Wikidna team final report
eblanconil at gmail.com
Wed Sep 23 02:38:52 CEST 2009
this is the IETF, it means engineers. Engineers know machines and
RFCs, not people and uses. For them telling some non-English speaker
"you can now eventually do what you expected and not only use
American" is candidly "reformating" him. They worked hard for ten
years to permit people to use their own scripts, but they still do not
want they use them in the way of their own language.
I know most outside the IETF think it is to support some capitalistic
strategy of their employer (read RFC 3869 by the IAB on that risk). I
do not think so in this case. The fundamental misunderstanding is they
confuse domain names with hackers' mnemonics (the hack is to better
remember a TM you saw on a Bus side because it is has to be written in
a different way).
Just be careful. We do not care about language. This is the IETF not
the UNESCO. DNs are neutral labels chains. The job of this WG is
precisely to make sure IDNs are also script neutral. This is for
people to use them as steganographic labels chains (such as these
famous possible menmonics) they must be neutral so they can support
any embedded semantic. Steganography depends on end to end (brain to
brain) orthotypography. Imposing an IETF English orthotypography that
bans 50% of the human character set kills most of the steganographic
interest of IDNA as a system for more linguistically complex
Never mind, this is a technically easy to solve layer violation (here,
at the IETF/LC or in the field deployment), between "engineering" and
"usering" layers. They deal with software, we deal with brainware. The
only thing we need is the RFCs to be eventually published so we build
on top. The pity is that we did a lot of work together and we are just
two lines of text and three lines of code a part from a common
consensus. Next time.
2009/9/22 Dominique Lacroix <domi.lacroix at gmail.com>:
> Gervase Markham <gerv at mozilla.org> wrote:
> since they
> started using the DNS and the Internet, which could have been up to
> around 25 years ago.
> After thinking, are you sure, dear Gervase, that such an argument (25 years
> of use) is an effective one to demonstrate that a choice is the right one?
> It could be 25 years of wrong use by users who have not any other choice...
> In French, we say: "fait accompli". I'm sure you understand it.
> And in Latin: "Errare humanum, perseverare diabolicum."
> Your fundamental misunderstanding is to not realise that DNS names are
> aides memoire, not words in a particular language, no matter how much
> they may resemble the latter.
> The double nature of DNS is indeed a great source of issues.
> IMHO, actually, domain names are *labels* (in French "étiquettes").
> And as digital labels, they are both: computing signs and linguistic signs.
> For the major part of users, reading a domain name when surfing deals mostly
> with language rules.
> As that double nature system is an original one, I believe the designers of
> the system are probably free to take in account more or less rules of each
> If IETF wished to better take in account the linguistic aspects of domains
> names it could be possible, don't you think?
> But perhaps the price could be more complicate computing rules?
> I wonder if this list is the best place for such a debate. Is there a
> better one?
> Best regards,
> @+, Dominique Lacroix
> Idna-update mailing list
> Idna-update at alvestrand.no
More information about the Idna-update