Update docs

docs/Development.md

@@ -82,7 +82,7 @@ Add a `.vscode/launch.json` file to launch the CLI in debug mode:
 			"program": "${workspaceFolder}/dist/cli/CLIStarter.js",
 			"outputCapture": "std",
 			"console": "integratedTerminal",
-			"runtimeArgs": ["--no-warnings", "--no-experimental-fetch", "--experimental-wasi-unstable-preview1", "--trace-uncaught"],
+			"runtimeArgs": ["--experimental-wasi-unstable-preview1", "--no-warnings", "--trace-uncaught"],
 
 			"args": ["speak", "Hello World!", "--debug"]
 		}

docs/Server.md

@@ -45,7 +45,7 @@ ws.on("open", async () => {
 
 ## Protocol details
 
-The protocol is based on binary WebSocket messages, for both request and response objects. Messages are encoded using the [MessagePack](https://msgpack.org/index.html) encoding scheme.
+The protocol is based on binary WebSocket messages, for both request and response objects. Messages are encoded using the [MessagePack](https://msgpack.org) encoding scheme.
 
 All messages are objects and have the basic structure:
 

docs/Tasklist.md

@@ -260,7 +260,7 @@
 
 ### Speech-to-text translation
 
-* Hybrid approach: recognize speech in its language using any recognition model, then translate the resulting transcript using a text-to-text translation engine, and then align the translated transcript to the original one using text-to-text alignment, and map back to the original speech using the recognition timestamps, to get word-level alignment for the translated transcript
+* Hybrid approach: recognize speech in its native language using any recognition model, then translate the resulting transcript using a text-to-text translation engine, and then align the translated transcript to the original one using text-to-text alignment, and map back to the original speech using the recognition timestamps, to get word-level alignment for the translated transcript
 
 ## Possible new engines or platforms