W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > July to September 2002

Re: Click here

From: Al Gilman <asgilman@iamdigex.net>
Date: Fri, 09 Aug 2002 08:56:00 -0400
Message-Id: <5.1.0.14.2.20020809083542.02198710@pop.iamdigex.net>
To: Lynn Alford <lynn.alford@jcu.edu.au>, <w3c-wai-ig@w3.org>

At 12:40 AM 2002-08-09, Lynn Alford wrote:

>Does anyone have or know of a good page that explains why "click here" is 
>bad linking practice?  I have tried various forms of gentle persuasion to 
>suggest that "click here" is not good.  I have mentioned that some 
>browsers can print a lovely table of urls and link text, at which point 
>"click here" is obviously useless.
>
>It's a bad habit of some coworkers who don't want to think about link text 
>so they always use "click here".  I need truly excellent reasons to show 
>them the light.

** once over lightly

Pages are mostly built too fat to make reading them from start to finish a 
practical
way of screen-reader-assisted browsing.  This is because visual users can 
skim the page.

So screen readers pretty much _must_ use link-skimming to skim the 
page.  This is achieved by tabbing or by using a screen-reader-generated 
"list of links" view to skim.  The list of links is roughly equivalent to 
tabbing through the page.  In the user habits of those who succeed in using 
the web with a screen reader, they predominantly use this as their primary 
mode of reading on entry to a new page.

In tabbing or in the list of links view, only the sensitive text of a link, 
the "link text" is grabbed to represent the link.  All contextual 
conditioning as to what "clicking here" will do is lost.

So if the link text is not standalone, the web fails in this link-skimming 
mode and is too tedious for words.  P2 grade failure is incurred -- useless 
for practical purposes.

** at more length

The problem on this point is that the shoe is really on the other foot.

"Click here" is not bad design, it is just inaccessible design.

A visual user of a web page doesn't read it all in order, they skim the 
page aided by layout and emphasis and pick a link to click or section to 
read at paragraph depth.

Users of screen readers don't have this luxury.  They get to hear stuff 
from the page read.  Either they listen to it all, which takes too long, or 
they tab through or use a functionally similar "list of links view" that is 
built for them by the screen reader.

But page authors pack pages with a lot of stuff because they are designing 
in a visual engagement where they can skim visually in random access of the 
screen.  So users of screen readers very heavily depend on the list of 
links to get as close to skimming as they can within the available adaptive 
modes of reading.  And this gets _just the link content proper_, the 
sensitive text and no context, for a link.

So if the link text taken out of context doesn't give some hint of where 
the link gets you, in a way that distinguishes this destination from the 
other destinations on the page, the "list of links" or tabbing mode does 
not work, the user is forced to listen their way through it all (try 
listening through the typical navbar preface to a page) which is P2-fashion 
inaccessible (unusable for practical purposes) and so they go away (mad).

Al


>Lynn
>
>--
>Lynn Alford, Multimedia Services        Tel     (07) 47 81 5525
>Multimedia and Print Services           Fax:    (07) 47 81 4018
>Corporate and Commercial                Email:  imla@jcu.edu.au
>James Cook University  QLD 4811  AUSTRALIA
Received on Friday, 9 August 2002 08:56:10 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 19 July 2011 18:14:05 GMT