>> A better "solution" than playing this cat-and-mouse game with
>> servers is to provide a mechanism where they can genuinely
>> be appropriately found. Say, for instance, that the search
>> services would agree on come common terminology for meta-tags,
>> and where the meta-tags go.
>
>They have. See above. And META tags positioning is determined by the HTML
>standard. They *ALWAYS* go in the HEAD of a document.
Could you point me (and the list) to your reference for the names and
content descriptions for the standard META tags that you describe? I
wasn't aware that there was any agreement for which tags robots will
capture nor what goes into them.
Let's keep in mind that various information domains use quite different
ontologies and thus will use different sets of META tags. For example, the
financial community will use quite a different set from the biology
community. This is why we have librarians at the standards meetings;
they've been struggling with these issues for years (and sometimes become
testy when we techno-dweebs oversimplify!)
Searching on fields and keywords can be a great complement to full-text
searching, but one shouldn't imagine that it produces accuracy.
Fundamentally, a search engine is trying to answer the searcher's question.
To do so accurately, it has to fully understand both the questions and the
answers. Even humans have trouble with that and the notion of which
documents "answer" which search queries is a matter of opinion.
Finally, I'd suggest that META tags can be a competitive issue among the
search services. If AltaSeek manages to get its tag schema adopted more
widely than InfoVista, then AltaSeek's searches will be more accurate,
which could help it win in the market.
Nick