Paul, This is great! We have the need to send UTF-8 encoded strings to an application and feel that SOAP is a little overkill for what we are trying to do. Hopefully this can get worked into CVS.
One suggestion though: I would like to have the option of specifying the encoding. Thanks! Kurt -----Original Message----- From: Paul Libbrecht [mailto:[EMAIL PROTECTED]] Sent: Wednesday, March 06, 2002 7:41 AM To: [EMAIL PROTECTED] Subject: Patch suggestions Hi there, Since a while happy users of the Helma XML-RPC library we have made a few patches. Two things that seem important for everyone and one that was important only for us: -> we wanted to have the ability to exchange UTF-8 encoded messages so the XMLWriter has been enriched and the MinML parser as well. XMLWriter now writes its messages with encoding either ASCII, ISO-8859-1 or UTF-8 depending on the highest character found when sending. MinML has been patched to support reading this (the patch is not really perfect, I expect John Wilson to present something better one day). -> we had Security Exceptions in the applets when run in any other applet VM than the InternetExplorer pilotted ones (tests were on MacOSX and Linux). This was due to the usage of Sun classes in the applet package. The Base64 class has been inserted there as well as the calls to it. Note that, I'd suppose a clean thing would be to quote the usage of this library (as the XML-RPC library) when using an applet. -> and the minor ones is that we have at least one parser, made for some Lisp, did not like the lack of the string element around textual values, so we've added it. The patched version is distributed under the same license as the previous ones (Apache for the org.apache.xmlrpc modifications, and BSD for the MinML modifications) and I would like if XML-RPC developers would consider including it in the XML-RPC distribution. http://www.activemath.org/~ilo/revisions/xmlrpc/ Thangs. Paul PS: the "ASCII" value for the encoding when no character over 127 is found is not very precise... that may need some workings PPS: our tests seem to indicate that this is working...
