On 13 Nov 2026, at 16:41, JINMEI Tatuya / 神明達哉 <[email protected]> wrote:

> I've read draft-jabley-dnsop-ordered-answer-section-02.

Thank you!

> 1. Should the requirement also apply to chains involving DNAME or
>    SVCB/HTTPS (for the latter, potentially including RRsets in the
>    Additional section)?

That is a reasonable question. 

We have not observed a problem with DNAME ordering, but then again DNAME is not 
so common and we haven't looked very hard. Perhaps we should. 

This seems like a good question for the working group, if this document is 
ultimately adopted. 

>    2. Section 5 states:
> 
>       Client implementations SHOULD accept responses where CNAME RRSets
>       appear in any order
> 
>   I'm not sure that "SHOULD" is the most appropriate requirement
>   here.

I understand your point. I am not sure whether MAY would be a reasonable 
alternative, though, since in this context it seems very similar to not 
providing guidance to clients at all (it would be very similar to "do what you 
want" which seems not very helpful.)

Declaring a long tail of crusty implementations that are unlikely ever to be 
upgraded to be non-conformant does not seem terrible to me. 


Joe
_______________________________________________
DNSOP mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to