W3C home > Mailing lists > Public > www-html@w3.org > September 2003

Re: Problem with LANG keyword

From: Yuval Rabinovich <yuval@lab.co.il>
Date: Wed, 24 Sep 2003 19:50:12 +0300
To: Reuven Nisser <rnisser@ofek-liyladenu.org.il>, ibi-l <ibi-l@topica.com>, www-html@w3.org
Cc: 'shaula haitner' <shaula@shaula.co.il>, 'Gertel Hasson' <gilagh@netvision.net.il>, "BIGELOW,JIM (HP-Boise,ex1)" <jim.bigelow@hp.com>
Message-id: <001c01c382bb$ecbbed20$230d84d9@YUVAL>

Hello, Reuven.

Most of the documents in the Internet are written in English. However, in
the future there will be many more English documents. Do you suggest the
LANG=EN-US to be mandatory for such documents?

I think not. There is nothing wrong with defaults, and a document can
default to English. The same applies for other languages. The user agent
should be able to apply English/Hebrew to the Windows-1255 code page,
English/Arabic to Windows-1256 and all three to UTF-8.

If a document is written in French and Yiddish, however, there is a reason
to override the defaults, because both Windows-1255 and UTF-8 can be used,
and the character set is the same as English/Hebrew.

This discussion reminds me of the DOCTYPE DTD attribute. I never understood
it completely, and there is probably no browser that needs it or even uses
it for anything. The insisting of the W3C validator that it must be present
in all documents may have prevented some website builders to write proper
HTML documents.

Planning for the future is a virtue, but using unnecessary limits because we
estimate it will be needed in the future puts a burden on present Internet
developers, that may not prove useful.

I still think we may omit declarations for defaults. Unless you do it, you
really should go over all web HTML documents and make sure they start with
<HTML DIR="LTR"> statement.


----- Original Message -----
From: "Reuven Nisser" <rnisser@ofek-liyladenu.org.il>
To: "Yuval Rabinovich" <yuval@lab.co.il>; "ibi-l" <ibi-l@topica.com>;
Cc: "'shaula haitner'" <shaula@shaula.co.il>; "'Gertel Hasson'"
<gilagh@netvision.net.il>; "BIGELOW,JIM (HP-Boise,ex1)" <jim.bigelow@hp.com>
Sent: Wednesday, September 24, 2003 10:33 AM
Subject: RE: Problem with LANG keyword

> Hello Yuval,
> First of all you are correct. We are talking about potential developments
> the future. However, what we will not do today will not exist later. Even
> today the number of Hebrew documents is very large and it's a problem to
> change the way people think. Every day which passes without guidelines
> cause a problem with additional site. To my opinion we need to decide now
> about the recommendations because later when the tools will exist it will
> a problem if the decisions are wrong.
> You know, there is a place in the USA where they gather rubber garbage.
> because it is possible to recycle it now but because they estimate that
> someday the technology will exist and then the rubber will be available.
> Now to my second point. There are standards from W3C which are dealing
> HTML, Accessibility and so on. Unless it is absolutely impossible we need
> stick with this standard. The main reason to stick with the standard is
> every program will use the standard as a basic. If we want such programs
> work with our sites as well we need to stick with the standard. Think
> Jaws, HTML readers and so on. All are standard tools which we want to use
> and if we stick with the standard chances say that we will be able to use
> them.
> Now, if you want to change the standard for the LANG usage, do it in a way
> complies with the standard. I think that asking everyone to add one META
> which states the language is not too much.
> Regards,
> Reuven Nisser
> Ofek Liyladenu
> -----Original Message-----
> From: Yuval Rabinovich [mailto:yuval@lab.co.il]
> Sent: Wednesday, September 24, 2003 12:56 AM
> To: Reuven Nisser; BIGELOW,JIM (HP-Boise,ex1); www-html@w3.org
> Cc: 'shaula haitner'; 'Gertel Hasson'
> Subject: Re: Problem with LANG keyword
> Greetings,
> Since I am not familiar with user agents that use speech synthesizers I do
> not know whether there is a real problem here. As far as I know there are
> automatic Hebrew speech synthesizers.
> So if we are discussing a potential future development here, I still think
> that in most cases the LANG attribute should remain optional. The default
> will be the server HTTP LANG and the user may have his/her own
> The use of the LANG attribute should be reserved for special cases, when
> defaults should be overridden.
> There is nothing wrong with a user agent that defaults to Hebrew content
> handling a code that occupies the 1488-1514 space. But if it happens to be
> Yiddish document, there the LANG=ji attribute should be used.
> Most of the documents in the Internet are written in English. Very few of
> them state it explicitly. The decision to have English as default for user
> agents seems natural.
> Regards,
> Yuval Rabinovich
Received on Wednesday, 24 September 2003 13:08:37 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:06:05 UTC