Hi everybody

I'm writing a post for the first time in this forum, but didn't find the response with my friend Google.

I have an issue with a message generated by BizTalk in the Windows-1252 encoding.

Here is the context: 

I'm receiving a flat file (CSV) and have created the corresponding schema with CodePage Windows-1252.

1) First half-flow : A map from the resulting XML to my Pivot message.

I have well a UTF-8 encoded message.

2) Second half-flow : A map from my Pivot message to B format ( a XML message ) 

 The generated message has again a Windows-1252 encoding


I played with the send pipelines, and I noticed that :

- With a PassThruTransmit pipeline, my generated message is as I would have expected (UTF-8)

- With a XMLTransmit pipeline, if I don't modify the default valued, it is Windows-1252

- With a XMLTransmit pipeline and TargetCharset UTF-8 (65001) set, still Windows-1252 but if I create a custom pipeline with the same configuration of XMLAssembler, that's fine, I have UTF-8 (BizTalk 2009 bug?)


So my question are:

- Why the former encoding is set with XMLTransmit ?

- How this encoding is retrieved ? A notion of BOM, maybe ?

- Could I "lost" characters by configuring PassThruTransmit ?


Thanks in advance, I'm trying to well understand encodings.