Prototype Progress

Last week, we created the foundation for our first functional prototype. This week, we’re focusing our efforts on both polishing and expanding it.

The way our prototype had been working so far was simple: a user could “scribble” a letter on their phone, and their scribble would be sent to an API for text conversion the second their finger left the screen. In theory, this sounded fine. In practice, we encountered a bit of an issue.

Hint: Letters like “t”, “i,” and “f” weren’t exactly getting converted correctly.

The way that we had constructed the initial prototype made it so that it was impossible to write letters that required you to lift up your finger. The process we had going was definitely *fast,* but the drawbacks of it couldn’t be ignored.

So, we’ve introduced a quarter-to-half-a-second waiting period after a user finishes scribbling a letter before we send their scribble to our API to deduce. If you lift your finger up and the waiting period hasn’t expired, your doodle won’t be sent. Now, it’s possible to write every letter with our tool. While speed has decreased a little, the added functionality is vital. In the future, we could introduce user-specific calibration tools to help users settle with a timer speed that’s just right for them.

On the Unity side of things, we’re just getting started with what will become a Bluetooth-driven text input demo. We decided to make the eventual “upgrade” from local HTTP to Bluetooth due to the eventual ease of setup, as well as the fact that Gear VR and Oculus Go both use Bluetooth to connect with their respective controllers. To help guide our implementation, we found an interesting (and well-timed, for us) piece online about reverse-engineering the Gear VR controller.

Our team was split up in the past week to work on non-programming assignments, but we expect to come back together in a push to integrate all the pieces before the end of the quarter. Stay tuned!