While some countries such as Germany have allowed the registration of 2LDs such as mercedes-benz.de, Australian practice has been to have a small number of 2LDs for various purposes (eg, .com.au for businesses, and .edu.au for education). Individuals and organisations can then register domains within those spaces.
If draft recommendations from auDA's 2015 names policy panel are accepted, that will change and the .au space will be opened to direct registrations.
I believe that would be a bad move.
Australia's original 2LD hierarchy recognised that the namespace was a public resource to be conserved, not strip mined. It also reflected the fact that different people and organisations could have an equal claim on a given name. So today, bonds.id.au has been registered by someone with the surname Bond (I'd argue that the association between 'bond' and 'bonds' isn't quite close enough, but we'll let that pass), bonds.com.au is used by the well-known clothing brand, and bonds.org.au appears to be unregistered.
What this really seems to be about - as have most of the significant changes to .au policy - is feathering the nests of registrars. With most of the 'good' names locked up, the only way to increase profits (other than raising prices, which is difficult in a competitive market when everyone has the same wholesale cost) is to increase the supply of names.
It is true that 2LDs would allow shorter and more memorable names, but it seems to me that the benefit accrues to registrants and registrars rather than the general public.
According to the draft recommendations, "the Panel thinks that the biggest benefit will be for individuals, who would be able to obtain an Australian domain name in a simple and straightforward way."
This seems to be because the proposal puts individuals and organisations on an equal footing in terms of a claim on any particular name, and does not (thank goodness) favour trademark holders, although it does seem to give an advantage to those who have registered a particular name in a different part of the namespace, which works contrary to the idea of expanding the overall namespace.
And while the draft isn't explicit on the matter, it seems that the plan is to keep prices for 2LDs similar to those charged for registrations within the current 2LDs "to make the new names widely accessible".
The draft recommendation is to keep the 'match' or 'close and substantial connection' criteria for allocating names to applicants, which is better than nothing, although auDA's idea of 'close and substantial' is a lot more liberal than mine.
So we could see someone called Archibald Neil Zelmann register anz.au and - as long as it wasn't his sole intention - subsequently transfer it to ANZ Bank for a significant sum. Or, since 'bond' and 'bonds' are considered sufficiently close, Nando Jones might do something similar with nandos.au and the chicken chain.
Does it seem right to you that someone should be able to make a windfall profit just because their name had some similarity with that of a company and they were lucky enough to secure a domain in a land rush? And anyone who has tried to buy tickets online for a very popular event knows that ordinary people are at a disadvantage to the pros - scalpers - in land rush situations. A lottery where everyone that has applied for a particular name gets a theoretically equal chance is slightly better, but it does nothing to address the pseudo-resale situation.
Do you think that's not going to happen? The domain name industry knows it will. Here's what Netfleet ("Australia's number one domain name trading platform") founder David Lye is saying: "If you don't act fast (once the change comes into effect), you could risk losing your domain name to someone else which could put you in the position of having to buy it back at an extremely high cost."
That phrasing says a lot about the industry: not "the domain name you wanted to register," but "your domain name." You don't 'buy' or 'own' a domain name, you licence it from auDA, which operates the space on behalf of the Australian government - which really means all of us.
So yes, I'm fundamentally opposed to allowing registrations in the .au space.
But if it does go ahead - and history suggests it will, because the interests of the domain industry usually seem to be put ahead of the public interest - here are some ways I think it could be made less bad.
• Registrants of names in .au are ineligible to register the same name in any of the existing .au 2LDs. That is, you can be smith.au or smith.com.au, but not both. This, along with a corresponding change to .com.au and .net.au and so on, is the only way to free up names so that as many people and organisations as possible can register their most appropriate name in one of the 2LDs.
• No priority for existing registrants. The registrants of smith.com.au, smith.net.au, smith.id.au and so on should all be on an equal footing with everyone else qualified to use the smith name.
• Restore the old list of generic names (eg, wine.com.au) as the two-year licences expire, and use them to provide a public 3LD with the same general rules as .com.au. Then allow businesses in those sectors to choose which one of (eg) smith.au, smith.com.au or smith.wine.com.au they want to register. This would be preferable to the introduction of a (probably premium-priced) wine.au which would not accommodate both businesses with geographical names and local oenophile groups (eg, Tahbilk is a winery and a town).
• Completely prohibit the resale or transfer of domain names. If an citizen or resident dies or no longer wishes to use a domain, it goes back into the pool. If the qualification for registration was based on an ABN or similar, then if the business is deregistered or ownership changes the name goes back into the pool. As soon as a different set of people have control of the business (beyond the everyday changes that occur as shares are bought and sold), they have no more right to the domain name than anyone else that meets the criteria.