInbox:changed and mailInbox:settings, and
?inbox=mail (the 'bookings waiting' notification) opens it.
- mailIngestRepo: the online-only path the views take, answers parsed
against the shared contracts.
- Dropped 'read with AI' and the unused mail-review contract: the server
reads a passing AI failure again on its own.
- mailInbox, mailAdmin and the new reservation, settings and notification
strings in all 28 locales; the pending notice is a plural group.
- The live flight line on the booking cards is one of the card's fields,
full width on the phone; the phone's live card keeps its map to the middle.
- Admin flight and mail cards sit in the left column.
The cron callbacks hand back nothing; the cursor's lookup is findCursor so it
no longer shadows the base repository's find(); the draft's suggested trip
uses the shared contract's type; endpoints spell out their optional columns;
the relay page's running count is typed. Prettier over every new file outside
the generated entities.
A plan without the import switches it off from outside the server, and an
admin can switch it off too; the plan promises a "mail import paused" note.
An hourly job sends mail_ingest_paused to each holder of a live address with
their own import on, notes the pause first, and forgets it once it is back.
The settings event was registered with no sender; renewing or revoking an
address and changing the switches now send it. The client lists both events
as not yet shown, which keeps the registry-parity test green until the mail
screens move them to their listener.
Describes TREK/deferred, which header names the address, the webhook's 401
and 429 answers and where the sender hint can show at all. The variables
table leaves an empty default blank instead of a dash.
The webhook refuses headers that cannot be signed right before reading a
byte, limits each address to 120 requests a minute and reads four bodies at
once. The admin save refuses a webhook secret under 32 characters, and the
MCP settings tool registers its input strict like every other tool.
Anyone could add an Authentication-Results header with dmarc=pass. The hint
now needs a DKIM pass signed by the From domain or a DMARC pass naming it,
and reads it only for IMAP, where the receiving server's header is topmost;
the Cloud relay, the webhook and uploads never show a sender as verified.
A co-member could move the forwarder's booking into a private trip, carrying
their mail along; undo stays open to editors. Move and filing refuse demo
accounts, and filing claims the entry before writing documents, putting it
back into the inbox when nothing could be filed.
An upload made an address and stored a mail the pipeline then put aside;
it now answers 409 instead. The run an upload asks for at once starts only
where the crons run, so it can no longer outlive a test's database.
The relay job reads the control plane's { messages, more, next } page, checks
each mail on its own, pages on while more wait and is idle on a 404. A mail
for several addresses is stored for each, and a passing refusal (inbox full,
daily share taken) leaves it with the relay or in IMAP's TREK/deferred.
/api/mail-ingest serves the settings card, the inbox, the mail preview, the
.eml upload and every action on an entry; trip addresses and the admin status
and connection test have routes of their own. The MCP tools act through the
same services and never hand out an address, and the trip assistant is not offered them.
The IMAP job reads the box every two minutes through the SSRF guard, keeps a
uid cursor and moves read mails to TREK/processed. The webhook takes raw MIME
signed with HMAC-SHA256 over a timestamp, the recipient and the mail, and is a
404 unless it is the transport. The relay job collects a managed install's mails.
The inbox pages a person's entries with the mail each came from and the trip
it suggests (named only when the reader can see it), shows a mail as plain
text, and takes an uploaded .eml, reading it at once. Move writes the booking
anew in the other trip and removes it from the first in one transaction.
The IMAP box, the webhook secret and the address template join the admin
settings form (masked, encrypted, held by the operator on a managed install),
and MAIL_INGEST_* variables override them through the mailIngestConfig token.
The audit row masks every masked key, and key rotation covers the new secrets.
Addresses per person and per trip (random base32 token, stored hashed and
encrypted), an intake every transport shares, and a job that reads each
mail once under a lease: filters without AI, machine-readable data always,
one AI request with the person's trips, the server's checks G1 to G9, then
one transaction per mail that adds what passed, keeps the rest in the inbox
with a reason, and notifies the forwarder and co-travellers. Undo within
7 days, manual apply, filing as documents, a prefilled form and a nightly
retention complete it; on a managed install the operator's switch decides.
mail_booking_added tells the forwarder (with keep or undo) and co-travellers
of a shared trip; mail_booking_pending points at the inbox. Both are listed
only while the mail import is on, with texts in all 28 locales.
registerResponseAction lets a domain name what a yes/no tap does without
this module importing it.
MailReadService sends one mail as one document: its subject, body and up to
three attachments inside data tags the prompt calls untrusted, with the
person's trips under references made up for the call. Answers keep only
offered trip references and known field names. BookingImportService reads
a mail's machine-readable parts and maps an AI reading to booking items.
Five tables carry the mail import: forwarding addresses (token hash plus the
encrypted token), a person's switches, the readers' cursor and lease, every
mail as it arrived, and each booking read from it with its checks and fate.
Repositories read the trips a booking may go to by column, so the import
weighs a booking without reaching into five other domains.
The Ollama router catches a failed model call and answered an empty list, which
the outcome read as a document without bookings. It now returns the outcome with
the list; failureOutcome and foundOutcome move to llm-outcome.ts so the router and
the service share them without importing each other.
The self-hosted import reads its catch-all mailbox over IMAP; imapflow is the
client the transport is written against. The lockfile follows with the install.
walkMail() reads a forwarded mail for the import: the envelope apart, a body
without From, To, Cc or forwarding header blocks, message/rfc822 and quoted
forwards unwrapped up to depth 10, and the PDF, pkpass, ics, JPEG, PNG and WebP
attachments whose bytes match their type. Nothing is rendered or fetched.
Zod contracts for the forwarding address, the person's switches, messages,
the bookings read from them, the inbox and every action on an entry, the upload,
the webhook and the admin status card. One reason enum names why a mail or booking
waits, mapped from every AI outcome; mailInbox:changed and mailInbox:settings are
registered as user-scoped events.
LlmParseService.parse() now says how a reading ended (ok, empty, no_text,
quota_exhausted, ai_disabled, user_cap, provider_error, too_large). The clients throw
LlmRequestError with status and body code, so the TREK Cloud gate's 429 QUOTA_SPENT
and HARD_CAP read as a spent allowance. The preview reports the outcome per file and
the assistant reads its empty reason from it rather than matching warning strings.
confirm() now answers { created, failed: [{ index, reason }] } instead of logging
and dropping the items it could not write, and checks the trip and reservation_edit
itself so the import route and the assistant get one answer. An item with a booking
reference records it as its origin (sync off) and a repeat import is a duplicate.
CreateReservationData and insertReservation take an optional external_source,
external_id and sync_enabled, so an importer can record where a booking came from;
without them every write keeps the column defaults. Client bodies on REST and the
plugin RPC have the origin stripped, so no caller can pose as an import or a link.
getReservationWithJoins now takes the trip and reads nothing outside it.
Reading the bridge's MaplibreGL export threw under the test mocks, which
define only the factory, so every mocked basemap counted as an unloaded
chunk. The guard now patches the class of the layer being attached, and
only a class that has _transitionEnd itself, so stand-in layers are left
alone.
The resize guard reached for the bridge's MaplibreGL class, which the test
mocks of the bridge do not provide, so every mocked basemap counted as an
unloaded chunk. The guard now only applies where the class exists. The
mini map hides place names through VectorBasemap's own hideLabelLayers
(a hideLabels prop) instead of a second copy of that logic, and the route
test no longer uses Array.prototype.at, which the client's lib target lacks.
The live flight card shows a small real map behind the pass: the great
circle between the airports, solid where flown and dotted ahead, the
aircraft on the line turned along the route. The route is framed in the
free middle, the map fades out under the airports and times laid over it,
and its place names are hidden. Data and map credits sit on the map as its
hover text instead of a footer line.
The card head now follows the booking dialog: a tinted band with pills for
the state, what comes next and the delay; the numbers of the moment use the
dialog's fields on the desktop and the sheet's stat boxes on the phone; both
times sit on equal tiles in one line. Without live data the card, the strips
and the plan line show nothing, and the pulsing dots are gone.
Also fixes the failed dev run: unit symbols and the leg position are listed
as NOT_PLURAL, the aircraft label uses logical sides, the flight store is in
the test reset registry, the sync loop stops once no flight lies ahead (no
endless timers under fake timers), and timeCell and widgetFlight drop below
the max-params and complexity limits. Guards the GL bridge's resize frame
against a layer removed before it runs, and keeps the booking code pill as
tall as its neighbours.
The transport page explains what TREK follows, where it shows and what open
data cannot give; the environment page documents the switch and what leaves
the server.
Android lint refuses a composition reading the Configuration through
LocalContext. The arc's mirroring now follows the locale's layout direction,
which is what the widget's language decides anyway.
A new Flight widget on iOS (small to large and the lock screen) and Android
shows the trip's next flight or the one in the air: both airports, the arc
with the aircraft, a countdown the system keeps running, and the delay. It
follows the feed's phase, so a late departure is not flown along the arc.
Around a flight the travel day's Live Activity becomes a boarding pass with
a running bar and countdown, in the Dynamic Island too.
The booking dialog and the phone's transport sheet show a flight's moment,
both airports with planned and expected times, the route as an arc with the
aircraft on it and the numbers of the flight; booking cards and day plan
rows get a one-line state with a progress bar. The trip map draws the
aircraft in the air on Leaflet and GL, and the admin settings carry the
switch. Translated into every language.
A new flight-live domain polls the open feeds adsb.lol and adsb.fi (no key,
no cost; adsb.fi only on self-hosted installs, its terms are non-commercial)
for every booked flight leg in its watch window, derives phase, take-off,
landing, delay and an arrival estimate, and pushes changes as flight:live.
GET /api/trips/:tripId/flights/live, the MCP tool get_trip_flight_status,
notifications for take-off, landing and delays, and the admin switch
flight_live_enabled (FLIGHT_LIVE_ENABLED wins) come with it.
"Vorhandene verknüpfen" broke onto two lines in the narrow half of the
booking dialog. The labels now truncate instead of wrapping, and the German
label is the shorter "Datei verknüpfen", matching "Datei anhängen".
