- From: Richard Ishida <ishida@w3.org>
- Date: Wed, 10 Aug 2011 13:31:36 +0100
- To: Chris Mills <cmills@opera.com>
- CC: "'public-evangelist@w3.org' w3. org" <public-evangelist@w3.org>, www International <www-international@w3.org>
Hello Chris, [cc www-international so that they know i have sent feedback, and in case others wish to comment] Here's some feedback on http://www.w3.org/wiki/The_HTML_head_element "The language codes may be two-letter codes, such as en for English, four-letter codes such as en-US for American English, or other, less common, codes. The two-letter codes are defined in ISO 639-1, although modern best practice dictates that you should use the IANA subtag registry for your language code definitions." I think this paragraph needs a fair bit of attention. [1] language codes => language tags (for consistency and clarity - codes was used in the past to refer to ISO language codes or region codes, but something like en-US is two such codes (though only one language tag)). (btw, en and US are both 'subtags' - be careful not to mix tags with subtags) [2] language subtags can be 2 or 3 letters, region subtags can be 2 or 3 alphanum characters, so the opening part of the paragraph is quite misleading. [3] i strongly urge to not refer people to ISO 639 - they should use the IANA registry to look things up (and you may want to point to http://rishida.net/utils/subtags/ which makes lookup a little more user friendly). [4] 'modern best practice': well actually its in the standards, so it's a little more than best practice [5] it may be better for this audience to link to http://www.w3.org/International/questions/qa-choosing-language-tags rather than http://www.w3.org/International/articles/language-tags/ "Don't worry too much about this for now. utf-8 is the universal character set, which includes pretty much any character that you might want to use on a web page, from any common human language, so it is a good idea to declare this to make sure you HTML has full international capabilities. In addition, you can avoid a serious Internet Explorer security risk by declaring it in the first 512 bytes of the page. So just below the <head> tag is fine. This is what all the below examples will do." [6] actually they need to worry about it at least enough to ensure that they are actually *saving their document* as UTF-8, not just changing the encoding declaration - otherwise, a doc saved as iso-8859-1 for example will fail to display properly when it comes to accented characters. They also need to be aware that the server may be overriding their declaration. I recommend that you step back a little in the wiki, add a brief description of what an encoding is and why it's important, and add some text to say that authors should ensure that their editor *saves the text* in utf-8, but, if not, they should ensure that the charset attribute should indicate what the actual encoding used is. We have some articles that can help people understand these concepts at http://www.w3.org/International/questions/qa-choosing-encodings http://www.w3.org/International/questions/qa-changing-encoding http://www.w3.org/International/questions/qa-setting-encoding-in-applications Hope that helps, RI On 04/08/2011 16:56, Chris Mills wrote: > UPDATE - 4th August 2011: I've updated http://www.w3.org/wiki/The_HTML_head_element to clean up language, add new HTML5 features, and add in a new section about doctypes, to replace Choosing the right doctype for your HTML documents (http://www.w3.org/wiki/Choosing_the_right_doctype_for_your_HTML_documents). The original article was a bit long winded, and needed a lot of updates to account for new thinking about doctypes, HTML5 doctype, etc. > > this is ready for proofing/translation now. > > QUESTION - should this big new doctype section be put into a new article? Does it make the article a bit too long? > > > > -- > > Chris Mills > Open standards evangelist and dev.opera.com editor > Opera Software > > * Try our browsers: http://www.opera.com > * Learn to build a better web, with the Opera web standards curriculum: http://www.opera.com/wsc > * Learn about the latest open standards technologies and techniques: http://dev.opera.com > > > > -- Richard Ishida Internationalization Activity Lead W3C (World Wide Web Consortium) http://www.w3.org/International/ http://rishida.net/ Register for the W3C MultilingualWeb Workshop! Limerick, 21-22 September 2011 http://multilingualweb.eu/register
Received on Wednesday, 10 August 2011 12:32:00 UTC