Davide Musella (
Thu, 9 May 1996 15:05:46 -0100 (GMT)

Date: Thu, 9 May 1996 15:05:46 -0100 (GMT)
From: Davide Musella <>
To: Charles Peyton Taylor <>
Subject: Re: META
In-Reply-To: <>
Message-Id: <Pine.SOL.3.91.960509145038.1432B-100000@jargo>

On Wed, 8 May 1996, Charles Peyton Taylor wrote:

> Hey, look: a Message that *isn't* about Wilbur :)
> Recently I installed a search engine and told people
> they could put keywords in their documents using the
> <meta> tag.  They told me they needed documentation
> for this.  So I began to write up documentation for 
> <meta> and <link> as well.
Yop.. there is my draft about tag META ... 

> and guidelines at Alta Vista's web site, I discovered
> that instead of HTTP-equiv, the search engines were 
> using NAME to designate keywords.  
> I'm not sure what to write in the documentation now.
> I'm thinking about dropping any mention of "http-equiv"
> in favor of "name", but I want to jive with the specs.
> Are any http servers reading (and presumeably, displaying 
> in HEAD requests) the HTTP-equiv sections of HTML files?
you have centered the problem!! It doesn't exist any HTTP-server that handle
Tag META... (maybe the WN server does it only wiht the keyowrds)
I'm working on a patch for Apache 1.0 but it will be ready in 4-5 weeks...
Try to read this paper about META info in HTML:

I don't concern with you when you say :"I'm thinking about dropping any 
mention of "http-equiv in favor of "name""
HTTP-EQUIV attribute can be really usefull in a cataloguing-context
there aren't any semantic difference between http-equiv and name
so your robot can use both if you retrive the entire doc.
But think if you can have the same info retriving only the HEAD HTTP part
 of a doc containing also the METATAG info. You and the NET could save a 
lot of band-width and time!!!!


Davide Musella
Institute for Multimedia Technologies, National Research Council, Milan, ITALY
tel. +39.(0)2.70643271