Re: [DNSOP] New Version Notification - draft-ietf-dnsop-dns-catalog-zones-09.txt

[Willem Toorop <[e-mail]>]

Op 09-02-2023 om 14:46 schreef Paul Wouters:
> On Feb 9, 2023, at 06:33, Willem Toorop <[e-mail]> wrote:
>> Op 07-02-2023 om 16:45 schreef Paul Wouters:> I find the valid use of the name "invalid" to be pretty horrible. An
>>> engineer looking at a catalog might quickly believe
>>> the invalid is a bug where it should have shown a real domain. Why not _catalog.arpa or something ?
>>
>> We, the co-authors, actually prefer producers to use a domain they own (because no chance on collisions with consumers from multiple producers). I've done a commit to express that more clearly. The new text is:
>>
>>    ``It is RECOMMENDED to use a domain name owned by the catalog producer if possible, or if that is not possible use a name under a suitable name such as "invalid." [RFC6761].''
> 
> A name under a suitable name such as invalid would then be “example.com.invalid” and not as it has now just “invalid” ?

Do you mean in the Catalog Zone Example appendix ( 
https://www.ietf.org/archive/id/draft-ietf-dnsop-dns-catalog-zones-09.html#appendix-A 
) ?

The name used there is "catalog.invalid.", not just "invalid.".
You prefer "example.com.invalid" over "catalog.invalid"?

> Or it could use “_catalog.example.com”  ?

Yes, if we add a sentence that the fictional organization producing this 
catalog is "example.com", then we could use that too yes.