Patch: RFC: InputStreamReader and OutputStreamWriter

[Tom Tromey]

>>>>> "David" == David Daney <ddaney@avtrex.com> writes:

David> You mean something like this?:
David> http://gcc.gnu.org/ml/java/2004-09/msg00174.html

David> As posted above my patch doesn't quite get us to having the
David> Charset constructors for InputStreamReader and
David> OutputStreamWriter, but it shows the way.

Wow, I'm sorry this slipped through the cracks.

This looks like a nice approach to me.  I'm going to follow up on this
next week.

David> The little test program in the refered to message shows us that
David> Sun uses (last time I checked) a hybrid approach similar to
David> this.  I think that there are performance wins to using the
David> java.nio.Charset infrastructure as little as possible.  We
David> should keep that in mind.

At FOSDEM we talked about charset converters for a while.  This is one
of the last frontiers for libgcj/classpath merging.  One idea that
came out of there was, instead of trying to pick the winning solution
up front, we would just set acceptance criteria (mostly performance as
I recall... my notes aren't readily accessible atm) and let people
experiment.

Tom