Till a few decades back, all computing was controlled through the central mainframes server. Multiple users could connect to the central host through unintelligent terminals which captured the keystrokes, sent the information to the host and displayed the text output. All the processing was done by the applications residing on the main central server.
Only large transaction-oriented applications were developed during that time. Business tasks such as accounts receivable, accounts payable, general ledger, credit account management and payroll that were repetitive and could be run as batch jobs were automated.
In these centralized computing models, the host provided both the data storage and processing power for the client systems. There was no support for graphical user interface or access to multiple databases from geographically dispersed sites.