Re: what's the language of a document ?

Hi Aryeh,

On Mon, February 8, 2010 05:01, Aryeh Gregor wrote:
> On Sat, Feb 6, 2010 at 8:55 PM, Andrew Cunningham <andrewc@vicnet.net.au>
> wrote:

>> lang="" could be required or not required. but language detection is a
>> poor reason for deciding.
>
> I think it's fair to say that right now, <html lang> is not so
> uniformly useful that authors need to be warned if they omit it.  If
> the *average* page (some random hit from Google, say) doesn't have any
> use for it, then I don't think it should raise a warning -- warnings
> should be useful to most authors, not only a small minority.
> Otherwise you're making warnings as a whole less useful.
>

I did say "could be required or not required". I have no preference either
way ;) The point of the post was

1) in certain situations language tagging can be critical, and if css3
fonts module goes the right way, the number of languages it is necessary
for will increase.

2) the language detection argument is an extremely poor reason for
deciding one way or other.

I suppose a few things are happening here, you propose that most authors
don't use it and don't need it in most cases, As a consequence warnings
aren't needed.

I suppose is a philosophical position.

I believe that if not generally, then at least in specific circumstances,
it is necessary. I don't find your argument convincing, if you used that
argument on language tagging you could use a similar methodology on all
other HTML tags. Although, after some thought i would probably agree that
a warning isn't necessary.

But I think I'm coming to the understanding that the HTML specs, the CSS
specs and validators rarely tell you all you need to know to create a web
page properly.

lets see if I can explain that in a practical sense.  I'm doing more and
more work with Burmese and S'gaw Karen at the moment. And probably teh cux
of the web development job is knowing

1) which HTML elements never to use
2) which HTML elements you need to do a complete overwrite of the default
presentation'
3) limitations in web browsers' font rendering and OpenType support

-- 
Andrew Cunningham
Research and Development Coordinator
Vicnet
State Library of Victoria
Australia

andrewc@vicnet.net.au

Received on Sunday, 7 February 2010 19:54:08 UTC