Commit Graph
6 Commits
Author SHA1 Message Date
bvandeusenandClaude Opus 5 fd1e4ae487 downloads: five clients, and the page leads with the one that fits you
CI & Build / Build now, or wait for Android? (push) Successful in 3s
Android / Build, or is the channel already serving this? (push) Successful in 3s
Desktop (Tauri) / Build, or is the channel already serving this? (push) Successful in 2s
CI & Build / Python lint (push) Successful in 3s
Android / Kotlin + Rust (APK) (push) Skipped
CI & Build / TypeScript typecheck (push) Successful in 6s
CI & Build / Python tests (push) Successful in 11s
CI & Build / integration (push) Successful in 19s
CI & Build / Build & push image (push) Successful in 36s
Desktop (Tauri) / Windows installer (cross-compiled) (push) Successful in 2m21s
Desktop (Tauri) / Tauri desktop (Linux) (push) Successful in 4m21s
Desktop (Tauri) / Update manifest (push) Successful in 5s
The Account page offered the APK and nothing else, because the APK was all
the server held. Step 3 baked in four more, so the single card had to become
a section — and five artifacts is exactly where a downloads page turns into
a table of filenames and stops being a product.

So it LEADS with what fits the machine asking, from the user agent, and keeps
the rest quiet but visible. A wrong guess costs nothing: nothing is behind a
disclosure and every other client is one click away.

Linux gets all three at once, because the UA says "Linux" and nothing about
dpkg or pacman — there is no better answer available. They are named for the
distro rather than the package format, since a person knows which system they
run and not necessarily which packaging it uses. The AppImage carries one
clause of its own: it is 95 MB against 3, and it is also the only bundle that
updates itself in place. Both facts belong to the same decision.

macOS and iOS lead with nothing and say so. There is no build for either, and
"There's no macOS build yet" is the difference between deliberate and broken.

The version renders `unknown` rather than blank, and the download stays
offered — not knowing which build it is, is not a reason to withhold it.

Two things this did NOT do, both deliberate:

The task asked for a Tauri case — do not offer the desktop app to someone
already running it. That case cannot be reached: `/account` redirects to the
board in the desktop app (requiresServer, router/index.ts), because device
tokens are a server-side concept. A branch for it would be dead code.

`.btn-link` mirrors BaseButton's declarations rather than replacing them.
BaseButton is a <button> and cannot carry an href, and unifying the two would
have put every button in the app into an operator pass that CI cannot check —
for a cosmetic gain. The comment names the pair.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01K3MMqUtzX1TJgA1oypvm1c
2026-08-31 07:58:17 -04:00
bvandeusen d77a79859c server: hand out the Android client this server syncs with (2726)
CI & Build / Python tests (push) Successful in 11s
CI & Build / Python lint (push) Successful in 3s
CI & Build / TypeScript typecheck (push) Successful in 6s
CI & Build / Build & push image (push) Successful in 50s
Desktop (Tauri) / Windows installer (cross-compiled) (push) Failing after 3m8s
Desktop (Tauri) / Tauri desktop (Linux) (push) Failing after 5m32s
Desktop (Tauri) / Update manifest (push) Skipped
Android / Kotlin + Rust (APK) (push) Successful in 8m8s
A self-hoster should not need an account on someone else's forge to get the app
for their own notes. The Fabled-Git instance is private — which is why
`install.sh` already cannot fetch for anyone but the operator — so a release page
is no use as a distribution point. The server holding the notes is something the
person already trusts and already reaches.

It also keeps the pair in step by construction. Client and server negotiate a
sync protocol version before linking, so a server that also serves the client
cannot hand out a phone it is unable to talk to.

**Two files, and both must be present**: `thoughtsync.apk` and a
`thoughtsync-android.json` sidecar carrying `{version_name, version_code, size,
sha256}`. The sidecar exists because an APK keeps its version in a binary AXML
manifest, which Python cannot read and which is not worth putting `aapt` on a
Quart server to reach. CI writes it beside the APK, where the values are already
known — including the digest, computed over the same bytes it uploads, so a
phone can tell a truncated download from a complete one before handing it to the
installer. Not a trust anchor; the signature is that.

