Effective pretty much immediately, the code that CongoCC generates will be targeted at JDK 17 or higher.
The current situation (about to change) is that the CongoCC tool itself requires JDK 17 or higher, but the code it generates could run on JDK 8 or higher.
Background
When I picked up my FreeCC work again at the very end of 2019, I decided to base it around JDK 8. I think that was the natural decision. The latest LTS release at that point was JDK 11, put out in autumn of 2018, not much more than a year earlier, so it was arguably still a bit bleeding edge. JDK 8 was the LTS release prior to that, released in early 2014. I myself was coming back to Java programming after a decade of inactivity and was catching up with all the new things in Java. So, JDK 8 already contained a lot of new stuff (for me!) like Lambdas and so on.
(Note that, within a month or two at most of my picking up this work again the Java parser inside of JavaCC21 could parse the most up-to-date Java and one could use it in code actions/injections. And that is still the case. For example, you can use JDK 25 level code in your own grammars if you want, say, but then your generated parser will require JDK 25.)
Well, I announce this, and people can object, and if they have really good objections, I might even heed them. But actually, I am kind of suspecting that if I didn't even anounce this and the tool just quietly started generating JDK 17 code, most people wouldn't even notice!