W3C home > Mailing lists > Public > public-i18n-geo@w3.org > August 2005

Labelling articles

From: Richard Ishida <ishida@w3.org>
Date: Thu, 11 Aug 2005 18:44:05 +0100
To: "GEO" <public-i18n-geo@w3.org>
Message-Id: <20050811174406.7BC234F33E@homer.w3.org>

See http://www.w3.org/International/questions/changing-encoding for a suggested location (suggested by Russ) and an example.  I and others on the call yesterday seemed to like the placement.

More problematic is what content to put there.

Current content contains short, descriptive text, that references both the type of person and the technologies.

Another suggestion made yesterday by Molly was to use instead tags.  These may be something like Flickr. For examples see the photo at http://www.flickr.com/photos/ishida/32399183/ and the list of tags at http://www.flickr.com/photos/ishida/tags/ )

I'm still not clear in my mind whether we should use just tags or tags plus descriptive text to label intended audience.  

I'm interested in the idea of tags, though.  I've been wanting for some time to have a way of getting to information via keywords, though I've been thinking more in terms of standard indices.  For example, we have an article that touches on the use of the select element.  This isn't an entry in the topic index at the moment, but people ought to be able to look for information about it.

Using tags seems to offer a non-labour-intensive way of doing this (which is important).  We may even be able to auto-generate an index of tags from the keywords we put at the top of articles already.

What kind of things should we capture in tags? I'm thinking:
- maybe some basic audience level (actually maybe only 'novice')
- technologies in the article (eg. XML, XHTML, CSS, XSL-FO, HTTP, etc.)
- specific things such as HTML elements discussed (eg. select, hreflang, xml:lang, etc.) or tasks (eg. changing encoding, declaring language, etc.)
- general topic areas (eg. character sets, language, bidirectional text, etc.)

How would it work?  What I've so far thought is:
- there would be a tag index, containing all the tags in alphabetical order on a single page
- if you clicked on a tag you would be presented with a page listing articles that contain that tag
- all tags relevant to a given article would be listed at the top of the article, as links
- you would be able to click on a tag link and get back a list of articles

What would be even more useful in my mind is to have a tag search box where you could specify multiple tags, eg. "novice declaring encoding CSS" and be presented with a list of docs that contain all or any of those tags.

Allowing people to add their own tags?  Hmm.  Not sure how that would work.

What do you think?


Richard Ishida

contact info:

W3C Internationalization:

Publication blog:
Received on Thursday, 11 August 2005 17:44:10 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:28:03 UTC