The whole-list sharing runs on usePackingCategorySharing in the packing panel
and the phone tab, its sheet is markup fed by the tab. The trips service reads
a stored costs_mode as one of TripsCostsMode, which the export check wants used.
feeds and trips import the calendar query DTO, so it lives in
calendar.contract.ts rather than as calendar's internal dto file. The places
controller tests pass the KML service in every call, and the revisit pill
test builds a whole list.
Inside the trip settings, Escape and the backdrop no longer close the dialog
while the date question is up, and the embedded form does not load the
members it does not show. Every trip open is timed afresh for the splash.
The tests follow the cover image card, the danger button, the delayed splash
and the open and download mocks.
oidc_auto_redirect is managed-locked in the registry too, costs_mode is read
as the mode it is, the days of a dated trip are read through the factories,
and the dashboard's country order moves out of the atlas service into a
module of its own, so the service keeps its size.
New warnings fixed at their source (refs written in effects, typed tile
options, no stray imports, smaller functions and components), the payer and
split heads hold their switch instead of letting it overhang, the baselines
lowered where the round removed warnings, and the tests updated to the
solo trip, the shares view and the revisit status.
The trip settings sheet opens on a list of its areas, each a page of its own
with a way back, and the journal there is drawn in the phone's own cards. A
new packing list is typed into a card with its add and cancel, selecting
places keeps its way out in view with the actions as icons below, and the
journey's satellite switch sits in its bar beside the menu.
Edit trip is an area of the settings dialog, beside People, Sharing and
Journal, instead of a second dialog. The side column names the areas only, the
dialog keeps one height, and every area uses the same cards, fields and
buttons; the trip form on its own takes the same head, cards and buttons. A
trip that loads quickly opens without the splash on the desktop too.
The trips entity takes costs_mode as the generator writes it, the plugin
output contract publishes it, and the hand-built test schemas carry it. The
new MCP fields come from @trek/shared, the link tool tests join the collab
file that already mocks the database, oidc_auto_redirect is the operator's
on a managed install, and the tests follow the KML service, the dated trip
days and the zod pipe.
The transport sheet's row wrapped once a ticket and the wallet were there.
Its actions are icons now, named by their labels, and the attachments sit in
their own rows with a download, like the transports list. The wallet is a
black circle, beside the close button in the booking editor. A failed
download on the phone now says so.
The Google link carried no mode, so a hike opened as a drive, and a stop on
the equator or the prime meridian was dropped. CoMaps routes a start and an
end only; a tour with stops between would reach it as unrouted pins.
With a category left out the total card still counted every expense, and in
the table view the only way back was to switch views; the chip now brings
everything back. The phone says what is left out once, under the total.
Linking a receipt offers pictures and PDFs only, like the upload.
The footer wrapped once a place had a day, a website and a collection. Now
only adding to the day, navigation and edit carry words, website and
collection are icons, and removing from the day and delete sit behind one
"…". The words drop out when the panel is narrow (a tablet) instead of the
row wrapping. Folding is kept for the place on show only (#3039).
DialogIconButton is the square footer button DeleteButton already was, for
any action named by its tooltip. A "…" menu near the bottom of the window
used to be pushed up over its own button; it now opens above it.
Safari before 16 has no AbortSignal.timeout, which cost the whole export; a
controller and a timer replace it. Photos decode at twice the thumbnail
height and three at a time, so a long trip no longer ends the web view.
The list menu on desktop and phone gets "Sharing for Whole List": pick
Shared, Personal or Shared with people once and every item of yours in
that list moves there through the existing per-item sharing write. Items
other people added stay as they are, and the dialog says so up front.
Each attached file row now has a download icon next to its name, on the
desktop booking cards and detail dialog and on the phone's bookings and
transports cards. On the phone the file rows move out of the card body
button so open and download are two real sibling buttons.
The place and endpoint reads behind the passport tile now carry the last day
of each started trip, and getTravelStats orders the countries by their latest
visit, with countries only marked by hand after them. The country set is
unchanged; repository, parity and service tests pin the order.
A new app setting oidc_auto_redirect (default on) is part of the admin form
and the public app config. Off, an instance without password login no longer
sends visitors straight to the IdP: the login page shows its SSO button,
also while the cached config is read. Admin toggle on desktop and phone.
Both feed routes and the one-off export.ics take ?days=0, validated by a
shared query contract, and then skip the all-day "Day N" event per day; the
default is unchanged. The subscribe dialog gets a Day overview events switch
that adds the parameter to its links, export_trip_ics takes dayEvents and the
feed tools explain the parameter.
list_collab_links, create_collab_link, update_collab_link and
delete_collab_link run the Links panel's REST routes as tools: the shared
body contracts, trip access, collab_edit on writes and the collab🔗*
broadcasts, behind the links sub-feature flag. Unit, e2e and broadcast
parity cases cover them; the wiki lists the tools.
dashboard_timezones joins the display-preference allow-list as a list of
IANA zone ids, each checked with Intl like the widget does, without repeats
and capped at 24. The write goes through saveMany, so open tabs refresh, and
both tool descriptions name the widget's fallback zones.
persistence.data.accessModes and persistence.uploads.accessModes default to
ReadWriteOnce, so existing installs render the same claims. ReadWriteMany lets a
backup pod mount the volumes; values.yaml and the docs warn that SQLite allows
only one writer of travel.db. CI renders the override too.
The tour detail gets a Maps button beside Navigation that hands the whole route
to the apps that follow one, in the tour's own mode (walking or cycling). The
stops are spread evenly along the line and capped at ten, the most a Google
Maps directions link opens. Ski, paddle and horse tours are not offered, since
a map app would route them by road; the GPX export stays the way for those.
On the road a stop's picture is often what decides whether it is worth the
detour. The phone's stop sheet shows the place's picture full width under its
name and the desktop stop dialog shows it in place of the kind tile; both open
the place's photos on tap. The picture is the place's own image, else the
provider photo for its Google or OSM id through the shared photo cache.
The journey map gets the planner's round satellite switch in its top corner,
on the desktop aside, the phone views and the public page. Leaflet takes the
vector basemap off while the imagery shows and reattaches it on the way back;
the GL renderer slides the imagery under the journey's own layers. The choice
is per view and opens on the map, so no setting is read or written.
A shared Immich or Synology album can hold hundreds of photos from other
trips. Once an album is open the picker now offers a from/to range that
filters the already loaded assets by the day the grid files them under, so
Select all only takes the photos in range. No server change.
The accept dialog opened on "not on a day yet" although the stay carries its
local date. It now picks the trip day whose date matches, on open and when the
trip is switched, and falls back to no day when the trip has none that matches.
Google Maps takes no list from outside, but My Maps imports a KML layer.
The trip export dialog and the phone export sheet get "Places for Google
My Maps" (GET places/export.kml, MCP export_trip_kml), and a list's export
menu gets KML next to GPX. The KML codec gains a pin writer; the list file
mapping moves to collections.helpers to keep the service within its size.
A read tool that answers one list a page at a time (100 by default, up to
500) as compact one-line JSON rows, filtered by status, label or "no label",
with the list's labels, the matching total and a cursor. The cursor is the
last place's position (sort_order, created_at, id), so labelling a page
while paging through unlabelled places skips none. getCollection orders by
the same keys, id last, so both tools agree on ties.
The update contract takes google_place_id, google_ftid, osm_id, website,
phone, price and currency like saving does, website with the same
normalisation and refusal. REST and update_collection_place share it, so an
agent can correct a place instead of re-saving it and losing its labels,
ratings, position and id. The assistant's change card names the new fields.
A fourth status after visited, for a place already seen that the user wants
to go back to. It joins the one-tap cycle, the filters, the pickers, the
assistant cards and the MCP status tool, and counts as visited wherever
visits are counted (isVisitedCollectionStatus), so "visited everywhere"
keeps it. The status pickers go two by two where four would not fit.
The phone toolbar gets a category chip beside the status and label chips,
fed by the same categoryOptions the desktop dropdown uses, and no longer
clears a category picked on desktop. Managing labels moves into the label
panel so the row keeps four chips at phone width.
Ctrl/Cmd+Enter in any field of a DialogShell presses its one primary footer
button, focusing it first so blur-committed fields keep their value. Dialogs
with no or several primaries leave the keys to the field, as does the journal
story editor, which opts out with data-keeps-ctrl-enter.
The print engine embedded every place photo at full resolution, so a trip of a
few hundred stops saved to a PDF of tens of megabytes. Same-origin photos are now
fetched six at a time, scaled to a 156 px short side and inlined as JPEG data
URLs; photos from other origins, failures and an aborted export keep the URL.
A third switch next to the page break and transport notes switches scales the
day content's type by 1.35 and darkens its greys, for readers who print the
plan. Off by default and remembered per browser like the other two; the day
content's sizes and greys now come from three CSS variables on <body>.
Mapbox Standard labels shops, restaurants and sights of its own; they look like trip
places but cannot be picked. A Hide Points of Interest switch next to the other Mapbox
options, on the desktop and the phone, turns them off on the trip map and the preview
without a rebuild. The Standard config handling moves into useStandardBasemapConfig.
A place saved with a typed address and no search pick, link or map point gets no
coordinates, so it never shows on the map, and nothing said why. The place form warns
under the address while that is so, and the inspector, the phone's place sheet and
both place lists mark such a place. No geocoding on save; the form stays unchanged
for a service stop, which explains a missing position itself.
On a landscape phone or a small tablet the inspector covered most of the map, and the
only way to clear it was closing it, which deselected the place. A chevron in the head
now folds the card to the avatar and name; the next place opens in full. The head band
moves to PlaceInspectorHead so the inspector stays inside its size budget.
The trip form was mounted in the planner but nothing opened it, so title, dates,
cover and currency could only be changed from My Trips. Trip settings now lists
Edit Trip at the foot of its areas for anyone with trip_edit; it swaps the dialog
for the trip form. The phone already has Edit Trip in the More sheet.
The category picker had no icon to tell a car park or a terminal apart from the
places a trip visits. Five transport icons join the shared table, which the admin
pickers, map pins and the PDF all read; the ship's tooltip now also says ferry.
With hotel bookends on, a day whose first stop was a hire car pick-up got no morning
bookend, so the drive from the stay to the desk was missing. Only a landing means the
traveller was not at the hotel; the pick-up is the first drive of the day.
