W3C home > Mailing lists > Public > www-international@w3.org > April to June 2008

Re: Language tag education and negotiation

From: Leif Halvard Silli <lhs@malform.no>
Date: Mon, 28 Apr 2008 05:10:14 +0200
Message-ID: <48154016.7000904@malform.no>
To: Andrew Cunningham <andrewc@vicnet.net.au>
CC: www-international@w3.org

[I hope you can play with Squirrel a bit, so that it sends out messages 
that are easier to read and work with. Also, please do not insert manual 
linebreaks.]

Andrew Cunningham 2008-04-28 04.20:
> >> For instance as indicated already in this thread, some users might
> >> wantto give a higher priority to Swedish, and web browsers
> >> automatically making assumptions on how to treat Norwegian issues
> >> may negatively impact on such users.
> > I must disagree strongly. The spesific example I mentioned was an
> > edge case [...]
>
> Lets see, a different example, then. Some of my friends speak a language Dinka (ISO-639-2 language code din). In ISO-639-3 it is represented by the language codes dip, diw, dib, dks, dik each of them will have preferences to which language code they'd prefer and what order they would be in.
>   

As I noted in an reply to an offlist message to you: Provided that those 
sublanguages of Dinka was specified as "sub codes" as well, then the 
issue woudl be rather simple:

"din" would cover all languages.
"din-dip" would cover that sublanguage

Then, if you as Dinka user, comes to a web server with a request to get 
a third variant, such as  "din-dik", but that server only operated with 
"din", then you woud get that, because of the "partial match" that John 
explained about.

Unfortunatly, the extended language tag system will not work like that. 
And thus, the perhaps simples way of offering fallback support, is gone.

(This is an argument against your claim that "what the world needs most, 
is educations, sweet education". I think the world most needs simple 
rules, which are simple to educate about/teach.)

> An added level of complexity would be fall back language; some would prefer "en" as a fall back; some would prefer  "ar" as a fall back. Some would oppose any "ar" content at any cost. Some  would want to fallback to "ar" then "en". some would prefer "sw-KE" as a fallback and then "en"
>   

Now you have switched to a subject which is no different from fallback 
from English to French or from French to English.

The question I have then is: How could we generalize the issue of making 
such choices, for the users?

Another thing to keep in mind is "where to draw the line". If you are 
not willing to identify certain useful settings, if you are only 
(claiming to be) thinking on individual solutions, then you might fail 
to come up with solutions that could actually help developing the 
society. For instance, it is outright destructive for the official 
status of Nynorsk as one of two official Norwegian langauges (a status 
that is more than 100 years long), that Mac OS X makes no association 
between 'nn' and 'no'.

Now, back to the issue of how we could generalize. Someone used the word 
'second language'. *That* is a keyword. When I open Firefox to configure 
its language settings, then it is impossible to find anything that 
resembles the word 'second language' or 'main language'.

Web browsers and Operative Systems should let users select language 
support using such words as 'main langauge' and 'second language'. 
Things that users, society, developers are familiar with. It should not 
just be one massive list of langauges to select from - such as in the 
case in Firefox now.

When we speak in such terms, then it becomes obvious - to developers, 
authors, users, that selecting a little used languag as main langauge, 
should not prevent you from getting a more used second language.

> It would be useful to set up a server or a web browser to have useful defaults. But in either case both the the server and browser configurations need to be customisable by individuals.
>   

Useful defaults which users can reconfigure are OK. But - except for 
Safari (on OS X at least), which doesn't let me set up any cascade of 
preferred langauges, the most important thing seems to be to get more 
useful defualts. Or - I should rather say: useful sets to choose between.

> For lesser used and minority languages, customisation is critical. The reality is
> when it comes to using the web the burden of knowledge is often placed on
> lesser used languages.
>   

What you say could also be taken as a mistrust in that the developers 
are able to get things right, and therefore we must at leaste have as 
much flexibility as possible.
-- 
leif halvard silli
Received on Monday, 28 April 2008 03:11:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 2 June 2009 19:17:17 GMT