[rfc-i] Comments on draft-iab-rfcformatreq

"Martin J. Dürst" duerst at it.aoyama.ac.jp
Tue Dec 18 00:49:05 PST 2012


On 2012/12/18 16:54, Brian E Carpenter wrote:
> On 18/12/2012 01:36, Dave Thaler wrote:

>> if the display character is normative (as opposed to U-xxxx being normative,
>> which is not confusable).    So the wording you (Heather) mention is still
>> pretty problematic compared to something precise like
>>
>> "Codepoints greater than U+007F can only appear in non-normative text"

I like that much better than "UTF-8 characters", because the issue isn't 
about the UTF-8 encoding, but about the character repertoire.


> Even that bothers me a bit, because what will be the effect (say) of
> some bi-di characters (e.g. Arabic) placed elsewhere in the document?

Well, this problem will only happen if you have normative and 
non-normative text in the same paragraph, which I think is a bad idea to 
start with.

> I suspect this cannot be boiled down to a few words, and needs to be
> described at greater length as draft-hoffman-utf8-rfcs attempted to do.

Yes. In particular, the above wording would allow U+007F (delete 
character), which is clearly a bad idea.

Regards,   Martin.


More information about the rfc-interest mailing list