The encoding implementation is not looking so great. The generated docs have UTF-8 hardcoded in the xml prologue, and yet in some cases non-ascii characters are emitted as entities. Worse, the code that has the UTF-8 value hard coded in it trusts its caller to provide a Writer instance which is actually set up with that encoding -- is that actually happening?
In my opinion this is indeed a serious bug, but I would not advocate accepting Jens' patch, as it just papers over the real problem. Yes, it will work, but it's neither cheap nor elegant (as I'm sure Jens would concede). my 2 cents, Eric -----Original Message----- From: Jens Schumann [mailto:[EMAIL PROTECTED] Sent: Wednesday, June 11, 2003 11:44 AM To: [EMAIL PROTECTED] Subject: Re: Status of Axis 1.1 Final On 6/11/03 08:29 PM Glen Daniels <[EMAIL PROTECTED]> wrote: > > Did you find a showstopper? I missed it if so...! > Maybe Europe isn't a target market for Axis anymore ;). I would call #15133 a major showstopper. The encoding for Strings is pretty much broken for special characters. Jens
