There is no process defined to vet who represents which country for
the purposes of URN NID registration, and until there is, there will
be no two letter country code NID registrations.
To counter that argument, solve ICANN's problems in ccTLD zone
administration authority, and then come back :-)
Less obliquely, there is the very real probability that, for eg, a
national library could claim to be responsible for setting up
permanent identifiers for resources, while the national passport
office turned around and defined something entirely different. How
is the IETF/IANA to determine what authority within the country is
going to arbitrate between them?
Hi Andy, I've cc'ed Leslie Daigle on this message, as she was quite
heavily involved in both setting up the registration procedures and
in working with New Zealand on arranging their NID. I've also cc'ed
Lisa Dusseault, who is the Area Director who currently processes most
URI-related drafts). The first question, at least from my
perspective, is what time frame do you need this in? The procedures
for getting a namespace identifier according to the current rules are
relatively clear cut and with sufficient coordination could probably
be done in a couple of months (get a draft written along the New
Zealand model, have it go through the required review by the urn-nid
list, approved by the IESG). The last step is the longest/least sure
since the IESG approves all IETF technical standards and many related
documents as well, so their calendar can get full. If you need it
relatively quickly, though, using the existing procedures is probably
the way to go. You're right that the 2-letter IDs were reserved in
3406, with the following text:
NOTE: ALL two-letter combinations, and two-letter combinations
followed by "-" and any sequence of valid NID characters are reserved
for potential use as countrycode-based NIDs for eventual national
registrations of URN namespaces. The definition and scoping of rules
for allocation of responsibility for such namespaces is beyond the
scope of this document.
To allocate a two-letter combination, the rules for the allocation
would have to be codified in an update to BCP33/RFC3406 and then a
document written that conformed to the new rules. That's pretty much
automatically going to be slower than getting a longer NID approved
(since it involves two passes through the IESG). Theoretically,
there could be some parallel processing, but realistically it will be
slower. The big reason it will be slower is that it is not entirely
clear how to create a process that allows the IETF/IANA to be assured
that a NID request comes from *the part of an individual government*
that should be assigned the NID. Since governments vary considerably
in their structure, we cannot simply say: "A request from the
government's CIO or equivalent to assign will be honored". In some
cases, there will be no such office; in others, the request might
most appropriately come from a governmental office charged with
library services (e.g. the U.S. Library of Congress). There are also
codes like AQ (for Antarctica) which are assigned but for which
identifying a single governmental entity as responsible is
difficult, and codes which ISO has re-assigned after the passage of
time (like SK). We would also have to deal with situations like
yours in which there is an assigned code (GB) and an exceptionally
reserved one (UK). Writing that document and getting agreement on it
is unlikely to be quick, and none of this is, honestly, in the IETF's
core area of concern or competence. That also makes things slow. If
you have no particular time constraints and have the energy to put
into proposals here, it may be possible to resolve this. But I do
believe it will be relatively slow to complete. regards, Ted Hardie
At 10:59 AM -0700 5/20/08, Andy Greener wrote:
Hello,
I recently joined this list so forgive me if this has been
discussed before (I did check the archive at www.nabble.com but
couldn't see anything of direct relevance recently).
As my sig indicates I'm a technical consultant working for HM
Revenue & Customs in the UK, and I'm also a member of a
sub-committee of the UK CTO Council's Architecture Review Board
that is considering the future of the UK GovTalk policies and
standards web site (www.govtalk.gov.uk). As part of this work we
are considering establishing a URN scheme for namespace naming of
persistent artefacts (XML Schemas, code lists, etc) across the UK
Government space, and naturally the subject of a "UK" NID came up.
I note that RFC3406 states that all two-letter combinations are
reserved for potential use as countrycode-based NIDs for eventual
national registrations of URN namespaces, but it hints at another
set of definition & scoping rules for such namespaces. I also note
that there are no existing two-letter NIDs on the IANA list, but
that at least one country (New Zealand) has already tackled this
issue and worked around it by using their ISO three-letter code
instead (not an option for us as the three-letter code for the UK
is "GBR", which we feel is inappropriate under the circumstances,
as well as being somewhat "politically incorrect").
Are we on a hiding to nothing if we wish to pursue the "UK" NID?
The registration would be made on behalf of the UK Govt by the
Govt's CTO, and it is unclear to me who else would have ultimate
authority to formally request this particular NID if it's not the
UK government. I expect I'm opening a can of worms here, but if you
don't ask you don't get! I'd be grateful for any guidance or advice
anyone can give me.
The best qualifying alternative we can come up with is "UKGOV", but
this is likely to be subsumed into any future "UK" NID anyway (not
a desirable attribute for a supposedly persistent naming mechanism!
- we'd like to do the "right" thing once, and for all).
Andy
-- Andy Greener Enterprise Architect Architecture Solutions &
Assurance IMS Strategy & Architecture, HMRC
andy.greener at hmrc.gsi.gov.uk / andy at gid.co.uk +44 7836 331933
Attachment converted: Macintosh HD:smime 1479.p7s ( / )
(007B1E4D)