The Web is not a library, but the random collective repository of the world's "digital print." (See March '97 Scientific American Special Report)

Order comes from classification. One can seek out "pre-ordered" environments (library or professional group repositories, to a lesser extent Yahoo) more suited to browsing, or use powerful search engines for "post-ordering", subject to personal search skills expressed within the possibilities of a given search engine environment.

Web search engines--there are many--are actually large indexes with references to millions of resources, with user query interfaces--details vary. The search engines use automatic tools, often referred to as worms, spiders, crawlers, and robots, to search websites worldwide, collect information, and store the information.

Search engine collection processes differ and accumulated data is always incomplete and partly outdated. Query results can vary greatly from one search engine to the next.

Automated tools have difficulty identifying the characteristics of a document, such as its "genre" (datasheet, poem, advertisement, resume, recipe), something a human indexer could do at a glance. Further, the Web lacks the standards that would facilitate automated indexing. There are no defined data fields for author, publication date, subject matter, etc., the attributes that make classical bibliographic databases (e.g. Compendex, INSPEC) so useful.

AltaVista, Infoseek, Open Text are a few examples of major search engines for "post-ordering", Yahoo is the major contender for a "pre-ordered" environment (uses AltaVista for searching and humans for classifying). On the other end of the search engine spectrum are smaller specialized ones.

You can also consult meta-guides such as AlphaSearch (a meta-meta site, i.e. an index of bibliographies, thus the highest level of entry into the topics it covers) and InterLinks.

In formulating Web queries, practice with modest-scale Boolean searches. The quality of your results will improve, but you will often not find precisely what you thought you wanted. But in rapidly changing technology situations, a certain fraction of your time should be spent in probing the larger context.

Meta-search engines, i.e. search engines that search across search engines (e.g. ProFusion) are increasingly used to cope with the increasing volume of online information. While powerful in interesting ways (e.g. Google) they introduce new unpredictabilities. See the June 1999 Scientific American Feature Report "Hypersearching the Web").