**Under DATA_DIR, not baked into the image.** Baking charges ~55 MiB to every
self-hoster including everyone who never touches Android. `/var/thoughtsync` is
already the mounted volume that holds attachments, so a build dropped there
survives container recreation.

**Absence is an ordinary state, not an error.** No APK means the key is absent
from `/api/config` — absent rather than null, so a client testing for it cannot
confuse "this server has no client" with "this server predates the field" — the
web UI hides the card instead of offering a button that 404s, and the metadata
route answers 404. A server whose owner does not use Android is not misconfigured.

**A mismatched pair also counts as no client.** If the sidecar's recorded size
does not match the file on disk, the two did not arrive together; serving one
build while advertising another is worse than serving none, because the phone
would compare versions against a promise the bytes do not keep. That makes the
copy order in docs/android-distribution.md load-bearing, and it is written down
there: APK first, sidecar last.

**The version is public, the bytes are not.** An updater has to be able to ask
"is there something newer?" cheaply and before it has done anything; 55 MiB is
not for anyone who can reach the port. `login_required` already accepts either a
session cookie or a device bearer token, so the browser and a linked phone both
work with no second auth path.

The Android lane now publishes both files to the same rolling `dev` release the
desktop bundles use, reusing `publish-release.sh` — its nullglob asset list was
already built for several jobs in separate workspaces publishing to one release,
which is exactly this. Signed builds only: publishing an unsigned APK would offer
people something they cannot install over what they already have.

Nine tests, DB-free like the rest of the suite — this lane runs no Postgres, so
the advertisement is asserted through `advertisement()` rather than through
`/api/config`, whose other half needs a database. Both routes ARE exercised,
because neither opens a session.
2026-08-20 20:11:32 -04:00
bvandeusenandClaude Opus 5 e64d67e904 Expire trash after 30 days, and make the deadline something you can see
CI & Build / Python lint (push) Successful in 2s
CI & Build / TypeScript typecheck (push) Successful in 6s
CI & Build / Python tests (push) Successful in 12s
CI & Build / Build & push image (push) Successful in 44s
Desktop (Tauri) / Tauri desktop (Linux) (push) Failing after 1m45s
Desktop (Tauri) / Windows installer (cross-compiled) (push) Successful in 2m12s
Trash had no end. A note sat in /trash until someone emptied it by hand, and
its attachment BYTES sat on disk the whole time — the pile-up the operator
asked about. Nothing purged; there was no scheduler at all.

Retention is server-owned: `trash_retention_days` (default 30, 0 = keep
forever) in the settings registry, so it lands in admin Settings with no
migration and takes effect without a restart. A background sweep started in
before_serving does the work. Clients learn about a purge the way they learn
about any deletion — as a tombstone on the delta feed.

An auto-purge nobody can see coming is data loss on a timer, so the window is
now visible: /api/config publishes it, notes carry `deleted_at`, Trash leads
with the policy, and each card counts down. The countdown rounds DOWN — saying
"1 day left" for a note with ten minutes on the clock is the one error here
that actually costs someone a note.

Three things this turned up on the way:

- `DELETE /api/notes/<id>` hard-deleted the row, leaving no tombstone at all.
  A permanent delete in the web UI never reached a linked device, which would
  keep its copy forever and push it back on the next edit. It now purges
  through the same path as everything else.
- The purge left `note_revisions` and `note_link_previews` behind. A revision
  holds the full body, so the text of a "permanently deleted" note was still
  sitting in the database.
- `deleted_at` now SURVIVES a purge instead of being cleared. It's still true,
  and it means every query that says "not trashed" excludes tombstones for
  free — without it a content-less row reads as a perfectly normal active note
  and shows up on the board as a blank card.

