- From: Olivier GENDRIN <olivier.gendrin@gmail.com>
- Date: Wed, 19 Aug 2009 14:51:31 +0200
- To: Julian Reschke <julian.reschke@gmx.de>
- Cc: "public-html@w3.org" <public-html@w3.org>
On Wed, Aug 19, 2009 at 1:45 PM, Julian Reschke<julian.reschke@gmx.de> wrote: > Hi, > > a few days ago, I came across a discussion about meta/@name=description and > how it's used by search engines to provide a summary > (<http://krijnhoetmer.nl/irc-logs/whatwg/20090817#l-607>). > > Contrary to what some people said, using it doesn't seem to be a waste of > time, as it *is* being used to generate a summary in the search results. > And, at least with the documents I am producing, it fails to generate a good > one automatically. > > A drawback is that it uses hidden metadata, and, even worse, is likely to > duplicate information from the page. It seems what's needed is to markup the > page so that a consumer can actually locate the description inside the page. > > One way to do that would be a link relation, for instance: > > <link rel="description" href="#desc"/> > > ... > > <div id="desc"> > <h2>Abstract</h2> > <p> > ...description text... > </p> > </div> > > > Feedback appreciated, OTOH, a robot could ask a server to send only the <head> part of a webpage to get these infos. -- Olivier G. http://identi.ca/lespacedunmatin http://www.lespacedunmatin.info/blog/
Received on Wednesday, 19 August 2009 12:52:31 UTC