6/143.4.2.3 Searching
Due to the enormous quantity of information available in distributed information systems, searching is very important, and often the only way to retrieve information. Searching has always been built into the Hyperwave server, in contrast to Gopher and WWW where searching is an add-on, and thus can differ from server to server.
There are two ways of searching in Hyperwave. First, you can search for attributes like Title, Keywords, Author, and CreationTime. For instance, you can search for all documents whose title includes a certain string and which have a creation time greater than a specific day. This is very useful retrieving documents which are not out of date. Second, indexed documents (text, HTML, PostScript in the future) are accessible via a full text search. When such a document is inserted to the server, an index is created which speeds up searching.
As mentioned in the paragraph about collections, you can limit the search to a subset of the collection hierarchy. Of course you can perform the search on the whole server, too. If collections contain documents from other servers, the servers will be queried in parallel.