Costs always converted to the display currency from Settings, so a trip
that moves on to another country could not be read in its new local
currency. A picker in the Costs bar (beside the Expenses heading on the
phone) now reads one trip in any currency; the server already settles in
any base. The pick is per trip and kept in this browser, nothing stored.
Seeing what a trip cost without, say, food meant adding it up by hand. A
click (or tap) on a category in the breakdown now leaves it out of the
total, your share and what you paid, dimmed in place with a line naming it
and Count all to undo. View only: nothing is saved, and what is owed, open
or settled stays whole. The desktop breakdown moves into its own component.
Travellers who track what a trip costs but never split it saw an
Unfinished warning on every expense and a settle-up they had no use for.
A trip now has a costs mode (appended migration, trips.costs_mode, split
by default); totals leaves payers, shares and settling up out of Costs on
desktop and phone through the same gate as a trip of one, and saves new
expenses without a payer. PUT /api/trips/:id and update_trip take it on
budget_edit. The split chips move into their own component.
With one traveler every share is the whole amount, yet Costs showed owe,
owed, outstanding, settle-up, balances and the final budget. A shared gate
(showsSplitting) now hides that side on desktop and phone, unless the ledger
still names someone else; the expense editor drops who paid and the split
while the expense holds nothing more. The payment dialog moves into its own
file to keep the panel under the size ratchet.
A booking confirmation that is also the invoice had to be uploaded a second
time to sit on the expense. The expense editor (desktop and phone) now lists
the trip's files to link, the booking's own first; the save sends the id in
receipt_file_ids, which the server already links. The desktop receipts block
moves into its own component.
A stay shared by people who slept a different number of nights had to be
worked out by hand. The custom split now takes shares next to amounts and
percent: five nights against four split the total 5/9 to 4/9, cent-exact by
largest remainder. Only the amounts are saved, so server and MCP are
unchanged; the phone sheet gains the same unit switch, percent included.
A new expense opened in the display currency, so every bill on the road
needed the trip currency picked from the full list. It now starts in the
trip's own currency (a scanned receipt keeps the one it read), and the
trip and display currencies lead the picker on desktop and phone.
The home assistant sat behind the More button on every page outside a
trip, so a dock with free slots showed "..." for a single entry. It is now
an entry of the dock list while its addon is on and it can run for the
user: an account that never sorted its dock gets it in a slot the addons
leave free, and Settings > Appearance > Mobile can move it between the bar
and More like any page.
A widget or Live Activity tap reached the UI as Capacitor's appUrlOpen, or
as the launch URL. When iOS had ended the web view's process in the
background, the event went to the old page's listeners and was lost, and
the reloaded page found the same link in getLaunchUrl that it had already
followed once and skipped it: the second tap on the packing widget opened
the start page.
The binary now takes com.liketrek.trek://navigate links itself and hands
them over like a quick action, held until the UI takes it and kept for a
page that is still loading. It reports the feature appLinks, and the UI
only follows the link through Capacitor for binaries without it.
"Obidos" did not find "Óbidos", nor "Sao Bras" "São Brás". The search of the
places list, a collection, the saved-place picker, the import from a trip
and the location picker now compare without case and Latin accents, with
letters such as ß, ø or ł folded to their plain spelling. Marks outside Latin
script stay, since there they change the word.
The place form guessed the category of a picked place from its kind, so a
place taken from a collection inside a trip lost the category it was saved
under, while adding it from the collection kept it. A picked place that
carries a category this server has now brings it along; a category chosen
in the form by hand still wins.
The server nests a saved place's category, while the markers and the hover
card read the flat fields a trip place carries, so every pin on a
collection's map was the grey default. The map now hands its places on with
the category's colour, icon and name where the markers look for them.
The check before a purchase read only the entitlements on the device. Those
can be empty while the subscription runs, so the app offered the plans and
the purchase changed the running subscription's plan in the App Store. The
status of the subscription group now decides when the device has none. A
renewal the device has not finished yet no longer counts as a purchase
waiting for its instance either, which also let the check through.
nativeApi goes to 3 with 5.1.4, whose UI follows a widget or Live Activity
tap that starts the app. The apps refuse an older server with the hint to
update, and check a server they remember once more.
The address and the key it may no longer carry now change together, as
they do for Dawarich, so the transaction ratchet drops the AirTrail entry.
The tests that pinned the old no-transaction shape now pin the new one.
Saving the AirTrail settings with a different URL and the key field left
masked or blank kept the stored key, so the next sync sent the old
instance's key to the new host. The key now goes with the host, as it
already did for the connection test and for Dawarich: a new origin without
a new key clears it, a path edit on the same origin keeps it. The key is
cleared before the new address is written.
A tap on a widget or a Live Activity hands the app
com.liketrek.trek://navigate/<path>. A running app gets it as appUrlOpen;
an app started by the tap gets it as its launch URL, which nothing read, so
the path was lost and the app opened its start page (the packing widget
landed on the trip's plan instead of its lists). The launch link is now
followed once per launch.
A flight created with a day and times but no endpoints (over MCP) showed
them in the form, but buildTransportPayload took a flight's span only from
rows with an airport, so a save sent day, end day and both times as null.
Like a train without stations, such a flight now takes its day, times,
airline, flight number and seat from the raw rows. Endpoints and legs still
come only from rows with an airport.
Fixes#3070
The edit dialog's component crossed the complexity limit with the section's
conditions; the section and its confirm step now live in SsoUnlinkSection,
which decides itself when it shows.
A sign-in the app starts runs through the server's handoff page in the
system browser. When the SSO callback failed (an expired or overwritten
state), the login page dropped the way back to that page, so the retry from
there was a plain web login and ended in the web app inside the app's
browser sheet instead of signing in the app. The login page now keeps a way
back that points at the handoff page, and only that, and renews its
fallback copy for a slow retry. State, nonce, PKCE and the handoff verifier
are unchanged; the handoff page still asks for a tap before it passes
anything on.
- The SSO unlink answers with a response contract (adminSsoUnlinkResponseSchema).
- It reuses linkOidcIdentity with nulls instead of a new repository method,
so UsersRepository keeps its method count and size.
- The client calls the route through apiClient, so api/client.ts does not
grow, and its confirm dialog takes no async callback.
- The pager's preview step and pagesFor keep their complexity.
A sign-in with another subject of the same provider is refused, so an
account whose provider re-issued the subject (a migrated or rebuilt IdP, a
changed sub format) was locked out: editing the email, as the wiki advised,
never touched the stored link. The edit dialog now offers "Remove SSO link"
for other accounts while SSO is set up (DELETE /api/admin/users/:id/sso,
audit admin.user_sso_unlink). The next SSO sign-in links again under the
rules of a first one, so it still needs a verified, not self-set email.
The log line names the remedy, and the wiki tells the two refusals apart.
Fixes#3065
Only the selected day and two on either side were drawn, and the selection
moves only once the strip comes to rest, which iOS takes a while for. A run
of fast swipes ran past the drawn pages into blank ones until it settled.
The drawn window now also follows the day the swipe is heading for, so the
next days are there before they slide in. At most two windows of five pages
are mounted, however long the trip.
Fixes#3067
The transport form re-seeded itself whenever the trip's costs arrived as a
new array, and the booking form whenever days, places or stays did. A
socket reconnect re-reads the whole trip, so behind a proxy that drops idle
sockets every 15 to 30 seconds everything typed so far was wiped. Both forms
now re-seed only for what opens a different form; the data is read when the
form is seeded, as before.
Fixes#3069
lint:i18n-keys took the spread in front of t() for a member call and did not
see native.provisioning.sub; the sentence is its own constant now. The
shell chrome test asks for the logical margins and their right-to-left swap.
- native-shell/amount.ts: the no-break spaces in the price pattern were
written out as characters; \s already matches them.
- widgets.ts: the Live Activity's lead time stays inside the module.
- i18n: the count strings of the new plan tiles replace the removed ones
on NOT_PLURAL, and the address placeholders of the non-Latin locales
are their own instead of English copies.
- mobile.css: the side safe-area margins of the floating bars use
margin-inline, swapped for right to left, instead of physical sides.
TREK is the application, not a thing one has a copy of: "I already have
TREK", not "a TREK". The same for the not-TREK messages in German and
Dutch and the agreement title in French, Spanish, Italian and Catalan.
The screens the app ships itself (welcome, plans, address, own server and
the states between) move from the dark card layout to a light design on an
animated aerial landscape drawn by a WebGL shader (native-shell/scenery.ts):
an island, a coastal town, farmland and mountains in one world.
- Welcome: the landscape with the TREK wordmark, three perks in one card,
"Explore cloud plans" and "I already have a TREK" as two buttons.
- Plans: one tab per plan; choosing one flies the camera to its place,
further inland with every plan. Each plan reads as a kind of trip, the
price sits in a card with a monthly/yearly switch and counts to its new
amount in the store's own format (native-shell/amount.ts), and a yearly
price shows what it makes a month. Restore purchases, the renewal terms
and the links stay as the store rules require.
- Address: the address writes itself on a glass card over the landscape;
free, the password rules and the agreement sit on white cards.
- Own server: the server address on the glass card; a failed check turns
the card red with what went wrong, proxy sign-in and client certificate
panels in the same style.
- States: server not reachable, certificate needed, already subscribed,
setting up (with its steps, then "Arrived") and the slow setup.
- iPad, an unfolded phone and a phone on its side get two columns, the
landscape beside the content; on a dual-screen device each half takes
one screen. Arabic and Hebrew mirror through logical sides only.
- The shader caps the pixel ratio at 2, drops to a lower resolution on a
slow device, stops in the background and is one still frame with
reduced motion.
- The sign-in and error page bars over a proxy's page (iOS and Android)
take the same look: a white card and a dark button.
- Plus Jakarta Sans is bundled for the text, with Montserrat standing in
for Cyrillic; the wordmark fonts and the old preview images are gone.
- New strings in all 28 languages; the keys of the old screens removed.
- The dev sample plans follow the cloud's plans (start, basis, plus, pro).
