More like 15. By 2016, Rails was supposedly dead and we were all going to be running the same code on the front end and back end in a full stack, MongoDB euphoria.
Nah, the good LLMs can generally web search and read documentation well enough that the fact that pre-training isn’t up to the minute is not a serious concern. Badly-documented projects are more of a concern, but they weren’t likely to get much pre-AI usage either.
10 years ago it was Heroku and Ruby on Rails*
but now Ruby on Rails is not a circus like how Next.js is.
see [0]: Rails security Audit Report
[0]: https://ostif.org/ruby-on-rails-audit-complete/
More like 15. By 2016, Rails was supposedly dead and we were all going to be running the same code on the front end and back end in a full stack, MongoDB euphoria.
New one coming in 5 years. Cycle repeats itself.
I don't think so, AIs are going to freeze the tooling to what we have today since that's what's in the training corpus, and it's self reinforcing.
Nah, the good LLMs can generally web search and read documentation well enough that the fact that pre-training isn’t up to the minute is not a serious concern. Badly-documented projects are more of a concern, but they weren’t likely to get much pre-AI usage either.