On Sat, Mar 12, 2005 at 03:56:16AM +0000, Miguel Mingo wrote:
> We have a search engine (external) that indexes the site's pages, in order
> to provide the search functionality inside the site. The way this search
> engine excludes part of the code is the main problem. It uses a special
> <meta> tag to do that.

> The meta tags inside body arent allowed, so the code from _all_ of out sites
> doesn't validate.
 
> We're looking for a solution, and the owners of the search engine have told
> us which tag we want to use instead of meta. I don't know which one choose
> (if there's one), but must be a valid one, and not deprecated (or near
> deprecated).

The two options that spring to mind are:

1. Continue using your non-standard data format, but strip out the
meta data from the body for user agents other then the search engine
indexer. You could do this based on the ip address the request comes
from, or by the User Agent header.

2. Use the class attribute on any element you like. For example:

<ul class="navigation noindex"> and <div id="mainContent" class="index">

The class attribute is designed to make elements members of custom
groups - and "content to be indexed by a search engine" and its
counterpart seem reasonable to me.

-- 
David Dorward                                      http://dorward.me.uk