--On Wednesday, 27 September, 2006 06:49 -0700 Dave Crocker
<dhc2@xxxxxxxxxxxx> wrote:
> Fred Baker wrote:
>>> Sorry, probably that's all obvious, but where is ["domain
>>> suffix"] defined ?
>>
>> At the Verisign site. It is the new-speak for use when all us
>> ancient geeky types would prefer "TLD".
>
>
> Not quite. A TLD is the right most (visible) field, like com,
> net, my or us, whereas a "suffix" is typically longer,
> constituting an "organizationl" base domain name string, like
> example.com, or mtview.ca.us.
Dave, unfortunately, if "suffix" is formally defined, I haven't
been able to find it. And it has been relatively common
practice among registrars, registries, and other participants in
"the domain name market" to use "suffix" as a synonym for TLD,
as Fred suggests. That terminology follows the practice in a
number of operating systems of referring to the second part of a
two-part file name as a "suffix".
It is, IMO, just not acceptable for a standards-track document
to leave the interpretation of terms like "suffix" and "item" up
to the imagination of the reader when their definitions are key
to the implementation of the specification.
> Hence:
>
> Stig Venaas wrote:
> > It's what a client might add to it's hostname to form an
> FQDN.
>
> is a particularly practical way of putting it.
Then that definition belongs in the specification. Moreover,
IMO, the authors of the specification should work with DNSext
and/or DNSops to determine whether there is other terminology
that might be more established and/or more clear and then use
it.
I've recently been reminded that the DHC charter requires
precisely that type of check and approval; I have reason to
believe that the charter requirement was not followed in this
case. To the extent to which a WG Charter is a contract between
the WG and the community, this particular WG has apparently
broken that contract.
john
_______________________________________________
Ietf@xxxxxxxx
https://www1.ietf.org/mailman/listinfo/ietf