shawjef3

@sschaef thanks for the work-around. I ran into this problem not with a generic rule, but a rule with two arguments. ``` scala import org.parboiled2._ class P(val input: ParserInput) extends...

I tried to use JDK_PRIMITIVE_FUNCTION, but ran into many places it didn't fit. For instance, in `Byte2BooleanMaps`, the collector needs `Object2ByteFunction` for the keys, and `Predicate` for the values. However,...

@vigna just a reminder to get eyes on this.

If you're going to support this, better to support passing an arbitrary Charset.

https://github.com/scala/bug/issues/11178 is closed.

An alternative thought I had was to make `mapZIO` emit a chunk at a time instead of emitting an element at a time.

ZIO could also create its threads with a required stack size. https://docs.oracle.com/en/java/javase/17/docs/api/java.base/java/lang/Thread.html#%3Cinit%3E(java.lang.ThreadGroup,java.lang.Runnable,java.lang.String,long)

Something like this might work (totally untested). https://github.com/shawjef3/zio/commit/890b6a585be06ce4c58d8da129bbf78c11eba683

I have a possible implementation for this. https://github.com/johnrengelman/shadow/pull/756

This is not a feature of SBT or Gradle. It's a feature of the shadowing plugins. In principle there's no need to change build tools to relocate Scala classes.