Panel members

• Nick Arnett, Verity Inc., Internet Evangelist (Moderator)
• Darren Hardy, Netscape, lead developer, Catalog Server
• Louis Monier, Digital Equipment Corp., Alta Vista Technical Lead
• Tim Bray, Open Text Corp., Senior VP, Technology
• Ron Murray, Microsoft Corp., Technical Evangelist

Issues: Redundancy

• Many spiders are performing essentially the same tasks.
• There is no way to ask an HTTP server "what has changed?"
• Complete detection of identical documents is difficult (symbolic links, slightly altered versions, etc.).

Issues: Resource management

• There is no formalized way to discover new resources.
• The lack of meta-information standards (document collection name, icon, author, maintainer, etc.)
• Complete rejection of irrelevant documents (RCS files, etc.) is difficult.
• "Word spam" -- the inclusion of text designed to spoof a search engine into ranking a page as more relevant than it actually is.

Issues: Standards

• A query-based updating mechanism would require a standard query language.
• Incremental full-text index (to update a local index from a remote source; applies to servers and browsers) would require standard index format.