6285d937904c35f0031601ad657e7d0fcbf27e00
4
Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
  mr.zeroandClaude Opus 5
|
6285d93790 |
Stores can be removed, and added from an address you type
Two gaps that were the same gap: the store list could only ever grow, and it could only grow from what the registry happened to offer. **Removing** uninstalls what the store installed, then deletes the store itself, in that order. The order is the whole of it: `state.json` is the only record of which payloads, icons and menu entries belong to a store, so deleting the home first would strip the one thing that knows — leaving files nothing could ever identify, least of all a later install of the same store into the same folder. The confirmation says how many titles will go, because that is the part nobody would otherwise expect. The token goes too; a credential for a store that is not here is a secret kept for nothing. The window names a *store*, never a path: the home is resolved against what a disk scan actually found before anything is deleted, and `removeHome` refuses anything else. That is the only guard between a bad argument and `rm -rf`, so it has a test. **Adding** moved to a + beside Refresh — both are actions on the whole store rather than on one of them, and the full-width button under the list read as a third store — and the picker now takes a catalog address as well as a listed one. A bare host is enough and the name comes from the address; nothing else about installing changes, which is why the typed path hands the same record to the same method instead of growing a second one. The picker also has a Cancel now: opening it with a store installed used to replace the grid with no way back. `make storetest` is new, and it earned itself immediately. Removal is the only code here that deletes a directory tree, which the smoke test cannot cover — it runs against the real machine and would have to delete a real store to prove anything. Two bugs on the first run: - `http://` was accepted and became a store called *http*. The trailing slashes were stripped before the scheme was checked, turning `http://` into `http:` and then into `https://http:`, whose hostname parses as "http". The URL is rebuilt from the parsed form now, which also settles the trailing slash in one place. - `STORE_ROOT` only *prepended* to the search path, so a "sandboxed" run still listed the real stores — despite the README saying "instead of the real one". Harmless while a sandbox could only add; not harmless now that it can delete. It replaces the search path. The self-test needed two changes, both of which are it working: the store row is a wrapper now, so clicking `.store-row` did nothing at all, and the footer icon check counted exactly two named controls when there are three. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
mr.zeroandClaude Opus 5
|
26c7aa9be1 |
A catalog that can say a title is not yours
ci/woodpecker/push/woodpecker Pipeline was successful
A store with paid titles had nothing to tell this client and no way for it to listen: the catalog carried no price, no entitlement and no sign-in, so a gated download could only come back 403 and leave the window guessing why. The knowledge belongs on the server, not here. This client serves whichever catalog a registry names, so anything it knew about a particular shop would be a rule that breaks every other one. WarpEngine 0.5 answers GET /api/service with what it offers and puts an `access` block on every entry; this reads both. There is no store name anywhere in the diff. - **0.5 is a dialect of its own**, the older shape with `access` added. The version list is exhaustive over the selector, so adding it was a compile error until somebody said what it reads like — which is what that switch is for. - **A card shows a price and a Buy button** when a title is not yours, opening the store's own page. Buying stays in a browser: a checkout rebuilt here would be a second place to get card handling wrong. - **Signing in is the device grant**: a short code, the person's own browser, and no password crossing this window. The token goes in the OS keychain through safeStorage — one per store — and where no keychain exists it is not stored at all rather than written out in the clear. - **Owned / To buy** join the categories, since owning something is not the same as having installed it. Three things worth stating about the shape: The bearer token stops at the origin that issued it. A gated download redirects to signed storage — often somebody else's host — and some object stores refuse a request outright when an Authorization header arrives alongside the signature. An absent access block is not "free". It is an engine too old to have an opinion, and only one of those two is a reason to offer somebody a sign-in, so the three states are kept apart all the way to the card. state.json does not carry entitlement. Whether somebody may download a title is the server's answer to a question asked now; a copy on disk would go stale on the next purchase or refund, and a stale yes is the dangerous direction. A store with no sign-in shows none, and every WarpEngine before 0.5 is such a store: no Account block, no prices, no new categories. The smoke test against the live catalog reports exactly that — `sign-in: not offered`, `access: open:13`. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> |
||
|
mr.zeroandClaude Opus 5
|
06f3f2a3b1 |
The store engine moves into the client, and Python goes with it
Reading the catalog, choosing the release that fits this machine, unpacking it,
writing the menu entry and remembering what went where all happen in process now.
There is no interpreter to find, no child process, and no JSON-lines protocol
between the two halves — `PythonEngineProcessRunner`, the runtime locator, the two
engine mappers and the version negotiation are all gone, and with them the one
unchecked cast this codebase had (engine stdout to a typed event).
