Posted by usenet+2004 on 09/23/06 08:41

Beauregard T. Shagnasty:

[<http://shirley.profectus.com.au>]

> I would change it from:
> You can view the notes <a href="text/PortraitNotes.pdf"> here</a>.
> to:
> <a href="text/PortraitNotes.pdf">View the notes</a>.

Talk about out of the frying pan and into the fire! You have gone from
'here', a word that merely fails to describe the resource, to 'View the
notes', a phrase that (1) dictates *how* the user should experience the
resource, or at least reveals how the author expects the user to
experience it, and (2) presumes that the action performed on the link
is a retrieval. Link text, I think, should say what the resource is,
in this case Shirley Bourne's notes.

Ideally, without recasting the paragraph, I would have the markup:

<P>Sometime, fairly recently, Shirley decided to document her
portraits. <A href="<URI>" title="Shirley Bourne's notes on her
portraits">The notes</A> are mainly reminiscences about the Segall
family.</P>

The link destination could then be an abstract resource, not a
particular representation. In other words, you could offer a PDF
version and an HTML version, allowing the browser's Accept header to
determine in the background which version the server should respond
with. Alternatively, you could have a normal link (to the abstract
resource) and a second link to a PDF version. The second link's URL
could have a suffix <.pdf> since it identifies a particular
representation of the resource.

> Aside: on pages of mine where PDF downloads are available, I include
> the usual Adobe Reader link and instructions, but also include same for
> alternative readers, usually FoxItReader, and explain that it is much
> more lightweight and less intrusive than Adobe.

These instructions violate the principle Don't Mention The Mechanics.
(Basil Fawlty's 'I mentioned it once but I think I got away with it'
won't cut it on the WWW.) I do anticipate the counter argument that
such instructions are acceptable because they are so common that users
are by now familiar with them, but I regard that argument as weak since
what is familiar is not necessarily what is best. After all, there are
alternative ways of presenting the same information that do not involve
mentioning the mechanics.

--
Jock

[Back to original message]