Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Tangential, but I think github's secret weapon of inertia is. . .(drumroll) github stars.

They're still seen by a lot of people as a sign of project maturity and use. My unfounded suspicion is if they all dissapeared tomorrow, people would be a lot more likely to try alternative code forges.

I've been using codeberg of late, more because of their politics than anything, but in all honesty the user experience between github/gitlab/codeberg/sourcehut/gitea is near identical.



I think it's a lot harder of an OSS project not hosted on GitHub to find contributors and gain traction in general. Network effect, as always.


It depends where your contributors are coming from. For example for Rust, the crates index is the discovery mechanism. Contributors will come to your repo by whatever link you put in your package's metadata. I've split my Rust packages between GitHub and GitLab and don't see a difference in participation.


It's the way I want it to be honest. Keeps the low effort garbage away for the moment.


It's one factor but I think they have more important "secret weapons":

1. Network effects; people already have an account.

2. Free CI, especially free Mac and Windows CI.


You can add two more things:

- 2000 minutes of free compute time with GitHub Actions

- free Docker Hub alternative with unlimited pulling (they say that you're limited to 500Mb but I currently have probably +20Gb of images on my Free account)

They have the community aspect AND the freebies


I never understood going by stars when there's a much stronger signal in how many issues are being tracked and closed. Very easy to see if its software people actually use


> a much stronger signal in how many issues are being tracked and closed

This is a strong signal, but what it signals is confused. How much of it is the nature of the user base in actually reporting issues? Suppose the project receives regular fixes and issues are promptly closed on average — how much of that is because the project has to constantly respond to external factors, and how much is due to developers doing constant fire-fighting on an intrinsically poor design and not getting around to new functionality? Suppose there are lots of outstanding issues — how many of them are effectively duplicates that nobody's bothered to close yet?


Yeah, this is the first thing I check, and also I verify a few closed ones to understand how they were handled.


There are websites to buy stars :D It's like fake reviews.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: