Currently only scripts defined in Unicode 2.0 are allowed in XML element and attribute names
All scripts defined in Unicode are allowed in element and attribute content
Unicode 3.0 adds:
Ethiopic (Amharic, Geez, etc.)
Burmese
Cambodian
Mongolian
Dvihehi
Yi syllabary
Also:
Cherokee
Canadian aboriginal languages
Perhaps:
Japanese
Cantonese
Unicode 3.1 adds:
Deseret
Old Italic
Gothic
Unicode 4.0 adds:
Linear B
Ugaritic
Shavian
Osmanya
Is this enough to justify breaking compatibility?