> Many trademarks are indeed more or less global.
And the vast majority are not, so this has no relevance.
> That's irrelevant - Apple Computers owns the Apple trademark (for certain industries).
And yet you yourself concede in this very same statement that it is not absolute. Even with one of the most famous examples from one of the largest companies in the world.
> It's not perfect and not guaranteed to be unique, yes. But people's names have many orders of magnitude more collisions.
For most people our data was very clear that for most names the number of collisions are in fact relatively low, further significantly mitigated by the combination of nicknames and ability to use middle names. Very few last names are highly frequent, and very few of those with common last names have an intersection of a very common first name and a very common last name. The vast majority of people fall in a long tail of last names held by a few thousand people, and first names held by a few thousand people.
For the webmail service that preceded .name, the vast majority of the two million accounts registered were for names for people who would have no way of getting their firstnam@lastname.<tld> addresses without it because of colissions. That some of those would not be able to get one of our addresses either because their specific combination was particularly common does not alter the objective fact that we massively broadened the availability.
> No, you'd at best get `??@firstname.lastname.name` as your email address. Which, again, is irrelevant, as it's impossible to say which of the many thousands of Firstname Lastname people this might belong to.
No, that was categorically not how it worked. I personally designed the system to handle this. We provided firstname@lastname.name e-mail forwarding to an address of the registrants choice, and provided a reference platform for registrars to support that, that I also designed, and managed the implementation of.
Given that you haven't even bothered to understand what it provided before criticising it, it's hard to assign much value to your arguments.
> I wasn't referring to a specific order, but to the cultural idea that people have 1 firstname and 1 lastname. In Spain, for example, people typically have 1 first name and 2 last names - but you couldn't get `a.b.c.name` as an address. Having a first name, a middle name, and a last name is also very common.
Spanish people are perfectly capable of using their firstname and their fathers first lastname when limited to one last name. They are also capable of using 2, 3, or 4 last names depending on context and preference. For any that wanted to use more, they had more flexibility, because they could similarly register b.c.name, and use a.b.c.name, or register any combination preferred for the last two labels and set up the rest themselves.
Same if you insisted on using your middle name. We only provided the increased ability to share among those with overlapping two last labels, but that did not change the ability to use regular DNS features.
> There are also cultures where people simply have one name, not first + last. For example, a person's full legal name in Tibet might just be "Woeser".
While you're right that exists, firstly it's a miniscule proportion of people. We collected stats on that two. If you fell in that group, your options were either to use another TLD, or use another indicator. In Europe as well, in cultures that used to use only one legal name, people would still regularly use distinguishing attributes, such as place names or occupations. In fact, both of those are the source of most legal lastnames in Europe today. E.g. my last name is the name of the farm my great great grandfather came from, and names like Baker and Smith are occupations.
There is no need to be able to be a perfect match to a given format for everyone to still meaningfully and significantly enlargen the available namespace.
This is a long way of saying you're being intentionally obtuse.
We overestimated the demand for paying for "vanity" addresses, but ability to register peoples names was not a problem - something we categorically proved with the webmail service that preceded it.