What that buys a person: on Windows and on a fresh Mac the app simply works. It
used to look for `python3`, `python` and `py -3` and draw a link to python.org
where none answered.
What lands on disk is unchanged, deliberately. `config.json` and `state.json` keep
the shell engine's snake_case shape, its `<scope>:<name>` keys and its file modes,
so a machine whose library was installed by the CLI keeps it — verified against the
Python engine on the same catalog: the same 13-title listing with zero field
differences, byte-identical payloads, identical modes and an identical Info.plist,
and a re-sync over a Python-installed home that writes nothing. Remove, prune,
prune-suppression on a named sync and the v1 state migration were each exercised.
Three things worth knowing about the new code:
- the zip reader is ~150 lines over `node:zlib`, because Node has none and this
application has no runtime dependencies. It restores the executable bit from
each entry's external attributes, without which nothing installed can start,
and it refuses zip64, unknown compression and paths that escape the
destination rather than guessing;
- `SUPPORTED_WARP_ENGINE_VERSIONS` names the engine versions this client is
written against, checked against the `WarpEngine-Version` header every
response carries. `selectCatalogDialect` switches over that list exhaustively,
so adding a version fails the build — type checker and linter both — until
somebody says what its catalog reads like. An absent header is read as the
oldest version, which is what an engine before 0.4.0 is;
- refresh and the language picker are icons at the foot of the side menu now,
both named for a tooltip and a screen reader, the picker still a real
`<select>` under its glyph.
The repository is free of Python as well: the Makefile, the CI check and the
release script read package.json and the forge's JSON with Node.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|
||
|
mr.zeroandClaude Opus 5
|
3d63c8a0b0 |
TypeScript, in layers, with a strict linter
The client was one main.js, one preload.js, three files in lib/ and one renderer
script. It is now a typed application whose imports point inward: domain (models,
ports, errors) knows nothing about Electron, Node or Python; application orchestrates
it through those ports; infrastructure holds the adapters — the Python CLI, HTTP, the
filesystem, Electron itself — and main, preload and renderer sit on top as hosts.
STRUCTURE.md is the map, and the deliverable as much as the code is: every layer, every
pattern in use (ports and adapters, repository vs gateway, service, DTO and mapper,
composition root, controller and router, single flight, observer streams, state store
with unidirectional flow, passive view, coded error hierarchy, frozen constant tables,
untrusted-data readers) and the naming rules — files, classes, and a verb vocabulary
for methods where find/require/read/list/apply/render/handle each state a contract.
Two properties fell out of the move, and they are why it was worth doing:
- The catalog can be driven with no window and no Electron at all. The smoke test
assembles the same services against the same ports in a plain Node process; it used
to be a script that reimplemented the bridge.
- The window never receives a filesystem path. A title crosses the bridge without
one, and launching is asked for by name, resolved in the main process from the
store's own state. Verified with a fake launcher: an unknown name answers false, a
native title resolves to its menu entry, a hosted one to its catalog URL.
Types are mandatory, including where inference would manage: explicit return,
parameter and property types, strict plus noUncheckedIndexedAccess,
exactOptionalPropertyTypes, noImplicitOverride and noPropertyAccessFromIndexSignature,
typescript-eslint strictTypeChecked and stylisticTypeChecked, exhaustive switches, no
any, no non-null assertions, and no casts on foreign data — engine stdout and the
registry go through readers that turn unknown into typed values. naming-convention
enforces the patterns rather than trusting them.
Two rule conflicts had to be decided rather than papered over. typedef and
no-inferrable-types disagree about `fallback: string = ''`: the annotation wins, since a
signature states its types. erasableSyntaxOnly is off, because it forbids constructor
parameter properties, which are how dependencies are declared here.
The preload and the renderer are bundled by esbuild into one file each: a sandboxed
preload may not require its own modules, and a module script over file:// is blocked by
the page's own origin rules. tsc compiles the rest. The package ships build/** and
package.json — 111 entries, no sources, no toolchain.
New targets: build, typecheck, lint, lint-fix, and check — typecheck, lint, then both
test suites, cheapest failure first. Every script that runs the app builds first, so a
stale bundle cannot be tested.
Nothing about the window changed: same side menu, same categories, same switcher, same
two languages. make check is clean, both test suites pass with one store and with two,
the packaged 1.3.0 bundle drives the real store, and the window was photographed before
and after — the two are the same picture.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
|