On 8 October 2014 11:39, Greg Keogh <g...@mira.net> wrote: > I've been looking for official REST documentation on this matter (that I > still can't find),
I think REST is more an idiom than a hard standard. The HTTP 1.1 spec is the closest thing you will find to a REST standard. RFC 2616 (HTTP 1.1) and RFC 2617 (HTTP authentication) are invaluable resources. > I didn't search the Unicode documentation, where it does > say > > "Where UTF-8 is used transparently in 8-bit environments, the use of a BOM > will interfere with any protocol or file format that expects specific ASCII > characters at the beginning" > > I'm guessing this applies to REST bodies, so I'm getting rid of the BOM. I'm > using a stream writer for the XML, so there'll be some options somewhere to > omit it. REST does not prescribe any particular text encoding. Indeed, a resource (the HTTP spec calls it an "entity") needn't be text at all. Whatever entity you send, just be sure to include a correct and detailed Content-Type header and always include the charset parameter if the media type is text. Note however that JSON's media type is "application/json", not text. I think it is generally agreed that JSON is always encoded in UTF-8. On 8 October 2014 11:04, Thomas Koster <tkos...@gmail.com> wrote: > The standard states quite explicitly that a BOM in UTF-8 is > unnecessary[1, 2] and discouraged[2]. -- Thomas Koster