The Live Activity started with the first sync of a travel day and stayed on
the lock screen all day, so a day whose first stop is an evening flight held
it from the early morning on. It now appears three hours before the day's
first timed stop and stays until the day is done; liveDayDue() decides.
A load that failed sent the app to the address screen with a general message only, and a tester's screenshot told nothing. The web view's own error (net::ERR_..., an HTTP status, NSURLErrorDomain and its code, a client certificate the proxy asked for) is kept natively and shown under the message once.
The cloud screen shows real screens of the app. The plans are a choice now: one tap marks a plan, the recommended one is marked, what every plan has is said once above them, storage says how many photos it holds, and the yearly switch shows what a year saves; one button at the foot carries the chosen plan on. The claim screen names the plan with a way to change it, shows the address as it will be once it is free, ticks off the password rules while typing, says in three words where the servers are, that it can be cancelled and that there are no ads, and puts the photo behind the form away.
One subscription per Apple ID or Google account. Choosing another plan in the app changed the running subscription on the App Store, and the redeem then failed with 'We could not set up your TREK'. The app now asks the store first and shows a screen of its own: the account already has TREK Cloud, a new one can be taken out once it has ended, with Open my TREK and the store's page to manage it. The native purchase refuses as well; a paid one that was never set up is still redeemed. On the address screens the link 'What is a TREK server?' is gone, and the small print under a purchase is set in the system text face at regular weight.
The iOS app hands the page the safe area it measures, a floor under env() where WebKit reports none, and on a phone the floating bars keep clear of the side insets as well.
A Samsung phone may end the app while the sign-in runs in the browser. The browser's return then started the app anew without the call that waited for it, the first page load failed while the network was still held back, and the user came back to 'server not reachable' and then to the sign-in page, in a loop. The return is now kept natively and redeemed by the sign-in page with the verifier kept until then, and a page that failed on the network is loaded once more before the app gives up.
The import was only offered once the check saw the proxy ask for a certificate during the handshake. A proxy that refuses without asking left no way in. The wiki also points to the certificate's new place in Settings and explains why a certificate from an iOS configuration profile cannot be used by the app.
Sonnet 5.5 hands back the words written before a tool call as a hidden thinking block. After the ask for the answer, an empty step got the general push with tools open again; the model answered, offered follow-ups once more, and the answer was lost a second time. That push now asks for the answer as text and stays text-only.
The dim faded in on the sheet's own parent, and opacity multiplies down the tree, so for its first frames even the opaque sheet let the list behind read through as ghost text (Android tester, expense editor). The dim now fades on a layer of its own beside the sheet, and the card is opaque halfway through its pop.
DayPlanSidebar grew past its size baseline with the Firefox drag fix. Its drag state (what is dragged, the drop line, the day under the drag, whether a drag happened) and the effect that ends a lost drag move into useDayPlanDragState, typed, which also lowers the file's strict-mode count; size and strict baselines are lowered with it.
vacay.service.test.ts went past the 2000-line limit once the Nager v4 cases were back in it. They move into vacay.holiday-provider.test.ts, which takes VacayService from a Nest testing module over a snapshot db, so neither the db mock nor a hand-built service is needed.
- iOS: the navigation proxy reports an error page through a callback the controller sets; it called the controller's method on itself and did not build.
- keyboard: the done key reads the field's text in the capture phase, before the field's own handler, so a list that adds an item on Enter and clears itself keeps the keyboard; the tests wire it as the app does.
- Ireland migration: the queued writes use addSql instead of awaiting inside loops; only table constants and integer ids reach the SQL.
- tests: the Dawarich GL test types its two layer specs once (strict baseline lowered), the collab test reads the id off the listed message, the reasoning test runs its four servers in parallel, and a wasm copy covers the precompressed branch without a charset.
- en wording: the deviations agreement no longer calls TREK self-hosted.
A tester could not tell which settings follow the account to every device and which stay on the phone, and looked for offline mode for a while. The section list and dropdown now stand under two headings, "Your account" and "This device". The server the app opens, the app lock and a client certificate move out of Account into a section of their own, "App", shown only in the app, next to Offline; About stands under TREK. Labels in all locales. TestFlight feedback, 5.1.3.
geoBoundaries has no county layer for Ireland: its ADM1 is the four
provinces and its ADM2 the 166 local electoral areas, so the Atlas showed
Ireland as four provinces next to a UK of counties.
The Atlas builder can now take a country's regions from another source.
Ireland's come from Tailte Éireann's 26 counties (CC BY 4.0), with Dublin,
Cork, Galway and Limerick each one county, under their ISO 3166-2:IE
codes. Those outlines come at full detail, so the builder simplifies them
itself and keeps the borders two counties share identical on both sides,
so no point between them falls into neither. ATLAS_ONLY rebuilds just the
named countries without downloading the whole world.
A migration clears Ireland's cached place regions so they are worked out
again as counties, moves visited rows that name a county onto its code,
and hides each county of a province a user had hidden. Provinces a user
marked as visited are kept as they are.
Fixes#2922
Azure AI Foundry reasoning deployments behind an "Own server" connection
reject function tools on /v1/chat/completions at their default reasoning
effort and answer HTTP 400 naming reasoning_effort. The Responses API
switch from #2792 only covers direct OpenAI keys, so the assistant could
not run on these deployments at all.
The chat client still never sends reasoning_effort, so servers that do not
know the field (Ollama, LM Studio, vLLM, OpenRouter) see unchanged
requests. When a request with tools is refused over reasoning_effort, it
goes out once more with reasoning_effort "none", and the rest of the turn
keeps sending it. The refused field is now carried on AiProviderError
instead of only being logged.
Fixes#3043
Firefox fires dragstart while it handles the mouse move that starts the
drag and delivers that move to the page afterwards, button still held.
The listener that clears a drag whose dragend never arrived took that
move as the end of the drag and dropped the drag data at pick-up, so a
stop could only be reordered with its arrows. The same held-button moves
come from the touch polyfill on hybrid laptops throughout a drag. Only a
move with no button held now counts as the drag being over.
The day list also switched its stagger animation back on whenever a drag
ended, which replayed the entrance of every day card ("weird reload").
It now plays once on mount and stays off after the first drag.
Fixes#3041
On iOS 27.0.x the web view started below the status bar, and the strip above it showed the view controller's black instead of the screen background. The colour behind the web view now follows the theme the UI reports through StatusBar.setStyle (the top colour of --m-scr), so the strip reads as part of the page on every release. TestFlight feedback, 5.1.3.
The two new test files mocked the db module and built VacayService by hand,
which the test-mocks and new-service ratchets refuse for files without a
baseline entry. The cases move back into vacay.service.test.ts and
vacay.test.ts unchanged; VACAY-037 asks for a year no other case caches. The
unrelated admin addon test change is reverted and two comments name the new
provider.
A multi-day hike restocks on the way: bread in the morning, food for
lunch and supper every day or two. Neither could be searched for.
- Explore pill: new `bakery` and `supermarket` chips. `supermarket`
already existed on the server (shop=supermarket|convenience) but no
pill offered it; `bakery` is new (shop=bakery).
- Along the tour: `bakery` and `supermarket` are offered to a hike,
after its existing kinds so a new hike still starts on the same three.
- TREK places index: `bakery` maps to food_and_drink>casual_eatery>bakery;
`supermarket` is narrowed to grocery and convenience stores, since the
whole food_and_beverage_store branch also brought tobacconists and
wine shops. Both paths checked against a live answer of the index.
As the more specific term, a bakery is no longer listed as a place to
eat when only `food` is asked along a tour.
- Category guess: a bakery or grocer picked from the map was preselected
as Bar/Cafe or Shopping even where the admin had made a Bakery or
Groceries category. The guess now looks for such a category first, by
name in the usual languages or by icon, and keeps the broad one when
there is none. The category icon picker offers Croissant and
ShoppingBasket, the icons the map chips use.
- Labels in every locale; MCP tool description names the new kinds.
A point just off a simplified coastline lies in no admin0 polygon, and the
fallback then took the candidate with the smallest bounding box. Nice and
Menton sit ~80 m outside France's coast and 20+ km from Italy, but Italy's
box is the smaller one, so they resolved to IT. The fallback now ranks the
candidates by distance to their border. A deferred loose box (PS/XK) keeps
its priority. The parity test's reference lookup mirrors the new rule.
Closes#3048
The places list and the trip bundle resolve the country of every place
without a cached region on each read, and that lookup was about 90% of
their CPU time: it ray-cast every polygon of each candidate country, which
is thousands of rings for a country that is mostly islands.
A polygon whose outer ring's box does not hold the point is now skipped
without walking it, and an answer is kept by coordinate pair (20,000 pairs,
oldest out first), since the borders never change while the process runs.
The answers are the ones the ray-cast alone gives; ATLAS-SVC-051 and
ATLAS-SVC-055 compare every country against the reference lookup.
Measured on a production build with a 360-place trip: GET places
22.7 ms -> 3.5 ms, GET bundle 27.5 ms -> 6.2 ms. A first lookup of a
pair is about 1.8 times faster (90 us -> 50 us over 20 cities).
Every file under assets/ was compressed again on each download, at the low
level a request path can afford. They carry their content hash in the name
and never change, so the client build now writes a Brotli copy beside each
compressible one (postbuild, scripts/precompress.mjs) and production points
a browser that accepts Brotli at that copy before express.static serves it.
Measured on a production build of 5.1.2, fetching the 652 precached files
the way a service worker install does: 16.3 MB -> 13.9 MB on the wire and
about 1.0 s -> 0.2 s of server CPU per install. A client without Brotli, a
range request and a build without copies are served exactly as before.
From Node 25 the permission model covers the network, so a plugin child
forked with --permission and no --allow-net gets ERR_ACCESS_DENIED on every
lookup and connect, and no plugin can reach a host it was granted. Node 24
is unaffected and rejects the flag as a bad option, so it is passed only
where process.allowedNodeEnvironmentFlags knows it.
Which hosts a plugin may reach is still decided by the egress guard inside
the child.
Fixes#3045
Every lucide icon is its own module, and the plugin icon lookup reaches all
of them by name from the navigation. The bundler therefore put each icon
another chunk also uses into a chunk of its own: 187 files of a few hundred
bytes, each one a request on a first visit and an entry in the precache.
