- Users want their email addresses to be memorable and transcribable
by recipients (who may use a language other than the sender's
native language)
And that's *exactly* the point. There exists a mass delusion among
today's "multiculturalists" that *any* system that contains *any*
prejudice towards a particular culture, language, or pattern of
thought is inherently "xenophobic". (That term has already been used
by victims of that delusion in this thread.)
If a bunch of rural Chinese farmers get together and come up with
their own means of communication, and I decide to join in with it, I
do *not* have any Divine Right to insist that they suddenly change all
their communication policies to accommodate the fact that I not only
don't speak their language, I can't even identify, much less
recognize, the most basic glyphs of their language.
Now, anyone have a US postage stamp handy? Take a look at it. Does
it say "USA" on it? Yes it does. It *has* to. Every country has to
put a *recognizable* identifier of its country on its stamp, and that
means recognizable to post-office workers around the world, and that
means the *Latin* alphabet.
(As far as I know, international mail is not permitted to have stamps
that have only *non-Latin* identification marks on it.)
There is, however, one exception to this rule: British postage stamps
need not identify themselves, because Britain (UK, England, whatever
it's called these days ;-) *invented* the postage-stamp system.
And correct me if I'm wrong, but when a letter arrives in the USA from
China, it is *not* permitted for the originating and destination
addresses to be written entirely in, say, Mandarin. We don't expect
our postal workers (or even our automated systems) to be able to
handle that. And, AFAIK, letters going in the other direction *can*
be addressed entirely in English, according to Western rules for such
addressing.
The Internet, and email, may *reach* into most every corner of the
globe, but it is fundamentally a Western "invention", its standards
and practices are based largely on specifications written in Western
language (primarily English), and its policies are expressed almost
100% of the time via the Latin alphabet. NG email is not going to
change that, any more than the Free Software Foundation (FSF)
choosing, years ago, to force "gender-neutral language" into one of
its standards documents changed the way free-software documentation
got written. (Instead, it caused some people, certainly myself, to
quit maintaining that document. I've got plenty of years as a
technical writer and project leader to know the difference between
writing good technical documentation in English and maintaining a
translation of such documentation into other languages, such as a
gender-neutral form of English; I wasn't about to take on both jobs at
the same time as an unpaid volunteer with no say in the matter.)
Is the Latin alphabet perfect, ideal, or even the best we have? Of
course not: we could do a much better job, today, based on our present
understanding of how the human visual cortex works and how children
learn, of designing a smallish (256 or fewer) set of glyphs through
which all fundamental computing and communication policies are
expressed, build a handful of similarly-derived languages on top of
that, and so on.
But that wouldn't do well with the "installed base" that already knows
Latin like the back of its hand. (Unless it looked and displayed a
lot like the screens on "The Matrix", or maybe like Klingonese,
whatever that looks like, few people would see any reason to make the
effort to learn the new glyphs.)
Even so, I'd personally prefer to "start over" by coming up with a set
of "computing glyphs" that starts with the Latin and other Western
characters most "everybody" knows and that are in common usage,
eliminating any close resemblances in favor of the most useful,
necessary, and/or recognized version (so, no lowercase "l", since the
digit "1" is more important; no letter "O" or "o", since the digit "0"
is more important; pick either "u" or "U", "c" or "C", "-" or "_",
either "." or ",", either ";" or ":", and so on), and build upward
and onward from there, perhaps using something like Esperanto for
basic computing-language terms.
This would allow leveraging familiarity among the installed base,
while dispensing with certain common problems (mainly, ambiguity,
either audio or visual -- what does the "oh" in "see three pee oh"
identify, the digit zero or the letter "O"?). All computing-related
entities needing precise representation even in human communications
(such as among sysadmins) could benefit from relying solely on such a
character set, which, aside from its *absences* (e.g. no lowercase
"l", it typically would get automatically "translated" to the digit
"1"), is already being taught worldwide in any school or training
situation whose students are expected to become part of the
international community.
But that's probably not going to happen either, so we either stick
with the basic Latin-oriented character set for the underlying
protocols (including email addresses), and let localization software
handle the necessary translations, or we risk deploying a fully
"multicultural" system that, due to operators inexperienced with the
vast array of glyphs they'll be seeing in email addresses, will either
not work or will rapidly dissociate into nationalized camps of mutual
recognizability and acceptance -- perversely *defeating* the goal of
fully internationalized communication that NG email promises and,
frankly, even today's email and web delivers, to at least some extent.
And if anyone thinks a "non-xenophobic" system can be deployed and
work via some kind of automated translation system that everyone will
have at their disposal, then just ask this question: why don't you
work on your mythical automated translation system, and have *it*
handle the Latin-based email addresses that will be universally
available in the NG email system *we* deploy?
Do *not* go down the rat hole of trying to use NG email to impose your
New Multicultural World Order. Otherwise it's a complete waste of
everybody's time and energy; while you're trying to get it to work and
be accepted, the rest of us will be happily using a *viable, working*
email system that uses Latin-based email addresses as its fundamental
unit of address.
(I'm sure this whole post will get me labeled as "xenophobic". I
don't give a rat's a**.)
--
James Craig Burley
Software Craftsperson
<http://www.jcb-sc.com>
--Fix qmail's qmail-smtpd so it doesn't crash on a big header line:--
<http://www.qmail.org/netqmail/>