On 12/18/2017 10:20 PM, Martin Hoffmann wrote:
That said, there seems to be a mistake in the text for the common
display format:

| so, in both English and C the first label in the ordered list is
| right-most

Previously this list was defined as being "ordered ordered by
decreasing distance from the root" which would make the first labal in
this list to be the left-most label in English.
Hmmm.
1. "Domain name" is defined as «an ordered list of one or more labels… identifying a portion along one edge of a directed acyclic graph» (presumably starting at the root). 2. In "Global DNS", "Composition of names" relies upon that same order: «in a fully-qualified domain name, the first label in the ordered list is 0 octets long… it is called the "root" or "root label"». 3. But then in "Format of names", the opposite order is specified: «The basic wire format for names in the global DNS is a list of labels /ordered by //*decreasing*//distance from the root/, with the root label last» and «The presentation format for names in the global DNS is a list of labels /ordered by *decreasing* distance from the root/» (emphasis mine) and «The common display format… is the same as the presentation format… in both English and C the first label in the ordered list is right-most». It makes sense if that last "ordered list" is the one from "Domain name" and "Composition of names", but not if it is the one from "presentation format", which I guess is the problem.

The initial graph-theoretic definition is accurate, but suggests (although notably does not require) an order backwards with respect to all three name formats. Perhaps everything would be more clear if those issues were eliminated right up front: «given a directed acyclic graph in which no node has in-degree greater than one, a domain name is a list of labels identifying a path in reverse (i.e., against the direction of the edges)» (the restriction on in-degree precluding child-to-parent directionality). For global DNS, the labels would then /always/ be ordered by decreasing distance from the root, hopefully eliminating confusion about how they appear in text or on the wire.
_______________________________________________
DNSOP mailing list
DNSOP@ietf.org
https://www.ietf.org/mailman/listinfo/dnsop

Reply via email to