A `vendor-icons` group collects them. The first load asks for 32 script
files instead of 224 (the bytes stay the same, 3.9 MB before compression),
and the service worker installs 460 files instead of 652. With 50 ms added
to every response the login page reaches DOMContentLoaded in 0.5 s
instead of 3.8 s.
Deleting a message set the flag and kept the text in the row and in every backup; only the client hid it. The delete now blanks the text with the flag, and a migration clears what earlier deletes left. The row stays for the placeholder.
The dashboard, journeys and notifications switched the iOS bounce off entirely so it would not race the pull. The pull now holds only a clearly downward move from the very top, from its first pixel, and the page bounces everywhere else, its end included. TestFlight feedback, 5.1.3.
The page took 100vh with no top inset inside the touch shell, so the plugin's header ran under the status bar and the frame reached past the screen and under the dock. It now takes the visible height (100dvh) less the dock, after the safe area and the rail, and the plugin's own document scrolls. TestFlight feedback, 5.1.3 (Trip Importer).
A thin, translucent dash over a nearly opaque white casing read as a pale trace and was barely visible on past trips. Full colour, a wider line and longer dashes, shared by the Leaflet and GL layers. TestFlight feedback, 5.1.3.
A field outside a form gets enterkeyhint="done", but a web view only sends Enter and nothing listened, so the key did nothing. When no handler took the Enter (same text, still focused a moment later) the field lets go of the focus; a field that acts on Enter and clears itself keeps it. TestFlight feedback, 5.1.3.
A proxy's block page or a refused client certificate (Cloudflare's "Sorry, you have been blocked") left the app on a page without TREK's UI and no way to change the server. Any 4xx main-frame page that the client-certificate path does not handle now shows the sign-in bar with its own hint and "Change server", on iOS and Android. TestFlight feedback, 5.1.2.
The claim screen loads the agreement text and its version from /api/iap/consent, shows it with its own checkbox and buys nothing until it is ticked. The redeem sends consent { acceptedDeviations, version }; a refusal for an outdated text reloads it and asks again without a second purchase.
The login card is a fixed light surface whose text colours are written out,
not tokens. Four of them moved with the two-factor step, so the baseline moves
them too: 58 to 54 for the page, 4 for the new component, the same total.
A partial update that left day_id out (a notes-only PUT, an MCP
update_reservation or update_transport call) re-derived the day from the
stored reservation_time whenever one was set. A time-only value such as
"12:00" resolves to no day, so the booking's day_id was written as null and
it vanished from its day in the planner. The end_day_id branch did the same
when a time-only reservation_end_time was sent.
The day is now derived only when the time carries a date that resolves to a
trip day; otherwise an omitted day_id or end_day_id keeps the stored value.
A dated time still re-pins the day, an explicit day_id null still clears it,
and a cleared end time still clears end_day_id. REST and both MCP update
tools share updateInTx, so the fix covers all three.
Tests: service unit cases (RESV-SVC-044..048), an e2e case per behaviour
on PUT /api/trips/:tripId/reservations/:id, and an update_reservation MCP
case, all seeded with the ORM factories.
Fixes#3025
Both files had grown past their size limit with this round's fixes. The
two-factor step moves to pages/login/MfaCodeStep.tsx and the reservation row
types to reservation-rows.ts, re-exported from the repository, and both size
baselines come down to match.
On iOS the keyboard plugin sizes the body and zeroes the scroll inset in a
frame update that runs after the keyboard is up, so the room made for it was
taken back. The inset is now kept while a sign-in page shows, the body gets its
height back after that update, and with the keyboard down the page can scroll
past the bar. On Android before 11 the keyboard went unseen in the insets
Capacitor passes on, so the bar stayed over the form; it is read off the
window's own insets now.
Behind Pangolin or authentik the app shows the proxy's sign-in page itself, and
nothing on it led back to the address screen: a failed passkey left users stuck
on every launch. A bar over that page now offers to change the server and says
that passkeys and security keys do not work there, as a web view cannot use them
for a domain the app is not tied to. It steps aside while the keyboard is up.
On iOS the page also gets the room the keyboard takes, so a button under it can
be scrolled into view. The address screen hands the bar its words in the user's
language when it connects.
The pager held the selected day and two either side, and moved that window
only once the strip had come to rest. Swiping on before the snap settled ran
into the end of the strip after two days and sprang back until the animation
was over. Every day now has its page, only the selection and two either side
draw their content, so a run of swipes goes as far as the trip does.
A booking with stopovers was one stop at its first departure, and a stop
counted as done the moment it began. With a car at one and a connecting flight
at two, the Live Activity said all done for today mid-afternoon. Each leg is
now a stop of its own, as the planner shows it, and a flight or anything else
with an end time counts as under way until it lands or ends.
The edit form fell back to the linked stay's times, the cards read only the
booking's metadata, so a hotel showed 15:00 and 11:00 when edited and no times
in the list. The reservation list now carries the stay's check-in window and
check-out, and the desktop facts and the phone card fall back to them. On the
phone they come first, so a price or a seat cannot push them off the card.
The prompt looked at the app's features once, when the user counted as signed
in. After the sign-in the page loads with the session already there, and on
iOS the bridge had not named its features yet, so the prompt gave up for good.
It now waits for them. It also counted as asked the moment it opened, so a
sheet that went under another one never came back; it counts once answered.
The code step drops the email field, so a password manager had no login to
match a one-time code to; the step now keeps the address in a hidden username
field. The code field names itself and sets its keyboard hints directly
(characters, no autocorrect), as autoFocus gave it focus before anything could
prepare it, and its placeholder is translated.
The file preview kept its data source in a stored property inside an
extension, which Swift does not allow, so every iOS build since failed. It is
a static now, in its own extension, and the Wallet comment it had split is
whole again.
The fixes that keep the app's sign-in in the app live in the server's UI,
so an updated app on a 5.1.2 server still ended in the browser. The server now
reports nativeApi 2 and the app refuses anything below it, naming 5.1.3.
A server the app remembers from before is not loaded unchecked: the first
launch after the update shows the address screen filled in with it, says why
and checks it right away. A server that is new enough loads again with its
session; an older one gets the update hint.
The strip re-centred only when its window of pages changed. On a trip of up
to five days a tap can keep the same pages, so the rail showed the new day
while the card below stayed on the old one.
The iOS app rewrites raster tiles to trek-tile:// and answers them from its
store. WebKit checks that scheme against img-src, which had only https:, so
every tile was refused and the app showed routes and markers on a blank map.
The app's browser sign-in opens /native-handoff, which sends the user on to
the identity provider by itself. The page was not on the list of public auth
pages, so the 401 from loading the user sent the browser to /login with the
hand-off as its return path first. On a server that also offers passwords the
user then had to pick SSO a second time before getting back to the app.
A PDF, image or document opened in the app went straight to the share sheet.
It now opens in Quick Look on iOS, to page through and zoom with its own share
button, and in the app that views its type on Android. An app without the
viewer (feature filePreview), or a type nothing on the device shows, gets the
share sheet as before.
The home assistant wrote its answer and offered the trip's own chat in the
same step. The loop asked the model once more after the card, and it wrote
the same answer a second time below it. A step that has its answer and only
hands the conversation on now finishes the turn; a hand-on without an answer,
or beside a read, still asks the model again.
The actions under a swiped row were square blocks to the row's edge, unlike
the rounded cards around them. They are now tiles of their own with the
card's radius and a gap, inset within a grouped card such as the packing
list, and show only their icon (the label names them for a screen reader).
An expense already shows edit and delete beside its card, so it no longer
swipes to a second, larger pair of the same buttons.
An app linked against an SDK older than iOS 27.1 runs letterboxed on the inner
display of an iPhone Duo. The iOS build and the release now run on GitHub's
xcode-27 image and select Xcode 27.1; the previews stay on Xcode 26.
From #3032, adjusted where it broke:
- react and react-dom are pinned in the root overrides; the group raised the
workspaces to 19.3.1 and left the override at 19.3.0, so two react-dom
versions met in every client test. The override moves along.
- @mikro-orm/nestjs kept the old @mikro-orm/core at the root while the server
got 7.2.8 of its own, two copies whose types do not match; the lockfile now
holds one 7.2.8.
- mapbox-gl stays at 3.31: 3.34 drops setRTLTextPlugin from its types, which
the map engine uses.
- the musl pins of rollup stay at 4.62, the rollup version the lockfile holds.
@capacitor/cli 8.6 drops the old uuid 7 through its xcode dependency (#3012).
vite-plugin-pwa 2 keeps the registerSW.js in public/ in place of its own, as
1.x did (#3033). @nestjs/schedule 12 still runs on Nest 11 (#3035).
The login on an iPhone scrolls in no pane of its own, and the keyboard plugin
holds the body's height, so the page had nothing to scroll by and the password
field stayed under the keyboard, most of all with an error shown above it. Such
a page now gets room as tall as the keyboard at its end and scrolls the field
above it; the room goes with the keyboard.
A browser that ran TREK before 5.1.1 keeps that version's app in its service
worker. When the app opened /native-handoff, the worker answered it from its
cache with an app that has no such page, which sent the user to /, and the
sign-in ended on the dashboard in the browser instead of in the app.
The app now opens /api/auth/native/start, which every TREK service worker
passes to the server. That page takes the workers off the browser and
forwards to the handoff page. Until the sign-in is handed back, the web app
registers no service worker and does not reload for a new version: a tab
still open on the older TREK keeps its worker alive, and a registration or a
reload in the middle of the sign-in brought it back to answer the next page.
A proxy that lets /api through (an Authelia or Authentik bypass for the app)
still sends the page itself to its sign-in. The app took that redirect for a
foreign link and handed it to the browser, so the sign-in happened there and
the app stayed black on every launch. A redirect away from the server while
its first page loads is now kept as the proxy's sign-in host, on Android and
iOS, and the iOS bridge guard is always in place for hosts learned that way.
Bookings in the Bookings tab, the day sheet, a place's sheet and the road trip
opened their editor straight away, while a transport showed its details with
a pencil to edit. Every booking now opens the same detail sheet; its edit and
delete follow the booking's own permission (reservation_edit for a stay, a
table or a ticket, day_edit for a transport). Notes keep opening for editing,
as on the desktop.
