Skip to content

Don’t believe everything you hear

Back in 1983, VisiCalc was unceremoniously defeated by Lotus 1-2-3 in about 15 minutes. There were many reasons for this. To be sure, VisiCalc was a seminal product (and so was 1-2-3), but development had been slow for a while, and the users VisiCalc had enabled needed much more power.

As Product Manager of SuperCalc, it was my job to keep abreast of market trends. It was clear that 1-2-3’s early success was driven by users with jobs in finance. They needed larger spreadsheets (VisiCalc’s spreadsheet was tiny), and they needed financial functions. Oh, and everyone said these finance types really needed graphics.

VisiCalc and SuperCalc were missing these capabilities. Adding them to SC didn’t look like it would be too difficult – well let me correct that: making a bigger spreadsheet wasn’t too bad now that the IBM PC could have lots of memory (not by today’s standards!). Internal Rate of Return (IRR) was more difficult, and in part it led to us adding iterative calculations.

Graphics – well that was an entirely different story. We simply couldn’t believe all the hoopla about 1-2-3’s graphics, because they were really primitive. I called their pie chart a lemon chart because it was oblong. And the user had to effectively run a separate program to produce graphics, which involved swapping floppy disks.

At Sorcim, we had been working on SuperChart for a while, and we decided to kill that as a separate project and use its graphics engine in SuperCalc. That wasn’t easy, since SC was written in a non-relocatable assembler and the graphics package was written in C. Martin Herbach came to our rescue again – he figured out a way to load the C program into the assembler space, which everyone thought was impossible. And our graphics were really beautiful (and didn’t require a disk swap). In fact, in one major competition, we came in third among all dedicated graphics programs.

While we were working on this project, one of the developers came to me and said there was something wrong with the cable HP supplied with their new low-cost pen plotter (HP7475?). Just for background, laser printers didn’t yet exist for PCs, and HP had just introduced its first ink jet, the ThinkJet, which did low-resolution character printing at a startling 30 characters per second. You either did graphics on a pen plotter or you used a daisy wheel printer (anybody remember them?) and printed a gazillion periods, moving the print head a little each time. We wore out a lot of print heads testing that…

It turned out the cable HP sold would only work in BASIC – if you booted your computer into MS-DOS, you simply couldn’t print to the plotter. He showed me how the cable should have been wired, and I contacted the product manager for the pen plotter and sent a cabling diagram. His response was “I wondered why we were getting all those support calls.”

The common belief was that you just had to support graphics if you wanted to sell a spreadsheet. But we knew the nasty little secret: all those finance types weren’t printing those graphs, and they probably weren’t even using them.

What’s the moral of the story? “Don’t believe everything you hear.”

Posted in Back in the day.

Tagged with , .