In Cairo I'm finally going to have scheduled ICANN minutes to speak about Indigenous Intellectual Property, the set of issues lacking from ICANN's original WIPO-2 IP framework. Finally. It was just last Spring (of 1999) that Bob Gough and I put this effort together.
[Goodness, I'd not yet walked out of Nokia Research, trying to get Nokia interested in name spaces for mobile devices and physical location as a network semantic -- a fancy way of saying that I wanted to buy tax-free in the 20 minutes it took me to get from Maine to Massachusetts. Now both are no-thinkums and the GeoPriv activity in the IETF is almost ten years old.]
The context is new generic top-level domains, if for cultural and linguistic purposes, what are the ways to make the senior rights of cultural custodians and language users prevail over the claims of corporations that have expropriated (stolen) words and built mass marketing around those stolen words? Who wins? The CNO+UKB+ECN or Chrysler, LLC? Who gets "Cherokee"?
Similar issues arise for cities as top level domains, where we're innovating this time around. Does the Times Corporation "own" all references to Times Square? Does whoever owns the Chrysler Building (unlikely to actually be a CAFE avoiding, muscle car, SUV and heavy pickup truck, auto manufacturer) also own all common spatial references that are relative to the Chrysler Building?
Rinse and repeat for regional top level domains, where we, and others, are innovating this time around. Which government (in Africa) can prevent the regional .africa registry from using geographic identifiers? Who "owns" the Nile or the Sahara? Naturally, there are governments ready to assert the obvious.
Anyway, the point isn't that I get to speak at ICANN, I do that all the time. The point is that with the Intellectual Property Constituency observing, I and others are going to be articulating the IP policies that are going to be in applications for linguistic and cultural, municipal, and regional applications, which will change ICANN's IP "law", relative to both trademark and government claims of priority.
As if that weren't enough fun, I've spent the past two days ruining my eyes reading Cree and writing Cherokee, with the goals of (a) once again preventing a standards body from banning Cherokee, and (b) being certain that "finals" (which is where the Y-Cree vs Th-Cree ... fun begins) aren't affected by a "labels MUST NOT (begin|end) with combining characters" rule, which is to say, the w-dot and dot-w problem plus the generalized finals problems, all made harder by Unicode having used the Canadian Government's "Unified Canadian Aboriginal Syllabics" project, for which I kick myself, having been present at the point of error and having thought it a very, very good thing (tm). All the Dene characters (several variants), the Inuktitut characters, the Cree characters (more variants), the Siksika characters (oki kokona!), plus some Ojibberish, are all smushed together into a glyph catalog in UCAS (now Unicode), making reading the standard reference for any specific language absurdly difficult.
You are in a maze of twisty passages that all look alike. This is what the Unicoders did to Chinese, on a significantly larger scale, with "Han Unification". Perhaps printer vendors are inherently evil, how else to explain their indifference to text-as-characters, not puddles of fungible glyphs, to be compressed so that dingbats and other bits of dreck will fit into factory installed ROM? They managed a 21% reduction in memory necessary to put Cree et al fonts on printers. Odd people to have handed over all human language to. Mercifully, that cabal will be brewing over a fire in California the week we actually struggle over IDN policy in Cairo. Competing clans of Neanderthals.
This bus stops at Minneapolis a week after the Cairo meeting, and there is a strong dependency as ICANN "consumes" a revised "International Domain Name (in Applications)" standard, and the IETF IDNAbis WG may, eventually, produce yet another work of profound brokenness.
I thank my stars I'm not responsible for Korean, as that's the whipping boy of the moment. In 2003 the Chinese were the people the standards bus ran over, and carefully backed up to run over a second time.
13A0..13F4 ; PVALID # CHEROKEE LETTER A..CHEROKEE LETTER YV
That's what success looks like. One line in a
table. PVALID, not DISALLOWED, for the entire range of the Cherokee codepoints in Unicode.
1401..166C ; PVALID # CANADIAN SYLLABICS E..CANADIAN SYLLABICS CAR
166D..166E ; DISALLOWED # CANADIAN SYLLABICS CHI SIGN..CANADIAN SYLLAB
166F..1676 ; PVALID # CANADIAN SYLLABICS QAI..CANADIAN SYLLABICS N
Only the Xtian "cross" symbol and the w-dot fix for punctuation, the little "x" sentence terminator character, are disallowed. Yes, I did kill Christianity. Proud of it. But making sure of the leading, and trailing dot and other "terminators" is the cause for my temporary blindness.
I can't express how absurd it is to have to scan over lines and lines of texts in all Northern languages in all scripts and writing conventions, to ensure that one code point has the associated semantics correctly defined, and that a rule intended to cure a defect in the Unicode algorithm for "dot" as a sentence terminator when "dot" is used as in infixed label separator in domain names, at least does no harm to Cree and other languages' writers and their readers who use syllabic rather than roman characters. If I hadn't killed Christianity I'm sure my reward would be "in heaven". The obverse is obvious.
Writing in Cherokee is hampered by the fact that the Unicode notion of the script isn't the CNO's notion of the script, and every time I write "do", instead of a "V" I see a lambda. That's a bother, and I suspect there are others.
The wayback machine found this bit of old cree-l at nisto for me:
Date: 7 Jul 1997 14:58:34 -0400
From: Eric Brunner <(suppressed)@opengroup.org>
Subject: Re: Making computers that work in Cree
How time flies.