RE: EBCDIC problem.

["Noel J. Bergman" <noel@ibm.net>]

> I have fixed a problem in the test suite:
> // JFC BS2000 EBCDIC.... http ASCII, so ASCII.
> BufferedReader r = new BufferedReader(new
InputStreamReader(conn.getInputStream(),"ASCII"));
> The InputStreamReader assumes that the input is "platform's default
encoding", but this is only OK if the machine encoding is ASCII, not when
the coding is EBCDIC!.

According to Sun's current documentation regarding supported character
encodings,
http://web2.java.sun.com/products//jdk/1.1/docs/guide/intl/encoding.doc.html
, the "ASCII" encoding is part if the International i18n.jar package, and
thus may not be part of US-only versions of the runtime environment.

Running a test program on systems here, the default encoding is Cp1252
(Windows Latin-1) for Windows and Cp858 (extended MS-DOS Latin-1) for OS/2.
I'd have to look at the source to know what differences exist.  Cp1252 is
guaranteed to be present in all environments, as is ISO8859_1.

So as long as we're changing the code to incorporate an encoding, should we
use one of the basic encodings, or "ASCII"?

	--- Noel