Path: utzoo!utgpu!watserv1!watmath!att!linac!uwm.edu!spool2.mu.edu!uunet!sunquest!alpha.sunquest.com!gavron From: gavron@alpha.sunquest.com (Ehud Gavron) Newsgroups: comp.protocols.tcp-ip.domains Subject: Re: definition of sub-sub-domains Message-ID: <26JAN91163912@alpha.sunquest.com> Date: 26 Jan 91 23:39:12 GMT References: <1131@nikhefh.nikhef.nl> Sender: news@sunquest.UUCP Reply-To: gavron@alpha.sunquest.com Organization: Sunquest VMS Internals, Tucson AZ Lines: 48 News-Software: VAX/VMS VNEWS 1.3-4SQ1 In article <1131@nikhefh.nikhef.nl>, e07@nikhefh.nikhef.nl (Eric Wassenaar) writes... # #Is it legal for a nameserver, authoritative for domain 'some.domain', #to define in its zone data file a sub-sub-domain (two levels deeper) #like # foo.bar.some.domain NS somehost.some.domain Yes, but add the trailing periods: foo.bar.some.domain. ns somehost.some.domain. # #especially when 'bar.some.domain' is not defined, i.e. there exists #neither a SOA record, nor a NS record (nor any other record) for #'bar.some.domain' ? Yes. Look at the nss records. For example boulder.co.nss.nsf.net. There are no co.nss.nsf.net domains/hosts, yet the subsub of nss.nsf.net exists. #I cannot find any information about this situation in the appropriate #RFCs. RFC1035 does not prohibit this. By its definition of and