The wallet button reads "Wallet" (its accessible name still says which one),
and "On map" reads "Map" in every language.
Mapbox GL measures its container on window resizes only. In the app on a
tablet the trip's plan pane closes beside the map without one: the box grew
to the full width while the canvas kept its old size, and the rest stayed
black. Every GL map built through buildGlMap now watches its container and
resizes once per frame while it changes; the observer goes with the map.
Closes#3023
Each photo's write follows its own network lookup, the repair's marker only
records that the pass ran, and a lookup may renew the Synology session on
its own; none of these can share one transaction.
A proxy in front of TREK (Cloudflare, nginx, Caddy, Traefik, HAProxy) can
let in only devices that show a client certificate. The apps could not, so
such a server was unusable from them.
- Address screen: the server check runs natively and notices when the
server asks for a certificate. Instead of "no server answered" it says so
and offers to add a .p12/.pfx file; a wrong password asks again, an
expired, not yet valid or keyless file is named as such.
- The certificate is shown only to the host and port it was added for.
iOS keeps it in the keychain (this device only), Android seals the key
with an Android Keystore key. The file's password is never stored.
- The web view answers the server's certificate request on both platforms;
WebSocket, service worker and API requests go through with it. The
offline copy, watch requests and the check use it too.
- When the server stops letting the app in (certificate removed, expired
or no longer trusted) while the page still comes from the service worker
or the offline copy, a native check sends the app to the address screen,
which offers to add another. A server that only asks and lets everyone
in is left alone.
- Cloudflare-style 403 and nginx-style 400 refusals without a certificate
lead to the same screen instead of staying up as an error page.
- Settings, Account: a Client certificate card (only with one stored)
shows subject, issuer, server and expiry, with Replace and Remove.
- Wiki: Native-Apps "Client certificates (mTLS)" with Cloudflare, nginx,
Caddy and Traefik setup and how to make a certificate for the app.
- The app icon on the unreachable screen is black on the light surface.
The repair job wrote its marker rows through the AppSettings repository,
which the settings domain owns. It now registers both keys and goes through
SettingsService, as the import boundary check asks.
Until #3019 a Synology photo's capture time and GPS came from the item after
it, which on many libraries is the next photo, and no one can edit a photo's
capture time to put it right. On the first start after this, every Synology
photo is asked again through its owner's library, held to the backfill's few
lookups per NAS, and what Synology confirms for that very photo replaces what
was stored; a photo it cannot confirm keeps its values. While a library could
not answer, the pass runs again on the next start, at most three times.
On "Almost there" nothing put the keyboard away: the web view keeps it open on
a tap beside the fields, and "Next" sent the half-filled form, whose check
focused a field again. A tap beside the fields or a drag of the screen now
closes it, as in an iOS app, on every first screen. "Next" walks from the
address to the email to the password, "Go" buys, and the keyboard goes as the
purchase starts.
The detent came only once the strip had settled, which on iOS is a good while
after the finger lifts, so it felt late. It now ticks as soon as another day
becomes the nearer page, and no second time as the snap lands.
The strip held the selected day and one on either side. A second swipe made
while the first still snapped ran into the end of the strip and sprang back,
so paging several days took a pause after each. It now holds two on either
side, and the second swipe carries on to the day after.
getSynologyAssetInfo asked Browse.Item for the unit id plus one, a fixed
offset that holds on some libraries and on others names the next photo, whose
date and GPS were then stored for the attached one. It now asks for the unit
id, then for the one after, and takes only an item whose own thumbnail belongs
to the requested unit; when neither does, it answers not found instead of a
neighbour's metadata.
Closes#3019
Packing items, tasks and expenses swipe sideways to their actions, as in
Mail or Reminders: done or open one way, delete and edit the other, the
first one run by a full swipe. They are second ways to what each row
already offers; an expense still asks before it goes.
A long press on a trip card lifts it and opens its menu with the card's own
actions. Plain confirmations in MConfirmSheet now also ask through the
system where the app can.
Lists, the atlas and a journey show grey placeholders in their shape while
loading instead of a spinner in an empty screen (MSkeleton); buttons keep
theirs. A picture still loading fades in once it is there; one from the
cache shows at once. In the iOS app the page bounces at its ends again,
except on screens with pull-to-refresh, whose pull it would race; Android
keeps its stretch off.
On iOS a long press on the app icon offers "New trip" and the trip coming
up by its name, set by the UI in its language (TrekShell.setQuickActions,
feature quickActions), and a tap opens it, also from a cold start.
The app's own CSS moves out of index.css into native/nativeApp.css, which
had outgrown its size budget.
After a tap, WebKit and Chrome keep :hover on what was tapped and send a
made-up mouseover, so buttons and rows stayed tinted until the next tap.
While a finger drives the app <html> carries data-touch-input: a PostCSS
step guards every :hover rule with it (no specificity change), and the
made-up mouseover and mouseout stop before any handler. A trackpad or mouse
clears the mark and gets hover back; a browser never sets it.
A control without a press of its own dims the moment a finger lands on it,
after a 60 ms wait so a scroll does not flash rows, and only the innermost
pressed control. A focus: ring no longer stays round a tapped button.
Fields get a fitting return key ("done", "go" in a form, "search" on the
search fields), and addresses, codes and numbers are no longer capitalised
or corrected. Ticking off packing items and tasks, a picker chip, a select
option, a picked date and dragging a plan row now give the haptics a native
list gives.
Confirmations ask through the system where the binary has the new
@capacitor/action-sheet and @capacitor/dialog plugins: a sheet with the
action in red when it destroys something, an alert otherwise, also for
"discard changes" in a sheet. An older app, a browser and a question with a
list under it keep the HTML dialog, as does a system that could not ask.
iOS no longer lifts a page preview on a long press on a link.
In the iOS and Android app a long press on a label, card or image started a text
selection, the magnifier or a "save image" menu, the clearest sign of a web page
in a frame. startNativeApp marks <html> with data-native-app, and index.css then
turns selection and the callout off for the UI. Inputs, textareas, selects and
editors stay selectable so typing, pasting and the caret work; text worth copying
opts in with .trek-selectable. A browser keeps its default.
The claim screen takes the password (TREK's policy, checked before anything is
bought) and sends it with the purchase; the control plane answers with a
one-time welcome code, which the app hands to setServer. iOS and Android open
the instance's welcome page with it once after the first page, which signs the
buyer in. Without the code (a restore, an older control plane) the app opens
the instance as before.
Empty in this repository, as the module itself. An install that replaces the
module adds its anonymous routes in the same file, and the gate still fails on a
route missing there or an entry that went stale.
npm run test --workspace=server # vitest run (server); also test:unit / test:integration / test:ws
npm run test:e2e # server e2e (boots Nest against a temp SQLite)
npm run test:pg --workspace=server # the Postgres lane (vitest.pg.config.ts; needs TREK_TEST_DB_URL, e.g. postgres://postgres:trek@127.0.0.1:55432/postgres)
npm run test:cov # coverage (lcov) for all four packages incl. plugin-sdk
cd client && npm run test# client vitest run
cd client && npm run e2e # Playwright (CI runs the public and app projects; the screenshot and help-media ones are local)
@@ -51,7 +52,9 @@ cd client && npm run lint:skips # no .only; skipped tests may
cd client && npm run lint:i18n-keys # every key the client names exists in en; unused en keys may only shrink (CI gate)
cd server && npm run lint:boundaries # no new import cycles or cross-domain reach-ins (CI gate)
cd server && npm run lint:dialect # SQLite-only SQL spellings may only shrink (CI gate; rules in server/CLAUDE.md)
cd server && npm run lint:engine-dispatch # engine checks outside src/db/dialect/ and src/db/engine/ may only shrink (CI gate)
cd server && npm run probe:pg # dialect helpers + the repository statements it reaches against Postgres (CI job; needs TREK_PG_PROBE_URL)
cd server && npm run check:pg-baseline # the Postgres baseline in server/assets/postgres/ matches the entities and the migrations (CI gate)
cd server && npm run lint:tx # no new method writing twice outside one transaction (CI gate)
cd server && npm run lint:test-sql # raw SQL fixtures in tests/ may only shrink; seed through tests/helpers/factories (CI gate)
cd server && npm run lint:test-mocks # vi.mock of config, realtime and db modules in tests/ may only shrink (CI gate)
@@ -134,13 +137,18 @@ These principles come out of a full-repo audit and shape all new code:
## Quality gates in CI
`.github/workflows/test.yml` runs, per package: typecheck (server also `typecheck:tests`, client also the strict-mode ratchet `lint:strict`), lint (client and server through `lint:warnings`, which also holds the warning count per rule), `lint:pages`, `lint:rtl`, `lint:size` (client and server), the server's `lint:strict`, `lint:boundaries`, `lint:tx`, `lint:dialect`, `lint:test-sql`, `lint:test-mocks`, `lint:test-new-service`, `lint:query-api`, `lint:mcp-zod`, `lint:service-http` and `lint:response-contracts`, the client's and server's `lint:format`, `lint:layers`, `lint:offline`, `lint:dup`, `lint:pairs`, `lint:skips`, `lint:i18n-keys` and `theme:lint`, shared's `contracts:open`, the repo-wide `lint:knip` (in server-lint-checks), `check:plugin-facts`, `i18n:parity:strict` (with the untranslated-strings ratchet), the S3 contract suite, the Postgres probe (`postgres-probe`: every dialect helper on SQLite and Postgres, and the statements every repository method reaches with sample arguments against a Postgres service; the SQL errors per method and the list of methods it cannot measure are held to a baseline that only shrinks and fails while it is still unmeasured; see `server/CLAUDE.md`), the browser end-to-end flows (Chromium and WebKit), a client build with `check:gl-split`, coverage for all four packages, then a SonarCloud scan. It also builds the Docker image and boots it until `/api/health/ready` answers (`docker-image`), and lints and renders the Helm chart against the Kubernetes schemas and runs `docker compose config` on every compose file (`deploy-manifests`). `lint-prettier.yml` additionally runs `lint` + `lint:format` on `shared/`, both check-only. `plugin-sdk` has no lint script and is outside the eslint gates. **Not in CI**: the screenshot and help-media Playwright projects; run them locally. There are no git hooks; nothing runs before a commit except you. Other workflows build Docker images, scan them, and publish `trek-plugin-sdk` on `plugin-sdk-v*` tags.
