Github and making the Commitment
Now it is time for the developer to commit the changes she has been working on, tested and pushed through the peer reviews. This step is pretty simple, however, we want to make sure we are doing it correctly.
Todo example commit
When the code is merged into the main or master repository, just like commitments, it will be accompanied by the log comments that go along with the log files.
Additionally, every commit will have a unique identifier, a big nasty looking hex-i-decimal string guaranteed to most likely be locally unique.
History in the Making
Well written comments, source diffs and a unique identifier come together to create a specific moment in the code life history.
Stringing all commitments together, one forms a very accurate history of the source development. If the comments are well written, that history is accompanied by a wonderful narrative told by the developers as the software is being written.
The Beginning of Visibility
This can provide the viewer with literally a line by line replay of the entire history of a software development project. The commitments can also be directly related to the particular github issues that have spurred the updates to the code in the first place.
This is a very powerful way for not only developers, but also users, management and marketing a deep insight into the process of software development.
Tagging - Releases
Git also provides a feature called tagging, which is in fact very similar to branching. Tags however are meant to label a specific reference of the source at a specific point in time.
Software releases are a good reason to use tags. For example: a tag v1.0.2 would be a typical tag that might be used to reference the version of code after v1.0.1.
This practice allows the team, or anybody interested, the ability to reference and re-create a specific past version of the software.
Automated Builds and Testing
Github provides a ton of goodies with it’s well defined API and the plethora of web-hooks it provides.
In the next section we’ll get into automatically building new images upon every committment, running the images through automated tests and finally sending only the tested images to a file server for general access.