edfcore 0.4.299 → 0.4.301

package/dist/constants.d.ts

@@ -109,5 +109,5 @@ export declare const SIGNAL_FIELD_BLOCK_OFFSETS: {
     readonly reserved: 224;
 };
 /** Published package version. Kept in sync with package.json by a test. */
-export declare const VERSION = "0.4.299";
+export declare const VERSION = "0.4.301";
 //# sourceMappingURL=constants.d.ts.map

package/dist/constants.js

@@ -79,5 +79,5 @@ export const SIGNAL_FIELD_BLOCK_OFFSETS = {
     reserved: 224,
 };
 /** Published package version. Kept in sync with package.json by a test. */
-export const VERSION = '0.4.299';
+export const VERSION = '0.4.301';
 //# sourceMappingURL=constants.js.map

package/docs/CHANGELOG.md

@@ -6,6 +6,37 @@ alone does not tell you whether you were affected.
 edfcore is pre-1.0. Patch releases have carried behaviour changes where the old behaviour was a
 defect; those are called out below.
 
+## 0.4.301
+
+- **Executed** the worked example on `concepts.md`, the page the site opens with and the README
+  calls "the mental model the rest of the API follows from". It is built almost entirely out of
+  arithmetic on one described file — a 768-byte header, thirty 544-byte records, 17,088 bytes
+  total, and a ten-record read of the narrow channel costing 5,440 bytes for 160 samples — and
+  every number was prose. A reader who works through it and gets a different answer from their own
+  file has no way to tell which of the two is wrong.
+- All of them are correct. What was missing is anything keeping them so: they follow from
+  `headerByteLength = 256 * (signals + 1)` and the record layout, and a change to either would
+  leave the page teaching the old ones. The fixture is built to the page's description with the
+  suite's own writer, which imports nothing from `src/`, so the numbers are checked against a file
+  assembled from the specification rather than against edfcore's idea of one.
+
+## 0.4.300
+
+- **Executed** the claim the error API is shaped around. `src/errors.ts` says class identity is
+  false across a realm boundary, `api-errors.md` repeats it, and `public-api.test.ts` files
+  `isEdfError` under a heading calling it the cross-realm discriminator — none of them showed it
+  happening. An API built entirely around a property nobody demonstrated is an API built around a
+  belief.
+- Two copies of the module rather than a `vm` realm, because two copies is the case that reaches
+  people: one dependency tree resolving edfcore twice, which npm does whenever two packages want
+  incompatible ranges. `instanceof` fails across them, `isEdfError` does not, and the
+  discriminator survives — while `instanceof` keeps working inside one copy, which is what makes
+  the failure invisible to every test a consumer writes against their own.
+- Writing it corrected my reading of `isEdfError`: it is a duck type, and a plain object with a
+  string `edfErrorKind` passes. That is the design rather than a hole — tightening it to
+  `instanceof Error` would reintroduce the problem, since `Error` identity is per-realm too, so
+  the check would fail on exactly the foreign errors it exists to recognise. The test says so.
+
 ## 0.4.299
 
 - **Extended** `verify:site` to check the generated markdown carries the page, not just its head.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "edfcore",
-  "version": "0.4.299",
+  "version": "0.4.301",
   "description": "Modern, typed, zero-dependency reader for EDF, EDF+, BDF and BDF+ biosignal files. Works in browsers and Node with true random access.",
   "keywords": [
     "edf",

package/src/constants.ts

@@ -93,4 +93,4 @@ export const SIGNAL_FIELD_BLOCK_OFFSETS = {
 } as const;
 
 /** Published package version. Kept in sync with package.json by a test. */
-export const VERSION = '0.4.299';
+export const VERSION = '0.4.301';