4 ms·
I think it is interesting idea. However, I think you need a better system than just the number of commits, or lines of code committed. I don't these are very ac
by gersh 14y ago
I think it is interesting idea. However, I think you need a better system than just the number of commits, or lines of code committed. I don't these are very accurate measurements for the degree which someone contributed to the project. You need a more sophisticated system to determine that.
- lefnire 14y agoThat's true, there could be a lot of time in research and offline action going behind a single commit. Any suggestions re: other Github-provided data?
- skram 14y agoPerhaps look into # of github issues closed.. but that means you need to somehow account for "major" bugs versus "feature requests", etc.. and make the tagging is as consistent as possible.
- japhyr 14y agoThis strikes me as quite similar to the difficulty in paying teachers by their performance. You can use a lot of objective data for your decision-making, but there probably needs to be some subjectivity in there as well.
- _delirium 14y agoIn other parts of software as well: the use of lines of code, and various modified LOC metrics, to quantify programmer productivity is fairly notorious.
- Spoom 14y agoI don't think quantifying contributor value can be automated. You'd do better with some sort survey among the contributors where they are asked what percentage of the revenue they would give to each other contributor.