Improve queue logic when outputting by chords

[rileykeen]

Here is the pseudocode. Essentially there should be a timer which resets on every API call for the length of that audio sequence. This will keep the text more in sync when you are typing really fast and also make the audio sound better due to increased inflection.

The biggest issue with pressing enter, period, comma, etc all the time is that I have to multitask a lot if I want to prevent pauses in audio output. I'm listening to things I've already typed, while I still have to keep typing new asynchronous content and then remember to trigger enter to make it to keep going every time the sound pauses. Imagine how hard it would be to talk if there was a 5 second delay from when the sound left your mouth to when it hit your ears. It is incredibly exhausting and it also greatly increases the keystrokes required, remember pressing enter basically takes the same time as chording an entire word.

There is a psychologic concept known as echoic memory which basically means your brain can effortlessly store and/replay sound within a 2-4 second interval immediately after the sound is made. I've found that as long as the audio remains in this short interval of what I'm thinking, and as long as the auto automatically outputs as I chord, I've found the streaming of thought still feels quite effortless. With the queue I could simply slow down if I start to reach the edge of the outer edge of that threshold and I wouldn't have to keep track of what words have or have not been submitted as is currently required when submitting sentences or longer phrases.

I understand some of these concepts may be difficult to grasp for those not practiced in chording that may not be able to experience this for themselves. If this does not make total sense please ask questions.