Yeah I know stars are a fairly worthless vanity metric, but it's the best quick proxy that I know of for community size on this sort of project. I usually check stars and skim open issues. Sometimes a code search to see how many projects and what sort of projects consume the library.
back
2 comments
> skim open issues
What do you go looking for with this? And what results make you consider/not consider the project based on what you find while skimming open issues?
Sort the issues by # of reactions, and then read the top few.
Just like reading Amazon reviews.
See what people are complaining about and then answer whether you care about those things
I usually look for duplicate issues that aren’t being addressed, or if there’s just a lot of seemingly non-trivial ones that go unanswered
For me the presence of a stale bot that automatically closes issues after 30 days - like Bulma - is a major red flag.
Sweeping issues under the carpet actively makes your library worse, only because one really values the vanity metric of "closed/open issue ratio".
Github star count is another useless metric, often inflated by "star this on github" widgets on the homepage.
I wish the auto-close bot could be considered a red flag for the exact reasons you stated, but unfortunately it’s in effect on tons of actually good projects, so it’s not useful as a signal.
Checking issues can be a little less useful than it used to be. I know that the FastAPI project plays games where they close issues and refile them as discussions so as to keep the apparent count low.
that's just keeping the issue tracker on topic for "issues" rather than random questions