W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > July to September 1999

Re: Visible, simple, accessible sites

From: Charles McCathieNevile <charles@w3.org>
Date: Sun, 19 Sep 1999 12:54:14 -0400 (EDT)
To: John Nissen <jn@tommy.demon.co.uk>
cc: mitake@alum.dartmouth.org, wai@tommy.demon.co.uk, w3c-wai-ig@w3.org
Message-ID: <Pine.LNX.4.10.9909191243560.26497-100000@tux.w3.org>
One of the problems with this approach is that it assumes all websites are,
or should be, hierarchical. This is simply not true - the purpose of  the web
was to enable linking of document points to other document points in a manner
which did not require hierarchy. In the case where there is a natural next,
previous, contents, main page relationship HTML provided a mechanism for the
browser to present that through the link element. Many browser manufacturers
decided, in their wisdom, not to implement that mechanism, and many users
decided in their wisodm to use those browsers anyway.

If you read the HTML 4.0 specification yo will notice that the hierarchy of
files, as reflected in the URI pattern of slash-separated names, does not
directly match the hierarchical structure (such as it is) of the document.
Howevr thre are navigation points provided which match the link elements
provided - if you have a decent browser you will be able to navigate the
structure automatically.

Getting to a "parent" page, where there is one present, is a useful
technique, as is navigating sequentially through pages in certain types of
collections. Until User Agents are made which allow the user to make use of
the authors encoding of the site structure, authors should continue to
provide such links. Terms like next, previous, index, contents, welcome, main
page, and home have meaning in pagees. Since some browsers call their initial
page a start page, and others call it a home page, and others probably call
it something else I suspect this is a concept which will simply need to be
learned by hte user - that the browser and the content are separate, and
there are two or more possible home pages depending on the context (which is
anyway true of web content itself, regardless of the browser).

"Out" seems like a bad choice - it gives me no idea where I am being sent.
Think of what it means to go out. Links should give some idea of what is at
the other end (See WCAG).

Charles McCN

On Sun, 19 Sep 1999, John Nissen wrote:

  Hello Mitake,
  You make a good point about entering a page other than the main page 
  (or front page as I prefer to call it).
  What is definitely confusing is if the buttons on the page have
  the same names as buttons on the browser.  So no button on the page 
  should labelled "home", but rather "main page" or "front page".  
  Likewise no button on the page should be labelled "back", but rather
  "parent page", "next outer page", or perhaps simply "out".
  However there are two solutions for getting to the parent:
  a) have a button on each page (except the front page) to its parent;
  b) work out the parent's URL from the current URL, and use this.
  The latter assumes that the URL naming adds a slash-something for each
  level in the hierarchy, so to go to a parent you simply strip the
  last slash-something off the current URL.  This could be done by the
  browser, with a Parent button.
  I assert that the more you can leave to the browser the better,
  so the browser keeps track of where you are in the structures,
  and builds up a model which you then navigate with its buttons.
  The danger of having buttons on the page is that this subverts the
  model being built by the browser, so the browser cannot in turn
  present the simplest possible model to the user.  OK, the buttons
  can be there, but allow the user to simply ignore them.  Of course
  at the same time we need more intelligent browsers and a strict
  URL naming convention corresponding to the site hierarchical
  structure as formed by internal links.
  The way forward towards more intelligent browsers is a matter for 
  W3C and the WAI user agent group. 
  Cheers from Chiswick,
  In message <1287516182.937590386@[]> 
  Mitake Holloman Burts writes:
  >I find site specific  back and home links to be extremely useful and
  >important when my initial introduction to a website is a page other than
  >the main page of the hierarchy, i.e. when I am using a search engine or a
  >reference page. It often gives me an easy way to understand the larger
  >context of a given page. I am generally frustrated by pages which don't
  >give me a way to get up to their parents.
  >> At 05:04 PM 9/17/99 GMT, John Nissen wrote:
  >>> 2.2  Don't have back and home buttons on the page
  >>> Again these are confusing, as the behaviour conflicts with that
  >>> of Back and Home buttons on the browser.
  >>> 2.3  Have a hierarchy, navigated top-down
  >>> Keep the site hierarchical, and encourage people to enter at top level
  >>> (by premoting the URL for the top level page, by always refering to the
  >>> site by this URL, by using it as link from other sites, etc.).
  >>> Have links only down the hierarchy, except for cross-links where 
  >>> they are natural (e.g. in an index, see 2.4).
  Access the word, access the world       Tel/fax +44 181 742 3170/8715
  John Nissen                             Email to jn@tommy.demon.co.uk
  Cloudworld Ltd., Chiswick, London, UK   http://www.tommy.demon.co.uk

--Charles McCathieNevile            mailto:charles@w3.org
phone: +1 617 258 0992   http://www.w3.org/People/Charles
W3C Web Accessibility Initiative    http://www.w3.org/WAI
MIT/LCS  -  545 Technology sq., Cambridge MA, 02139,  USA
Received on Sunday, 19 September 1999 12:54:17 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:05 UTC