Computing's first 'killer app' changed everything
bbc.co.uk
bbc.co.uk
If you've never spreadsheeted on paper you may not realize the glory of even the most primitive spreadsheet app. I used to use erasers until my fingers went numb or white-out by the quart.
Up hill both ways. You young whippersnappers have no idea how good you have it.
I sat with the prof to show him my first attempt, with a half dozen formulas in it. He looked over my shoulder and showed me that every. fucking. formula. was wrong. I felt humiliated. But we fixed them as we went, and watching the changes flow into the cells was magic. My force had been multiplied.
The skills from VisiCalc have transferred to other spreadsheets over the years very well. It's amazing how much Mr. Bricklin got right. I wonder how much was his brilliance and how much was the inevitability of the design.
He said "well... ok, standards have gone down over the years", then he looked in the amphitheater where he had a section of chairs that had missing desks and finally said: " though I really shouldn't, because you're all engineers, and just look what engineers that make "small" errors in calculations" accomplish".
Probably I could not have become an engineer 100 years ago.
Whatever existed prior to VisiCalc, was not similar to the spreadsheet as we know it and define it today. It's a path that VisiCalc blazed, to universal credit by everyone that was in the industry at the time and helped pioneer software.
It wasn't the visual paradigm of a tabular presentation. It wasn't the two-formed concept of cells having both values and formulas, or that the formulas could be stated in terms of other cell's values.
It wasn't that you could do what accountants had done for centuries using the same tools. It wasn't that the same type of calculations could be done using a computer.
So what we're left is that he took a visual concept and implemented it on a screen. I don't think that qualifies as "invention", myself. It is re-implementation, copying, or something that doesn't involve a conceptual leap.
He was in the right place at the right time to spot the technological opportunity, and we are all better for that.
It would have had the same effect if only one formula was left intentionally wrong :)
Couple that with the fact that you can literally use a Google Sheet as the database backend for small apps, and I think it's a good case that spreadsheets are still killer.
More than any of the iterations of Windows, Excel is Microsoft's great contribution to software, and I say that as an old-school MS hater.
I think people too often forget that "useful" is the key ingredient in killer software.
Another fun bit of his talk, about the max digits in VisiCalc:
The number of digits that we stored for precision, it was stored as binary-coded decimal, just like a calculator, it was not stored as a binary number. It was stored as the digits, so that it would be exactly the same that you would see on your calculator, without funny round-off errors and stuff like that.
But we had to figure out what the maximum number of digits would be. And we decided, let’s - ha-ha. We thought, here us, two little programmers trying to figure out if anyone would use it. Let’s take the budget of the United States of America, in dollars. That’s the maximum, at the time. So that’s the maximum number of digits we had. Sure enough, within a few years, people were using it, according to the Wall Street Journal, for calculating the budget of the United States of America, but only in dollars, not in pennies, because that’s all it could do without going to scientific notation.
[1] https://scs.hosted.panopto.com/Panopto/Pages/Viewer.aspx?id=...
UIs typically recapitulate neoteny: from the beginning, competitive spreadsheets needed to be UI-compatible with Visicalc (even today Excel accepts most Visicalc controls, and has, I have been told by Frankston+, VisiCalc-compatible bugs). Notably, spreadsheets or spreadsheet-like interfaces that don't confirm people's assumptions struggle in the market (e.g. Apple Numbers).
+ Bob Frankston who was the implementor of Visicalc, and later worked at Microsoft, though not on Excel. You can thank him that Windows had TCP rather than some proprietary system...for the same path-dependency / network reasons.
>the Jennifer Unit, an earpiece that directs warehouse pickers to collect products by breaking down instructions into the most mindless, idiot-proof steps.
That makes me think of the story of Manna, which predicted something like this. http://marshallbrain.com/manna1.htm
I have to disagree with Marshall Brain's "solution" to the problem. When I read those final chapters, I couldn't help but think the Vertebrane system was subtly and insidiously even worse than Manna.
There are parallels in many areas of customer service for instance. Anywhere where you can do a better job of helping the customer with knowledgeable staff that don't have to be doing so much mundane stuff.
Whether the economy could have grown to 4x the size without productivity advances like spreadsheets is a related question.
Sumerian accountants seem to have invented cuineform, which evolved into the (AFAIK) earliest widespread writing system. Italian accountants (including Fibonacci's dad) first adopted Indo–Arabic numerals, because they made ledgering easier than roman numerals.. and opened the door to arab mathematics systems like algebra
For example, the Mark I mechanical computer[1] could, when pointed at a target, measure its distance, altitude and heading, its own ship's speed, pitch, and heading, current windspeed - and combine that with the chosen projectile type, weight, propellant type, and current temperature - all to compute a firing solution that had the best chance of sinking ships adorned with swastikas, or a rising sun.
Not a general-purpose computation machine, but work on these sorts of devices heavily overlapped with work on programmable computers.
[1] https://en.wikipedia.org/wiki/Mark_I_Fire_Control_Computer