`.github/workflows/test.yml` runs, per package: typecheck (server also `typecheck:tests`, client also the strict-mode ratchet `lint:strict`), lint (client and server through `lint:warnings`, which also holds the warning count per rule), `lint:pages`, `lint:rtl`, `lint:size` (client and server), the server's `lint:strict`, `lint:boundaries`, `lint:tx`, `lint:dialect`,`lint:engine-dispatch`,`lint:test-sql`, `lint:test-mocks`, `lint:test-new-service`, `lint:query-api`, `lint:mcp-zod`, `lint:service-http` and `lint:response-contracts`, the client's and server's `lint:format`, `lint:layers`, `lint:offline`, `lint:dup`, `lint:pairs`, `lint:skips`, `lint:i18n-keys` and `theme:lint`, shared's `contracts:open`, the repo-wide `lint:knip` (in `server-lint`), `check:plugin-facts`, `i18n:parity:strict` (with the untranslated-strings ratchet), the S3 contract suite, TREK on Postgres (`server-postgres`, once per supported major, 14 and 18, each in three slices: first the Postgresprobe (in the first slice), every dialect helper on SQLite and Postgres and the statements every repository method reaches with sample arguments, its SQL errors per method and the methods it cannot measure held to a baseline that only shrinks and fails while it is still unmeasured; then the Postgres lane, every suite that boots the app on the real database module, found by `server/tests/postgres/lane.ts` rather than listed, plus PARITY-PG and the write lock; see `server/CLAUDE.md`), the browser end-to-end flows (Chromium and WebKit, in the Playwright container, in two slices), a client build with `check:gl-split`, coverage for all four packages, then a SonarCloud scan. It also builds the Docker image and boots it until `/api/health/ready` answers, on SQLite and on Postgres 18, and lints and renders the Helm chart against the Kubernetes schemas and runs `docker compose config` on every compose file (both in `docker`, each part when its area changed). `lint-prettier.yml` additionally runs `lint` + `lint:format` on `shared/`, both check-only. `plugin-sdk` has no lint script and is outside the eslint gates. **Not in CI**: the screenshot and help-media Playwright projects; run them locally. There are no git hooks; nothing runs before a commit except you. Other workflows build Docker images, scan them, and publish `trek-plugin-sdk` on `plugin-sdk-v*` tags.
- **Job names read "package: what"** (`Server: Types`, `Client: Tests 3/7`), and the list at the top of `test.yml` says what each job holds. A new job follows the scheme and joins that list, and the job-count note there says when a long job needs a slot freed first.
- **Tests run beside the gates, in shards.** The client and server suites run in slices on parallel runners (`client-tests`, `server-tests`, through each package's `vitest.shard.config.ts`), and `client-coverage`/`server-coverage` merge their blob reports and check the thresholds against the whole. A red slice is usually a failing test, a red merge job usually a threshold.
- **`ci-ok` is the required check.** The branch ruleset requires `ci-ok` (plus `lint` and `check-target` from their own workflows), and `ci-ok` needs every job above except the scan. It fails if any of them failed or was cancelled and passes if they passed or were skipped. A new gate job counts only once it is listed in `ci-ok`'s `needs:`.
- **Job names read "<Area> <what>[ <variant>][ i/n]"** (`Server tests 3/10`, `Server Postgres 14 2/3`, `Docker and Helm`): the area first, one of Changes, Shared, Server, Client, Docker or Sonar, then lower case, with no colon, no parentheses and no middle dot, and a slice only at the end, so the checks list sorts by area. The workflow is named `CI`, so a check reads `CI / Server tests 3/10`. Job ids are the names in kebab-case (`server-tests`), the jobs stand in `test.yml` in area order, and the list at the top of `test.yml` says what each job holds. A new job follows the scheme and joins that list; `scripts/ci/workflow-jobs.test.mjs` fails on a name outside it. `ci-ok`, `lint` and `check-target` keep their names: the ruleset requires them by name.
- **What the folded jobs hold.** `Shared tests` is the plugin SDK's typecheck and coverage, run before the root install so the SDK resolves only its own dependencies, then i18n parity and shared's coverage. `Client types` is the typecheck and the strict ratchet; `Client checks` the source ratchets, then the production build with `check:gl-split`. `Docker and Helm` builds and boots the image when the image area changed and checks the chart and compose files when the deploy area changed. Inside a folded job each group of steps runs after a failure in another, so one red check shows every gate that failed; the step names say which.
- **Tests run beside the gates, in shards.** The client and server suites run in ten slices each on parallel runners (`client-tests`, `server-tests`, through each package's `vitest.shard.config.ts`), the Postgres lane in three per major, and the end-to-end flows in two, and `client-coverage`/`server-coverage` merge their blob reports and check the thresholds against the whole. A red slice is usually a failing test, a red merge job usually a threshold. A sliced job states its slice count in its matrix's shard axis, the `i/n` in its name and its `--shard=i/n`; `scripts/ci/workflow-jobs.test.mjs` fails until all three agree, and on any job naming `matrix.shard` or `--shard` whose axis is not the one-line`shard: [1, ..., n]` it reads.
- **`ci-ok` is the required check.** The branch ruleset requires `ci-ok` (plus `lint` and `check-target` from their own workflows), and `ci-ok` needs every job above except the scan. It fails if any of them failed or was cancelled and passes if they passed or were skipped. A new gate job counts only once it is listed in `ci-ok`'s `needs:`, and `scripts/ci/workflow-jobs.test.mjs` (in `test:ci-scripts`) fails while a job other than the scan is missing there.
- **The release pull request mirrors the push run.** On the pull request from `dev` into `main` (head branch in this repository), `changes` sets `mirror` and turns every area off (`scripts/ci/changed-areas.mjs`), so no test, gate or scan job runs, and `ci-ok` runs `scripts/ci/mirror-push-run.mjs`: it waits for the `ci-ok` of the push run of `test.yml` on `dev` for exactly the PR's head commit and passes or fails with it. No such push run, one that ended without a verdict, or a wait past 17 minutes fails with a re-run hint, never passes. A push to `dev` lists its changes against the merge base with `main`, not against the previous push, so the run it mirrors covers every area the pull request brings in, even when an earlier push's run was cancelled or red. Testing that commit twice doubled the jobs past the 60 slots.
- **Guides name paths that exist**: the `changes` job runs `scripts/ci/doc-paths.mjs`, which resolves every backtick path in the CLAUDE.md files, `README.md`, `server/src/nest/README.md` and the wiki against the tracked files; the missing ones are listed in `scripts/ci/doc-paths-baseline.json`, which only shrinks (`--update` after a fix).
- **What runs is decided by `changes`, not by `on.paths`**: `scripts/ci/changed-areas.mjs` maps the changed files to the `code`, `image` and `deploy` areas (tests: `npm run test:ci-scripts`). A path filter on the workflow would leave `ci-ok` unreported on a docs-only PR and block it. Add a new top-level file the build reads to the right area there.
- **What runs is decided by `changes`, not by `on.paths`**: `scripts/ci/changed-areas.mjs` maps the changed files to the `code`, `image` and `deploy` areas (tests: `npm run test:ci-scripts`). A path filter on the workflow would leave `ci-ok` unreported on a docs-only PR and block it. Add a new top-level file the build reads to the right area there. A change to `test.yml`, to `scripts/ci/` or to a local action under `.github/actions/` runs everything.
- **Setup goes through one local action.** Every job that installs dependencies uses `.github/actions/setup/action.yml` (`scope` full, client or shared; `build-shared`) instead of its own `npm ci`; `workflow-jobs.test.mjs` fails on a job that installs on its own (plugin-sdk's own install in its directory excepted). The action restores node_modules on an exact key (runner image, scope, Node ABI, `.nvmrc`, the lockfile and every workspace `package.json`, so npm ci still refuses a `package.json` out of step with the lockfile; no restore-keys) and shared/dist on a key over shared's sources and build config, and runs `npm ci` or the shared build on a miss. Only pushes to dev and main save, the npm download cache included; pull requests only read.
- **End-to-end runs in the Playwright container**, `mcr.microsoft.com/playwright:v<version>-noble` pinned by digest, so no step installs browsers or system libraries. Dependabot bumps only the lockfile: on every Playwright PR move the tag and its digest in `client-e2e` by hand (`docker buildx imagetools inspect mcr.microsoft.com/playwright:v<version>-noble`); `workflow-jobs.test.mjs` fails until the tag names the lockfile's `@playwright/test` version.
- **Budget: `ci-ok` within ten minutes.** One code run starts at most about 35 long jobs (gates, slices, Postgres, e2e, Docker), so two PR runs fit in the organisation's 60 concurrent jobs. Every job sets `timeout-minutes` of at most 20 (15 for the long ones, 10 for the short ones), which `workflow-jobs.test.mjs` holds. A new long job, slice or shard needs a reason that outweighs its slot; the job-count note at the top of `test.yml` has the numbers.
- **Runners and Node**: jobs run on `ubuntu-24.04`, not `ubuntu-latest`, and every `setup-node` reads `.nvmrc`. Third-party actions (anything outside `actions/*`) are pinned by commit SHA with the tag in a comment; Dependabot keeps both current.
- **Security scans are advisory.** `security.yml` runs `npm audit` (what the image installs, plus `plugin-sdk`) on every PR and push into dev and main, and Docker Scout on the built image for main. Neither is required: a newly published advisory would turn every open PR red at once. Fixes are held as floors in the root `package.json``overrides`; npm ignores an `overrides` block in a workspace `package.json`, so never put one there.
- **Releases** are dispatched by hand and refuse a commit whose `ci-ok` is not green. `docker-dev.yml` (any branch, usually dev) is the canary: it builds the dispatched commit as the next minor or major `-pre.N` and moves only `:latest-pre` and `:<major>-pre`. `docker.yml` (main only) bumps the version, commits and tags it, builds both platforms from exactly that commit and moves `:latest`, `:<major>`, `:<major>.<minor>` and `:<version>`. Its `bump=auto` only finalizes a prerelease in flight; without one it stops and asks for patch, minor or major. Ship a release that migrates or changes a lot as a prerelease first, and as a minor, so installs pinned to `:<major>.<minor>` can wait it out.
