Hello community,
I have the following situation:
- Partner has custom IDOCs to exchange
- Encoding settings for the Interface: Non-Unicode: "Detect client code page, based on logon at runtime"
- EDI Client with latest JCO 3.x
Problem:
- EDI Client sends 1100 and gets IDOC with the data partly encoded as 1100 and partly as unicode

Correct behavior:
- All segments are delivered to communication partner in 1100
Workaround:
Cast fixlength IDOC as unicode and then convert it to target LATIN1 Encoding. Has potential risks of unconverted chars, thus suboptimal.
Any idea what could be the reason?
Thanks,
Yefym.