- From: John Franks <john@math.nwu.edu>
- Date: Fri, 19 Jul 1996 15:08:51 -0500 (CDT)
- To: "David W. Morris" <dwm@shell.portal.com>
- Cc: jg@zorch.w3.org, Paul Leach <paulle@microsoft.com>, 'Gisle Aas' <aas@a.sn.no>, "'Roy T. Fielding'" <fielding@liege.ICS.UCI.EDU>, "'http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com'" <http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com>, http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
On Fri, 19 Jul 1996, David W. Morris wrote: > > On Fri, 19 Jul 1996, John Franks wrote: > > > In fact, I would find an argument to make it mandatory persuasive. > > I wouldn't be happy about any change to http which made analysis of the > content of a file mandatory. We should not demand that servers > analyze HTML. > > And in current practice, my experience as the designer of a search > tool is that titles are used to provide meaningful captions on browser > screens not sumarize the document content which would be useful for > search engines. > > I haven't looked at the mime spec so it may already agree with me ... but > if HTTP were to include a header relating to title or document content > it should be clearly a document description which might be provided by > a server and might (NOT mandatory) be provided by content analysis or > meta data stored by the server external to the document. It should not > be directly specified as being a particular element of a particular > content type. > I wouldn't really argue that it be mandatory. I was venting my pique that it was removed (even from an appendix) without discussion. I agree with everything you say and I hope that a header with this functionality can make its way back into HTTP/1.2. The Title header (its name is not important) as implemented in the WN server has many of the aspects you suggest. It is provided from meta data stored external to the document which can be set by the maintainer, but is by default automatically extracted from HTML documents if not set (this extraction is done once when the meta data is "compiled" not every time the document is served). Non HTML documents which have no title set get the filename by default. I would really like to see not only Title, but also Keywords. Abstract would be nice too, but perhaps that could be done with "Link". John Franks Dept of Math. Northwestern University john@math.nwu.edu
Received on Friday, 19 July 1996 13:15:17 UTC