Some notes and learnings when working with Tailr.fit, including some python and things I needed to learn for this project.
I had an idea on how I was going to approach this project. First though, I was going to start off with some python and scripts to capture images using a usb camera. BAM!
Starting off I ran into scaffolding. Coming from JS and or some other langs I've been dabbling in I was looking for some corrolations. I thought I'd need a particular directory structure, some dependancy management and the unknown unknowns. Turns out the unknown uknowns started quick, so by default pip usage should use a virtual env, this is a per python project thing, and you boot the venv and install within it. But even after installing within the venv, there was no dep list like in js e.g. package.json and its lock file. You can use something like the pip freeze > requirements.txt but versioning becomes cumbersome. So I opted for pipenv which creates a virtual env if it doesn't find one and adds the niceness of a Pipfile and a Pipfile.lock.
Once that scoffolding was set up I was able to install opencv-python and activate the venv pipenv shell (next to the py file). Check out Real python: pipenv guide
I started on the camera capturing service. That was actually easier than expected. I had script to start off from and the help of AI so I had something up in running for a single camera in about 10 minutes. Afterwards, I removed the singular portion of the py script and used the script to capture the the array of camera's (2) frames.