Re: [Ltru] matching comments: use of prefixes

[Harald Tveit Alvestrand <harald at alvestrand.no>]

--On onsdag, september 28, 2005 10:17:34 -0700 Addison Phillips 
<addison.phillips at quest.com> wrote:

> The use of prefixes (and thus basic language ranges) simply allows the
> use of a prefix if this is the case. --
>
> ...the final sentence seems particular lame.

I think this sentence (without the parenthesized remark) dates right back 
to RFC 1766, and I still remember the debate that got us there.

What it's trying to prevent is dialogues like this:

Registrant: I register (or use) sgn-GB and sgn-FR
Searcher: I want something in "sgn".
System: OK, here is one in "sgn-FR". Enjoy!
Searcher: I got back "sgn-FR", and don't understand it. The system is
         broken!
Registrant: Which part of 'might not all be mutually intelligible' did
         you fail to understand?
Searcher: But I search for "en" every day and get back "en-US"....
Registrant: I repeat....

There will always be some ranges that it doesn't make sense to search over. 
And the standard is not going into the business of saying which those are.
When Addison suggests:

> The use of prefixes (and thus basic language ranges) allows languages to
> be selected as if this were always the case. In most cases, the language
> range and the set of language tags associated with it are mutually
> intelligible enough to be useful and this form of matching has
> historically been to most prevalent.

I think that's going a little far in the direction of "this works most of 
the time, you can safely act as if it works all the time".
But it's a minor distinction.






_______________________________________________
Ltru mailing list
Ltru at ietf.org
https://www1.ietf.org/mailman/listinfo/ltru