Subject: Re: [DNSOP] Is there a concise and comprehensive definition of a "zone file"? Date: Fri, Feb 20, 2015 at 01:24:20PM -0800 Quoting Paul Hoffman ([email protected]): > On Feb 20, 2015, at 1:08 PM, Francis Dupont <[email protected]> wrote: > > > > In your previous mail you wrote: > > > >> Okay, well, maybe something new. Like - should the master files be in > >> ASCII or accommodate, say UTF-8? Just a teaser of what would be part of a > >> definition. (I have seen some zone files put IDNA U-Labels in the owner > >> field.) > > > > => at least one of the IDNA DNS RFCs is pretty clear about the zone files > > (aka master files) are in ASCII. > > The IDNA RFC in question simply relied on the assumption that RFC 1035, in > turn, assumed ASCII. Note that RFC 1035 never explicitly required ASCII for > the format: it simply didn't say how to interpret non-ASCII octets, but did > say how to escape characters.
Given, - that RFC20 is standard now, ;-) - that we've got IDNA, the not-yet-proposed scheme by Ed might propose that, - while the wire-format is 8-bit, - the protocol /as such/ does not say ASCII, ...we still very much would like STD80 charset in the canonical file. Non-ASCII characters then probably should be encoded using IDNA. (comments can of course be anything that does not interfere with reading of the file, like NULL. The rationale for this is that IMNSHO a very desirable property of the canonical file format is that it is loadable into the absolute maxiumum of DNS zonefile parsers. One can argue that this can be the over-arching rationale for such a specification. A procurement specifiation for a piece of DNS software could then state "The software MUST be capable of reading and parsing a zonefile compliant with draft-dnsop-lewis-zonefile-format-07". -- Måns Nilsson primary/secondary/besserwisser/machina MN-1334-RIPE +46 705 989668 does your DRESSING ROOM have enough ASPARAGUS?
signature.asc
Description: Digital signature
_______________________________________________ DNSOP mailing list [email protected] https://www.ietf.org/mailman/listinfo/dnsop
