Message-Id: <6.1.2.0.2.20050328230617.049c0430@mail.jefsey.com>
Date: Mon, 28 Mar 2005 23:58:44 +0200
To: John Cowan <jcowan@reutershealth.com>
From: "JFC (Jefsey) Morfin" <jefsey@jefsey.com>
Subject: Re: [Ltru] Re: XML registries now available
In-Reply-To: <20050328172934.GQ24325@skunk.reutershealth.com>
References: <CFEE79A465B35C4385389BA5866BEDF00C7B00@mailsrvnt02.enet.sharplabs.com>
	<20050327214939.GN24325@skunk.reutershealth.com>
	<6.1.2.0.2.20050328121911.037a4eb0@mail.jefsey.com>
	<20050328172934.GQ24325@skunk.reutershealth.com>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; format=flowed
Cc: ltru@ietf.org
Precedence: list
Sender: ltru-bounces@lists.ietf.org
Errors-To: ltru-bounces@lists.ietf.org

At 19:29 28/03/2005, John Cowan wrote:
>JFC (Jefsey) Morfin scripsit:
>
> > I assume you consider the situation where:
> > 1. the application gets a langtag
> > 2. consider what do to with it.
> > 3. there are three possible solutions?
> >     - it knows the langtag from it being built-in.
> >     - it does not know it but it has a filtering/negociation strategy
> > built-in or based upon non IANA parameters
> >     - all this fails and its filtering/negociation strategy resort to a
> > IANA input before concluding?
>
>In practice, the second and third strategies are a matter of using a
>matching algorithm (which need not be standardized and can rely on local
>knowledge as well) to reduce the unknown langtag to a known one.  With an
>absolutely unknown langtag, nothing useful can be done except to display
>the full name of the language, which is useful but not very useful.
>Without translators, spelling checkers, and other such tools, one can
>do little with such a document.

Yeap. But the first and second strategy has no big load on the IANA.
The first point is to know if the langtag is registered. Then to know its 
environment. With the maximum of speed and the lowest load for the database 
system. The rest is a matter of other services as you say.

>This is quite different from the DNS case, where given the mapping of
>a computer's name to an IP address, one can indeed interact with that
>computer in useful ways.  If I know how to reach newmachine.somedomain.fi,
>I can try to discover what resources it provides.  If I receive a document
>langtagged "sms", the most I could do with it is label it "Skolt Sami".

I do not quote the DNS as a way to reach an IP (what can also be the use 
for). The IP address given by the DNS is just an information as another. 
The DNS is first a distributed database system, which can easily supports a 
billion of information at low load and low cost. With a distributed 
management system.

> > I would say this is similar to anti-spam or to anti-virus strategies.
>
>Languages aren't constantly being created:

In term of load, the problem is the same. What motivates a query is the 
change. Whatever it is. Actually this is not totally true: the number of 
_documented_languages_ is going to increase.

>on the contrary, they are
>constantly going extinct.  The list of living languages will probably
>shrink by 90% over the next century or two.

Who knows? The change over the last 50 years when a cultural development 
policy is engaged is not necessarily that way. The debate I follow right 
now is for example to know if a regional language (Occitan) is to be 
registered as one (as today in the 639-3 draft) or as six languages.

I see young kids in my family having more Breton than French. 20 years ago 
it was unthinkable. IMHO, Multilingual Internet MAY (I mean nothing more) 
lead to a change in the current trend. We start having statistics (it will 
still take time to have significant results in number of people and in 
number of languages and cultures). They seem to show that bilingual people 
have a slightly higher IQ and that "apparent IQ" (when you test the same 
person's IQ in different languages) is lower in English (common language) 
than in their mother tongue.

> > Obviously codes change les frequently than spamming hosts and viruses, but
> > the simple fact they change make the load similar (the load is not because
> > there are a lot of changes, but because there is a check if nothing
> > changed).
>
>In practice this table will probably be updated only when the operating 
>system is.

Probably yes, and this is in line with/below the figures I gave. Question 
is from where: from the OS CD or from IANA?
jfc


_______________________________________________
Ltru mailing list
Ltru@lists.ietf.org
https://www1.ietf.org/mailman/listinfo/ltru