On Sun, 30 Jan 2022 at 20:36, Steven D'Aprano <steve@pearwood.info> wrote:
On Sun, Jan 30, 2022 at 08:36:43AM -0800, Jelle Zijlstra wrote:
Agree, the count of 1.6k open PRs is not a good look for new contributors.
How does that compare to other comparable open source projects?
How it compares is a separate question from how it looks to someone coming from "outside". This is a common problem in community-supported open-source but it is still a problem whichever way you look at it: most significant open source projects have large numbers of unresolved attempts at improvements. The problem is that an issue/PR tracker becomes meaningless when it's mostly filled with no chance of success suggestions/changes.
Number of open PRs per KLOC seems like a more useful metric than just the number of open PRs.
Maybe but either way it's better if there are fewer unresolved PRs sitting there. As Irit says above it's common (in CPython and also many other projects) that something sits there as unresolved even though it's obvious to the experienced that it will never go anywhere. The difficulty is that closing a PR entails awkwardness that is well described by Irit as "emotional labour". It's immediately easier to leave something to languish than to close it off regardless of what is better in the long term. Taking an alternate approach though: if a community cannot cope with the number of already incoming attempts to make changes then maybe it doesn't really benefit from signalling to new contributors that making changes is easy... -- Oscar