Re: CVS commit: src/lib/libc/net

[Steffen Nurpmeso <sdaoden%yandex.com@localhost>]

Robert Elz <kre%munnari.OZ.AU@localhost> wrote:
 |Which standard?

RFC 1035 2.3.1 defines hostnames.

 |ps: do go read section 11 of rfc2181 while you're pondering all of this.

I think RFC 6895 would then be for the better:

3.3.1.  Label Types

   At the present time, there are two categories of label types: data
   labels and compression labels.  Compression labels are pointers to
   data labels elsewhere within an RR or DNS message and are intended to
   shorten the wire encoding of NAMEs.

   The two existing data label types are sometimes referred to as Text
   and Binary.  Text labels can, in fact, include any octet value
   including zero-value octets, but many current uses involve only
   printing ASCII characters [US-ASCII].  For retrieval, Text labels are
   defined to treat ASCII uppercase and lowercase letter codes as
   matching [RFC4343].  Binary labels are bit sequences [RFC2673].  The
   Binary Label type is Historic [RFC6891].

Thank you.

--steffen

 |ps: do go read section 11 of rfc2181 while you're pondering all of this.

For my own taste all those other RFCs from all those honorable
people are not outweighed by a single sentence of an RFC that
otherwise pushes forward that completely refusable DNSSEC system
that i personally did not implement consciously back in 2004+.
And even a thousand more IETF announcements which read something
like "DNSSEC now enters widespread usage [and here is another one
which aids in supporting another part of the game]" won't change
that.
Instead i'm personally thrilled by draft-ietf-dnsop-5966bis, and
then on top of TLS and a library which gets used a billion times
a time.  But anyway the second half of that sentence.