The hard part of feature design…
…what to leave out. Dan Bricklin
1969/1970 – Dan Bricklin and Bob Frankston both worked on the MULTICS project, the time-sharing system, at MIT On leaving they continued in different aspects of the developing PC business, but they had identified a desire to start a small business together.
Bricklin decided to study for an MBA at the Harvard Business School believing this to be a sound preparation for founding a business. While there he conceived the electronic spreadsheet. Watching a professor develop a financial model on a blackboard, he noted that when parameters were erased or changed it was necessary to follow through and modify data across the rest of the model.
Douglas Engelbart of ARC had recently demonstrated his mouse at Harvard and maybe as a result of this Bricklin considered equipping his TI calculator with a mouse to move between cells of a financial model.
The term spreadsheet already existed. ‘The Dictionary for Accountants’ by Eric L Kohler first published in 1952 described a spreadsheet as a matrix of columns and rows of numbers completed manually. Richard Mattessich pioneered computer spreadsheets in the early 1960s. In ‘Simulation of the Firm through a Budget Computer Program’ he included illustrations and even a FORTRAN program.
When Bricklin outlined his thoughts to various professors he received limited interest. One of them introduced him to Dan Fylstra, a student a year ahead of him. Before attending Harvard Fylstra studied computing at MIT and worked as an associate editor at Byte before becoming the founding editor of Computer Dealer magazine.
1977 – Fylstra realised there was a need for an operation to support software authors through the development process and to forge relationship with new computer stores and traditional retailers.
Fylstra founded Personal Software, one of the earliest PC software publishing operations. He convinced Bricklin to develop his software for Apple rather than a DEC minicomputer. He even loaned Bricklin an Apple to advance the development of his concept.
1979 – Bricklin asked Frankston, by now an experienced computer consultant, to help develop his ideas. They formed Software Arts Inc and evolved Calcu-Ledger, the spreadsheet now renamed VisiCalc, for visible calculator, to automate the frustrating manual cross-casts. By using a software interpreter the arithmetic would be assured.
The early limitations of available PCs, their very simple screens and even simpler printers necessitated a whole series of innovations. One of these was the referencing of columns as letters and rows as numbers. But perhaps more significant was the need to develop windows or panes to view different parts of the spreadsheet simultaneously on the inadequate screens.
They tried to fit the software code for VisiCalc into the less expensive 16K Apple II but found it needed the larger 32K version. Its great attraction was that anyone, even without programming experience, could grasp the processes speedily and be writing and operating ‘programs’ within seconds. The first advertisement for VisiCalc appeared in the May 1979 issue of Byte Magazine
Fylstra represented VisiCalc with a 37.5% royalty on direct sales and 50% on any other equipment manufacture business. VisiCalc was formally launched at the 1979 West Coast Computer Faire and the National Computer Conference in New York. At $99.50 per copy, VisiCalc took off and took Apple with it. This was the first killer-app; it sold over 700,000 copies in six years.