W3C home > Mailing lists > Public > www-html@w3.org > July 2004


From: Jukka K. Korpela <jkorpela@cs.tut.fi>
Date: Wed, 28 Jul 2004 20:42:06 +0300 (EEST)
To: www-html@w3.org
Message-ID: <Pine.GSO.4.58.0407282026410.28284@korppi.cs.tut.fi>

On Wed, 28 Jul 2004, Anne van Kesteren wrote:

> Someone might have published a PNG and you have the text on your page
> and markup like:
>   <blockquote cite="http://example.org/reports/2004/content.png"
>               citetype="image/png"
>   >
>    ...
>   </blockquote>

But what use would the citetype="..." attribute have? There are some
imaginable uses, like helping the browser tell the user that the reference
probably cannot be followed - if the browser is not capable of rendering
PNG for example. This in turn might be wrong information. The URL
could in fact refer to an image file in any format, or any other data, and
this could even be changed without notice. For example, the author of the
cited document might decide to change the image format from PNG to GIF
or KNIF (Kewl New Image Format). This might cause some practical problems
if browsers mistakenly stare at the .png suffic or incorrectly give the
citetype attribute preference over HTTP headers, but I think we need to
hope that XHTML 2.0 user agents won't do such things.

> Perhaps:
>   <blockquote>
>    <p>...</p>
>    <p>...</p>
>    <cite href="http://example.org/poems/34" hreflang="nl">...</cite>
>   </blockquote>
> ... but I think this is rather ugly since it mixes both inline (CITE)
> and block level (P) elements within the same container. And while it is
> mostly desired to be rendered within BLOCKQUOTE, it isn't part of the quote.

The <blockquote> element is too unstructured. A reference to the source of
a quotation is an inherent part of quoting (and normally a legal
requirement, by copyright legislation) and this should be reflected in the
content model of <blockquote>. But if <cite> is used for this purpose,
then the legacy <cite> element probably needs to be renamed, since legacy
<cite> can be used freely for things like book titles, without implying
that the books are quoted.

> With QUOTE, as in inline quotations, this is probably easier, since you
> don't want to include the resource within the element there. Although it
> would be nice if there was a way to connect the quote with the resource.

Indeed - both for structural reasons (e.g., to make it possible to
generate a list of quoted sources) and for presentation, since the
specification of the source of a quotation should appear (even in the
absence of any CSS) in a style different from the quoted text, yet as
associated with it.

> So you consider HREFLANG as not useful? And HREFTYPE? And TYPE of
> embedding attributes module? Since TYPE only says something about the
> linked resource (SRC).

Yes, I think they have too little potential use to justify the added
complexity, and they would introduce pitfalls - authors would have
difficulties in figuring out what they really mean.

Jukka "Yucca" Korpela, http://www.cs.tut.fi/~jkorpela/
Received on Wednesday, 28 July 2004 13:42:17 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:06:08 UTC