SuperCalc

~ the eulogy ~

supercalc is the spreadsheet nobody chose. that is not entirely fair to it, but it is close enough to be the reason this one is worth burying. most of the people who ran supercalc got it because it was already inside the box their computer came in, and they learned it because it was what was sitting there.

it stopped shipping in 1992, when i was eight and six years off my first modem. the first spreadsheet i ever opened was excel, and it did not occur to me that it had ever been a choice either.

supercalc got where it got by accident, because a licensing deal fell through. in late 1980 adam osborne was building a portable computer and needed a spreadsheet to put in it. he went after visicalc, which had invented the category on the apple ii the year before, and got nowhere. bill godbout introduced him to richard frank of sorcim at a poker party. sorcim was a months-old company of ex control data engineers in sunnyvale, and it agreed to write a visicalc competitor in time for the west coast computer faire in april 1981.

they made the date. supercalc went into the osborne 1 as part of a software bundle that also carried wordstar, and osborne sold the machine on the arithmetic that the software alone was worth more than the price of the computer around it. that made supercalc the spreadsheet a very large number of people owned without ever deciding to. it became the default for cp/m mostly because visicalc was not there, and it was on ms-dos by 1982.

i came up on bundled software too, twenty years later and at the other end of it. the modem came with a disk. the isp came with a disk. what shipped in the box was not something you evaluated, it was the thing you had, and you learned it because it was there. that part of supercalc's story i understand without having touched the product.

~ the dossier ~

~ the number it could do ~

supercalc was one of the first spreadsheets that could resolve a circular reference by iteration. a circular reference is when cell a depends on cell b while cell b depends back on cell a, which most spreadsheets treat as an error and refuse outright. supercalc ran the sheet over and over until the numbers stopped moving. that is how you model anything that feeds back into itself, interest accruing on a balance that the interest is being added to, and it is not a small thing to have working in 1982. excel did not ship iterative calculation for more than ten years after.

~ the four percent ~

computer associates bought sorcim in the mid-eighties. the wikipedia article on supercalc says 1985 and the article on sorcim says spring 1984 with one reference giving june, so the sources do not agree and this grave is not going to pretend they do. either way supercalc became ca's first personal computer product and it kept coming out. release 4 landed in july 1986 and release 5 in 1989. version 5.1 followed in 1991, with a copyright line you can still read on its startup screen.

the sales did not follow the releases. release 3 claimed a million users by march 1987. that same year an estimate of the fortune 1000 market for pc financial analysis put lotus at eighty-five percent and computer associates at four. sorcim's own history says the company never again reached prominence after the acquisition, which is a polite way to describe a product that kept shipping into a market that had stopped asking for it.

~ how it ended ~

there is no shutdown notice for supercalc. release 5.5 came out in 1992 and nothing came after. a computing history museum in cambridge holds a 5.5 box dated to that year, which is as close to a death certificate as this product has ever had. by then windows had arrived and excel was on it. a dos spreadsheet from a company whose real business was mainframe systems software had nothing left to argue with.

~ what it proved ~

the thing supercalc proved is still true and nobody says it out loud. the spreadsheet that wins is not the best one. it is the one already on the machine when you sit down. supercalc won that way for four years, then lost the same way for eight more.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »