On Nov 16, 2010, at 7:02 AM, Tony Finch wrote: > On Tue, 16 Nov 2010, David Conrad wrote: >> 1123 says: > in the informative text
True, but back when 1123 was written (and when code was developed based on it), I suspect the distinction wasn't made. >> Note that it says 'will be alphabetic' not 'will contain an alphabetic'. > The prediction has already been broken by IDNA. Yes, hence the draft being discussed. > I think the raw syntax should remain as simple > as possible and all restrictions such as alphabetic-only should be left to > the policy layer. I'd think that if this were done, it wouldn't be a minimal change and would open up a larger discussion. While (as I indicated) I think that this discussion will happen eventually, I'd personally like to fix the current specification brokenness caused by the introduction of IDNs in the root now and deal with the more general problem later. Regards, -drc _______________________________________________ DNSOP mailing list [email protected] https://www.ietf.org/mailman/listinfo/dnsop
