[rfc-i] UTF-8 and Unicode examples

Henning Schulzrinne hgs at cs.columbia.edu
Tue May 4 15:19:27 PDT 2004


Currently, there is no documentation. We don't need something that has 
the force of law ("if you don't do this, your RFC won't get published"), 
but rather provide guidance to people. If somebody has a good reason to 
choose something else, great.

Kurt D. Zeilenga wrote:

> At 07:19 AM 5/4/2004, Henning Schulzrinne wrote:
> 
>>Thanks for the pointer. I would suggest that the following convention be adopted:
> 
> 
> Before we attempt to discuss the merits of any particular convention,
> I think we should discuss whether it appropriate or not to adopt
> a particular convention.
> 
> I doubt that one could devise a convention that would work
> for all RFCs.  As I demonstrated in the examples I posted, often
> documents need to use multiple conventions.
> 
> We never defined conventions for ASCII examples (whose
> representations are ambiguous or otherwise unclear).  We got
> along there without adopting a convention for all RFCs to
> follow.
> 
> Sure, Unicode contains a lot more characters which are
> problematic when used in examples... but I don't see why
> the approaches we've used with problematic ASCII characters
> wouldn't work with Unicode characters.
> 
> But, more to the point, I think our operational experience
> with ASCII shows that one convention is insufficient and the
> current practice of allowing documents to define appropriate
> conventions for their needs works just fine.
> 
> Just as we do with ASCII escaping, we should encourage reuse
> of established conventions (such as Unicode conventions).  But
> I don't think we need to be BCP'ing anything here.
> 
> Kurt



More information about the rfc-interest mailing list