W3C home > Mailing lists > Public > www-tag@w3.org > April 2010

Re: Deep linking barriers in the UK: The Royal Mail

From: Paul Libbrecht <paul@activemath.org>
Date: Mon, 26 Apr 2010 06:17:56 +0200
Cc: John Kemp <john@jkemp.net>, Tim Berners-Lee <timbl@w3.org>, Henry S. Thompson <ht@inf.ed.ac.uk>, www-tag@w3.org
Message-Id: <CF7556B9-0BCC-4558-934C-E13835E3222D@activemath.org>
To: Pat Hayes <phayes@ihmc.us>

Le 25-avr.-10 à 06:59, Pat Hayes a écrit :

> This seems rather disingenuous to me. Sure, a link is a link. But I  
> read the Royal Mail request (as quoted by Dan C.) simply a plea, or  
> warning, that their 'internal' links are labile and liable to be  
> changed without warning, and that therefore it would be unwise to  
> use them as if they were stable links. In contrast to the links to  
> the 'top' of their internal hierarchies, which are promised (or  
> assumed) to be more stable.

No debate. A policy to advise on link stability is a useful thing.

I believe the royalmail case and several others use the word "you may  
not link" and this, to me, is the too hard part.

Useful would be: "please do not link to parts a,b,c, these links are  
likely to change". For example, it still makes sense to post to a  
public forum such a temporary link to discuss the page or links to it,  
purposely knowing that the link is dangerous. In the current  
expression, this is just prohibited which is kind of non-sense.

I agree the royal-mail case is better than other conditions of use  
which prohibit any link other than the top page because they specify  
an algorithm.

Le 25-avr.-10 à 08:51, Tex Texin a écrit :
> Deeplinking is not about depth. It refers to content that is a  
> component
> that supports a page or other larger entity. Owners may object to  
> direct
> linking to these components if context, advertisements, or license or
> copyright information is lost, if the reference is for an unintended
> purpose, or because it drives up hosting costs without returning  
> value.

This is also understandable but it cannot be fully enforced, in  
particular in communications and bookmarks systems which both tend  
come in the public web, not to count all the "automated links  
production" such as histories, teleconferences log of shared  
browsing, ... It probably is also an explanation of why this issue is  
called deep-linking. Another name would still be desirable, such as  
"restricted linking".

The problem here is the legal terms are simplistic instead of being  
advises to users supported, maybe, by technical means (as dumb as  
using http_referers). The fact that such conditions of use make people  
fear and remove their web-content is, to me, a vicious effect of such  
a simplistic legal term.

Maybe the solution to this issue would be to express recommendations  
of best practice for policy makers?
And maybe also to have supporting systems that can services me quick  
"yes, that link is safe", "someone needs to read the conditions of  

Received on Monday, 26 April 2010 04:18:32 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:33:06 UTC