[Ietf-calsify] Status of simplification work

[Helge Hess]

On 26. Mai 2006, at 18:11 Uhr, George Sexton wrote:
>> For that matter, if Java uses UTF-16, there is no such thing as an  
>> "array of characters" there either.  Characters outside the BMP in  
>> UTF-16 are represented by surrogates that occupy two array positions.
> This is totally false. Consider a typical Java declaration:
>
> char[] buf=new char[1024];
>
> declares an array of Unicode 16 characters.

Are you aware of the difference between UTF-16 and UCS-2? AFAIK Java  
(like Cocoa) uses the former which makes Mark's statement totally  
correct. A Java character (char basetype) is not the same like a  
Unicode character.
But thats Java/Unicode 101 and doesn't really belong on this list?

> Given what the spec says about wrapping lines on arbitrary byte  
> boundaries, what is the solution?
>
> I haven't heard you argue for removing the wrapping requirement.

I'm kinda confused. Its a SHOULD, not a requirement?

Anyway, if the item is sent over a transport which doesn't support  
longer-than-75-lines (like email) you can still apply content- 
transfer-encodings?

Summary: I think the 75-char requirement should be dropped, its of no  
use? iCalendar is the content type and transport requirements are  
already covered by the transports (eg MIME).

Helge
-- 
http://docs.opengroupware.org/Members/helge/
OpenGroupware.org