40 Years Later, Lessons From the Rise and Quick Decline of the First 'Killer App' (wsj.com) 77
It was the first killer app, the spark for Apple's early success and a trigger for the broader PC boom that vaulted Microsoft to its central position in business computing. And within a few years, it was tech-industry roadkill. From a report: The story of VisiCalc, a humble spreadsheet program that set the tech world ablaze 40 years ago, has reverberated through the industry and still influences the decisions of executives, engineers and investors. Its lessons include the power of simplicity and the difficulty of building a hypergrowth company in a hypergrowth industry. Indeed, its lessons have been so internalized by today's tech titans that they have significantly inoculated themselves against that sort of tumultuous, competitive dynamism -- aka disruption.
VisiCalc was unveiled on June 4, 1979, and shipped that October. Dan Bricklin first dreamed it up in a classroom at Harvard Business School -- the room now bears a plaque commemorating his idea -- and partnered with Bob Frankston, who coded VisiCalc and collaborated in its design. When users opened VisiCalc, they would see a character-based grid where numbers or text could be manipulated. It was handy for budgeting, financial projections, bookkeeping and making lists. Today it's instantly recognizable as a spreadsheet, as familiar to us as a blinking cursor, but at the time it was a novel idea that had to be experienced to be understood. Initially VisiCalc ran only on the Apple II, a then-revolutionary new personal computer and Apple's first major consumer product. While some Apple II models had just 4 kilobytes of RAM, VisiCalc demanded a whopping 32KB. (Even the cheapest of today's iPhones have tens of thousands of times as much RAM.) Further reading: VisiCalc Turns 25, Creators Interviewed (June, 2004).
VisiCalc was unveiled on June 4, 1979, and shipped that October. Dan Bricklin first dreamed it up in a classroom at Harvard Business School -- the room now bears a plaque commemorating his idea -- and partnered with Bob Frankston, who coded VisiCalc and collaborated in its design. When users opened VisiCalc, they would see a character-based grid where numbers or text could be manipulated. It was handy for budgeting, financial projections, bookkeeping and making lists. Today it's instantly recognizable as a spreadsheet, as familiar to us as a blinking cursor, but at the time it was a novel idea that had to be experienced to be understood. Initially VisiCalc ran only on the Apple II, a then-revolutionary new personal computer and Apple's first major consumer product. While some Apple II models had just 4 kilobytes of RAM, VisiCalc demanded a whopping 32KB. (Even the cheapest of today's iPhones have tens of thousands of times as much RAM.) Further reading: VisiCalc Turns 25, Creators Interviewed (June, 2004).
Re:I don't get it (Score:5, Interesting)
Fast forward two years and Lotus 123 hit the streets and basically cut Visicalc off at the knees.
Fast forward another nine years and Lotus chose not to port to Windows 3.0, leaving the door open for Microsoft to release Excel, which then cut Lotus off at the knees.
Re: I don't get it (Score:4, Insightful)
MacPaint was the early hook for the Mac, Word and Excel gave it some more momentum, but the killer app was WYSIWYG DTP.
Re: (Score:1)
Insightful? Maybe, but all that was eons after Visicalc (in computer terms, that is).
I was there. Visicalc was back then what Excel is today. Business were starting to pick up its importance but hardware certainly hindered many applications -- most things were 8-bit back then, 16-bit would soon appear and make life more amenable. After Visicalc, Wordstar would rise to fame and become even more of a killer application than Visicalc itself.
Back then Microsoft had a correct view with bad products (Multiplan+Wo
Re: (Score:2)
Well, it took them a while, but when they came out with it their 8086/8 C compiler was a lot better than the competition. And they didn't lose their ability to improve their software and develop new programs, that was the fate of many of their office competitors like Lotus and Ashton-Tate (dBASE, showing a useful database could run on CP/M 8080/Z-80 systems, which had more computing power than early
Re: (Score:2)
That was a WordPerfect technical management problem. Unlike Ashton-Tate, Borland, and Lotus, WordPerfect got a non-crashing Windows version to the market. It was somewhat late I think, probably in part because they didn't get with the program and insisted on using their own printer drivers which they were so proud of from the DOS days, but it
Re: (Score:2)
Between Petzold and the SDK documentation, I didn't find it difficult to get Windows documentation, and I have to wonder if that was a problem for companies not being able to develop large programs than didn't GP fault. That I interpret as these companies, aside from WordPerfect, and perhaps Borland (don't know their story) having lost their ability to write software that works, a very common failure of that era. And a "secret" of Microsoft's success, which until Ballmer had a really good programmer at t
Re: (Score:2)
If Mr. Turbo Pascal had been in charge of strategic decisions, due diligence in buying companies like Ashton-Tate which had utterly failed by then, etc., maybe. But as far as I can tell, he continued to work on development tool
Re: (Score:2)
Excel was first released and highly successful on the Mac. In fact, it was the second time that early release on an Apple computer 'saved' that computer from obscurity.
