My little interview to Mr. Bob Frankston 18 May 1999
Me: I'd like to know
something behind the scene, who wrote the program, who got the
idea, where you met Bricklin?
Bob Frankston: Behind
the scene? Dan and I met at MIT in 1970 and we'd talked about
starting a company for years. I continued at MIT and did some consulting
including 6502 programming. Dan went to DEC where he developed screen-based
word-processing products and then went on the Harvard Business School where
we confronted tedious calculations. He developed the basic ideas including
the prototype. I did the programming since he was in school. This allowed
him to focus on design without the burden of the details of programming
which was an important part in balancing the product creativity against the
Me: Why you started to
develop a program like this?
BF: The reason for the program should be covered on Dan's site. But, looking
more generally, while there were some interactive editors, the idea of a
screen-based interactive and calculating piece of paper or surface wasn't
there. Given both of our experience with screen-based editing (such as Emacs
and Dan's work at DEC (now Compaq), we were, in a sense, prepared. For Dan,
dealing with non-recalculating spreadsheets, putting the interactivity
together with recalculation was an obvious match. At least, obvious by
hindsight. The availability of the personal computer at just this time made
such interactivity very affordable. After he wrote his prototype in Basic, I
started on a framework in 6502 Assembler. The framework set out the basic
architecture which we were able to evolve over time as we "learned by doing"
so we could rapidly improve the design as we were developing it.
Me: Which problems you
find with the hardware of that time?
BF: The biggest problem was the lack of memory. We wound up requiring 32
kilobytes (more than 1000 times smaller than today's PC's). Floppies were
just becoming available and we had a version that didn't require the floppy
(but chose not to market it). We also had to write everything including the
disk I/O and file system, make sure we polled the keyboards on a system that
didn't provide keyboard events. The challenge was to create the product
despite the many hardware limitations.
Me: Why the
program was discontinued?
BF: Discontinued? Complex business issue but the main factor was the
publisher's (Visicorp) mistaken belief that they could do a next generation
product on their own.
Obviously there's a lot more to say and, over time, I'm working with Dan to
provide more of this on his web site.