[Prev][Next][Index][Thread]

Re: What happened to: More notes from meta/link session



Dan,

I asked either Christian or Stu to merge the two reports, so that there
would be just one report from each session, but neither had time to do the
merge.  So, I left it out of the report.  If someone wants to merge the
reports, it's fine with me to update the report
 - Mike

>
>I can't find these in the workshop report. Does anybody know
>what happened to them?
>
>------- Forwarded Message
>
>From: Dan Connolly <connolly@w3.org>
>To: schwartz@home.net
>Subject: More notes from meta/link session
>Date: Fri, 07 Jun 1996 03:57:56 -0400
>Sender: connolly@beach.w3.org
>
>
>These were transcribed by Christian Morgansen, who handed them
>to me cuz he couldn't find you and I didn't duck fast enough :-)
>
>They should be considered in the preparation of the workshop report.
>
>Sorry for the delay in forwarding them.
>
>=====================
>Group 2: Stu Wiebel
>	Christian Morgensen
>
>
>Problems addressed:
>	* Describing a page's content; i.e. providing metadata in HTML
>	* Packaging and describing collections of documents
>
>Related Standards:
>	* HTML 2.0 META and LINK tags
>	* REL/REV draft: 
>ftp://ftp.ietf.cnri.reston.va.us/internet-drafts/draft-ietf-html-relrev-00.txt
>
>
>Goal: avoid introducing new tags into HTML (since folks won't
>be able to create them with existing authoring tools)
>
>Scope: Proposing and infrastructure and suggesting some basic terms
>for content providers to use.
>
>Time Frame: less than three months to implement
>
>
>Describing Pages
>- ----------------
>
>Use the meta tag to provide meta data.
>
>Problem: NAMEs are a flat namespace
>
>Solution: partition NAMEs into sevarl spaces, using a dot separator
>	e.g.: <meta name="dc.author" content="...">
>
>where "dc" is the schema, "author" is the field.
>
>Users can provide data for sevaral schemes without conflict.
>
>Resolve schema names using the LINK tag.
>
>e.g.: <link rel="schema.dc" href="http://oclc.org/dublin_core">
>
>The page contains a description of the fields used in the scheme
>(human readable or machine readable)
>
>Suggestion: the field descriptions are named anchors, so that search
>engines can link to the specific field descr.
>
>The schema URL is the unique key the index uses to determine the exact
>schema. The OCLC has offered to host a schema registry. [note from Dan:
>using URLs obviates the need for this registry]
>
>Action item: settin up a dublin core description page
>	Dan Connolly to write proposal for inline marking of metadata
>	(out of scope for today)
>
>
>
>Packaging
>- ---------
>
>Use the LINK tag to create typed links. Use REL and REV type HREFs on
>existing and new content.
>
>in HEAD: <link rel="title-page" href="../index.html">
>in text: <a href="../index.html" rel="title-page">title page</a>
>
>Some suggested REL types:
>
>[Dan's note: see original list at:
>	http://www.w3.org/pub/WWW/MarkUp/Relationships ]
>
>parent: useful for navigating clusters [Dan recalls that lycos rep
>	said it could be used to heuristically determine the priority
>	of indexing a page]
>
>icon_for: for putting customized icons on results lists
>
>title-page: a papaer front page, TOC, etc.
>home-page: site's front page
>
>A subsidiary page LINK REL="title-page" to it table of contents.
>The title page has several LINK REV="title-page" to each of
>its constituent pages. It also has a LINK REL="icon-for" to a GIF
>of the book cover image.
>
>The <meta name="site-name" content="..."> allows home pages to
>publish an index-friendly name in addition to the usual <TITLE>.
>Allows indexer to use something more descriptive than site DNS name.
>
>Action item: educate content providers, tools builders
>	Document what indexers look for and use
>
>- ---
>
>syntax of description scheme:
>
>	<meta name="schemename.fieldname" content="free text ...">
>
>	<link rel="schema.schemaname" href="uri">
>
>	(alternate suggestion was to put the schemaname in
>	a separate attribute of LINK, but this would require
>	changing the authoring tools. The above syntax
>	has the advantage of working with existing tools
>	like Hotmetal)
>
>- -------------------
>
>Daniel W. Connolly        "We believe in the interconnectedness of all things"
>Research Scientist, MIT/W3C     PGP: EDF8 A8E4 F3BB 0F3C  FD1B 7BE0 716C FF21 
><connolly@w3.org>                  http://www.w3.org/pub/WWW/People/Connolly/
>
>------- End of Forwarded Message
>
>