Initially Microsoft's spreadsheet entry for the Mac was named MultiPlan. Excel was a rebranding update.
Re: (Score:2)
I agree with that assessment.
The problem with fast successes is a general fear of changing what already works. The decision isn’t easy. The IBM PC XT was an expensive computer that for many aspects is inferior to other computers at the time. The rise of the IBM Compatible with MS DOS was a surprise to the industry.
Then during the time to upgrade to a windowing system OS2 had a lot of buzz while more expensive then windows had a lot of cool features. Also MS Dos was the big player and Lotus for DOS
Re: (Score:2)
Re: (Score:2)
Lotus 123 had a distinct improvement though, in that it had some relatively easy to naviage help that was quickly accessed with a single key and without referring to a printed manual. As such it was very easy to learn. In many ways I think it's better than modern Excel for that reason which has a really lousy help system and too many complicated ways to doing things that should be simple.
Re: (Score:2)
What's the lesson here? (Score:2)
That the editors don't?
VisiCalc on a Trash Eighty (Score:5, Interesting)
Dad brought home a TRS-80 just to run VisiCalc. That same week there was a Basic programming book at the school book fair.
Even at age 5, my software ambition immediately led me into memory issues. Little did I know that's what I'd be spending a good decade of my adulthood doing.
It was awesome watching the home garages of Northern California swap hot-rod, automobile culture for micro computer culture. I remember very vividly being able to ride my BMX around the neighborhood and see Altairs, Apple IIs and- oh wow a Vic 20. And the Dads would show you a little bit of something on each.
There's lots of things to love about our everyone-owns-a-mobile-device-that-runs-javascript world, but I can't help but think something went terribly wrong. I dunno. Get off my lawn.
Re:VisiCalc on a Trash Eighty (Score:5, Insightful)
Something did go terribly wrong. The fact that here on Slashdot, the editors feel the need to tell us things that we already know.
We're not idiots, msmash. We know the difference between 32KB, 32MB and 32GB.
Re: (Score:2)
We're not idiots, msmash.
However, we do appear to not recognize a quote.
Can't read the WSJ story, but here's what happened (Score:5, Interesting)
Mod Up (Score:2)
I hope the drain of the legal battles and power of distractions to derail a project is the lesson internalized from most companies...
I don't know at this point how many companies are even thinking or looking that far back though. It seems unlikely.
Re:Can't read the WSJ story, but here's what happe (Score:5, Informative)
Recompiling?? In this era of 8 and 16 bit computers, assembly was king, both original VisiCalc on the 6502 and Lotus 1-2-3 on the 8088 were written in assembly, as well I'm sure VisiCalc ports to the 8080 family ending with the Z-80. Lotus in fact attempted a C rewrite but they'd long lost their ability to develop significant software, a common fate of companies of that era and a "secret" to Microsoft's success.
As long as you were stuck with 65KiB minus whatever the system took, 48KiB was a very common DRAM budget, for both code and the spreadsheet's data, "new features" weren't such a big thing, that came with the fielding of 8086/8 and 68000 based computers that could address a lot more memory.
Knowing some of the people involved including Bob Frankston, and some of the things they did in addition to VisiCalc, your general thesis is very very wrong.
Re: (Score:2)
Dan Bricklin first dreamed it up in a classroom at Harvard Business School...
I think I can see why it ultimately failed.
Re: (Score:2)
insane and unsustainable royalty rate, according to Wikipedia 36-50%
You know, 30% is considered "standard" nowadays, if you ask the likes of Apple and Valve. On consoles it can be all over the place, I recall Microsoft used to take 50% for games sold through Xbox Live Arcade. (Or at least through one of their related publishing programs? I don't know, I never had an Xbox.)
Re: (Score:2)
That's more than sustainable for a publisher who can primarily market on the Web, do fulfillment using a website and it and email for primary communications with users or their custom app, and send the software's bits over the Internet.
In the bad old days, fulfillment was vastly more expensive, multiple physical SKUs (for example, Apple II and TRS-80)
Now fits on a single DVD! (Score:1)
Re:Now fits on a single DVD! (Score:4, Informative)
My dad actually used VisiCalc to create spreadsheets for calculating feed ratios for hundreds of cattle at a dairy operation. He could tweak some numbers here and there, recalculate, and then bang out a copy on paper to put on a clipboard for the guys who would be dumping yards of ingredients into a mixing truck in order to get the nutrition right.
Beats the hell out of using an adding machine, or whatever was done before.
