Rasmus, Now do 5 or even 10+ years and commits to Zend and APC. We are talking > about a core language feature here, so commits to the code most affected > is what you should be looking at and when I talk about maintenance I > talk about code we are fixing 10 years from now. Commits in the past > year doesn't really reflect that very well. >
True, but if someone hasn't been active with even a single commit in the past year, I don't think they should be counted as an active maintainer. I ran the numbers back to 2011. And they actually shift more towards Yes: Total Commits: No: 2011 Yes: 1877 All but top 2: No: 996 Yes: 1011 And for 2010 (past 3 years): No: 2455 Yes: 2395 All but top 2: No: 1440 Yes: 1028 This is pointless though. The point is pretty well proven that within reason the activity level of both groups is about even. And when measuring a feature against "number of maintainers", I honestly believe that only current active maintainers should count for that ranking. It's not about "discrediting" prior contributors. Not in the least. I'm not suggesting their votes should count less. I'm not suggesting that they should lose voting rights or anything like that. But to count active maintainers against a list which contains people who's last commit was in 2006 isn't fair. Anthony