Re: Endianness in UTF-16 encoding

[Tom Tromey <tromey@redhat.com>]

>>>>> "Marcus" == Marcus Better <marcus@better.se> writes:

Marcus> According to Sun's API docs (Charset class), the UTF-16
Marcus> encoding is supposed to default to big-endian.

Interesting.

Marcus> What confuses me is that I checked the Classpath source for
Marcus> OutputStreamWriter, and it does the right thing, i.e., big-endian:

Classpath and libgcj differ a bit here.
Most likely this is happening because libgcj is using the glibc iconv
to do this conversion, and glibc is probably picking the native
endianness.
 
Could you file this in the gcc bugzilla?

Tom