Another abandoned server code base... this is kind of an ancestor of taskrambler.
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
 

269 lines
10 KiB

<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta name="generator" content=
"HTML Tidy for Mac OS X (vers 31 October 2006 - Apple Inc. build 13), see www.w3.org" />
<title>
The Implications of Links -- Axioms of Web architecture
</title>
<link href="di.css" rel="stylesheet" type="text/css" />
<meta http-equiv="Content-Type" content="text/html" />
</head>
<body bgcolor="#DDFFDD" text="#000000">
<address>
Tim Berners-Lee
<p>
Date: April 1997
</p>
<p>
Status: personal view only. Editing status: first draft.
</p>
</address>
<p>
<a href="./">Up to Design Issues</a>
</p>
<h3>
Commentary on Web Architecture
</h3>
<hr />
<h1>
Links and Law
</h1>
<h3>
<i>Preface</i>
</h3>
<p>
This personal note I have put into the set of web
architectural notes as it expresses fundamental
understandings upon which the practical use and power of the
web rest.
</p>
<p>
The questions addressed are about the relationship of the
hypertext forms of <i>linked</i> and <i>embedded</i> material
to the social concepts involved such as attribution,
endorsement, and ownership of information.
</p>
<p>
Links in hypertext are new in that they can be followed
automatically, but the concepts of reference and inclusion of
material predate paper. There should not therefore be much
confusion about what links imply, but as there have been some
strange suggestions recently which would seriously damage the
web, I write this note.
</p>
<h3>
<a name="Abstract" id="Abstract">Abstract</a>
</h3>
<p>
Normal hypertext links do not of themselves imply that the
document linked to is part of, is endorsed by, or endorses,
or has related ownership or distribution terms as the
document linked from. However, embedding material by
reference (sometimes called an embedding form of hypertext
link) causes the embedded material to become a part of the
embedding document.
</p>
<h2>
<a name="sorts" id="sorts">Two sorts of link</a>
</h2>
<p>
Basic HTML has three ways of linking to other material on the
web: the hypertext link from an anchor (HTML "A" element),
the general link with no specific source anchor within the
document (HTML "LINK" element) and embedded objects and
images (IMG and OBJECT). Let's call A and LINK
"<b>normal</b>" links as they are visible to the user as a
traversal between two documents. We'll call the thing between
a document and an embedded image or object or subdocument
"<b>embedding</b>" links.
</p>
<p>
This distinction is an old one in hypertext. Some systems
such Peter Brown's original "Guide" worked only by expanding
links inline, and some (such as HTML before the IMG tag was
introduced) worked only with normal links.
</p>
<h2>
<a name="Normal" id="Normal">Normal Links</a>
</h2>
<table border="1" cellpadding="2">
<tbody>
<tr>
<td>
<b>The intention in the design of the web was that
normal links should simply be references, with no
implied meaning.</b>
</td>
</tr>
</tbody>
</table>
<p>
A normal hypertext link does NOT necessarily imply that
</p>
<ul>
<li>One document endorses the other; or that
</li>
<li>One document is created by the same person as the other,
or that
</li>
<li>One document is to be considered part of another.
</li>
</ul>
<p>
Typically when the user of a graphical window-oriented Web
browser follows a normal link, a new window is created and
the linked document is displayed in it, or the old document
is deleted from its window and the linked document displayed
in its place. The window system has a user interface metaphor
that things in different windows are different objects.
</p>
<h3>
Meaning in content
</h3>
<p>
So the existence of the link itself does not carry meaning.
Of course the contents of the linking document can carry
meaning, and often does. So, if one writes "See Fred's web
pages (link) which are way cool" that is clearly some kind of
endorsement. If one writes "We go into this in more detail on
our sales brochure (link)" there is an implication of common
authorship. If one writes "Fred's message (link) was written
out of malice and is a downright lie" one is denigrating
(possibly libellously) the linked document. So the content of
hypertext documents carry meaning often about the linked
document, and one should be responsible about this. In fact,
clarifying the relative status of the linked document is
often helpful to the reader.
