On 9/14/26 19:37, Bill Woodcock wrote:
On 9/14/26 12:49, Willem Toorop wrote:
Hi Bill and Patrik,

I made a start with implementing the types in NSD but failed, because of the use of (unquoted and unescaped) parentheses in the Value RDATA fields for both types. The zone parser for NSD uses (unquoted and unescaped ) parentheses to group data, potentially across lines (see Section 5.1 of RFC1035: https://www.rfc-editor.org/info/rfc1035/ #section-5.1 ). At least for NSD it would be quite cumbersome to implement the exception for this RDATA field type. Unbound and ldns also treat parentheses this way and will have the same issue.

Could you consider another way to indicate the precision? Perhaps square brackets instead of parentheses?

Hi, Willem.  I am perfectly fine with that in principle...  Anyone have any objection or other suggestion?

Actually, square-brackets are already used to indicate optional things, so perhaps just using "+-" would be better, since it reads correctly to a human?

precision = "+-" ( "0" / posnum / "?" )

@ IN ISO 4217 2400000+-50000 EUR "insured value, hull and machinery"
@ IN ISO 4217 780000+-? CHF "salvage award, assessment ongoing"


        -Bill

Attachment: OpenPGP_signature.asc
Description: OpenPGP digital signature

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

Reply via email to