GitHub reliability keeps sliding, and people are noticing
GitHub had another incident today affecting pull requests, issues, Git operations, and API requests. The HN thread is short but pointed: one comment is 'Are they running paid marketing campaigns for Gitlab?' and another is simply 'I'm so done with GitHub.' This is not an isolated complaint. The pattern of GitHub incidents has become frequent enough that it is now a running joke in the community, and the jokes are starting to carry real frustration.
The practical problem for founders and small teams is that GitHub is load-bearing infrastructure. When it goes down, CI/CD pipelines break, deployments stall, and remote teams lose their collaboration surface entirely. The fact that this keeps happening without visible improvement from Microsoft, which has owned GitHub since 2018, is a real organizational signal about prioritization.
Gitlab keeps getting mentioned as the alternative, and this thread is another small push in that direction. The self-hosted option is increasingly appealing to teams who have been burned enough times. This is a slow burn, not a crisis, but every outage is a small conversion event.
So what?
If your deployment pipeline has a single point of failure at GitHub, today is a good day to audit that dependency. At minimum, make sure your CI can fail gracefully and your team has a documented fallback. If you have been on the fence about mirroring to Gitlab or running a self-hosted Forgejo instance, the calculus keeps tilting in that direction.