If you need it pointed out to you that John Berryhill's submission was perhaps partly in jest, then I guess you didn't get the joke. OTOH, it can be read seriously, as the IDN claims that end users might be confused by dots in front of their eyes, where is the analysis to show that end-users can tell the difference between, for example, .pn and .php? While I, and presumably you, and John, can hopefully tell the difference most of the time, if such esoterica was common knowledge, URL semantic attacks would go nowhere.|
I imagine John, and ICANNWatch (with whom I am not connected except that I sometimes provide submissions they are kind enough to carry), can take care of themselves. I don't presume to speak for them, but I think John's article was bang on, lacking only a mention of the dot as also used in subdomains, EG: forum.icann.org, another potential area of confusion.
Here's the definition for keywords given by the IDN Committee:
DEFINITIONAs you seem to know something about this, perhaps you could explain the above so that it can be fully understood by even an average internet user.
An Internet "keyword" can be defined as an alias (possibly localized) for a fully-qualified domain name or URL. In many of these systems, one keyword matches exactly one DNS name. Hence, there is no categorization function (in the traditional sense) involved, nor is there a search process that could yield more than one DNS domain name.
The practical effect of the IDN's position is to lay sole claim to the period, period. This isn't just about XTNS, there are many other services offering either IDN.TLD or IDN.IDN, including Verisign/RealNames, and Walid with a patented method, and there are more all the time. There are also various methods in use, client based, server based, and client-server based, and each of these aren't all done in the same way.
Let's look at the practical effect of the IDN committee's claim. Is the Verisign IDN testbed, with Walid and RealNames, and others, a keyword system? Is new.net, reportedly soon to offer IDN.IDN, a keyword system? As Walid does, and new.net plans to, market directly to ccTLDs, I think this has the IDN Committee, and ICANN, somewhat worried. A DNS based, ICANN controlled, IDN.IDN system is some time off, Walid says 10 years. I think that's too pessimistic, but given how many years it has taken ICANN to create seven new Latin gTLDs, to say nothing of the subsequent problems, perhaps not.
In the meantime solutions will be created by others. Let's turn your statement around. If the IDN's position is limited to keywords then it doesn't address some client plugins, server enhancements, et cetera, rendering their position ineffective, if not moot. If the IDN's position is not limited to keywords (and even if it is today that could change tomorrow with mission creep, or what I call ICANN incrementalism), then the IDN Committee and perhaps ICANN (we may get a sense of that at Accra) seem to be engaged in an act of desperation.
The ICANN controlled Latin namespace is shrinking, internet use by those who prefer non-Latin is growing, to the point that it will soon be the more dominant of the two. There's a real chance that ICANN won't be first to, or ever have control over, that IDN.IDN space. I think they're running more than a little scared. -g