After missing the schedule and more than two months later than planned, the prototype is ready for hardware behavior coding. The code structure already works exchanging information with the web interface, which has been (by far) the most delicate part of the process. For the tests I have only developed two types of device: The controller itself and a binary digital input.

I don’t think I can recover almost anything from the previous version. It will be faster and more efficient to recode from the new structure. Something I did wrong last time was not having a good benchmark for the hardware. Thinking that the code for the Arduino ports would be simple, I made the mistake of doing the tests on the installed device itself. That’s how I discovered “operation by chance”. Trial and error might work to turn on a simple light bulb or detect a switch press, but I couldn’t get the blind controller to work right.

This time I’ll have a bunch of real stuff plugged into the insert board so I can see with the computer in front of me how they really work and do all the refactoring without leaving the developing table. Considering the time I’ve already wasted, it’s probably worth spending a little more if the result has the quality I’m needing. From now on, I will make more updates as the progress will be significant.