So, I've been working on developing a little first application for myself, to learn some of Construct's capabilities.
When the main layout starts, it immediately prompts for continuous interim Speech Recognition. I have a lot of events that seek to analyze the interim transcript, and perform actions based on audio input. Simple things, like [Object][Command]... "Square. Up." for example.
I have been trying to make the entire game run off of a single request for continuous mode Speech Recognition to not annoy the user with repeated prompts. I've run tests and it mostly works, set up as follows:
On speech recognition result, if SpeechRecog.InterimTranscript matches regex "Square" ("gi"), then Dictionary set key "ActiveShape" with value "SQUARE".
However, I notice from time to time that Speech Recognition stops on its own. I haven't figured out if this is because I have alt-tabbed to a different window, if the mic has been inactive for a duration, or if "on speech recognition result" has not been triggered for a period of time. I am then required to re-request mic usage. Why is this? Is there a way to ensure the mic remains active until specifically told to stop?
As I require speech recognition, I have been testing in Chrome and not my default Firefox which apparently does not support speech recognition.