foreword
Apache Iceberg solves a problem that is nearly 50 years old, from a time when computing looked very different from what it does now. There was no internet. There were no laptops. We were just beginning to get a glimpse of what PCs might become. It was hard to know at the time, but the long reign of the mainframe was about to come, if not to an end, at least to a middle. Databases, which used to be centralized, were scattered all over the enterprise.
Once computing was cheap enough to put a useful machine on every desktop, it also became economically feasible to put a computer under every desktop. The emerging client–server architecture in the late ’80s and early ’90s took the data that was once centralized in the mainframe, where applications and reporting could share the same data, and scattered it in dozens or hundreds of departmental databases that were physically and logically dispersed throughout complex, global organizations. It was no longer possible to see the state of the increasingly computerized business in one place.