@skyl practically provoked me to write this answer so please mind the redundancy.
NCName stands for “non-colonized name”. NCName can be defined as an XML Schema regular expression [\i-[:]][\c-[:]]*
…and what does that regex mean?
\i and \c are multi-character escapes defined in XML Schema definition.
http://www.w3.org/TR/xmlschema-2/#dt-ccesN
\i is the escape for the set of initial XML name characters and \c is the set of XML name characters. [\i-[:]] means a set that consist of the set \i excluding a set that consist of the colon character :. So in plain English it would mean “any initial character, but not :“. The whole regular expression reads as “One initial XML name character, but not a colon, followed by zero or more XML name characters, but not a colon.”
Practical restrictions of an NCName
The practical restrictions of NCName are that it cannot contain several symbol characters like :, @, $, %, &, /, +, ,, ;, whitespace characters or different parenthesis. Furthermore an NCName cannot begin with a number, dot or minus character although they can appear later in an NCName.
Where are NCNames needed
In namespace conformant XML documents all names must be either qualified names or NCNames. The following values must be NCNames (not qualified names):
- namespace prefixes
- values representing an ID
- values representing an IDREF
- values representing a NOTATION
- processing instruction targets
- entity names