One of the problems with PGN is that it requires ASCII, while names of players often contain non-ASCII characters. (And of course, many names are not written in the Latin alphabet ...)
The background to this question is the rendering of non-ASCII characters in Arena 3.5.1 (on Windows 7), as in the image below. The names for white and black are
Mittelsträß, Jürg and
王, 玥, respectively. Arena reads the UTF-8-encoded file as if it were encoded as Windows Codepage 1252. Editors such as Notepad++, SublimeText and JEdit identify the file as UTF-8 (which is the default encoding of all my plain text files, anyway).
Here's an example PGN for testing:
[Event "Office Armageddon"] [Site "Böblingen GER"] [Date "??.??.??"] [Round "?"] [White "Mittelsträß, Jürg"] [Black "王, 玥"] [Result "*"] [SetUp "1"] [Mode "OTB"] [Termination "unterminated"] 1. d2-d4 d7-d5 2. e2-e3 c7-c5 3. Bf1-b5+ Bc8-d7 4. a2-a4 Ng8-f6 5. d4xc5 e7-e5 6. b2-b4 Nb8-c6 7. Ng1-f3 Qd8-c7 8. Nb1-c3 a7-a6 9. *
(SCID assumes a system encoding if a PGN is encoded in UTF-8 without a byte order mark (BOM), but the BOM makes no difference in Areana 3.5.1.)
So I'd like to know if there is a Unicode-compatible alternative (ideally UTF-8) that is supported by some chess programs. (If not, PGN needs an overhaul.)