Re: [idn] proposed i18n naming rules

[Mark.Andrews@isc.org]

> 
> ben wrote:
> 
> > It was first brought up by Eric H in this thread who said, " U+002E
> > (FULL STOP) is ONLY valid for use as a separator when the IHI is
> > written out".  I thought it will be a good time to draw the same
> > attention to the "Chinese Full Stop" for this WG to consider.
> 
> DNS doesn't store the separator. The only place the separator appears is
> in domain names that are written-out and stored in data streams. And since
> the separators are not encapsulated in the labels (they are separators not
> data), they are not subject to encoding, and therefore would always appear
> between encoded labels (including ACE labels). So really you are asking
> about domain names as they appear in protocol and/or application data. You
> would have to update every protocol and application that processes domain
> names in order for Chinese Full Stop to be usable as a label separator
> everywhere that a domain name was encoded or encapsulated.
> 
> -- 
> Eric A. Hall                                        http://www.ehsco.com/
> Internet Core Protocols          http://www.oreilly.com/catalog/coreprot/
> 

	DNS doesn't store the seperator but it can and does store periods
	in labels.

	e.g.
		mark\.andrews.isc.org
		
	which is 4 labels (including the root) when encoded on the wire.

		"mark.andrews"
		"isc"
		"org"
		"."
	
	Just because it is not used in labels in *hostnames* is no reason
	not to consider it.

	If you are going to drop characters in IHNs you have to also drop
	the in IDNs.  If you are going to map characters in IHNs you have
	to map the in IDNs.  If you map several characters to periods
	that then become seperators you have to map the same characters
	to periods that appear in labels.

	Mark
--
Mark Andrews, Internet Software Consortium
1 Seymour St., Dundas Valley, NSW 2117, Australia
PHONE: +61 2 9871 4742                 INTERNET: Mark.Andrews@isc.org