[hocuspocus]

Java could have easily caught up with C# in the early 2000s. In 2001 Pizza (Scala's ancestor) showcased generics, higher-order functions, ADTs and pattern matching. Interestingly enough only generics made it to Java in a reasonable timeframe, the rest came only much later.

Aside from early versions being rushed, I feel that Java's success and adoption were the bigger issue. While Microsoft could iterate quickly and break backwards compatibility with major versions of C# and the .NET runtime, Java was deliberately moving at a much slower pace.

[int_19h]

In fact, Java 1.5 got generics in 2004, one year before C# 2.0 added them in 2005. OTOH C# 2.0 had anonymous delegates, although syntactically they were still very verbose (no type inference for arguments) and thus only marginally better than Java's anonymous classes.

It was really from 2007 on (.NET 3.5 / C# 3.0) that C# started to get major features at an ever increasing pace while Java significantly stagnated for quite a long time.

[kllrnohj]

C#'s runtime got generics in 2005 whereas Java still just has compiler-only sugar. It's a rather significant difference in the quality of the implementation. Not just for things like reflection but also means List<T> ToArray actually can return T[] instead of Object[].

It also means the .NET ecosystem didn't need the inline+reified kludge that Kotlin came up with https://kotlinlang.org/docs/inline-functions.html#reified-ty...

[hocuspocus]

Thanks, my recollection of C# evolution was blurry, I stand corrected.

So really, Sun and Oracle could have definitely moved faster around Java 6 and 7, the Java 8 release took a long time given the feature set.

I feel that records could have come quicker, their implementation isn't exactly ground breaking. Avoiding the async/await route was a smart call though, and Loom could probably not have happened much earlier.

Valhalla is another can of worms entirely