Hacker News Reader

A lot of work still needs to be done on the libraries and other tools before it's useful for end users. We've spent years migrating to "reactive frameworks" or being stuck in "legacy".

Virtual threads isn't just something to "enable". You do need to adapt existing codebases somewhat e.g. use of synchronization.

The upcoming future is likely a mix of reactive and virtual threads where appropriate. Virtual threads is still very good for short lived tasks.

The application that we are working on uses thousands of platform threads today (split over a handful of applicationservers). It’s a humongous banking system. I have been working on performance related improvements for years. I am now curious if these new virtual threads might be beneficial for us in some places. Need to read up on them a bit more

Indeed. It is the curse of the Stroustrup law.

The top level comment is someone gushing over Java so perhaps Java advocates bring this on themselves?