W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > January to March 2004

Re: link to us: Is there a recommendation to provide a graphic for external linking? if so where?

From: Charles McCathieNevile <charles@sidar.org>
Date: Mon, 19 Jan 2004 17:53:42 +0100
Message-Id: <09B7D8C8-4AA0-11D8-9125-000A958826AA@sidar.org>
Cc: WAI Interest Group <w3c-wai-ig@w3.org>, tina@greytower.net
To: Jonathan Chetwynd <j.chetwynd@btinternet.com>

On 18 Jan 2004, at 23:37, Jonathan Chetwynd wrote:

> Chaals & Tina,
> I do find it sad that I have neither the wit, patience, ability or 
> charm to express myself more clearly, but thanks again.

Nor do I. Sadly I am even worse as a graphic artist. This is why I 
write so much :(

> BTW Chaals how about your views on AAA conformance and the need to 
> include a 'link to us' graphic?
> the 'obvious' point being that if everyone has a go, and link images 
> are royalty free (did I forget to mention that?)
> then pretty soon there will be a public and rich semantic graphical 
> directory, which could be a great help....
> images as images are great, but images as links are different kettle 
> of fish, one which has little history, and much to offer.

Well, images as links have a long history - but it seems that there has 
been relatively little work done in WAI to try and collect the things 
we have learned. I think that such work would be valuable. (I don't 
mean just starting from scratch - we should look at the existing work).

> is "شارل مككثينبل" "click here" possibly?

Nope, it's my name written in arabic in Unicode. (hint: try looking at 
the text in Lynx). It isn't an easy way to find me though because I 
don't use a consistent transcription of my name. (Then again, others 
don't for latin characters either. Google is semantic-web-like enough 
for me to find out who Charles McCathy-Neville is, and eikeon.com/foaf 
has worked out that people write my names in ways that most systems 
can't use to allege a match, but it's language logic isn't as good for 

> next page, or play and speaker or sound are other well known and 
> popular graphics....
> 'Link to us' graphics are similar to trade marks, and favicons. They 
> publicise a product, however because they are intended to be used by 
> others, they will generally be royalty free. This means there is the 
> possibility of their being used in other contexts. Furthermore like 
> those very irritating logos on TV stations they may become ubiquitous, 
> like toilet signs. This is where they become extremely useful. SVG 
> would be great ~:"

They may be royalty free, but are not always condition free. EasyJet is 
a site that provides a handful of link graphics - 
http://www.easyjet.com/en/contact/link.html - where they specify the 
condition that such links are made to their home page (I suppose they 
would like this link to be to their home page too, but I would be 
surprised if they cut me off as a customer for not doing so).

> http://www.gnote.org/svg-images/Diner.svg may not be as well known, 
> but may have more interesting references?
> For me 'history' might be a pyramid, alternatively 
> http://www.symbolworld.org/images/learning/history.jpg is a ~17th 
> century ship.
> So this could be a confusing morass, or just conceivably a consensus 
> could arise.
> Abstract topics are extremely difficult to illustrate well.
> They need a broad and speedy evaluation, which the web is eminently 
> suited to provide.

There are some interesting issues here about how language is created 
and used. In general it happens because people use the same thing - 
whether that's because it came built in to their software, or because 
their friend invented a word they like. I expect similar approaches to 
determine whether graphics are good on the Web for linking, but some 
careful study and thinking is required to work out what are the factors 
that determine success and failure so we can say something better than 
"If you are McDonalds, you can make it easy to find you on the Web...".


Received on Monday, 19 January 2004 11:55:23 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 13 October 2015 16:21:27 UTC