Open Source September 25, 2026 bearish ⇧ 474 pts across 2 threads

Rails Identity Crisis: DHH's Vision Leaves Builders Confused

The Rails World 2026 opening keynote sparked genuine confusion in the HN thread, with commenters noting that Hey, DHH's own product, is no longer a Ruby on Rails app. The keynote seemed to gesture toward an AI-agent future without clearly explaining what Rails becomes in that world. One commenter asked flatly: 'So it's not Ruby on Rails anymore. What is it?' Another interpreted DHH's talk as an indirect admission that Ruby is slow, pointing to performance improvements that came from moving away from it.

A follow-up thread titled 'What About Rails?' drew a pointed critique of DHH's take on the current tech cycle, with the top comment arguing Rails needs to be forked. Several commenters dismissed both Rails and PHP as having very little to offer going forward, calling out Ruby's speed as a ceiling that compounds every year AI workloads grow.

The counterpoint is that Rails still has enormous productivity advantages for certain product categories, and the framework has survived death notices before. But the signal here is different from past cycles: DHH himself seems to be quietly acknowledging the ground has shifted, and the community doesn't have a clear answer for what comes next.


So what?

If you're running a Rails shop, the question isn't whether Rails works today, it's whether the ecosystem will maintain enough momentum to keep dependencies maintained and talent available over a three to five year horizon. The fork talk is a yellow flag worth watching. Founders evaluating new stack choices should treat 'proven but plateauing' as a real category and price in switching costs now rather than later.

Read these