>Thanks Christof,
>
>One final question to do with which conversions to use, even though the app is only used in North America (and from what I know, we only need the french and sometimes spanish characters), it seemed to me that I needed to always do both conversions e.g.
>
>
>
>STORE STRCONV(lcXMLString,1) TO lcXMLString
>STORE STRCONV(lcXMLString,9) TO lcXMLString
>
>
>
>That's because the help file says for STRCONV() for nConversionSetting #9
>
>"Converts double-byte characters in cExpression to UTF-8"
>
>So I took it to mean that I had to first convert to DBCS because it would not operate properly on single byte high ascii characters? Is that correct?
Yes, simply because we don't have a simple 8-bit to UTF-8 conversion, so this DBCS (which is, AFAIK, the full 16-bit Unicode set) is a necessary intermediate step. UTF-8 is simply a way to encode Unicode strings so that they pass as 8-bit strings, but thanks to the lead bytes are properly recognized and interpreted by the presentation layer. Two bytes header and occasionally two bytes per character surely beats six or seven bytes per character when inserting one of these in HTML or in a character field.