Re:A Non Story (Score:4, Interesting)
Re: (Score:3)
Today I'd hate to use anything less than Excel '97 because I've been spoiled by modern interfaces. But I used to have a 5150 with 448kB RAM (384kB on an ISA card) and Lotus 1-2-3 1.0; the PC-1 was a mere 4.77 MHz, though 16-bit. I think Lotus might have been capable of running in 64kB... nope [wikipedia.org]. It required 192kB. Scrolling around was poky, but maybe that was partly because I had the original text card at the time. (It was slow, but at least it would do 80x25.) I took a class in Lotus, and the textbook came w
Huh? (Score:5, Insightful)
The power of simplicity? I can't see how that is relevant today. It certainly doesn't influence any decision maker I know of. Today it's all about making things as convoluted as possible by shoving in every bell, whistle, nut, and bolt. Forget about opening a piece of software in a second and using it. Now it's all about signing in, waiting for background data transfers, synchronization, and finally system stabilization.
As for the hypergrowth portion, that is in play. In fact, that seems to be the only thing in play. How fast can one grow to burn through at least $1 billion yet claim you're producing something?
Like everything else, because VisiCalc was usable, it had to be killed. Just like the most perfect word processing program, WordPerfect 5.1 for DOS. Nothing compares to the speed and ease of use of that program, and never will. Why? Because simplicity is no longer something to aspire to.
Re:Huh? (Score:4, Informative)
Re: (Score:3)
"WordPerfect" was a laughable misnomer for that pile of shit. There was no ease of use, but rote memorization of row upon row of arcane keyboard shortcuts. Your layout would change every time you switched printers. It wasn't WYSIWYG so your layout would be guesswork and lots of iteration. The vaunted 'code view' was not a feature, but a necessity for troubleshooting because WP would place formatting codes willy-nilly and screw up your document.
MS Word 1.0 put it into the ground, and for good reason: it was
Re: (Score:3)
"WordPerfect" was a laughable misnomer for that pile of shit. There was no ease of use, but rote memorization of row upon row of arcane keyboard shortcuts.
So, you are saying it was EMACS...
Re: (Score:2)
"WordPerfect" was a laughable misnomer for that pile of shit. There was no ease of use, but rote memorization of row upon row of arcane keyboard shortcuts.
So, you are saying it was EMACS...
No, vi.
We have easy fast text-processors today Re:Huh? (Score:1)
Just like the most perfect word processing program, WordPerfect 5.1 for DOS. Nothing compares to the speed and ease of use of that program, and never will.
Not true.
For very simple things that stay local, I use pencil and paper or if I don't have it, the note-taker tool on my phone.
For things slightly more complex I use the bare-bones text editor of whatever computer or device I have handy. They are fast, they allow me to fix typos, do cut-and-paste within the document and to email or other places, and if the device supports it, print.
It's only when I need things like tabs, bold, etc. that I might step up to the next level. Heck, on some platforms, the bare-
Died of self-inflicted wounds (Score:5, Interesting)
VisiCalc killed itself long before Excel showed up on the scene. Basically, the owners of the company decided to invest in a GUI system called VisiOn to replay MSDOS. Oops. The story is fascinating.
https://www.filfre.net/2018/06/doing-windows-part-1-ms-dos-and-its-discontents/ [filfre.net]
Re: (Score:2)
And don't forget that Lotus 1-2-3 was the intermediate king of the spreadsheet hill.
Re: (Score:1)
I just object to the lead-in, which is suggestive of a 'flash in the pan' story.
- Killer Apps became something of a mythical unicorn thing, but not necessarily in a bad way. People understood that the purpose was to snag a customer for the computer on the first application. Once they had the computer, more applications would follow;
- Although VisiCalc died, it begat Lotus 1-2-3, Multiplan, Quattro, Excel, and a thousand others. The entire spreadsheet category is indispensable today. There are multiple g
A link to the original article... (Score:1)
It may be paywalled but that's never stopped /. before...
https://www.wsj.com/articles/4... [wsj.com]
How to defeat the WSJ paywall (Score:1)
Use private browsing or purge your cookies and private data before visiting the link.
At least for now the WSJ is letting you have at least one free article to read.
It wouldn't surprise me if this article goes behind the "never free" paywall soon.
4kB? (Score:2)
Holy jebus, 4kB must have hurt bad. I was going to say that I've never seen an Apple 2 series machine with less than 64kB, but I've seen an actual Apple ][ (once) so it couldn't have had any more than 48kB. Looking at the numbers [stackexchange.com], it seems like virtually all Apple 2s had at least 16kB (the least configuration for the ][+).
Re: (Score:3)
Oh the memories. I wish I hadn't donated all that old equipment when I made a cross country move, it would be fun to
Re: (Score:2)
Cheap Cheap (Score:2)
Even the cheapest of today's iPhones have tens of thousands of times as much RAM.
Cheapest iPhones? Oh, so you're saying an overpriced phone has a lot more RAM? Interesting.
Re: (Score:3)
Even the cheapest of today's iPhones have tens of thousands of times as much RAM.
Cheapest iPhones? Oh, so you're saying an overpriced phone has a lot more RAM? Interesting.
They should have said "entry level". The words "cheap" and "iPhone" (or any Apple product) do not belong in the same sentence.