The WorldWideWeb uses a universal addressing format to identify documents and anchors across different naming sub-schemes and network protocols.
Address Syntax
For most internet-based access methods, the naming format follows this structure:
scheme : // host.domain : port / path / path # anchor
[Image of the anatomy of a URL showing scheme, host, port, path, and anchor]- Anchor Identifier (#): An optional suffix that refers to a specific location within a document.
- Search Query (?): A suffix followed by keywords (separated by +) used to search an index.
- Relative Names: References between documents with similar names can be abbreviated, though this requires specific path representation.
Common Examples
file://cernvax.cern.ch/usr/lib/WWW/default.html#123A fully qualified file name referring to a document on a specific internet node and a specific anchor (#123) within it.
#gregRefers to the anchor "greg" within the current active document.
Addressing Sub-Schemes
Each scheme typically utilizes a different network protocol. While formats vary, most internet-based schemes share a common node and port structure.
- http
- Access to hypermedia via the HTTP protocol. Internal addressing is mapped directly onto a W3 path.
- file
- Access to local or remote files via browser-specific or gateway-mediated methods.
- news
- Access to Usenet news articles and newsgroups, typically via the NNTP protocol.
- gopher
- Access via the Gopher protocol, which uses distinct menus and text files rather than integrated hypertext.
- telnet
- Interactive sessions for existing online systems that have not yet been mapped onto the W3 space.
- wais
- Access using the WAIS adaptation of the Z39.50 protocol, often via a gateway.
Gateways and Interoperability
Systems not directly accessed by W3 servers can be reached through gateways. In these cases, the original document address is encoded within the HTTP address. This allows even simple clients to follow links through complex systems like X.500 name servers.