Browser support and manual checks

GlyphFlow's initial target is desktop Chrome with English text and a voice that emits word boundary events. Voice behavior also depends on the operating system and installed voice.

Environment	Status	What to expect
Desktop Chrome, English, word-boundary voice	Initial target	Speech and the inside-word wipe work when the browser starts the voice and emits boundaries.
Desktop Chrome, voice without word boundaries	Limited	Speech may play, but the synchronized wipe is unavailable; GlyphFlow reports unsupported when playback ends.
Other desktop browsers	Not yet characterized	Web Speech behavior varies; verify the selected browser and voice manually.
Mobile browsers and non-English text	Outside v0.1 support target	No compatibility guarantee yet.

At startup, call getSpeechSynthesisSupport() to identify environments without the Web Speech API. This only checks API availability. A browser can expose the API but fail to start a particular utterance; GlyphFlow waits for the browser's start event and reports error if it does not arrive within 10 seconds. A voice also needs usable word boundaries for synchronized captions.

Speech synthesis is a shared browser queue. cancel() calls the browser's global speechSynthesis.cancel(), which may also stop speech started by other code on the page. Applications that use multiple speech features should coordinate ownership of that queue.

Run npm run demo with the browser and voice you intend to support. At rate 1.0, check:

Check	Expected result
Start playback	Audible speech begins and state changes from starting to speaking.
Inside-word motion	Colour advances continuously through each glyph, not as a whole-word flash.
Pause and resume	The current wipe holds, then resumes without a jump.
Text fidelity	Repeated words, punctuation, spaces, and line breaks remain aligned.
Missing boundaries	Playback ends with unsupported, rather than implying measured synchronization.
No browser start	State becomes error after 10 seconds instead of remaining speaking.