You’ve finally found a GitHub repository that fits your needs. You are planning to use the project for a longer time. Maybe even for productive usage. So how do you figure out wether the overall quality is good enough? How do you know how active it is? What about the maintainers? Are they still active?
Basics
About stars as GitHub metrics
The most popular GitHub metric people are using is GitHub stars. The more people stared the repository, the better it is - that’s at least what most people are assuming.
It is important to understand, that GitHub stars are actually just bookmarks. So when relying on stars, you are actually relying only on how many people are following that repository. Not necessarily a good indicator for quality, isn’t it?
What is a good GitHub metric to use?
The particular GitHub metric to use, depends on the aspect that is most important for you.
When searching for a smart home software on GitHub, security and stability might be the two most important aspects for you. Whereas if you would be searching for an email software, it might be privacy. Certainly in both cases, the project should be as active as possible, since you don’t want to develop bugfixes on your own.
We at reposcore call this personal set of criterion ‘metric profile’. The metric profile is your individual set of metrics that you use to make decisions.
Where to find the GitHub Metrics
Metric | Where to find it |
---|---|
Commit history timestamps | “Commits” button in the GitHub repository |
Last commit date | Repository homepage, check the dates on the right side |
Issue processing time | Klick on the issue and compare the first to the last timestamp |
Commits per day | Look at the GitHub history, and do the maths |
Maybe we stop here. As you can already see, extracting reliable GitHub metrics manually, quickly becomes tiresome. Extracting data manually just to do the math afterwards, is obviously nothing you’d do for a repository. It is just way too time consuming.
Conclusion
Evaluating the quality of a GitHub repository shouldn’t only be based on GitHub stars. Much more metrics and data is required to get qualified insights. Whereas it is theoretically possible to do this manually, reposcore provides an automated way to do that and much more.
Try it out now, it is free!