XML 1.1 was published in 2004 (second edition
2006). It is rarely used; the fifth edition of XML 1.0 (2008) took over its broader rules
for names, leaving XML 1.1 with the changes below. A document says it is XML 1.1 with
<?xml version="1.1"?>; XML 1.1 documents must have an XML declaration,
and a parser that does not support 1.1 must reject it.
What is different
| Area | XML 1.0 | XML 1.1 |
|---|---|---|
| Characters [2] | #x9, #xA, #xD, #x20–#xD7FF, #xE000–#xFFFD, #x10000–#x10FFFF | #x1–#xD7FF, #xE000–#xFFFD, #x10000–#x10FFFF |
| Restricted characters [2a] | None | #x1–#x8, #xB–#xC, #xE–#x1F, #x7F–#x84, #x86–#x9F may appear only as character references |
| Character references | Only to legal 1.0 characters | Any of the above, including #x1–#x1F; never #x0 |
| Line ends (2.11) | #xD #xA and #xD are normalised to #xA | Also #x85 (NEL) and #x2028, and #xD #x85, are normalised to #xA |
| Names [4] [4a] | The same character classes, since the fifth edition (older editions used Appendix B tables) | The classes on the syntax page |
| Normalisation (2.13) | Not addressed | Text should be fully normalised; parsers may check it |
| Namespaces | Namespaces 1.0 | Namespaces 1.1: IRIs, and prefixes can be undeclared |
In short, XML 1.1 lets a document carry control characters as references, recognises mainframe line ends, and asks for normalised text. Everything else, including every well-formedness constraint, is unchanged.