Bytes do not have “storage” representation in JVM,...

# announcements

[orangy]

Bytes do not have “storage” representation in JVM, and they are signed, so they are loaded into “int”-size locals (or fields) with I2B instruction, which is sign-aware. Thus, if you have 0xFA.toByte() you actually get byte with -6 value. Since it is stored in int-slot, it is stored as 0xFFFA, and thus when converted to char with I2C bytecode instruction happens to be \uFFFA and not \u00FA as you would expect.