Three Minutes: Godfathers of the Spreadsheet
As VisiCalc turns 25, originators Dan Bricklin, Bob Frankston, and Dan Fylstra recall when people wondered what a spreadsheet was.
Eric Bender, special to PC World
Three Minutes: VisiCalc Co-Creator Bob Frankston
After writing VisiCalc, Bob Frankston went on to pioneer early e-mail at Lotus, pen computing applications at Slate, and "IP everywhere" networking at Microsoft. Now he'd like to reinvent the Internet.
PC World:What was your background before VisiCalc?
Bob Frankston:I'd been programming since 1963. A lot of years of programming expertise went into it. I'd worked on machines of all sizes. The biggest surprise was that I could write a small program.
PCW:Okay, how small was it?
Frankston:The goal was 16KB, with a 4K operating system. Actually, it took 32KB. Every byte counted. We dropped the interactive Help to stay inside 32KB. These days, you can have a 100MB Help system.
PCW:How do you describe your approach to non-programmers?
Frankston:It was like origami. Fold the paper one way and get an elephant, fold it another way and get something else. There were lots of ways of using the same mechanisms.
I guess VisiCalc was very hard to write; it took people a lot of time to clone it.
PCW:How much did it build on previous financial tools?
Frankston:The key breakthrough in VisiCalc is the grid, which actively reduced interactivity [compared to earlier financial modeling tools that depended on writing commands]. The reduction of features made it feasible.
Some of its features are still here. I think "/" is still a command key in Excel. [Frankston fires up his notebook to confirm this.]
PCW:What's it like now compared to the early days?
Frankston:I just jumped ship to the freedom of inventing things. The ship is now much taller, and jumping is farther, but there's still exciting stuff. And today's PC is so much bigger than the old mainframe.
But the PC has been around for a long time. It's pretty obsolete and we're stuck with junky systems. We've brought back the mainframe and it's not really what we want. You try to do something like look up a contact in Outlook, and you just wait for all the system stuff to load. People forget that slow kills. These lessons have to be relearned each time.
I'm actually far more interested in networking stuff in these days. The Internet is in pretty sorry shape. The word 'pathetic' comes to mind. You look at the Web and you see a whole traffic jam. I want to look at all these devices and how you connect them, what you do with them and how you can build things. We need more basic [and highly programmable] tools.