Desktop keeps its own clock only when there's nobody else to keep one: the
sweep runs at startup on an UNLINKED device and refuses otherwise. A linked
client that expired notes on its own schedule could destroy something the
server was deliberately keeping, then push that delete upstream. Local policy
must never outrank the server's — so it also adopts the server's window for
the countdown rather than showing its offline default.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01SreJkbxB4gx8pPsu8QbLPi
2026-07-26 16:20:13 -04:00
bvandeusenandClaude Opus 4.8 20cf15c99c desktop M10.3: frontend data-source adapter seam (repo interface + rest.ts)
CI & Build / Python lint (push) Successful in 3s
CI & Build / TypeScript typecheck (push) Successful in 7s
CI & Build / Python tests (push) Successful in 9s
CI & Build / Build & push image (push) Successful in 29s
Extract a typed repository interface (adapters/repo.ts) from the scattered
store/view -> api.* calls, backed by adapters/rest.ts (verbatim HTTP mapping)
and selected through adapters/index.ts. Every store and the notes-facing views
now depend on `repo`, never the HTTP client directly -- the seam the offline
local source (M10.5, over Tauri invoke) plugs into next.

Behavior-preserving for web: rest.ts maps each semantic method to the exact
endpoint the code called before; query-string and multipart building moved out
of the stores/views into rest.ts (the one place that knows the URL shape).
Client-side logic (reconcile/sort/optimistic reorder/toasts) stays in the
stores. GraphView + admin SettingsView keep direct api calls -- out of the
offline-core scope (M10.5 is board/editor/capture/search/filter/labels/
checklists/reminders).

Task 1992.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FRgehjoz7Yv8LkUfADxACm
2026-07-24 22:31:00 -04:00
bvandeusenandClaude Opus 4.8 69bf04e948 M6 1901: URL capture with link-preview unfurl (SSRF-hardened)
CI & Build / Python lint (push) Successful in 3s
CI & Build / Python tests (push) Successful in 12s
CI & Build / TypeScript typecheck (push) Successful in 34s
CI & Build / Build & push image (push) Successful in 33s
Paste a link → fetch its OpenGraph/meta preview (title, description, image,
site) and show a rich card. User-triggered + persisted (never auto-fetches;
cached so it never re-fetches). Opt-in via a new admin setting
enable_url_unfurl (default on, rule 26).

Security (the whole point of this task): a new dependency-free unfurl.py
does the fetch with layered SSRF defenses — http/https only; resolve the
host and reject EVERY non-public address (private/loopback/link-local/
reserved/multicast/unspecified — blocks 169.254.169.254 etc.); connect to
the vetted IP with SNI so DNS-rebinding can't slip through; ≤3 redirects
each re-validated; 5s timeout; 512 KB cap; text/html only; blocking IO in a
worker thread. No server-side image fetch — the og:image URL is loaded by
the browser.

- note_link_previews table (migration 0020), one per (note, url); serialized
  inline on notes (+ rides the sync pull feed read-only).
- POST /api/notes/<id>/unfurl {url} (owner-scoped, setting-gated, 502 on
  fetch failure); DELETE /api/notes/<id>/previews/<id>.
- enable_url_unfurl exposed in public config so the UI hides the affordance
  when disabled.

Frontend: LinkPreview.vue card; editor detects URLs in the body and offers a
"Preview <domain>" chip per un-previewed link (ensureDraft first), renders
preview cards with remove; card shows previews read-only. New link icon;
notes-store unfurl()/deletePreview().

Tests (DB-free): is_public_ip range blocking, validate_url scheme/parts,
extract_preview (OG + <title> fallback + relative-image resolve), endpoint
auth-guards.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FRgehjoz7Yv8LkUfADxACm
2026-07-23 08:05:24 -04:00
bvandeusenandClaude Opus 4.8 68f7b260de M1.5 frontend: admin Settings UI + config store + register gating
CI & Build / Python lint (push) Successful in 3s
CI & Build / Python tests (push) Successful in 8s
CI & Build / TypeScript typecheck (push) Successful in 10s
CI & Build / Build & push image (push) Successful in 31s
- config store (public /api/config: site_name, allow_registration, version),
  loaded in the router guard; site name drives the board header.
- session User gains is_admin.
- /settings route with requiresAdmin guard + admin-only gear link in the header;
  SettingsView renders grouped typed fields (text/number/toggle), saves via
  PATCH /api/settings with saved/error feedback, refreshes public config.
- Register link hidden + /register route blocked when registration is closed.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FRgehjoz7Yv8LkUfADxACm
2026-07-19 18:28:56 -04:00