This PR contains the new module for the GPS hat, removes some unused code, updates the hardware docker build to better test the hardware code, and adds our pull request template file.
The gpsPi module code adds functionality that allows a raspberry pi to interpret data that is being fed via the GPS Hat and then pass that along to the commPi to be sent to the server.
The updated docker build modifies the Dockerfile so that the container built actually runs the commPi software, better indicating whether changes to it have broken the build.
Additionally we added in a .dockerignore file, which similar to a .gitignore prevents specified fields from being sent with the docker context. This has the benefit of restricting the files that are included in the docker cache comparison to just files that are actually relevant to the build, instead of the entire project.
This has 2 potential benefits, one being that the docker context that is communicated each time a docker build happens, only ~100kb of data is sent for comparison vs ~250mb. Also since the context is smaller, this should lead to fewer rebuilds and shorter test durations.
Added hardware entry point for different hardware roles. Removed unnecessary code and other bug fixes.
This PR contains the new module for the GPS hat, removes some unused code, updates the hardware docker build to better test the hardware code, and adds our pull request template file.
The gpsPi module code adds functionality that allows a raspberry pi to interpret data that is being fed via the GPS Hat and then pass that along to the commPi to be sent to the server.
The updated docker build modifies the Dockerfile so that the container built actually runs the commPi software, better indicating whether changes to it have broken the build. Additionally we added in a
.dockerignorefile, which similar to a.gitignoreprevents specified fields from being sent with the docker context. This has the benefit of restricting the files that are included in the docker cache comparison to just files that are actually relevant to the build, instead of the entire project.This has 2 potential benefits, one being that the docker context that is communicated each time a docker build happens, only ~100kb of data is sent for comparison vs ~250mb. Also since the context is smaller, this should lead to fewer rebuilds and shorter test durations.
Added hardware entry point for different hardware roles. Removed unnecessary code and other bug fixes.
This PR closes #459, #566 #280 #552.