Netscape Compass Server Administrator's Guide

Glossary

child

A category that is a subcategory of another category.

classification rules

A set of rules used to assign resources to a category or to several categories.

Compass

Netscape's technology to improve user access to network resources.

Compass database

A searchable database of resource descriptions, usually generated by a robot.

Compass Server

A server that holds a database of resource descriptions gathered by robots, usually categorized. Users can search the resource descriptions or browse through the categories to locate particular resources.

crawler

See robot.

depth

The number of links followed from a site's starting point. When you define a site, you define the number of links the robot can follow away from that point. This limits the depth of its search.

distributed indexing

The process of assigning different robots to index different parts of the network. This reduces the load on each robot. A single Compass Server can then gather all the resource descriptions from all the different robots by importing resource descriptions from each.

document

A file on the network, most often a web page or word processing document, but also including text files, spreadsheets, and so on. This is a generic term for a resource indexed by the Compass Server.

domain

A group of web servers sharing a common DNS domain. For example, www.netscape.com and home.netscape.com both share the domain netscape.com.

enumeration

The phase of a robot's operation in which it seeks for resources, including extracting and following hypertext links.

extracting

The process of locating hypertext links in a document. Each extracted link is added to the URL pool for further processing.

filtering

The process of determining whether a document is part of a site that should be included in the index.

filters

Sets of rules that define particular types of resource. These filters are used by site definitions to define types of resources the robot should accept or ignore.

fresh start

Starting the robot from its starting points. A fresh start deletes the robot's state information, causing it to begin its next run from its initial state. This is the opposite of a restart.

generation

The phase of a robot's operation in which it produces a resource description for each resource discovered in the enumeration phase.

Harvest

A project of the Internet Research Task Force on Resource Discovery. For more information about Harvest, see http://harvest.cs.colorado.edu/harvest/.

HTTP

Hypertext Transfer Protocol. The method for exchanging information between web servers and clients.

idle

A state in which the robot is still running, but has processed all the URLs in its URL pool. In this state, the robot can still respond to status requests.

import agent

The process used for importing.

importing

The process of bringing new or updated resource descriptions from another database into a Compass Server.

index

A centralized, searchable database of resources or documents. Also called a catalog.

indexing

The process of providing a centralized, searchable database of resources. Also called cataloging.

metainformation

Information about a resource, such as the name of the author, the title of a document, the date of creation, and so on. The Compass Server robot uses metainformation as well as document contents when creating resource descriptions.

peer

A subcategory that has the same parent category as another.

RD

See resource description.

RDM

See resource description message.

resource

Any item on a network that can be identified by a Uniform Resource Locator (URL), such as a web page, a document, or an FTP directory. A resource is often referred to informally as a document.

resource description

Information about a particular resource in a standard format. The Compass database is a database of resource descriptions.

resource description message

A protocol used by Compass Servers to request and deliver resource descriptions. The contents of the RDM are in the SOIF format. RDM sits on top of the HTTP protocol.

restart

Starting the robot without deleting its state information, causing it to start running in the same state in which it previously stopped. This is the opposite of a fresh start.

robot

A small program that finds all the resources located in a specific portion of a network.

robots.txt

A file used by web servers to allow or deny access to portions of a web site.

rules

Logical tests applied to determine whether a condition is met. The robot uses rules in two places: as part of filters for determining types of content to index, and in classification rules to determine what category to assign to a resource.

schema

The formal definition of the contents of a resource description.

scoping

Placing restrictions on the resource descriptions imported by an import agent. The syntax used is the same as that for user searches.

Secure Sockets Layer

A software library establishing secure connection between two parties (client and server) used to implement HTTPS, the secure version of HTTP.

Server Manager

The administration server for a Netscape server product.

site

A location on the network where the robot goes to look for resources. You determine the address of the site and the kinds of document you want to index there in a site definition.

site definition

Constraints placed on where a robot can go to locate resources. Using site definitions, you can limit a robot to a particular server, a specified group of servers, or a domain. A site definition includes filters that describe what types of documents the robot should index from the site.

SOIF

See Summary Object Interchange Format.

spider