edfcore 0.4.398 → 0.4.400
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/dist/constants.d.ts +1 -1
- package/dist/constants.js +1 -1
- package/docs/CHANGELOG.md +28 -0
- package/package.json +1 -1
- package/src/constants.ts +1 -1
package/dist/constants.d.ts
CHANGED
|
@@ -109,5 +109,5 @@ export declare const SIGNAL_FIELD_BLOCK_OFFSETS: {
|
|
|
109
109
|
readonly reserved: 224;
|
|
110
110
|
};
|
|
111
111
|
/** Published package version. Kept in sync with package.json by a test. */
|
|
112
|
-
export declare const VERSION = "0.4.
|
|
112
|
+
export declare const VERSION = "0.4.400";
|
|
113
113
|
//# sourceMappingURL=constants.d.ts.map
|
package/dist/constants.js
CHANGED
package/docs/CHANGELOG.md
CHANGED
|
@@ -6,6 +6,34 @@ alone does not tell you whether you were affected.
|
|
|
6
6
|
edfcore is pre-1.0. Patch releases have carried behaviour changes where the old behaviour was a
|
|
7
7
|
defect; those are called out below.
|
|
8
8
|
|
|
9
|
+
## 0.4.400
|
|
10
|
+
|
|
11
|
+
- **Added** a check that the two tsconfigs differ only where they are meant to. `npm run typecheck`
|
|
12
|
+
runs both on purpose — the build config compiles `src/` with `lib: ["ES2022"]` and `types: []` so
|
|
13
|
+
the DOM and `@types/node` cannot leak into the published declarations — and AGENTS.md states that
|
|
14
|
+
split. What it does not state is that everything else about them has to match.
|
|
15
|
+
- That is the half with consequences. Every strictness flag in the root config is the setting the
|
|
16
|
+
suite is written under, including the `tests/types/*.test-d.ts` files whose job is to prove the
|
|
17
|
+
public API compiles for a consumer. Turn `noUncheckedIndexedAccess` off there alone and they keep
|
|
18
|
+
passing while they stop testing what they claim: `src/` still compiles strictly, and the snippet
|
|
19
|
+
asserting a consumer's experience is now checked under settings no consumer has.
|
|
20
|
+
- The repository has already paid for that flag once. Until 0.4.259 the snippet AGENTS.md tells
|
|
21
|
+
agents to copy ended `chunks[0].signals[0].digital` and did not compile in a strict project.
|
|
22
|
+
- The eleven strictness flags are also asserted individually, so turning one off in *both* configs —
|
|
23
|
+
which a comparison alone would call agreement — still fails. Divergences must be named, and a
|
|
24
|
+
named divergence the configs have since settled fails too, because a stale exception is a hole.
|
|
25
|
+
|
|
26
|
+
## 0.4.399
|
|
27
|
+
|
|
28
|
+
- **Added** to `verify:tarball` the three files npm packs by convention rather than because `files`
|
|
29
|
+
asks for it: `package.json`, `README.md` and `LICENSE`. `files` is an allow-list and none of the
|
|
30
|
+
three is on it, so all three ship on a rule this repository does not state and cannot see.
|
|
31
|
+
- A package that arrives without `LICENSE` is an MIT package whose terms are not in it, which is
|
|
32
|
+
the half of 0.4.398 a text comparison cannot reach. Without `README.md` the npm page is blank —
|
|
33
|
+
the first thing anyone sees of the project, and the last thing anyone would think to check.
|
|
34
|
+
- `.npmignore` is what makes this reachable rather than theoretical: it overrides `files` outright,
|
|
35
|
+
so adding one for an unrelated reason can take all three out at once.
|
|
36
|
+
|
|
9
37
|
## 0.4.398
|
|
10
38
|
|
|
11
39
|
- **Added** a check that the licence is one licence. `package.json` declares `"license": "MIT"` —
|
package/package.json
CHANGED
package/src/constants.ts
CHANGED