</p>
<h2>
<a name="Embedded" id="Embedded">Embedded Material</a>
</h2>
<p>
The relationship between a document and an image embedded in
that document is quite different from normal link. (In some
designs it is still refered to as a sort of link).
</p>
<table border="1" cellpadding="2">
<tbody>
<tr>
<td>
<b>Images, embedded objects, and background sounds and
images are by default to be considered part of the
document.</b>
</td>
</tr>
</tbody>
</table>
<p>
If I say, "To understand this you only have to read this
article", or "This is the agreement between us", I am talking
about a particular document. It is important that we have a
clear picture of what is part of that document and what
isn't. Embedded images clearly are part of the embedding
document. The author of a document has responsibility for the
content, even if the images he or she includes are from
another web site.
</p>
<p>
(There are issues of expectations to be set about
availability and security from corruption of remote material,
but I do not address these here. Here I just emphasize is
that embedded images should be considered part of a document,
but documents connected by a normal link should be regarded
as separate documents.)
</p>
<p>
We compose documents out of parts, and the finished work
comprises contributions from the parts and also from the
arrangement. It is very important that we can include remote
parts by reference without having to make a separate local
copy. When an embedded image (or sound) is included by
reference to its original address (URI) this allows an
inquirer to know that address, and hence know the current
version of the image. It allows the owner of the image to to
a certain extent to know and possibly to control who has
access to that image. Also I expect in that in the future it
will allow one to find out the owner and licence terms for
distribution of that image, which is important for
intellectual property rights to be respected on the Web.
</p>
<h4>
Explict distinction
</h4>
<p>
Advertising provides an exception to this rule: a case in
which the embedded image is <b>not</b> part of the document.
&nbsp;At risk of making ittoo easy for users to turn
&nbsp;off advertizing, it would be ideal if the distinction
were make in the markup between embeeded information which is
or is not part of the document. &nbsp;This would allow, for
example, a border to be places around an advertizement to
allow the user to realize that it does not come from the same
source as the text. &nbsp;I personally feel that this would
be an important step forward in the integrity &nbsp;of the
web. A flag like
</p>
<pre>
&lt;IMG src="banner-ad.gif" foreign&gt;
</pre>
<p>
would be fine.
</p>
<h2>
<a name="User" id="User">User Interface</a>
</h2>
<p>
When Web documents are presented to people, most current
browsers (1997) make a clear distinction between embedded
images, which are presented in the same window as the
embedding document at the same time, and linked documents
which never are. The window system's concept of a "Window" is
used to convey when things are part of the same document. It
is important for many reasons, some of which were mentioned
above, that user interfaces continue to make this
distinction.
</p>
<h4>
Frames
</h4>
<p>
The "frames" of HTML unfortunately provide an interface which
is less clear. The parts of the document do appear with the
same window, but because within a single frame (subsection of
a window) one can follow hypertext links replacing content
with a separate document, it is easy to create the impression
that the owner of the surrounding frames is in fact
responsible for the defining document. It is possible that
work by the HTML community can produce explict markup (such
as the "foreign" flag above) for conveying, when frames are
used, which parts of the screen are considered to be the same
document. In the mean time, it is appropriate for content
providers so make efforts to ensure by the design of (and/or
statements on) their web pages that users are not left with
the illusion that information within an embedded frame is
part of their document when it is really not.
</p>
<p>
<i>Next: Some dangerous <a href="LinkMyths.html"><b>Myths</b>
about Links</a></i>
</p>
<hr />
<p>
<i>A reminder that this is personal opinion, not related to
W3C or MIT policy. I reserve the right to rephrase this if
misunderstandings occur, as its always difficult to express
this sort of thing to a mixed and varied audience.</i>
</p>
<hr />
<p>
<a href="Metadata.html">Next: &nbsp;Metadata architecture</a>
</p>
<p>
<a href="Overview.html">Up to Design Issues</a>
</p>
<p>
<a href="../People/Berners-Lee">Tim BL</a>
</p>
</body>
</html>