On Apr 4, 2014, at 9:09, Ted Lemon <ted.le...@nominum.com> wrote: > On Apr 4, 2014, at 8:53 AM, Antoin Verschuren <antoin.verschu...@sidn.nl> > wrote: >> I don't considder these other names with dots in them inferior, but >> they are simply not domain names. > > Whether you are right or not, I think Stephane's interpretation is > technically correct. I don't mean that it _is_, I just mean that I think it > is. So we aren't going to get to the truth of the matter by casting slogans > at each other, because it's just a matter of opinion. It's not wrong to > describe the names the way Stephane has, and it's not wrong to describe them > the way you have. So we should choose based on what meaning more clearly > expresses the policy we ultimately decide on, rather than trying to drive the > policy choice on the basis of what name we prefer.
My observation is that there is a world out there and name are applied to things (in the sense, the answer to the questions a toddler asks “what is that?”). These names transcend oral and written forms, are not always universal, and are some times applied with a heavy cultural bias. In short, the names can be in conflict with each other, there’s no sense of order to naming. And that there’s this DNS which enables a representation of the world. The DNS contains a mapping of the world into a hierarchy of domain names (and I’ll keep the adjective here to “scope” what is meant by “name"). Mapping means a transference of one reality onto another, it doesn’t guarantee a perfect copy, but a representation for a particular purpose. The point is, domain names are a proper subset of all names. It’s tempting to also comment on the quality of the mapping. The DNS itself “is what it is” - a system which imposes various and sometime seemingly arbitrary impingements on a smooth mapping. With it today we can only do so much. That shouldn’t stop us, but, it makes the work harder up to the point that it becomes necessary to overhaul the system. DNSOP WG has been convened around the DNS. The expertise present is expected to be knowledgeable about the protocol as it stands (and where it can “go”) and the set of requirements that are placed against it[0]. To this end, the work is to manage how domain names are operated, what is delegated and what kind of data is associated with each node. So, I think it is a non-starter to get into “inferior” names and whatever that means. We should stick to working with domain names and how they are operated. For example, in some cases, domain names could be barred from the protocol. These might be names that fail some test of what names should be mapped into domain names. You might think of content censorship here or you might think of visibly confusing combinations of different writing scripts (those are ratholes the WG is arguable not prepared to discuss). In other cases, some domain names can be restricted or scoped because in contexts, two names might mean different things and result in confusion and delay[1] or worse yet, a security concern. Oh, and as PS - in direct response to Antoin’s message, “Will.i.am” is a singer, not a domain name. ;) That is meant as a joke. However, I do not think a domain name is label.label.label. (Yes, despite STD 13 documenting the presentation format.) That is just a textual representation of the “byte count” “value” “byte count” “value” … “byte count=0” you see "on the wire.” I regret that people think of the “dot” as the official, burned-in-to-the protocol separator - it’s just what is thrown into representations in the Latin character set world (because back then, that’s all that was considered in the documents). There’s no reason why a special character native to the Han glyph set could be used in Chinese-Japanese-Korean representations of domain names, and when it comes to right-to-left scripts like Arabic, the “problem” of directionality would be solved. Sigh, the DNS “is what it is." [0] I say this having spent time trying to determine the official name of some governments in various, local scripts. Although this is “naming” it is not anything like understanding the DNS. [1] After a few Thomas the Tank Engine books, you can’t get that phrase out of your head. _______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop
