In trying to deal with a codebase that was originally proprietary and then open sourced, new contributors are coming along. Each file has names on it (though usually the main author's).
At this point, the codebase is drifting considerably. Maintaining a list of names in each file seems to clutter it, and introduces an unnecessary awkwardness of "when a change is significant enough to be on the list". There's also an issue of accuracy, because the name is generally not the person to try and contact about issues.
I was thinking that it would be best to just strike the names entirely, rely on the git blame log, and switch to a CONTRIBUTORS file. (This is along with a general push to take stuff repeated in each file and move it out to coding guidelines documents, etc.)
This project is on GitHub and is:
- mostly C code
- has a
README.mdas its entry point for describing the project (80-column limited)
- has a
LICENSEin text file format (Apache2)
So what sort of model might one use for a contributors file? An all-caps CONTRIBUTORS seems like a long name. Wondering what people call it and what would go in it.
It hit me right after asking that CREDITS.md would be a good short name (somewhat obvious, and I've even used it in the past I believe). But I wasn't acknowledging individual contributors to the code when I did this before, I was acknowledging the sources drawn from (libraries, tools, etc.) to build it.
So it still strikes me that these might be different files. But maybe they could both go in one? Then CONTRIBUTORS would be a section heading in the CREDITS.md file.