@@ -146,7 +146,7 @@ Most of what follows is an addon an admin switches on or off. Lists, Costs, Docu
#### 🗺️ Atlas, Vacay and Collections
- **Atlas**: countries and regions visited on [geoBoundaries](https://www.geoboundaries.org/) outlines, a bucket list, travel stats, and a tour activities tab with planned against recorded totals
- **Vacay**: leave calendar with half days, public holidays from date.nager.at, school holidays for 16 European countries, company holidays, carry-over between years, and joint or read-only sharing
- **Vacay**: leave calendar with half days, public holidays from [nagerholidays.com](https://nagerholidays.com), school holidays for 16 European countries, company holidays, carry-over between years, and joint or read-only sharing
- **Collections** (off by default): a place library outside any trip with lists, labels, ratings, a status per place, cover images, GPX and file export, copy into a trip, and sharing by invite
- **Dawarich** (off by default): link a Dawarich instance for recorded tracks, suggested stays and countries for the Atlas
@@ -276,6 +276,8 @@ Open `http://localhost:3000`, or `http://<server-ip>:3000` from another machine.
`COOKIE_SECURE=false` is only for trying TREK over plain HTTP: without it the browser drops the session cookie and the login does not stick. Once TREK sits behind HTTPS (see [Reverse Proxy](#reverse-proxy)), remove it.
**TrueNAS SCALE:** Install TREK directly from the Community catalog. See the [TrueNAS SCALE installation guide](https://github.com/liketrek/TREK/wiki/Install-TrueNAS-SCALE).
> [!TIP]
> **No server yet?** TREK's own public services run on [**dataforest cloud**](https://cloud.dataforest.net/en?mtm_campaign=trek), our infrastructure sponsor. A small seed is plenty for a TREK instance and is ready within a minute.
@@ -319,6 +321,11 @@ docker compose up -d
See [Install with Docker Compose](https://github.com/liketrek/TREK/wiki/Install-Docker-Compose)
for the full walkthrough.
TREK stores its data in SQLite by default. Set `TREK_DB_URL` to run it on a PostgreSQL 14 to 18 database instead. The
[`docker-compose.postgres.yml`](docker-compose.postgres.yml) example starts one beside TREK, and the
[Postgres guide](https://github.com/liketrek/TREK/wiki/Database-Postgres) covers the requirements, the backups and
how to move an existing SQLite install to Postgres with the command TREK ships for it.
<br />
<h2 id="helm-kubernetes">Helm (Kubernetes)</h2>
@@ -348,7 +355,7 @@ TREK then launches fullscreen with its own icon, just like a native app.
## Updating
See [Updating](https://github.com/liketrek/TREK/wiki/Updating) — Docker Compose,
Docker run, Helm, Portainer, Unraid and Proxmox, plus the encryption-key note.
Docker run, Helm, Portainer, Unraid, TrueNAS SCALE and Proxmox, plus the encryption-key note.
<br />
@@ -437,8 +444,9 @@ Every variable, its default and what it does: see
The Atlas map's country and sub-national (province/county) boundaries come from
[**geoBoundaries**](https://www.geoboundaries.org/) (Runfola et al., 2020), licensed
[CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). See [NOTICE.md](NOTICE.md)
for full third-party attributions.
[CC BY 4.0](https://creativecommons.org/licenses/by/4.0/); Ireland's counties come from
[**Tailte Éireann**](https://data-osi.opendata.arcgis.com/), also CC BY 4.0. See
[NOTICE.md](NOTICE.md) for full third-party attributions.
@@ -43,8 +43,10 @@ See `values.yaml` for more options.
## Notes
- Ingress is off by default. Enable and configure hosts for your domain.
- PVCs use the cluster's default StorageClass. Set `persistence.data.storageClassName` and/or `persistence.uploads.storageClassName` to bind a specific class.
-To use your own PVCs, set `persistence.data.existingClaim` and/or `persistence.uploads.existingClaim`. The other values for that volume (size, storageClassName, annotations) are then ignored.
-Both PVCs are `ReadWriteOnce`. Set `persistence.data.accessModes` and/or `persistence.uploads.accessModes` to override, for example `[ReadWriteMany]` so a backup pod can mount the volume. On the data volume such a pod may only read: SQLite corrupts `travel.db` when a second process writes it over a network filesystem.
- To use your own PVCs, set `persistence.data.existingClaim` and/or `persistence.uploads.existingClaim`. The other values for that volume (size, storageClassName, accessModes, annotations) are then ignored.
- With `persistence.enabled: false`, the data and uploads volumes use an `emptyDir` — storage is ephemeral and lost on pod restart. Intended for testing only.
-`database.url` runs TREK on a PostgreSQL database instead of the SQLite file. The chart writes it into its Secret as `TREK_DB_URL`; `database.existingSecret` and `database.existingSecretKey` read it from a Secret you manage. Leave both empty for SQLite. Keep `replicas: 1` either way. See the wiki page `Database-Postgres`.
-`JWT_SECRET` is managed entirely by the server — auto-generated into the data PVC on first start and rotatable via the admin panel (Settings → Danger Zone). No Helm configuration needed.
-`ENCRYPTION_KEY` encrypts stored secrets (API keys, MFA, SMTP, OIDC) at rest. Recommended: set via `secretEnv.ENCRYPTION_KEY` or `existingSecret`. If left empty, the server falls back automatically: existing installs use `data/.jwt_secret` (no action needed on upgrade); fresh installs auto-generate a key persisted to the data PVC.
- Liveness and readiness probes are configurable under `probes`. Overrides are merged over the defaults, so `--set probes.liveness.initialDelaySeconds=60` keeps the `httpGet` handler. When switching to `exec` or `tcpSocket`, clear `httpGet` in the same override. Set `probes.liveness: null` to remove a probe.
{{- if and .Values.database.url (not .Values.database.existingSecret) .Values.existingSecret }}
{{- fail "database.url is written into the chart's own Secret, which existingSecret replaces. Put TREK_DB_URL in your Secret and set database.existingSecret instead." }}
{{- end }}
apiVersion:apps/v1
kind:Deployment
metadata:
@@ -5,11 +8,13 @@ metadata:
labels:
app:{{include "trek.name" . }}
spec:
# One replica. Its rate limits, pending codes, permission cache and WebSocket
# connections live in the process's memory, so a second pod would not share them. That
# holds with database.url (PostgreSQL) too; see updateStrategy in values.yaml.
replicas:1
# TREK is a single-writer SQLite app on a ReadWriteOnce PVC, so the default
# RollingUpdate would start a second pod holding the same volume before the old one
# exits — a Multi-Attach deadlock, or two processes on one travel.db. Recreate tears
# the old pod down first. Keep it, see updateStrategy in values.yaml.
# The default RollingUpdate would start a second pod on the same ReadWriteOnce PVC
# before the old one exits: a Multi-Attach deadlock, or two processes on one travel.db.
# Recreate tears the old pod down first. Keep it, see updateStrategy in values.yaml.
@@ -6,7 +6,7 @@ Scope: the **`@trek/client`** workspace (React 19 + Vite + Zustand + Tailwind PW
```bash
npm run dev # Vite dev server; proxies the API/ws/uploads/MCP/OAuth paths → http://localhost:3001
npm run build # prebuild generates PWA icons, then vite build
npm run build # prebuild generates PWA icons, then vite build; postbuild writes a Brotli copy beside each hashed asset for the server to hand out
npm run typecheck # tsc --noEmit (CI)
npm run lint:strict # tsc over tsconfig.strict.json (strictNullChecks + noImplicitAny): errors per file may only shrink against scripts/strict-baseline.json, a new file starts at zero (CI gate; --list shows them, --update lowers it)
npm run lint # eslint . (CI runs lint:warnings instead, which fails on the same errors)
<stringname="widget_bookings_description">Οι επόμενες πτήσεις, τα τρένα, τα ξενοδοχεία και τα τραπέζια σου, και πότε είναι.</string>
<stringname="widget_flight_name">Πτήση</string>
<stringname="widget_flight_description">Η επόμενη πτήση σου ζωντανά: αντίστροφη μέτρηση για την αναχώρηση, το αεροσκάφος καθ’ οδόν, καθυστέρηση και άφιξη.</string>
<stringname="widget_flight_description">Penerbangan berikutnya secara langsung: hitung mundur keberangkatan, pesawat dalam perjalanan, keterlambatan, dan kedatangan.</string>
<stringname="widget_todos_name">Tugas</string>
<stringname="widget_todos_description">Yang masih tersisa untuk perjalanan dan sejauh mana persiapan barang.</string>
<stringname="widget_bookings_description">I tuoi prossimi voli, treni, hotel e tavoli, e quando sono.</string>
<stringname="widget_flight_name">Volo</string>
<stringname="widget_flight_description">Il tuo prossimo volo in tempo reale: conto alla rovescia per la partenza, l’aereo in viaggio, ritardo e arrivo.</string>
<stringname="widget_todos_name">Da fare</string>
<stringname="widget_todos_description">Cosa manca ancora per il viaggio e a che punto è la valigia.</string>
??"Passkeys and security keys do not work on this page in the app. Sign in with a password, or let the app through your proxy."
:labels["errorPageHint"]??"This page is not from TREK: the server or a proxy in front of it refused the app. Check your proxy or certificate, or change the server."
"widget.today.description" = "El pla del dia com a línia de temps, amb el temps i la propera parada.";
"widget.bookings.name" = "Reserves";
"widget.bookings.description" = "Els teus propers vols, trens, hotels i taules, i quan toquen.";
"widget.flight.name" = "Vol";
"widget.flight.description" = "El teu proper vol en directe: compte enrere de la sortida, l’avió en camí, retard i arribada.";
"widget.todos.name" = "Tasques";
"widget.todos.description" = "El que encara queda per al viatge i com va l'equipatge.";
"widget.packing.name" = "Equipatge";
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.