WHO WE ARE WHAT WE DO HOW WE WORK
This is the report of the Policy Advisory Board from its sub-committee meeting of 4 September 2001. The views expressed here are not necessarily the views of Nominet UK.
Introduction
We met in September 2001 to consider the general policy issues of geographic domains within the .uk namespace. Though this was triggered by the ScotNom proposal, it was not limited to any one concept. We were not there to invent SLDs, but to generate policy guidelines.
We came up with the following general principles that should apply to any regional naming system:
- understandable / recognisable to the man in the street (useful as a regional "brand");
- appropriate and consistent over time (unlike county names);
- naturally unique as far as possible (i.e. none or very few conflicts);
- should be largely non-overlapping (though some fuzziness would be acceptable);
- the set should effectively cover the entire UK, not just some parts;
- maximisation of imparted information;
- each SLD is large in terms of potential registrants (this wouldn't apply if the regions were at the third level).
"Four Nations"
It is regularly suggested that there should be one or more geographic SLDs: .scot.uk and .wales.uk (possibly with other spellings) are the most commonly mentioned. This, of course, immediately leads to suggestions that "the complete set" should be provided, meaning the addition of .eng.uk and .ni.uk. For convenience this was known as the "Four Nations" proposal.
At first glance, this proposal is attractive. However, while Scotland and Wales have obvious regional identities that would benefit from reinforcement through the DNS, England does not to the same extent - it is the regions of England that require that reinforcement. We felt that the Four Nations system only solves the problem for 10% of the UK and makes things worse for the rest by preventing any regional solution.
Side comments:
- We compared the situation with Germany, where there are 16 roughly equal Lander, each with its own strong identity. Were we the .de body rather than the .uk one, we probably *would* recommend the provision of 16 geographic SLDs.
- It is far from clear whether a Wales domain should have a name based on "Wales" or on Cymru.
- We were not at all sure whether Northern Ireland would gain or lose from a single identity ending in "UK".
There are too many regions in England, and they are not rigidly enough defined, to easily justify the provision of SLDs for all of them. We were unable to identify a useful set of N divisions of the UK where all the members are obvious and well-known *and* N is small enough that the number of SLDs would be reasonable. We felt that 10 is about the limit.
We do not feel that we can recommend the provision of just some domains (e.g. .scot.uk, .wales.uk, and .anglia.uk). Our remit was to solve the whole problem, not just part of it.
While there are many partial sets of regional names, none of them are strong candidates: the few complete sets are either not well-known or do not appear to be useful in terms of identity. [Examples: common tourist regions such as "Lake District" and "West Country" do not form complete sets. Sets such as Economic Planning Regions are not strong brand identities, and those such as "North", "West", "Southeast" are not well defined.]
If the PAB could find a suitable set of regions, this approach should be considered. Otherwise we recommend against having a set of geographic SLDs.
Other approaches
In the absence of a solution based on multiple SLDs, there are two other options open.
- Do nothing. While there are arguments for geographical domain names, nobody found them so compelling as to rule out this option.
- Have a single geographic SLD with a set of rules that allow for a sufficiently large number of geographical identities.
Our remit as a subcommittee did not extend to generating a proposal for such an SLD, but we felt it beneficial to explore this approach further to determine whether it was feasible or whether there were any "show-stoppers". Any formal proposal should take account of the issues we discuss below.
Single geographic SLD
A number of names were suggested (including .loc.uk and .geog.uk), but the most popular was .in.uk. This name is used as a placeholder in this report.
The .in.uk domain would be used for geographical-based names. It would not be limited to the Four Nations or the regions, but could expand to cover counties, districts, towns, and even villages and casually defined areas ("the Fens", "the Brecons").
We felt that there were two obvious restrictions on names:
- Given that they are for branding purposes, it should not be possible for one registrant to have a "lock" on a location name.
- Registrations need to have a clear location element to prevent the domain just becoming a free-for-all.
Based on this, we did not think that third level registration should be allowed. Rather, registration should be at the fourth level, with names of the form:
[anything].[location-identifier].in.uk
Any formal proposal should explain how the permitted values for the "location-identifier" will be determined and enforced.
Summary
Specific or multiple geographic SLDs are not recommended.
A single geographic SLD with registration at the fourth level (".in.uk") could be possible.
The relative merits of doing nothing and creating .in.uk have yet to be determined. This would be a necessary part of any proposal for the latter.
