W3C home > Mailing lists > Public > www-html@w3.org > January 2005

Re: Marking elements as 'volatile'

From: Denis Defreyne <amonre@amonre.org>
Date: Fri, 14 Jan 2005 23:17:59 +0100
Message-Id: <25FF1820-667A-11D9-B3FE-0011242E113E@amonre.org>
To: www-html@w3.org

> Checking for updates in this context is the job of the server, as a
> browser would have to read the whole page anyway, before  it could 
> check.
> As such such markup could be done in a server only namespace, or even
> private language.  Server side includes is an example of such a 
> language,

I do not get your point. Using a server-only namespace or even a 
private language is not an elegant solution at all. IMHO there is no 
easier way than to tag elements in XHTML files as 'volatile' and check 
whether other elements have been updated. Neither the server nor the 
script knows what elements on the page are 'important'; therefore they 
can't decide whether the page has been updated or not. The server 
should never make assumptions of whether elements are 'important' or 
not based on whether they're embedded using SSI or similar technique

> Or better yet, the exact opposite.  Some attribute or metadata 
> property which tags
> the part of the document which is the real content.
>
> The majority of web sites out there look like this:
>
>     header + sidebar + content + sidebar + footer
>
> Google only needs the content, so it would be cheaper to simply tag 
> the bit which
> you _do_ want indexed.
>
> Perhaps a known metadata property using the "property" attribute would 
> be enough
> here.  Or perhaps it really does deserve its own attribute.

I agree. Search engines should consider 'real' content more important, 
giving it more weight. This seems like a good way to enhance search 
engines.

Having a separate element, such as `<content>`, would make it quite a 
bit easier for parsers to get the actual content of the page. (The name 
of the `<content>` element should probably be changed; theoretically, 
anything is part of the content of an XHTML document, which could lead 
to confusion.)

Using an attribute seems more reasonable. The easiest way would 
probably be to have an `importance` attribute. (Again, I'm not sure 
about the name of this attribute.) `<div importance="important">` makes 
sense and is easy to read. Legal values would be `important`, 
`unimportant` and `unspecified` (default--99% of all web sites use this 
right now and will continue to). Search engines could give more weight 
to important elements--but not if the whole page is marked important.

Regards,

amon-re

--
mail: amonre@amonre.org
web: http://www.amonre.org/
Received on Friday, 14 January 2005 22:18:05 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 27 March 2012 18:16:01 GMT