Hi On Sun, Feb 19, 2006 at 02:13:45PM +0200, Oded Shimon wrote:
On Sun, Feb 19, 2006 at 10:47:03AM +0100, Michael Niedermayer wrote:
Hi
currently Names are UTF-8, IMHO we should limit them to [a-zA-Z0-9] at least or maybe just [a-z] furthermore its a little tempting to either use the unused bytes for common syllables or store things in 5-6bit per char or does that syllable table need to be extendible but not storeable too? in which case iam obviously not agreeing
You mean, names of fields in info packets?..
yes
i think not limit them, allow the user to have whatever extra weird fields he wants, there's no big disadvantage in allowing full utf-8...
id say require names and types to be ASCII and in english or at least suggest it if possible, otherwise we might end up with X-?????? stuff written in your favorite asian or whatever language values should be UTF-8 where appropriate and ASCII where not (Author,Title, ... should be UTF-8 while Language and Disposition just need ASCII) but i dont have a strong oppionion on these, making all UTF-8 is fine for me too [...] -- Michael