Re: PIDF - using xml schema

Adrian and Graham,

> On 22 April 2002 11:08, Graham Klyne wrote:
> > Speaking for myself, I supported the idea of mandating use of UTF-8 
> > because
> > I couldn't really see any value for allowing any other charset
> encoding in 
> > PIDF defined as part of a protocol suite, and minimizing the choices
> seems 
> > to be a better bet for promoting interoperability.  I am open to
> arguments 
> > why formats other than UTF-8 should be allowed in the presence
> protocol 
> > payload.
> 
> That was my thought also favouring interop, with a slight reservation
> that for some languages, UTF-8 might explode the data a little too much.

There are still used lots of character encodings other than UTF-8.  Even for
Japanese characters, we have several widely used encodings, such as 
ISO-2022-JP, Shift_JIS, and EUC-JP.   Those would be continuously used 
for several decades even if UTF-8 become popular according to wide
acceptance of XML.  

> > Maybe this document is not the right place to mandate use of UTF-8?  
> > Maybe,
> > the protocol/service document that says use to this XML format should
> also 
> > say that UTF-8 coding is required?  This would leave the PIDF
> specification 
> > free to be used as a generic XML format in other contexts with
> possibly 
> > different requirements?   (Just a thought.)
> 
> Thinking in terms of where we are with CPIM, etc. this sounds like a
> valid argument. If we adopt this view, we should point out in the text
> that protocol developers adopting this spec must make a conscious
> decision about which character encoding(s) to allow and perhaps make a
> recommendation behind UTF-8.

Fair argument.  I think this is a right way to go.  Thanks. 


-- Hiroyasu Sugano



  [reminder: [email protected] for non-technical discussions, please]